Skip to content

web search done right

Ftr/customized ranking (#38)

Mikkel Denkerc2b851bb4430
diff --git a/Cargo.lock b/Cargo.lock
index 85511793..0000b770 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -349,6 +349,12 @@ dependencies = [
"pkg-config",
]
+[[package]]
+name = "cast"
+version = "0.3.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "37b2a672a2cb129a2e41c10b1224bb368f9f37a2b16b612598138befd7b37eb5"
+
[[package]]
name = "cc"
version = "1.0.73"
@@ -405,6 +411,17 @@ dependencies = [
"libloading",
]
+[[package]]
+name = "clap"
+version = "2.34.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "a0610544180c38b88101fecf2dd634b174a62eef6946f84dfc6a7127512b381c"
+dependencies = [
+ "bitflags",
+ "textwrap 0.11.0",
+ "unicode-width",
+]
+
[[package]]
name = "clap"
version = "3.1.18"
@@ -419,7 +436,7 @@ dependencies = [
"lazy_static",
"strsim",
"termcolor",
- "textwrap",
+ "textwrap 0.15.0",
]
[[package]]
@@ -503,6 +520,42 @@ dependencies = [
"cfg-if",
]
+[[package]]
+name = "criterion"
+version = "0.3.6"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b01d6de93b2b6c65e17c634a26653a29d107b3c98c607c765bf38d041531cd8f"
+dependencies = [
+ "atty",
+ "cast",
+ "clap 2.34.0",
+ "criterion-plot",
+ "csv",
+ "itertools",
+ "lazy_static",
+ "num-traits",
+ "oorandom",
+ "plotters",
+ "rayon",
+ "regex",
+ "serde",
+ "serde_cbor",
+ "serde_derive",
+ "serde_json",
+ "tinytemplate",
+ "walkdir",
+]
+
+[[package]]
+name = "criterion-plot"
+version = "0.4.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "2673cc8207403546f45f5fd319a974b1e6983ad1a3ee7e6041650013be041876"
+dependencies = [
+ "cast",
+ "itertools",
+]
+
[[package]]
name = "crossbeam-channel"
version = "0.5.4"
@@ -624,7 +677,8 @@ dependencies = [
"bincode",
"bzip2",
"chrono",
- "clap",
+ "clap 3.1.18",
+ "criterion",
"csv",
"encoding_rs",
"flate2",
@@ -650,10 +704,10 @@ dependencies = [
"regex",
"reqwest",
"rocksdb",
+ "rust-stemmers",
"serde",
"serde_json",
"serde_urlencoded",
- "strum",
"tantivy",
"thiserror",
"tokio",
@@ -1803,6 +1857,12 @@ dependencies = [
"loom",
]
+[[package]]
+name = "oorandom"
+version = "11.1.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "0ab1bc2a289d34bd04a330323ac98a1b4bc82c9d9fcb1e66b63caa84da26b575"
+
[[package]]
name = "openssl"
version = "0.10.40"
@@ -2039,6 +2099,34 @@ version = "0.3.25"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "1df8c4ec4b0627e53bdf214615ad287367e482558cf84b109250b37464dc03ae"
+[[package]]
+name = "plotters"
+version = "0.3.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "716b4eeb6c4a1d3ecc956f75b43ec2e8e8ba80026413e70a3f41fd3313d3492b"
+dependencies = [
+ "num-traits",
+ "plotters-backend",
+ "plotters-svg",
+ "wasm-bindgen",
+ "web-sys",
+]
+
+[[package]]
+name = "plotters-backend"
+version = "0.3.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "193228616381fecdc1224c62e96946dfbc73ff4384fba576e052ff8c1bea8142"
+
+[[package]]
+name = "plotters-svg"
+version = "0.3.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "f9a81d2759aae1dae668f783c308bc5c8ebd191ff4184aaa1b37f65a6ae5a56f"
+dependencies = [
+ "plotters-backend",
+]
+
[[package]]
name = "png"
version = "0.17.5"
@@ -2382,6 +2470,15 @@ version = "1.0.10"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "f3f6f92acf49d1b98f7a81226834412ada05458b7364277387724a237f062695"
+[[package]]
+name = "same-file"
+version = "1.0.6"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "93fc1dc3aaa9bfed95e02e6eadabb4baf7e3078b0bd1b4d7b6b0b68378900502"
+dependencies = [
+ "winapi-util",
+]
+
[[package]]
name = "schannel"
version = "0.1.20"
@@ -2468,6 +2565,16 @@ dependencies = [
"serde_derive",
]
+[[package]]
+name = "serde_cbor"
+version = "0.11.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "2bef2ebfde456fb76bbcf9f59315333decc4fda0b2b44b420243c11e0f5ec1f5"
+dependencies = [
+ "half",
+ "serde",
+]
+
[[package]]
name = "serde_derive"
version = "1.0.137"
@@ -2611,28 +2718,6 @@ version = "0.10.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "73473c0e59e6d5812c5dfe2a064a6444949f089e20eec9a2e5506596494e4623"
-[[package]]
-name = "strum"
-version = "0.24.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "063e6045c0e62079840579a7e47a355ae92f60eb74daaf156fb1e84ba164e63f"
-dependencies = [
- "strum_macros",
-]
-
-[[package]]
-name = "strum_macros"
-version = "0.24.3"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "1e385be0d24f186b4ce2f9982191e7101bb737312ad61c1f2f984f34bcf85d59"
-dependencies = [
- "heck",
- "proc-macro2",
- "quote",
- "rustversion",
- "syn",
-]
-
[[package]]
name = "syn"
version = "1.0.96"
@@ -2791,6 +2876,15 @@ dependencies = [
"winapi",
]
+[[package]]
+name = "textwrap"
+version = "0.11.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "d326610f408c7a4eb6f51c37c330e496b08506c9457c9d34287ecc38809fb060"
+dependencies = [
+ "unicode-width",
+]
+
[[package]]
name = "textwrap"
version = "0.15.0"
@@ -2884,6 +2978,16 @@ dependencies = [
"crunchy",
]
+[[package]]
+name = "tinytemplate"
+version = "1.2.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "be4d6b5f19ff7664e8c98d03e2139cb510db9b0a60b55f8e8709b689d939b6bc"
+dependencies = [
+ "serde",
+ "serde_json",
+]
+
[[package]]
name = "tinyvec"
version = "1.6.0"
@@ -3130,6 +3234,12 @@ dependencies = [
"tinyvec",
]
+[[package]]
+name = "unicode-width"
+version = "0.1.9"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "3ed742d4ea2bd1176e236172c8429aaf54486e7ac098db29ffe6529e0ce50973"
+
[[package]]
name = "unicode-xid"
version = "0.2.3"
@@ -3194,6 +3304,17 @@ version = "0.9.4"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "49874b5167b65d7193b8aba1567f5c7d93d001cafc34600cee003eda787e483f"
+[[package]]
+name = "walkdir"
+version = "2.3.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "808cf2735cd4b6866113f648b791c6adc5714537bc222d9347bb203386ffda56"
+dependencies = [
+ "same-file",
+ "winapi",
+ "winapi-util",
+]
+
[[package]]
name = "want"
version = "0.3.0"
diff --git a/Cargo.toml b/Cargo.toml
index 823388a6..d8599537 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -51,16 +51,25 @@ once_cell = "1.13.1"
regex = "1.6.0"
http = "0.2.8"
kuchiki = "0.8.1"
-strum = { version = "0.24.1", features = ["derive", "strum_macros"] }
tower-http = { version = "0.3.4", features = ["compression-gzip"] }
lalrpop-util = { version = "0.19.8", features = ["lexer"] }
serde_urlencoded = "0.7.1"
+rust-stemmers = "1.2.0"
[dev-dependencies]
+criterion = "0.3.6"
maplit = "1.0.2"
[build-dependencies]
-lalrpop = "0.19.8"
+lalrpop = { version = "0.19.8", features = ["lexer"] }
[profile.release]
debug = true
+
+[[bench]]
+name = "search-preindexed"
+harness = false
+
+[[bench]]
+name = "search-preindexed-goggle"
+harness = false
diff --git a/benches/search-preindexed-goggle.rs b/benches/search-preindexed-goggle.rs
new file mode 100644
index 00000000..b7514378
--- /dev/null
+++ b/benches/search-preindexed-goggle.rs
@@ -0,0 +1,35 @@
+use criterion::{criterion_group, criterion_main, Criterion};
+use cuely::{index::Index, searcher::Searcher};
+
+const INDEX_PATH: &str = "data/index";
+
+macro_rules! bench {
+ ($query:tt, $searcher:ident, $goggle:ident, $c:ident) => {
+ let mut desc = "search '".to_string();
+ desc.push_str($query);
+ desc.push('\'');
+ desc.push_str(" with goggle");
+ $c.bench_function(desc.as_str(), |b| {
+ b.iter(|| {
+ $searcher
+ .search($query, None, Some($goggle.to_string()), None)
+ .unwrap()
+ })
+ });
+ };
+}
+
+pub fn criterion_benchmark(c: &mut Criterion) {
+ let index = Index::open(INDEX_PATH).unwrap();
+ let searcher = Searcher::new(index, None, None);
+ let goggle = include_str!("../testcases/goggles/quickstart.goggle");
+
+ for _ in 0..10 {
+ bench!("the", searcher, goggle, c);
+ bench!("dtu", searcher, goggle, c);
+ bench!("the best", searcher, goggle, c);
+ }
+}
+
+criterion_group!(benches, criterion_benchmark);
+criterion_main!(benches);
diff --git a/benches/search-preindexed.rs b/benches/search-preindexed.rs
new file mode 100644
index 00000000..72c9f05e
--- /dev/null
+++ b/benches/search-preindexed.rs
@@ -0,0 +1,29 @@
+use criterion::{criterion_group, criterion_main, Criterion};
+use cuely::{index::Index, searcher::Searcher};
+
+const INDEX_PATH: &str = "data/index";
+
+macro_rules! bench {
+ ($query:tt, $searcher:ident, $c:ident) => {
+ let mut desc = "search '".to_string();
+ desc.push_str($query);
+ desc.push('\'');
+ $c.bench_function(desc.as_str(), |b| {
+ b.iter(|| $searcher.search($query, None, None, None).unwrap())
+ });
+ };
+}
+
+pub fn criterion_benchmark(c: &mut Criterion) {
+ let index = Index::open(INDEX_PATH).unwrap();
+ let searcher = Searcher::new(index, None, None);
+
+ // for _ in 0..10 {
+ bench!("the", searcher, c);
+ bench!("dtu", searcher, c);
+ bench!("the best", searcher, c);
+ // }
+}
+
+criterion_group!(benches, criterion_benchmark);
+criterion_main!(benches);
diff --git a/configs/indexer/local.toml b/configs/indexer/local.toml
index 088331bd..f5e628ca 100644
--- a/configs/indexer/local.toml
+++ b/configs/indexer/local.toml
@@ -1,6 +1,8 @@
warc_paths_file = "warc.paths"
-centrality_store_path = "./data/centrality"
-limit_warc_files = 1
+centrality_store_path = "./data/centrality/host"
+limit_warc_files = 80
+download_images = false
+batch_size = 8
[warc_source]
# type = "Local"
diff --git a/configs/indexer/profile.toml b/configs/indexer/profile.toml
index dec9e255..1ef3c6f3 100644
--- a/configs/indexer/profile.toml
+++ b/configs/indexer/profile.toml
@@ -1,9 +1,9 @@
warc_paths_file = "warc.paths"
centrality_store_path = "./data/centrality"
-limit_warc_files = 1
+limit_warc_files = 3
download_images = false
[warc_source]
type = "Local"
folder = "./data/warc_files"
-names = ["CC-MAIN-20220116093137-20220116123137-00049.warc.gz"]
+names = ["CC-MAIN-20220116093137-20220116123137-00049.warc.gz", "CC-MAIN-20220116093137-20220116123137-00049_1.warc.gz", "CC-MAIN-20220116093137-20220116123137-00049_2.warc.gz"]
diff --git a/frontend/src/layouts/Layout.astro b/frontend/src/layouts/Layout.astro
index 705ddb47..0ba6345c 100644
--- a/frontend/src/layouts/Layout.astro
+++ b/frontend/src/layouts/Layout.astro
@@ -184,7 +184,7 @@ const { title } = Astro.props;
}
b {
- font-weight: 600 !important;
+ font-weight: 500 !important;
}
pre > code {
diff --git a/frontend/src/pages/privacy-and-happy-lawyers.mdx b/frontend/src/pages/privacy-and-happy-lawyers.mdx
index 64a8bed4..7f6cfe2c 100644
--- a/frontend/src/pages/privacy-and-happy-lawyers.mdx
+++ b/frontend/src/pages/privacy-and-happy-lawyers.mdx
@@ -37,10 +37,6 @@ We use the information we receive through the Services as described in our Priva
The source code for Cuely is distributed under the license described in the [git repository.](https://github.com/Cuely/Cuely/blob/main/LICENSE.md)
-#### Proprietary Rights
-
-Cuely does not grant you any intellectual property rights in the Services that are not specifically stated in these Terms. For example, these Terms do not provide the right to use any of Cuely's copyrights, trade names, trademarks, service marks, logos, domain names, or other distinctive brand features.
-
#### Indemnification
You agree to defend, indemnify and hold harmless Cuely, its contractors, contributors, licensors, and partners, and their respective directors, officers, employees and agents ("Indemnified Parties") from and against any and all third party claims and expenses, including attorneys' fees, arising out of or related to your use of the Services.
diff --git a/justfile b/justfile
index 733721a9..61a101e6 100644
--- a/justfile
+++ b/justfile
@@ -24,7 +24,7 @@
@local:
rm -rf data/index/
- RUSTFLAGS="-C target-cpu=native" cargo run --release -- indexer local configs/indexer/local.toml
+ cargo run --release -- indexer local configs/indexer/local.toml
mv data/index/CC-MAIN-*/* data/index/
just frontend
diff --git a/src/bangs.rs b/src/bangs.rs
index a41025e8..8d959b9b 100644
--- a/src/bangs.rs
+++ b/src/bangs.rs
@@ -125,7 +125,7 @@ mod tests {
use tantivy::tokenizer::TokenizerManager;
- use crate::{ranking::signal_aggregator::SignalAggregator, schema::create_schema};
+ use crate::{ranking::goggles::SignalAggregator, schema::create_schema};
use super::*;
diff --git a/src/entrypoint/indexer.rs b/src/entrypoint/indexer.rs
index 8446cfee..9168d4c0 100644
--- a/src/entrypoint/indexer.rs
+++ b/src/entrypoint/indexer.rs
@@ -135,7 +135,6 @@ impl Map<IndexingWorker, FrozenIndex> for Job {
debug!("{:?}", err);
}
}
- index.commit().unwrap();
if self.download_images {
info!("downloading images");
index.download_pending_images();
@@ -144,6 +143,8 @@ impl Map<IndexingWorker, FrozenIndex> for Job {
std::fs::remove_file(file).ok();
}
+ index.commit().unwrap();
+ index.inverted_index.merge_all_segments().unwrap();
info!("{} done", name);
@@ -250,7 +251,7 @@ impl Indexer {
config.webgraph_path.clone(),
);
- warc_paths
+ let mut index = warc_paths
.into_iter()
.take(config.limit_warc_files.unwrap_or(usize::MAX))
.chunks(config.batch_size.unwrap_or(1))
@@ -265,6 +266,12 @@ impl Indexer {
.unwrap_or_else(|| "data/index".to_string()),
})
.collect_vec()
+ // .into_iter()
+ // .map(|job| job.map(&worker))
+ // .fold(None, |acc: Option<Index>, elem: FrozenIndex| match acc {
+ // Some(acc) => Some(acc.reduce(elem)),
+ // None => Some(elem.into()),
+ // });
.into_par_iter()
.panic_fuse()
.map(|job| job.map(&worker))
@@ -279,6 +286,10 @@ impl Indexer {
},
);
+ if let Some(index) = index.as_mut() {
+ index.inverted_index.merge_all_segments().unwrap();
+ }
+
Ok(())
}
}
diff --git a/src/frontend/search.rs b/src/frontend/search.rs
index 25966423..e6a2920e 100644
--- a/src/frontend/search.rs
+++ b/src/frontend/search.rs
@@ -269,6 +269,16 @@ pub async fn route(
let skip_pages = params.get("p").and_then(|p| p.parse().ok());
+ let mut goggle = None;
+
+ if let Some(url) = params.get("goggle") {
+ if let Ok(res) = reqwest::get(url).await {
+ if let Ok(text) = res.text().await {
+ goggle = Some(text);
+ }
+ }
+ }
+
let selected_region = params.get("gl").and_then(|gl| {
if let Ok(region) = Region::from_gl(gl) {
Some(region)
@@ -279,7 +289,7 @@ pub async fn route(
match state
.searcher
- .search(query.as_str(), selected_region, None, skip_pages)
+ .search(query.as_str(), selected_region, goggle, skip_pages)
{
Ok(result) => match result {
SearchResult::Websites(result) => {
diff --git a/src/index.rs b/src/index.rs
index 998b6aaf..0021831f 100644
--- a/src/index.rs
+++ b/src/index.rs
@@ -45,7 +45,7 @@ const REGION_COUNT_FILE_NAME: &str = "region_count.json";
const IMAGE_WEBPAGE_CENTRALITY_THRESHOLD: f64 = 0.0;
pub struct Index {
- inverted_index: InvertedIndex,
+ pub inverted_index: InvertedIndex,
favicon_store: FaviconStore,
primary_image_store: PrimaryImageStore,
favicon_downloader: ImageDownloader<String>,
@@ -105,7 +105,6 @@ impl Index {
}
pub fn commit(&mut self) -> Result<()> {
- self.inverted_index.merge_all_segments()?;
self.spell_dictionary.commit()?;
self.inverted_index.commit()?;
self.region_count.commit();
@@ -292,7 +291,7 @@ impl From<Index> for FrozenIndex {
#[cfg(test)]
mod tests {
- use crate::ranking::{signal_aggregator::SignalAggregator, Ranker};
+ use crate::ranking::{goggles::SignalAggregator, Ranker};
use super::*;
diff --git a/src/inverted_index.rs b/src/inverted_index.rs
index 9bc94d65..2564d39c 100644
--- a/src/inverted_index.rs
+++ b/src/inverted_index.rs
@@ -26,6 +26,7 @@ use crate::image_store::Image;
use crate::query::Query;
use crate::schema::{Field, ALL_FIELDS};
use crate::snippet;
+use crate::webpage::region::Region;
use crate::webpage::{StoredPrimaryImage, Webpage};
use crate::Result;
use crate::{schema::create_schema, tokenizer::Tokenizer};
@@ -64,7 +65,7 @@ impl InvertedIndex {
.tokenizers()
.register(tokenizer.as_str(), tokenizer);
- let writer = tantivy_index.writer(10_000_000_000)?;
+ let writer = tantivy_index.writer_with_num_threads(1, 4_000_000_000)?;
let merge_policy = NoMergePolicy::default();
writer.set_merge_policy(Box::new(merge_policy));
@@ -137,6 +138,7 @@ impl InvertedIndex {
&page.body,
&page.dirty_body,
&page.description,
+ &page.region,
&searcher,
)?;
}
@@ -148,16 +150,23 @@ impl InvertedIndex {
}
pub fn merge_all_segments(&mut self) -> Result<()> {
- let segment_ids: Vec<_> = self
+ let segments: Vec<_> = self
.tantivy_index
.load_metas()?
.segments
.into_iter()
- .map(|segment| segment.id())
.collect();
- if !segment_ids.is_empty() {
+ if segments.len() > 1 {
+ let segment_ids: Vec<_> = segments.iter().map(|segment| segment.id()).collect();
self.writer.merge(&segment_ids[..]).wait()?;
+
+ let path = Path::new(&self.path);
+ for segment in segments {
+ for file in segment.list_files() {
+ std::fs::remove_file(path.join(file)).ok();
+ }
+ }
}
Ok(())
@@ -247,6 +256,7 @@ pub struct RetrievedWebpage {
pub favicon: Option<Image>,
pub primary_image: Option<StoredPrimaryImage>,
pub updated_time: Option<NaiveDateTime>,
+ pub region: Region,
}
impl From<Document> for RetrievedWebpage {
@@ -305,24 +315,32 @@ impl From<Document> for RetrievedWebpage {
}
}
}
- Field::StemmedAllBody => {
+ Field::AllBody => {
webpage.dirty_body = value
.value
.as_text()
- .expect("Stemmed all body field should be text")
+ .expect("All body field should be text")
.to_string()
}
+ Field::Region => {
+ webpage.region = {
+ let id = value.value.as_u64().unwrap();
+ Region::from_id(id)
+ }
+ }
Field::BacklinkText
| Field::Centrality
- | Field::Host
+ | Field::Site
| Field::StemmedTitle
| Field::CleanBody
| Field::Domain
| Field::DomainIfHomepage
| Field::IsHomepage
| Field::NumTrackers
- | Field::Region
- | Field::AllBody
+ | Field::NumCleanBodyTokens
+ | Field::NumDescriptionTokens
+ | Field::NumTitleTokens
+ | Field::NumUrlTokens
| Field::FetchTimeMs => {}
}
}
@@ -336,7 +354,7 @@ mod tests {
use maplit::hashset;
use crate::{
- ranking::{signal_aggregator::SignalAggregator, Ranker},
+ ranking::{goggles::SignalAggregator, Ranker},
webpage::{region::RegionCount, Link},
};
diff --git a/src/lib.rs b/src/lib.rs
index 1fa5191b..58828ed8 100644
--- a/src/lib.rs
+++ b/src/lib.rs
@@ -40,13 +40,13 @@ mod exponential_backoff;
mod frontend;
mod image_downloader;
mod image_store;
-mod index;
+pub mod index;
mod kv;
mod query;
mod ranking;
mod schema;
mod schema_org;
-mod searcher;
+pub mod searcher;
mod snippet;
mod spell;
mod tokenizer;
@@ -188,6 +188,9 @@ pub enum Error {
#[error("Unknown region")]
UnknownRegion,
+
+ #[error("String is not float")]
+ ParseFloat(#[from] std::num::ParseFloatError),
}
pub(crate) type Result<T> = std::result::Result<T, Error>;
diff --git a/src/query/intersection.rs b/src/query/intersection.rs
new file mode 100644
index 00000000..381d5b8b
--- /dev/null
+++ b/src/query/intersection.rs
@@ -0,0 +1,132 @@
+// Cuely is an open source web search engine.
+// Copyright (C) 2022 Cuely ApS
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program. If not, see <https://www.gnu.org/licenses/>.
+
+use tantivy::query::Scorer;
+use tantivy::{DocId, DocSet, Score};
+
+/// Creates a `DocSet` that iterate through the intersection of two or more `DocSet`s.
+pub struct Intersection<TDocSet: DocSet, TOtherDocSet: DocSet = Box<dyn Scorer>> {
+ left: TDocSet,
+ right: TDocSet,
+ others: Vec<TOtherDocSet>,
+}
+
+fn go_to_first_doc<TDocSet: DocSet>(docsets: &mut [TDocSet]) -> DocId {
+ assert!(!docsets.is_empty());
+ let mut candidate = docsets.iter().map(TDocSet::doc).max().unwrap();
+ 'outer: loop {
+ for docset in docsets.iter_mut() {
+ let seek_doc = docset.seek(candidate);
+ if seek_doc > candidate {
+ candidate = docset.doc();
+ continue 'outer;
+ }
+ }
+ return candidate;
+ }
+}
+
+impl<TDocSet: DocSet> Intersection<TDocSet, TDocSet> {
+ pub(crate) fn new(mut docsets: Vec<TDocSet>) -> Intersection<TDocSet, TDocSet> {
+ let num_docsets = docsets.len();
+ assert!(num_docsets >= 2);
+ docsets.sort_by_key(|docset| docset.size_hint());
+ go_to_first_doc(&mut docsets);
+ let left = docsets.remove(0);
+ let right = docsets.remove(0);
+ Intersection {
+ left,
+ right,
+ others: docsets,
+ }
+ }
+}
+
+impl<TDocSet: DocSet, TOtherDocSet: DocSet> DocSet for Intersection<TDocSet, TOtherDocSet> {
+ fn advance(&mut self) -> DocId {
+ let (left, right) = (&mut self.left, &mut self.right);
+ let mut candidate = left.advance();
+
+ 'outer: loop {
+ // In the first part we look for a document in the intersection
+ // of the two rarest `DocSet` in the intersection.
+
+ loop {
+ let right_doc = right.seek(candidate);
+ candidate = left.seek(right_doc);
+ if candidate == right_doc {
+ break;
+ }
+ }
+
+ debug_assert_eq!(left.doc(), right.doc());
+ // test the remaining scorers;
+ for docset in self.others.iter_mut() {
+ let seek_doc = docset.seek(candidate);
+ if seek_doc > candidate {
+ candidate = left.seek(seek_doc);
+ continue 'outer;
+ }
+ }
+ debug_assert_eq!(candidate, self.left.doc());
+ debug_assert_eq!(candidate, self.right.doc());
+ debug_assert!(self.others.iter().all(|docset| docset.doc() == candidate));
+ return candidate;
+ }
+ }
+
+ fn seek(&mut self, target: DocId) -> DocId {
+ self.left.seek(target);
+ let mut docsets: Vec<&mut dyn DocSet> = vec![&mut self.left, &mut self.right];
+ for docset in &mut self.others {
+ docsets.push(docset);
+ }
+ let doc = go_to_first_doc(&mut docsets[..]);
+ debug_assert!(docsets.iter().all(|docset| docset.doc() == doc));
+ debug_assert!(doc >= target);
+ doc
+ }
+
+ fn doc(&self) -> DocId {
+ self.left.doc()
+ }
+
+ fn size_hint(&self) -> u32 {
+ self.left.size_hint()
+ }
+}
+
+impl<TScorer, TOtherScorer> Scorer for Intersection<TScorer, TOtherScorer>
+where
+ TScorer: Scorer,
+ TOtherScorer: Scorer,
+{
+ fn score(&mut self) -> Score {
+ self.left.score()
+ + self.right.score()
+ + self.others.iter_mut().map(Scorer::score).sum::<Score>()
+ }
+}
+
+impl<TDocSet: DocSet> Intersection<TDocSet, TDocSet> {
+ pub(crate) fn docset_mut_specialized(&mut self, ord: usize) -> &mut TDocSet {
+ match ord {
+ 0 => &mut self.left,
+ 1 => &mut self.right,
+ n => &mut self.others[n - 2],
+ }
+ }
+}
diff --git a/src/query/mod.rs b/src/query/mod.rs
index 4f8b2839..18984cf2 100644
--- a/src/query/mod.rs
+++ b/src/query/mod.rs
@@ -14,14 +14,18 @@
// You should have received a copy of the GNU Affero General Public License
// along with this program. If not, see <https://www.gnu.org/licenses/>.
-use crate::{ranking::signal_aggregator::SignalAggregator, Result};
+use crate::{
+ ranking::goggles::{Goggle, SignalAggregator},
+ Result,
+};
use std::{collections::HashMap, sync::Arc};
use tantivy::{
- query::{AllQuery, BooleanQuery, Occur},
+ query::{AllQuery, BooleanQuery, Occur, QueryClone},
schema::Schema,
tokenizer::TokenizerManager,
};
+pub mod intersection;
pub mod parser;
use parser::Term;
@@ -76,6 +80,14 @@ impl Query {
})
}
+ pub fn set_goggle(&mut self, goggle: &Goggle, schema: &Schema) {
+ let mut subqueries = vec![(Occur::Must, self.tantivy_query.box_clone())];
+
+ subqueries.append(&mut goggle.as_tantivy(schema));
+
+ self.tantivy_query = Box::new(BooleanQuery::new(subqueries))
+ }
+
pub fn simple_terms(&self) -> Vec<String> {
self.terms
.iter()
diff --git a/src/query/parser.rs b/src/query/parser.rs
index 1ef5417a..03730c5c 100644
--- a/src/query/parser.rs
+++ b/src/query/parser.rs
@@ -22,7 +22,7 @@ use tantivy::{
use crate::{
bangs::BANG_PREFIX,
- ranking::signal_aggregator::FieldBoost,
+ ranking::goggles::FieldBoost,
schema::{Field, ALL_FIELDS},
};
@@ -185,7 +185,7 @@ impl Term {
.filter(|(field, _)| {
matches!(
ALL_FIELDS[field.field_id() as usize],
- Field::Domain | Field::Host
+ Field::Domain | Field::Site
)
})
.into_iter()
diff --git a/src/ranking/bm25.rs b/src/ranking/bm25.rs
new file mode 100644
index 00000000..7c45a5b9
--- /dev/null
+++ b/src/ranking/bm25.rs
@@ -0,0 +1,160 @@
+// source: https://github.com/quickwit-oss/tantivy/blob/main/src/query/bm25.rs
+
+use serde::{Deserialize, Serialize};
+
+use tantivy::fieldnorm::FieldNormReader;
+use tantivy::query::Explanation;
+use tantivy::{Score, Searcher, Term};
+
+const K1: Score = 1.2;
+const B: Score = 0.75;
+
+pub(crate) fn idf(doc_freq: u64, doc_count: u64) -> Score {
+ assert!(doc_count >= doc_freq, "{} >= {}", doc_count, doc_freq);
+ let x = ((doc_count - doc_freq) as Score + 0.5) / (doc_freq as Score + 0.5);
+ (1.0 + x).ln()
+}
+
+fn cached_tf_component(fieldnorm: u32, average_fieldnorm: Score) -> Score {
+ K1 * (1.0 - B + B * fieldnorm as Score / average_fieldnorm)
+}
+
+fn compute_tf_cache(average_fieldnorm: Score) -> [Score; 256] {
+ let mut cache: [Score; 256] = [0.0; 256];
+ for (fieldnorm_id, cache_mut) in cache.iter_mut().enumerate() {
+ let fieldnorm = FieldNormReader::id_to_fieldnorm(fieldnorm_id as u8);
+ *cache_mut = cached_tf_component(fieldnorm, average_fieldnorm);
+ }
+ cache
+}
+
+#[derive(Clone, PartialEq, Debug, Serialize, Deserialize)]
+pub struct Bm25Params {
+ pub idf: Score,
+ pub avg_fieldnorm: Score,
+}
+
+#[derive(Clone)]
+pub struct Bm25Weight {
+ idf_explain: Explanation,
+ weight: Score,
+ cache: [Score; 256],
+ average_fieldnorm: Score,
+}
+
+impl Bm25Weight {
+ pub fn boost_by(&self, boost: Score) -> Bm25Weight {
+ Bm25Weight {
+ idf_explain: self.idf_explain.clone(),
+ weight: self.weight * boost,
+ cache: self.cache,
+ average_fieldnorm: self.average_fieldnorm,
+ }
+ }
+
+ pub fn for_terms(searcher: &Searcher, terms: &[Term]) -> tantivy::Result<Bm25Weight> {
+ assert!(!terms.is_empty(), "Bm25 requires at least one term");
+ let field = terms[0].field();
+ for term in &terms[1..] {
+ assert_eq!(
+ term.field(),
+ field,
+ "All terms must belong to the same field."
+ );
+ }
+
+ let mut total_num_tokens = 0u64;
+ let mut total_num_docs = 0u64;
+ for segment_reader in searcher.segment_readers() {
+ let inverted_index = segment_reader.inverted_index(field)?;
+ total_num_tokens += inverted_index.total_num_tokens();
+ total_num_docs += u64::from(segment_reader.max_doc());
+ }
+ let average_fieldnorm = total_num_tokens as Score / total_num_docs as Score;
+
+ if terms.len() == 1 {
+ let term_doc_freq = searcher.doc_freq(&terms[0])?;
+ Ok(Bm25Weight::for_one_term(
+ term_doc_freq,
+ total_num_docs,
+ average_fieldnorm,
+ ))
+ } else {
+ let mut idf_sum: Score = 0.0;
+ for term in terms {
+ let term_doc_freq = searcher.doc_freq(term)?;
+ idf_sum += idf(term_doc_freq, total_num_docs);
+ }
+ let idf_explain = Explanation::new("idf", idf_sum);
+ Ok(Bm25Weight::new(idf_explain, average_fieldnorm))
+ }
+ }
+
+ pub fn for_one_term(
+ term_doc_freq: u64,
+ total_num_docs: u64,
+ avg_fieldnorm: Score,
+ ) -> Bm25Weight {
+ let idf = idf(term_doc_freq, total_num_docs);
+ let mut idf_explain =
+ Explanation::new("idf, computed as log(1 + (N - n + 0.5) / (n + 0.5))", idf);
+ idf_explain.add_const(
+ "n, number of docs containing this term",
+ term_doc_freq as Score,
+ );
+ idf_explain.add_const("N, total number of docs", total_num_docs as Score);
+ Bm25Weight::new(idf_explain, avg_fieldnorm)
+ }
+
+ pub(crate) fn new(idf_explain: Explanation, average_fieldnorm: Score) -> Bm25Weight {
+ let weight = idf_explain.value() * (1.0 + K1);
+ Bm25Weight {
+ idf_explain,
+ weight,
+ cache: compute_tf_cache(average_fieldnorm),
+ average_fieldnorm,
+ }
+ }
+
+ #[inline]
+ pub fn score(&self, fieldnorm_id: u8, term_freq: u32) -> Score {
+ self.weight * self.tf_factor(fieldnorm_id, term_freq)
+ }
+
+ #[inline]
+ pub(crate) fn tf_factor(&self, fieldnorm_id: u8, term_freq: u32) -> Score {
+ let term_freq = term_freq as Score;
+ let norm = self.cache[fieldnorm_id as usize];
+ term_freq / (term_freq + norm)
+ }
+
+ pub fn explain(&self, fieldnorm_id: u8, term_freq: u32) -> Explanation {
+ // The explain format is directly copied from Lucene's.
+ // (So, Kudos to Lucene)
+ let score = self.score(fieldnorm_id, term_freq);
+
+ let norm = self.cache[fieldnorm_id as usize];
+ let term_freq = term_freq as Score;
+ let right_factor = term_freq / (term_freq + norm);
+
+ let mut tf_explanation = Explanation::new(
+ "freq / (freq + k1 * (1 - b + b * dl / avgdl))",
+ right_factor,
+ );
+
+ tf_explanation.add_const("freq, occurrences of term within document", term_freq);
+ tf_explanation.add_const("k1, term saturation parameter", K1);
+ tf_explanation.add_const("b, length normalization parameter", B);
+ tf_explanation.add_const(
+ "dl, length of field",
+ FieldNormReader::id_to_fieldnorm(fieldnorm_id) as Score,
+ );
+ tf_explanation.add_const("avgdl, average length of field", self.average_fieldnorm);
+
+ let mut explanation = Explanation::new("TermQuery, product of...", score);
+ explanation.add_detail(Explanation::new("(K1+1)", K1 + 1.0));
+ explanation.add_detail(self.idf_explain.clone());
+ explanation.add_detail(tf_explanation);
+ explanation
+ }
+}
diff --git a/src/ranking/centrality_store.rs b/src/ranking/centrality_store.rs
index 5b89898f..f2c10677 100644
--- a/src/ranking/centrality_store.rs
+++ b/src/ranking/centrality_store.rs
@@ -43,6 +43,10 @@ impl CentralityStore {
self.insert(key, value);
});
+ self.flush();
+ }
+
+ pub fn flush(&self) {
self.inner.flush();
}
}
diff --git a/src/ranking/goggles/ast.rs b/src/ranking/goggles/ast.rs
new file mode 100644
index 00000000..54079016
--- /dev/null
+++ b/src/ranking/goggles/ast.rs
@@ -0,0 +1,278 @@
+// Cuely is an open source web search engine.
+// Copyright (C) 2022 Cuely ApS
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program. If not, see <https://www.gnu.org/licenses/>.
+
+use crate::Error;
+use crate::Result as CrateResult;
+use lalrpop_util::lalrpop_mod;
+use regex::Regex;
+
+lalrpop_mod!(pub parser, "/ranking/goggles/parser.rs");
+
+pub static PARSER: once_cell::sync::Lazy<parser::BlocksParser> =
+ once_cell::sync::Lazy::new(parser::BlocksParser::new);
+
+#[derive(Debug, PartialEq, Eq)]
+pub enum Target {
+ Signal(String),
+ Field(String),
+}
+
+#[derive(Debug, PartialEq, Eq)]
+pub struct RawAlteration {
+ pub target: Target,
+ pub score: String,
+}
+
+#[derive(Debug, PartialEq, Eq)]
+pub struct RawGoggle {
+ pub comments: Vec<Comment>,
+ pub instructions: Vec<RawInstruction>,
+ pub alterations: Vec<RawAlteration>,
+}
+
+impl From<Vec<GoggleBlock>> for RawGoggle {
+ fn from(blocks: Vec<GoggleBlock>) -> Self {
+ let mut alterations = Vec::new();
+ let mut comments = Vec::new();
+ let mut instructions = Vec::new();
+
+ for block in blocks {
+ match block {
+ GoggleBlock::Comment(comment) => comments.push(comment),
+ GoggleBlock::Instruction(instruction) => instructions.push(instruction),
+ GoggleBlock::Alteration(alteration) => alterations.push(alteration),
+ }
+ }
+
+ RawGoggle {
+ comments,
+ instructions,
+ alterations,
+ }
+ }
+}
+
+#[derive(Debug, PartialEq, Eq)]
+pub enum GoggleBlock {
+ Comment(Comment),
+ Instruction(RawInstruction),
+ Alteration(RawAlteration),
+}
+
+#[derive(Debug, PartialEq, Eq)]
+pub enum Comment {
+ Basic(String),
+ Header { key: String, value: String },
+}
+
+#[derive(Debug, PartialEq, Eq)]
+pub struct RawInstruction {
+ pub patterns: Vec<RawPatternPart>,
+ pub options: Vec<RawPatternOption>,
+}
+
+#[derive(Debug, PartialEq, Eq, Clone)]
+pub enum RawPatternPart {
+ Raw(String),
+ Wildcard,
+ Delimeter,
+ Anchor,
+}
+
+#[derive(Debug, PartialEq, Eq)]
+pub enum RawPatternOption {
+ Site(String),
+ InUrl,
+ InTitle,
+ InDescription,
+ InContent,
+ Action(RawAction),
+}
+
+#[derive(Debug, PartialEq, Eq, Clone)]
+pub enum RawAction {
+ Boost(String),
+ Downrank(String),
+ Discard,
+}
+
+pub fn parse(goggle: &str) -> CrateResult<RawGoggle> {
+ let newlines = Regex::new(r"[\n]+").unwrap();
+ let clean = newlines.replace_all(goggle.trim(), "\n").to_string();
+ let clean = clean.trim().replace('\n', ";").replace('\r', ";");
+
+ match PARSER.parse(clean.as_str()) {
+ Ok(blocks) => Ok(RawGoggle::from(blocks)),
+ Err(_) => Err(Error::Parse),
+ }
+}
+
+#[cfg(test)]
+mod tests {
+ use super::*;
+
+ #[test]
+ fn simple() {
+ let goggle = parse(
+ r#"
+ ! name: test
+ ! this is a normal comment
+ @host_centrality = 3
+ ! this is a normal comment
+ @bm25 = 100
+ @field_url = 2
+ /this/is/*/pattern
+ /blog/$site=example.com
+ "#,
+ )
+ .unwrap();
+
+ assert_eq!(
+ goggle.alterations,
+ vec![
+ RawAlteration {
+ target: Target::Signal("host_centrality".to_string()),
+ score: "3".to_string(),
+ },
+ RawAlteration {
+ target: Target::Signal("bm25".to_string()),
+ score: "100".to_string()
+ },
+ RawAlteration {
+ target: Target::Field("url".to_string()),
+ score: "2".to_string()
+ }
+ ]
+ );
+
+ assert_eq!(
+ goggle.comments,
+ vec![
+ Comment::Header {
+ key: "name".to_string(),
+ value: "test".to_string()
+ },
+ Comment::Basic("! this is a normal comment".to_string()),
+ Comment::Basic("! this is a normal comment".to_string()),
+ ]
+ );
+
+ assert_eq!(
+ goggle.instructions,
+ vec![
+ RawInstruction {
+ patterns: vec![
+ RawPatternPart::Raw("/this/is/".to_string()),
+ RawPatternPart::Wildcard,
+ RawPatternPart::Raw("/pattern".to_string())
+ ],
+ options: vec![]
+ },
+ RawInstruction {
+ patterns: vec![RawPatternPart::Raw("/blog/".to_string()),],
+ options: vec![RawPatternOption::Site("example.com".to_string())]
+ },
+ ]
+ )
+ }
+
+ #[test]
+ fn advanced_urls() {
+ let goggle = parse(
+ r#"
+ https://www.example.com?@hej
+ "#,
+ )
+ .unwrap();
+
+ assert_eq!(
+ goggle.instructions,
+ vec![RawInstruction {
+ patterns: vec![RawPatternPart::Raw(
+ "https://www.example.com?@hej".to_string()
+ ),],
+ options: vec![]
+ },]
+ );
+
+ let goggle = parse(
+ r#"
+ https://www.example.com?@hej$site=https://www.example.com
+ "#,
+ )
+ .unwrap();
+
+ assert_eq!(
+ goggle.instructions,
+ vec![RawInstruction {
+ patterns: vec![RawPatternPart::Raw(
+ "https://www.example.com?@hej".to_string()
+ ),],
+ options: vec![RawPatternOption::Site(
+ "https://www.example.com".to_string()
+ )]
+ },]
+ );
+ }
+
+ #[test]
+ fn ignore_consecutive_newlines() {
+ let goggle = parse(
+ r#"
+ |pattern1|
+
+
+
+
+ pattern2^
+ "#,
+ )
+ .unwrap();
+ assert_eq!(
+ goggle.instructions,
+ vec![
+ RawInstruction {
+ patterns: vec![
+ RawPatternPart::Anchor,
+ RawPatternPart::Raw("pattern1".to_string()),
+ RawPatternPart::Anchor,
+ ],
+ options: vec![]
+ },
+ RawInstruction {
+ patterns: vec![
+ RawPatternPart::Raw("pattern2".to_string()),
+ RawPatternPart::Delimeter,
+ ],
+ options: vec![]
+ },
+ ]
+ )
+ }
+
+ #[test]
+ fn quickstart_parse() {
+ assert!(parse(include_str!("../../../testcases/goggles/quickstart.goggle")).is_ok());
+ }
+
+ #[test]
+ fn hacker_news_parse() {
+ assert!(parse(include_str!(
+ "../../../testcases/goggles/hacker_news.goggle"
+ ))
+ .is_ok());
+ }
+}
diff --git a/src/ranking/goggles/mod.rs b/src/ranking/goggles/mod.rs
new file mode 100644
index 00000000..20563f16
--- /dev/null
+++ b/src/ranking/goggles/mod.rs
@@ -0,0 +1,590 @@
+// Cuely is an open source web search engine.
+// Copyright (C) 2022 Cuely ApS
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program. If not, see <https://www.gnu.org/licenses/>.
+
+mod ast;
+mod pattern_query;
+mod signal;
+
+use std::convert::TryFrom;
+
+use crate::{schema::Field, tokenizer, Result};
+use itertools::Itertools;
+use tantivy::{
+ query::{BooleanQuery, BoostQuery, Occur, PhraseQuery, QueryClone, TermQuery},
+ schema::{IndexRecordOption, Schema},
+ tokenizer::TextAnalyzer,
+};
+
+pub use self::signal::*;
+use self::{
+ ast::{RawAction, RawGoggle, RawInstruction, RawPatternOption, RawPatternPart},
+ pattern_query::PatternQuery,
+};
+
+pub fn parse(goggle: &str) -> Result<Goggle> {
+ let raw_goggle = ast::parse(goggle)?;
+
+ Goggle::try_from(raw_goggle)
+}
+
+impl TryFrom<RawGoggle> for Goggle {
+ type Error = crate::Error;
+
+ fn try_from(raw: RawGoggle) -> Result<Self> {
+ let mut instructions = Vec::new();
+
+ for inst in raw.instructions {
+ instructions.push(Instruction::try_from(inst)?);
+ }
+
+ Ok(Self {
+ aggregator: SignalAggregator::try_from(raw.alterations)?,
+ instructions,
+ })
+ }
+}
+
+impl TryFrom<RawInstruction> for Instruction {
+ type Error = crate::Error;
+
+ fn try_from(value: RawInstruction) -> Result<Self> {
+ let mut patterns = Vec::new();
+
+ for pattern in value.patterns {
+ patterns.push(pattern.into());
+ }
+
+ let mut options = Vec::new();
+
+ for option in value.options {
+ options.push(option.try_into()?);
+ }
+
+ Ok(Instruction { patterns, options })
+ }
+}
+
+impl From<RawPatternPart> for PatternPart {
+ fn from(value: RawPatternPart) -> Self {
+ match value {
+ RawPatternPart::Raw(text) => PatternPart::Raw(text),
+ RawPatternPart::Wildcard => PatternPart::Wildcard,
+ RawPatternPart::Delimeter => PatternPart::Delimeter,
+ RawPatternPart::Anchor => PatternPart::Anchor,
+ }
+ }
+}
+impl TryFrom<RawPatternOption> for PatternOption {
+ type Error = crate::Error;
+
+ fn try_from(value: RawPatternOption) -> Result<Self> {
+ let res = match value {
+ RawPatternOption::Site(site) => PatternOption::Site(site),
+ RawPatternOption::InUrl => PatternOption::InUrl,
+ RawPatternOption::InTitle => PatternOption::InTitle,
+ RawPatternOption::InDescription => PatternOption::InDescription,
+ RawPatternOption::InContent => PatternOption::InContent,
+ RawPatternOption::Action(action) => PatternOption::Action(action.try_into()?),
+ };
+
+ Ok(res)
+ }
+}
+
+impl TryFrom<RawAction> for Action {
+ type Error = crate::Error;
+
+ fn try_from(value: RawAction) -> Result<Self> {
+ let res = match value {
+ RawAction::Boost(boost) => Action::Boost(boost.parse()?),
+ RawAction::Downrank(down_boost) => Action::Downrank(down_boost.parse()?),
+ RawAction::Discard => Action::Discard,
+ };
+
+ Ok(res)
+ }
+}
+
+#[derive(Debug, PartialEq, Eq)]
+pub struct Instruction {
+ pub patterns: Vec<PatternPart>,
+ pub options: Vec<PatternOption>,
+}
+
+#[derive(Debug, PartialEq, Eq, Clone)]
+pub enum PatternPart {
+ Raw(String),
+ Wildcard,
+ Delimeter,
+ Anchor,
+}
+
+#[derive(Debug, PartialEq, Eq)]
+pub enum PatternOption {
+ Site(String),
+ InUrl,
+ InTitle,
+ InDescription,
+ InContent,
+ Action(Action),
+}
+
+#[derive(Debug, PartialEq, Eq, Clone, Copy)]
+pub enum Action {
+ Boost(u64),
+ Downrank(u64),
+ Discard,
+}
+
+#[derive(Debug, Default)]
+pub struct Goggle {
+ pub aggregator: SignalAggregator,
+ instructions: Vec<Instruction>,
+}
+
+impl Goggle {
+ pub fn as_tantivy(&self, schema: &Schema) -> Vec<(Occur, Box<dyn tantivy::query::Query>)> {
+ if self
+ .instructions
+ .iter()
+ .any(|instruction| instruction.is_empty_discard())
+ {
+ vec![(
+ Occur::Must,
+ BooleanQuery::from(
+ self.instructions
+ .iter()
+ .map(|instruction| instruction.as_tantivy(schema))
+ .collect_vec(),
+ )
+ .box_clone(),
+ )]
+ } else {
+ self.instructions
+ .iter()
+ .map(|instruction| instruction.as_tantivy(schema))
+ .collect()
+ }
+ }
+}
+
+fn process_tantivy_term(
+ term: &str,
+ analyzer: TextAnalyzer,
+ tantivy_field: tantivy::schema::Field,
+) -> Vec<tantivy::Term> {
+ let mut terms: Vec<tantivy::Term> = Vec::new();
+ let mut token_stream = analyzer.token_stream(term);
+ token_stream.process(&mut |token| {
+ let term = tantivy::Term::from_field_text(tantivy_field, &token.text);
+ terms.push(term);
+ });
+
+ terms
+}
+
+fn process_site(site: &str, field: tantivy::schema::Field) -> Box<dyn tantivy::query::Query> {
+ let mut terms = process_tantivy_term(
+ site,
+ TextAnalyzer::new(tokenizer::Normal::default(), vec![]),
+ field,
+ );
+
+ if terms.len() > 1 {
+ Box::new(PhraseQuery::new(terms)) as Box<dyn tantivy::query::Query>
+ } else {
+ let term = terms.pop().unwrap();
+ Box::new(TermQuery::new(
+ term,
+ IndexRecordOption::WithFreqsAndPositions,
+ ))
+ }
+}
+
+impl Instruction {
+ pub fn as_tantivy(&self, schema: &Schema) -> (Occur, Box<dyn tantivy::query::Query>) {
+ let mut subqueries = Vec::new();
+
+ let mut field = None;
+ let mut action = None;
+
+ for option in &self.options {
+ match option {
+ PatternOption::Site(site) if field.is_none() => {
+ field = Some(schema.get_field(Field::Site.as_str()).unwrap());
+
+ let domain_field = schema.get_field(Field::Domain.as_str()).unwrap();
+ let site_field = schema.get_field(Field::Site.as_str()).unwrap();
+
+ subqueries.push((
+ Occur::Must,
+ BooleanQuery::new(vec![
+ (Occur::Should, process_site(site, domain_field)),
+ (Occur::Should, process_site(site, site_field)),
+ ])
+ .box_clone(),
+ ));
+ }
+ PatternOption::InUrl if field.is_none() => {
+ field = Some(schema.get_field(Field::Url.as_str()).unwrap())
+ }
+ PatternOption::InTitle if field.is_none() => {
+ field = Some(schema.get_field(Field::Title.as_str()).unwrap())
+ }
+ PatternOption::InDescription if field.is_none() => {
+ field = Some(schema.get_field(Field::Description.as_str()).unwrap())
+ }
+ PatternOption::InContent if field.is_none() => {
+ field = Some(schema.get_field(Field::CleanBody.as_str()).unwrap())
+ }
+ PatternOption::Action(pattern_action) if action.is_none() => {
+ action = Some(*pattern_action)
+ }
+ _ => {}
+ }
+ }
+
+ let action = action.unwrap_or(Action::Boost(1));
+ let field = field.unwrap_or_else(|| schema.get_field(Field::Url.as_str()).unwrap());
+
+ if !self.patterns.is_empty() {
+ let query = self.pattern_query(field);
+ subqueries.push((Occur::Must, query));
+ }
+
+ match action {
+ Action::Boost(boost) => (
+ Occur::Should,
+ BoostQuery::new(
+ BooleanQuery::from(subqueries).box_clone(),
+ boost as f32 + 1.0,
+ )
+ .box_clone(),
+ ),
+ Action::Downrank(boost) => (
+ Occur::Should,
+ BoostQuery::new(
+ BooleanQuery::from(subqueries).box_clone(),
+ 1.0 / (boost as f32 + 1.0),
+ )
+ .box_clone(),
+ ),
+ Action::Discard => (Occur::MustNot, BooleanQuery::from(subqueries).box_clone()),
+ }
+ }
+
+ fn pattern_query(&self, field: tantivy::schema::Field) -> Box<dyn tantivy::query::Query> {
+ PatternQuery::new(self.patterns.clone(), field).box_clone()
+ }
+
+ fn is_empty_discard(&self) -> bool {
+ self.patterns.is_empty()
+ && self.options.len() == 1
+ && matches!(
+ self.options.first(),
+ Some(PatternOption::Action(Action::Discard))
+ )
+ }
+}
+
+#[cfg(test)]
+mod tests {
+ use crate::{index::Index, schema::create_schema, searcher::Searcher, webpage::Webpage};
+
+ use super::*;
+ const CONTENT: &str = "this is the best example website ever this is the best example website ever this is the best example website ever this is the best example website ever this is the best example website ever this is the best example website ever";
+
+ #[test]
+ fn discard_and_boost_sites() {
+ let mut index = Index::temporary().expect("Unable to open index");
+
+ index
+ .insert(Webpage::new(
+ &format!(
+ r#"
+ <html>
+ <head>
+ <title>Website A</title>
+ </head>
+ <body>
+ {CONTENT}
+ example example example
+ </body>
+ </html>
+ "#
+ ),
+ "https://www.a.com",
+ vec![],
+ 0.0,
+ 500,
+ ))
+ .expect("failed to parse webpage");
+ index
+ .insert(Webpage::new(
+ &format!(
+ r#"
+ <html>
+ <head>
+ <title>Website B</title>
+ </head>
+ <body>
+ {CONTENT}
+ </body>
+ </html>
+ "#
+ ),
+ "https://www.b.com",
+ vec![],
+ 0.0001,
+ 500,
+ ))
+ .expect("failed to parse webpage");
+
+ index.commit().expect("failed to commit index");
+ let searcher = Searcher::from(index);
+
+ let res = searcher
+ .search("website", None, None, None)
+ .unwrap()
+ .into_websites()
+ .unwrap()
+ .webpages
+ .documents;
+
+ assert_eq!(res.len(), 2);
+ assert_eq!(res[0].url, "https://www.b.com");
+ assert_eq!(res[1].url, "https://www.a.com");
+
+ let res = searcher
+ .search(
+ "website",
+ None,
+ Some(
+ r#"
+ $discard,site=b.com
+ "#
+ .to_string(),
+ ),
+ None,
+ )
+ .unwrap()
+ .into_websites()
+ .unwrap()
+ .webpages
+ .documents;
+
+ assert_eq!(res.len(), 1);
+ assert_eq!(res[0].url, "https://www.a.com");
+
+ let res = searcher
+ .search(
+ "website",
+ None,
+ Some(
+ r#"
+ $boost=10,site=a.com
+ "#
+ .to_string(),
+ ),
+ None,
+ )
+ .unwrap()
+ .into_websites()
+ .unwrap()
+ .webpages
+ .documents;
+
+ assert_eq!(res.len(), 2);
+ assert_eq!(res[0].url, "https://www.a.com");
+ assert_eq!(res[1].url, "https://www.b.com");
+ }
+
+ #[test]
+ fn quickstart_as_query() {
+ parse(include_str!("../../../testcases/goggles/quickstart.goggle"))
+ .unwrap()
+ .as_tantivy(&create_schema());
+ }
+
+ #[test]
+ fn example_goggles_dont_crash() {
+ let mut index = Index::temporary().expect("Unable to open index");
+
+ index
+ .insert(Webpage::new(
+ &format!(
+ r#"
+ <html>
+ <head>
+ <title>Website A</title>
+ </head>
+ <body>
+ {CONTENT}
+ example example example
+ </body>
+ </html>
+ "#
+ ),
+ "https://www.a.com/this/is/a/pattern",
+ vec![],
+ 0.0,
+ 500,
+ ))
+ .expect("failed to parse webpage");
+ index
+ .insert(Webpage::new(
+ &format!(
+ r#"
+ <html>
+ <head>
+ <title>Website B</title>
+ </head>
+ <body>
+ {CONTENT}
+ </body>
+ </html>
+ "#
+ ),
+ "https://www.b.com/this/is/b/pattern",
+ vec![],
+ 0.0001,
+ 500,
+ ))
+ .expect("failed to parse webpage");
+
+ index.commit().expect("failed to commit index");
+ let searcher = Searcher::from(index);
+
+ let _ = searcher
+ .search(
+ "website",
+ None,
+ Some(include_str!("../../../testcases/goggles/quickstart.goggle").to_string()),
+ None,
+ )
+ .unwrap()
+ .into_websites()
+ .unwrap()
+ .webpages
+ .documents;
+
+ let _ = searcher
+ .search(
+ "website",
+ None,
+ Some(include_str!("../../../testcases/goggles/hacker_news.goggle").to_string()),
+ None,
+ )
+ .unwrap()
+ .into_websites()
+ .unwrap()
+ .webpages
+ .documents;
+ }
+
+ #[test]
+ fn empty_discard() {
+ let mut index = Index::temporary().expect("Unable to open index");
+
+ index
+ .insert(Webpage::new(
+ &format!(
+ r#"
+ <html>
+ <head>
+ <title>Website A</title>
+ </head>
+ <body>
+ {CONTENT}
+ example example example
+ </body>
+ </html>
+ "#
+ ),
+ "https://www.a.com/this/is/a/pattern",
+ vec![],
+ 0.0,
+ 500,
+ ))
+ .expect("failed to parse webpage");
+ index
+ .insert(Webpage::new(
+ &format!(
+ r#"
+ <html>
+ <head>
+ <title>Website B</title>
+ </head>
+ <body>
+ {CONTENT}
+ </body>
+ </html>
+ "#
+ ),
+ "https://www.b.com/this/is/b/pattern",
+ vec![],
+ 0.0001,
+ 500,
+ ))
+ .expect("failed to parse webpage");
+ index
+ .insert(Webpage::new(
+ &format!(
+ r#"
+ <html>
+ <head>
+ <title>Website B</title>
+ </head>
+ <body>
+ {CONTENT}
+ </body>
+ </html>
+ "#
+ ),
+ "https://www.c.com/this/is/c/pattern",
+ vec![],
+ 0.0001,
+ 500,
+ ))
+ .expect("failed to parse webpage");
+
+ index.commit().expect("failed to commit index");
+ let searcher = Searcher::from(index);
+
+ let res = searcher
+ .search(
+ "website",
+ None,
+ Some(
+ r#"
+ $discard
+ $site=a.com,boost=6
+ $site=b.com,boost=1
+ "#
+ .to_string(),
+ ),
+ None,
+ )
+ .unwrap()
+ .into_websites()
+ .unwrap()
+ .webpages
+ .documents;
+
+ assert_eq!(res.len(), 2);
+ assert_eq!(res[0].url, "https://www.a.com/this/is/a/pattern");
+ }
+}
diff --git a/src/ranking/goggles/parser.lalrpop b/src/ranking/goggles/parser.lalrpop
new file mode 100644
index 00000000..8f061e10
--- /dev/null
+++ b/src/ranking/goggles/parser.lalrpop
@@ -0,0 +1,85 @@
+use crate::ranking::goggles::ast::*;
+
+grammar;
+
+Sep<S, T>: Vec<T> = {
+ <mut xs:(<T> S)*> <x:T> => {
+ xs.push(x);
+ xs
+ }
+};
+
+pub Blocks: Vec<GoggleBlock> = <Sep<";", Block>>;
+
+Block: GoggleBlock = {
+ <RawAlteration> => GoggleBlock::Alteration(<>),
+ <Comment> => GoggleBlock::Comment(<>),
+ <RawInstruction> => GoggleBlock::Instruction(<>),
+}
+
+
+RawAlteration: RawAlteration = {
+ "@" <target:Target> "=" <score:Ident> => RawAlteration {<>}
+}
+
+Target: Target = {
+ <Name> => {
+ if let Some(name) = <>.strip_prefix("field_") {
+ Target::Field(name.to_string())
+ } else {
+ Target::Signal(<>)
+ }
+ },
+}
+
+Name: String = Ident;
+
+Comment: Comment = r"![^;]*" => {
+ if let Some((key, value)) = <>.split_once(':') {
+ let key = key.strip_prefix("!").unwrap_or(key).trim();
+ let value = value.trim();
+
+ if matches!(key, "name" | "description" | "public" | "author" | "homepage" | "issues" | "transferred_to" | "avatar" | "license") {
+ Comment::Header { key: key.to_string(), value: value.to_string() }
+ } else {
+ Comment::Basic(<>.to_string())
+ }
+ } else {
+ Comment::Basic(<>.to_string())
+ }
+};
+
+RawInstruction: RawInstruction = {
+ <patterns:RawPatterns> "$" <options:RawPatternOptions> => RawInstruction { <> },
+ <patterns:RawPatterns> => RawInstruction { <>, options: Vec::new() },
+}
+
+RawPatterns: Vec<RawPatternPart> = <RawPatternPart*> => <>;
+
+RawPatternPart: RawPatternPart = {
+ "*" => RawPatternPart::Wildcard,
+ "^" => RawPatternPart::Delimeter,
+ "|" => RawPatternPart::Anchor,
+ Ident => RawPatternPart::Raw(<>.to_string()),
+}
+
+RawPatternOptions: Vec<RawPatternOption> = <Sep<",", RawPatternOption>> => <>;
+
+RawPatternOption: RawPatternOption = {
+ "site=" <site:Ident> => RawPatternOption::Site(site.to_string()),
+ "inurl" => RawPatternOption::InUrl,
+ "intitle" => RawPatternOption::InTitle,
+ "indescription" => RawPatternOption::InDescription,
+ "incontent" => RawPatternOption::InContent,
+ <RawAction> => RawPatternOption::Action(<>),
+}
+
+RawAction: RawAction = {
+ "boost=" <Ident> => RawAction::Boost(<>),
+ "boost" => RawAction::Boost("1".to_string()),
+ "downrank=" <Ident> => RawAction::Downrank(<>),
+ "downrank" => RawAction::Downrank("1".to_string()),
+ "discard" => RawAction::Discard,
+}
+
+Ident: String = <r"[a-zA-Z0-9_/][a-zA-Z_/:\.\?@0-9&\-]*"> => <>.to_string();
\ No newline at end of file
diff --git a/src/ranking/goggles/pattern_query.rs b/src/ranking/goggles/pattern_query.rs
new file mode 100644
index 00000000..c35d450e
--- /dev/null
+++ b/src/ranking/goggles/pattern_query.rs
@@ -0,0 +1,459 @@
+// Cuely is an open source web search engine.
+// Copyright (C) 2022 Cuely ApS
+//
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU Affero General Public License as
+// published by the Free Software Foundation, either version 3 of the
+// License, or (at your option) any later version.
+//
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+// GNU Affero General Public License for more details.
+//
+// You should have received a copy of the GNU Affero General Public License
+// along with this program. If not, see <https://www.gnu.org/licenses/>.
+
+use tantivy::{
+ fastfield::{Column, DynamicFastFieldReader},
+ fieldnorm::FieldNormReader,
+ postings::SegmentPostings,
+ query::{EmptyScorer, Explanation, Scorer},
+ schema::IndexRecordOption,
+ tokenizer::Tokenizer,
+ DocId, DocSet, Postings, Score, SegmentReader, TantivyError, TERMINATED,
+};
+
+use crate::{
+ query::intersection::Intersection,
+ ranking::bm25::Bm25Weight,
+ schema::{Field, ALL_FIELDS},
+ tokenizer,
+};
+
+use super::PatternPart;
+
+#[derive(Debug, Clone)]
+pub struct PatternQuery {
+ patterns: Vec<PatternPart>,
+ field: tantivy::schema::Field,
+ raw_terms: Vec<tantivy::Term>,
+}
+
+impl PatternQuery {
+ pub fn new(patterns: Vec<PatternPart>, field: tantivy::schema::Field) -> Self {
+ let mut raw_terms = Vec::new();
+
+ for pattern in &patterns {
+ match pattern {
+ PatternPart::Raw(text) => {
+ let mut stream = tokenizer::Normal::default().token_stream(text);
+
+ while let Some(token) = stream.next() {
+ let term = tantivy::Term::from_field_text(field, &token.text);
+ raw_terms.push(term);
+ }
+ }
+ PatternPart::Wildcard => {}
+ PatternPart::Delimeter => {}
+ PatternPart::Anchor => {}
+ }
+ }
+
+ Self {
+ patterns,
+ field,
+ raw_terms,
+ }
+ }
+}
+
+impl tantivy::query::Query for PatternQuery {
+ fn weight(
+ &self,
+ searcher: &tantivy::Searcher,
+ scoring_enabled: bool,
+ ) -> tantivy::Result<Box<dyn tantivy::query::Weight>> {
+ let bm25_weight = Bm25Weight::for_terms(searcher, &self.raw_terms)?;
+
+ Ok(Box::new(PatternWeight {
+ similarity_weight: bm25_weight,
+ scoring_enabled,
+ raw_terms: self.raw_terms.clone(),
+ patterns: self.patterns.clone(),
+ field: self.field,
+ }))
+ }
+
+ fn query_terms<'a>(&'a self, visitor: &mut dyn FnMut(&'a tantivy::Term, bool)) {
+ for term in &self.raw_terms {
+ visitor(term, true);
+ }
+ }
+}
+
+enum SmallPatternPart {
+ Term,
+ Wildcard,
+ Delimeter,
+ Anchor,
+}
+
+struct PatternWeight {
+ similarity_weight: Bm25Weight,
+ scoring_enabled: bool,
+ patterns: Vec<PatternPart>,
+ raw_terms: Vec<tantivy::Term>,
+ field: tantivy::schema::Field,
+}
+
+impl PatternWeight {
+ fn fieldnorm_reader(&self, reader: &SegmentReader) -> tantivy::Result<FieldNormReader> {
+ if self.scoring_enabled {
+ if let Some(fieldnorm_reader) = reader.fieldnorms_readers().get_field(self.field)? {
+ return Ok(fieldnorm_reader);
+ }
+ }
+ Ok(FieldNormReader::constant(reader.max_doc(), 1))
+ }
+
+ pub(crate) fn pattern_scorer(
+ &self,
+ reader: &SegmentReader,
+ boost: Score,
+ ) -> tantivy::Result<Option<PatternScorer>> {
+ let similarity_weight = self.similarity_weight.boost_by(boost);
+ let fieldnorm_reader = self.fieldnorm_reader(reader)?;
+ let mut term_postings_list = Vec::new();
+
+ for term in &self.raw_terms {
+ if let Some(postings) = reader
+ .inverted_index(term.field())?
+ .read_postings(term, IndexRecordOption::WithFreqsAndPositions)?
+ {
+ term_postings_list.push(postings);
+ } else {
+ return Ok(None);
+ }
+ }
+
+ let small_patterns = self
+ .patterns
+ .iter()
+ .map(|pattern| match pattern {
+ PatternPart::Raw(_) => SmallPatternPart::Term,
+ PatternPart::Wildcard => SmallPatternPart::Wildcard,
+ PatternPart::Delimeter => SmallPatternPart::Delimeter,
+ PatternPart::Anchor => SmallPatternPart::Anchor,
+ })
+ .collect();
+
+ let num_tokens_reader = match &ALL_FIELDS[self.field.field_id() as usize] {
+ Field::Title => reader.fast_fields().u64(
+ reader
+ .schema()
+ .get_field(Field::NumTitleTokens.as_str())
+ .unwrap(),
+ ),
+ Field::CleanBody => reader.fast_fields().u64(
+ reader
+ .schema()
+ .get_field(Field::NumCleanBodyTokens.as_str())
+ .unwrap(),
+ ),
+ Field::Url => reader.fast_fields().u64(
+ reader
+ .schema()
+ .get_field(Field::NumUrlTokens.as_str())
+ .unwrap(),
+ ),
+ Field::Description => reader.fast_fields().u64(
+ reader
+ .schema()
+ .get_field(Field::NumDescriptionTokens.as_str())
+ .unwrap(),
+ ),
+ field => Err(TantivyError::InvalidArgument(format!(
+ "{} is not supported in pattern query",
+ field.as_str()
+ ))),
+ }?;
+
+ Ok(Some(PatternScorer::new(
+ similarity_weight,
+ term_postings_list,
+ fieldnorm_reader,
+ small_patterns,
+ num_tokens_reader,
+ )))
+ }
+}
+
+impl tantivy::query::Weight for PatternWeight {
+ fn scorer(
+ &self,
+ reader: &tantivy::SegmentReader,
+ boost: tantivy::Score,
+ ) -> tantivy::Result<Box<dyn tantivy::query::Scorer>> {
+ if let Some(scorer) = self.pattern_scorer(reader, boost)? {
+ Ok(Box::new(scorer))
+ } else {
+ Ok(Box::new(EmptyScorer))
+ }
+ }
+
+ fn explain(
+ &self,
+ reader: &tantivy::SegmentReader,
+ doc: tantivy::DocId,
+ ) -> tantivy::Result<tantivy::query::Explanation> {
+ let scorer_opt = self.pattern_scorer(reader, 1.0)?;
+ if scorer_opt.is_none() {
+ return Err(TantivyError::InvalidArgument(format!(
+ "Document #({}) does not match",
+ doc
+ )));
+ }
+ let mut scorer = scorer_opt.unwrap();
+ if scorer.seek(doc) != doc {
+ return Err(TantivyError::InvalidArgument(format!(
+ "Document #({}) does not match",
+ doc
+ )));
+ }
+ let fieldnorm_reader = self.fieldnorm_reader(reader)?;
+ let fieldnorm_id = fieldnorm_reader.fieldnorm_id(doc);
+ let phrase_count = scorer.phrase_count();
+ let mut explanation = Explanation::new("Pattern Scorer", scorer.score());
+ explanation.add_detail(self.similarity_weight.explain(fieldnorm_id, phrase_count));
+ Ok(explanation)
+ }
+}
+
+struct PatternScorer {
+ similarity_weight: Bm25Weight,
+ fieldnorm_reader: FieldNormReader,
+ intersection_docset: Intersection<SegmentPostings, SegmentPostings>,
+ pattern: Vec<SmallPatternPart>,
+ num_query_terms: usize,
+ left: Vec<u32>,
+ right: Vec<u32>,
+ phrase_count: u32,
+ num_tokens_reader: DynamicFastFieldReader<u64>,
+}
+
+impl PatternScorer {
+ fn new(
+ similarity_weight: Bm25Weight,
+ term_postings_list: Vec<SegmentPostings>,
+ fieldnorm_reader: FieldNormReader,
+ pattern: Vec<SmallPatternPart>,
+ num_tokens_reader: DynamicFastFieldReader<u64>,
+ ) -> Self {
+ let num_query_terms = term_postings_list.len();
+
+ Self {
+ intersection_docset: Intersection::new(term_postings_list),
+ num_query_terms,
+ similarity_weight,
+ fieldnorm_reader,
+ pattern,
+ left: Vec::with_capacity(100),
+ right: Vec::with_capacity(100),
+ phrase_count: 0,
+ num_tokens_reader,
+ }
+ }
+ fn phrase_count(&self) -> u32 {
+ self.phrase_count
+ }
+
+ fn pattern_match(&mut self) -> bool {
+ self.phrase_count = self.perform_pattern_match() as u32;
+
+ self.phrase_count > 0
+ }
+
+ fn perform_pattern_match(&mut self) -> usize {
+ {
+ self.intersection_docset
+ .docset_mut_specialized(0)
+ .positions(&mut self.left);
+ }
+
+ let mut intersection_len = self.left.len();
+ let mut out = Vec::new();
+
+ let mut current_right_term = 1;
+ let mut slop = 1;
+ let num_tokens_doc = self.num_tokens_reader.get_val(self.doc() as u64);
+
+ for (i, pattern_part) in self.pattern.iter().enumerate().skip(1) {
+ match pattern_part {
+ SmallPatternPart::Term => {
+ {
+ self.intersection_docset
+ .docset_mut_specialized(current_right_term)
+ .positions(&mut self.right);
+ }
+ out.resize(self.left.len().max(self.right.len()), 0);
+ intersection_len =
+ intersection_with_slop(&self.left[..], &self.right[..], &mut out, slop);
+
+ slop = 1;
+
+ if intersection_len == 0 {
+ return 0;
+ }
+
+ self.left = out[..intersection_len].to_vec();
+ out = Vec::new();
+ current_right_term += 1;
+ }
+ SmallPatternPart::Wildcard => {
+ slop = u32::MAX;
+ }
+ SmallPatternPart::Delimeter => {}
+ SmallPatternPart::Anchor if i == 0 => {
+ if let Some(pos) = self.left.first() {
+ if *pos != 0 {
+ return 0;
+ }
+ }
+ }
+ SmallPatternPart::Anchor if i == self.pattern.len() - 1 => {
+ {
+ self.intersection_docset
+ .docset_mut_specialized(self.num_query_terms - 1)
+ .positions(&mut self.right);
+ }
+
+ if let Some(pos) = self.right.last() {
+ if *pos != (num_tokens_doc - 1) as u32 {
+ return 0;
+ }
+ }
+ }
+ SmallPatternPart::Anchor => {}
+ }
+ }
+
+ intersection_len
+ }
+}
+
+impl Scorer for PatternScorer {
+ fn score(&mut self) -> Score {
+ let doc = self.doc();
+ let fieldnorm_id = self.fieldnorm_reader.fieldnorm_id(doc);
+ self.similarity_weight
+ .score(fieldnorm_id, self.phrase_count())
+ }
+}
+
+impl DocSet for PatternScorer {
+ fn advance(&mut self) -> DocId {
+ loop {
+ let doc = self.intersection_docset.advance();
+ if doc == TERMINATED || self.pattern_match() {
+ return doc;
+ }
+ }
+ }
+
+ fn seek(&mut self, target: DocId) -> DocId {
+ debug_assert!(target >= self.doc());
+ let doc = self.intersection_docset.seek(target);
+ if doc == TERMINATED || self.pattern_match() {
+ return doc;
+ }
+ self.advance()
+ }
+
+ fn doc(&self) -> tantivy::DocId {
+ self.intersection_docset.doc()
+ }
+
+ fn size_hint(&self) -> u32 {
+ self.intersection_docset.size_hint()
+ }
+}
+
+/// Intersect twos sorted arrays `left` and `right` and outputs the
+/// resulting array in `out`. The positions in out are all positions from right where
+/// the distance to left_pos <= slop
+///
+/// Returns the length of the intersection
+fn intersection_with_slop(left: &[u32], right: &[u32], out: &mut [u32], slop: u32) -> usize {
+ let mut left_index = 0;
+ let mut right_index = 0;
+ let mut count = 0;
+ let left_len = left.len();
+ let right_len = right.len();
+ while left_index < left_len && right_index < right_len {
+ let left_val = left[left_index];
+ let right_val = right[right_index];
+
+ // The three conditions are:
+ // left_val < right_slop -> left index increment.
+ // right_slop <= left_val <= right -> find the best match.
+ // left_val > right -> right index increment.
+ let right_slop = if right_val >= slop {
+ right_val - slop
+ } else {
+ 0
+ };
+
+ if left_val < right_slop {
+ left_index += 1;
+ } else if right_slop <= left_val && left_val <= right_val {
+ while left_index + 1 < left_len {
+ // there could be a better match
+ let next_left_val = left[left_index + 1];
+ if next_left_val > right_val {
+ // the next value is outside the range, so current one is the best.
+ break;
+ }
+ // the next value is better.
+ left_index += 1;
+ }
+ // store the match in left.
+ out[count] = right_val;
+ count += 1;
+ right_index += 1;
+ } else if left_val > right_val {
+ right_index += 1;
+ }
+ }
+ count
+}
+
+#[cfg(test)]
+mod tests {
+ use super::*;
+
+ fn aux_intersection(left: &[u32], right: &[u32], expected: &[u32], slop: u32) {
+ let mut out = vec![0; left.len().max(right.len())];
+
+ let intersection_size = intersection_with_slop(left, right, &mut out, slop);
+
+ assert_eq!(&out[..intersection_size], expected);
+ }
+
+ #[test]
+ fn test_intersection_with_slop() {
+ aux_intersection(&[20, 75, 77], &[18, 21, 60], &[21, 60], u32::MAX);
+ aux_intersection(&[21, 60], &[50, 61], &[61], 1);
+
+ aux_intersection(&[1, 2, 3], &[], &[], 1);
+ aux_intersection(&[], &[1, 2, 3], &[], 1);
+
+ aux_intersection(&[1, 2, 3], &[4, 5, 6], &[4], 1);
+ aux_intersection(&[1, 2, 3], &[4, 5, 6], &[4, 5, 6], u32::MAX);
+
+ aux_intersection(&[20, 75, 77], &[18, 21, 60], &[21, 60], u32::MAX);
+ aux_intersection(&[21, 60], &[61, 62], &[61, 62], 2);
+
+ aux_intersection(&[60], &[61, 62], &[61, 62], 2);
+ }
+}
diff --git a/src/ranking/signal_aggregator/mod.rs b/src/ranking/goggles/signal.rs
similarity index 57%
rename from src/ranking/signal_aggregator/mod.rs
rename to src/ranking/goggles/signal.rs
index fba1c132..b3958857 100644
--- a/src/ranking/signal_aggregator/mod.rs
+++ b/src/ranking/goggles/signal.rs
@@ -14,12 +14,9 @@
// You should have received a copy of the GNU Affero General Public License
// along with this program. If not, see <https://www.gnu.org/licenses/>.
-mod ast;
+use crate::Result;
+use std::{array, convert::TryFrom, sync::Arc};
-use crate::{Error, Result};
-use std::{collections::HashMap, sync::Arc};
-
-use strum::{EnumIter, IntoEnumIterator};
use tantivy::{
fastfield::{Column, DynamicFastFieldReader},
DocId, Score, SegmentReader,
@@ -30,9 +27,9 @@ use crate::{
webpage::region::{Region, RegionCount},
};
-use self::ast::{Alteration, Target, PARSER};
+use super::ast::{RawAlteration, Target};
-#[derive(Debug, Clone, PartialEq, Eq, Hash, EnumIter)]
+#[derive(Debug, Clone, PartialEq, Eq, Hash)]
pub enum Signal {
Bm25,
HostCentrality,
@@ -43,6 +40,16 @@ pub enum Signal {
Region,
}
+const ALL_SIGNALS: [Signal; 7] = [
+ Signal::Bm25,
+ Signal::HostCentrality,
+ Signal::IsHomepage,
+ Signal::FetchTimeMs,
+ Signal::UpdateTimestamp,
+ Signal::NumTrackers,
+ Signal::Region,
+];
+
impl Signal {
fn from_field(field: &Field) -> Option<Self> {
match field {
@@ -60,14 +67,16 @@ impl Signal {
!matches!(self, Signal::Bm25)
}
+ #[allow(clippy::too_many_arguments)]
fn value(
&self,
doc: DocId,
bm25: Score,
fastfield: Option<&DynamicFastFieldReader<u64>>,
region_count: &Arc<RegionCount>,
- current_timestamp: f64,
+ current_timestamp: usize,
selected_region: Option<Region>,
+ aggregator: &SignalAggregator,
) -> f64 {
match self {
Signal::Bm25 => bm25 as f64,
@@ -76,14 +85,29 @@ impl Signal {
}
Signal::IsHomepage => fastfield.unwrap().get_val(doc as u64) as f64,
Signal::FetchTimeMs => {
- let fetch_time_ms = fastfield.unwrap().get_val(doc as u64) as f64;
- 1.0 / (fetch_time_ms + 1.0)
+ let fetch_time_ms = fastfield.unwrap().get_val(doc as u64) as usize;
+
+ if fetch_time_ms >= aggregator.fetch_time_ms_cache.len() {
+ 0.0
+ } else {
+ aggregator.fetch_time_ms_cache[fetch_time_ms]
+ }
}
Signal::UpdateTimestamp => {
- let update_timestamp = fastfield.unwrap().get_val(doc as u64) as f64;
+ let update_timestamp = fastfield.unwrap().get_val(doc as u64) as i64;
+
+ if current_timestamp as i64 - update_timestamp <= 0 {
+ return 0.0;
+ }
+
let hours_since_update =
- (current_timestamp - update_timestamp).max(0.000001) / 3600.0;
- 1.0 / ((hours_since_update + 1.0).log2())
+ ((current_timestamp as i64 - update_timestamp).max(1) / 3600) as usize;
+
+ if hours_since_update < aggregator.update_time_cache.len() {
+ aggregator.update_time_cache[hours_since_update]
+ } else {
+ 0.0
+ }
}
Signal::NumTrackers => {
let num_trackers = fastfield.unwrap().get_val(doc as u64) as f64;
@@ -105,13 +129,13 @@ impl Signal {
fn default_coefficient(&self) -> f64 {
match self {
- Signal::Bm25 => 3.0,
- Signal::HostCentrality => 3200.0,
- Signal::IsHomepage => 1.0,
- Signal::FetchTimeMs => 1.0,
- Signal::UpdateTimestamp => 1500.0,
- Signal::NumTrackers => 200.0,
- Signal::Region => 25.0,
+ Signal::Bm25 => 1.0,
+ Signal::HostCentrality => 1024.0,
+ Signal::IsHomepage => 0.1,
+ Signal::FetchTimeMs => 0.1,
+ Signal::UpdateTimestamp => 80.0,
+ Signal::NumTrackers => 20.0,
+ Signal::Region => 60.0,
}
}
@@ -137,34 +161,70 @@ fn fastfield_reader(segment_reader: &SegmentReader, field: &Field) -> DynamicFas
}
#[derive(Debug, Clone)]
-pub struct FieldBoost(HashMap<Field, f64>);
+pub struct FieldBoost(Vec<Option<f64>>);
#[derive(Debug, Clone)]
-pub struct SignalCoefficient(HashMap<Signal, f64>);
+pub struct SignalCoefficient(Vec<Option<f64>>);
impl SignalCoefficient {
pub fn get(&self, signal: &Signal) -> f64 {
self.0
- .get(signal)
+ .get((*signal) as usize)
.copied()
+ .flatten()
.unwrap_or_else(|| signal.default_coefficient())
}
+
+ pub fn new(scores: impl Iterator<Item = (Signal, f64)>) -> Self {
+ let mut fast_scores = Vec::new();
+
+ for (signal, score) in scores {
+ let idx = signal as usize;
+
+ while idx >= fast_scores.len() {
+ fast_scores.push(None);
+ }
+
+ fast_scores[idx] = Some(score);
+ }
+
+ Self(fast_scores)
+ }
}
impl FieldBoost {
pub fn get(&self, field: &Field) -> f64 {
self.0
- .get(field)
+ .get((*field) as usize)
.copied()
+ .flatten()
.or_else(|| field.boost().map(|s| s as f64))
.unwrap_or(1.0)
}
+
+ pub fn new(scores: impl Iterator<Item = (Field, f64)>) -> Self {
+ let mut fast_scores = Vec::new();
+
+ for (field, score) in scores {
+ let idx = field as usize;
+
+ while idx >= fast_scores.len() {
+ fast_scores.push(None);
+ }
+
+ fast_scores[idx] = Some(score);
+ }
+
+ Self(fast_scores)
+ }
}
pub struct SignalAggregator {
- readers: HashMap<Signal, DynamicFastFieldReader<u64>>,
+ readers: Vec<Option<DynamicFastFieldReader<u64>>>,
signal_coefficients: SignalCoefficient,
field_boost: FieldBoost,
+ fetch_time_ms_cache: [f64; 1000],
+ update_time_cache: Vec<f64>,
}
impl std::fmt::Debug for SignalAggregator {
@@ -178,36 +238,56 @@ impl std::fmt::Debug for SignalAggregator {
impl Default for SignalAggregator {
fn default() -> Self {
- Self::new(HashMap::new(), HashMap::new())
+ Self::new(Vec::new().into_iter(), Vec::new().into_iter())
}
}
impl SignalAggregator {
- pub fn new(coefficients: HashMap<Signal, f64>, boosts: HashMap<Field, f64>) -> Self {
- let signal_coefficients = SignalCoefficient(coefficients);
- let field_boost = FieldBoost(boosts);
+ pub fn new(
+ coefficients: impl Iterator<Item = (Signal, f64)>,
+ boosts: impl Iterator<Item = (Field, f64)>,
+ ) -> Self {
+ let signal_coefficients = SignalCoefficient::new(coefficients);
+ let field_boost = FieldBoost::new(boosts);
+
+ let fetch_time_ms_cache = array::from_fn(|fetch_time| 1.0 / (fetch_time as f64 + 1.0));
+
+ let update_time_cache = (0..(3 * 365 * 24))
+ .map(|hours_since_update| 1.0 / ((hours_since_update as f64 + 1.0).log2()))
+ .collect();
Self {
- readers: HashMap::new(),
+ readers: Vec::new(),
signal_coefficients,
field_boost,
+ fetch_time_ms_cache,
+ update_time_cache,
}
}
pub fn new_like(other: &SignalAggregator) -> Self {
Self {
- readers: HashMap::new(),
+ readers: Vec::new(),
signal_coefficients: other.signal_coefficients.clone(),
field_boost: other.field_boost.clone(),
+ fetch_time_ms_cache: other.fetch_time_ms_cache,
+ update_time_cache: other.update_time_cache.clone(),
}
}
pub fn register_readers(&mut self, segment_reader: &SegmentReader) {
+ self.readers.clear();
+
for field in &ALL_FIELDS {
if let Some(signal) = Signal::from_field(field) {
if signal.has_fast_reader() {
- self.mut_readers()
- .insert(signal, fastfield_reader(segment_reader, field));
+ let idx = signal as usize;
+
+ while idx >= self.readers.len() {
+ self.readers.push(None);
+ }
+
+ self.readers[idx] = Some(fastfield_reader(segment_reader, field))
}
}
}
@@ -218,19 +298,26 @@ impl SignalAggregator {
doc: DocId,
bm25: Score,
region_count: &Arc<RegionCount>,
- current_timestamp: f64,
+ current_timestamp: usize,
selected_region: Option<Region>,
) -> f64 {
- Signal::iter()
+ ALL_SIGNALS
+ .into_iter()
.map(|signal| {
+ let reader = match self.readers.get(signal as usize) {
+ Some(Some(reader)) => Some(reader),
+ _ => None,
+ };
+
self.coefficients().get(&signal)
* signal.value(
doc,
bm25,
- self.readers().get(&signal),
+ reader,
region_count,
current_timestamp,
selected_region,
+ self,
)
})
.sum()
@@ -243,43 +330,47 @@ impl SignalAggregator {
pub fn field_boosts(&self) -> &FieldBoost {
&self.field_boost
}
+}
+#[derive(Debug, PartialEq)]
+pub struct Alteration {
+ pub target: Target,
+ pub score: f64,
+}
- pub fn mut_readers(&mut self) -> &mut HashMap<Signal, DynamicFastFieldReader<u64>> {
- &mut self.readers
- }
+impl TryFrom<RawAlteration> for Alteration {
+ type Error = crate::Error;
- pub fn readers(&self) -> &HashMap<Signal, DynamicFastFieldReader<u64>> {
- &self.readers
+ fn try_from(raw: RawAlteration) -> Result<Self> {
+ Ok(Alteration {
+ target: raw.target,
+ score: raw.score.parse()?,
+ })
}
}
-impl From<Vec<Alteration>> for SignalAggregator {
- fn from(alterations: Vec<Alteration>) -> Self {
- let mut coefficients = HashMap::new();
- let mut boosts = HashMap::new();
+impl TryFrom<Vec<RawAlteration>> for SignalAggregator {
+ type Error = crate::Error;
+
+ fn try_from(alterations: Vec<RawAlteration>) -> Result<Self> {
+ let mut coefficients = Vec::new();
+ let mut boosts = Vec::new();
for alteration in alterations {
+ let alteration = Alteration::try_from(alteration)?;
match alteration.target {
Target::Signal(name) => {
if let Some(signal) = Signal::from_string(name) {
- coefficients.insert(signal, alteration.score);
+ coefficients.push((signal, alteration.score));
}
}
Target::Field(name) => {
if let Some(field) = Field::from_string(name) {
- boosts.insert(field, alteration.score);
+ boosts.push((field, alteration.score));
}
}
}
}
- Self::new(coefficients, boosts)
- }
-}
-
-pub fn parse(program: &str) -> Result<SignalAggregator> {
- match PARSER.parse(program) {
- Ok(alterations) => Ok(SignalAggregator::from(alterations)),
- Err(_) => Err(Error::Parse),
+ Ok(Self::new(coefficients.into_iter(), boosts.into_iter()))
}
}
diff --git a/src/ranking/initial.rs b/src/ranking/initial.rs
index 24660e6a..79c1312c 100644
--- a/src/ranking/initial.rs
+++ b/src/ranking/initial.rs
@@ -21,7 +21,7 @@ use chrono::Utc;
use tantivy::collector::{ScoreSegmentTweaker, ScoreTweaker};
use tantivy::{DocId, Score, SegmentReader};
-use super::signal_aggregator::SignalAggregator;
+use super::goggles::SignalAggregator;
pub(crate) struct InitialScoreTweaker {
region_count: Arc<RegionCount>,
@@ -46,7 +46,7 @@ impl InitialScoreTweaker {
pub(crate) struct InitialSegmentScoreTweaker {
aggregator: SignalAggregator,
region_count: Arc<RegionCount>,
- current_timestamp: f64,
+ current_timestamp: usize,
selected_region: Option<Region>,
}
@@ -57,7 +57,7 @@ impl ScoreTweaker<f64> for InitialScoreTweaker {
let mut aggregator = SignalAggregator::new_like(&self.aggregator);
aggregator.register_readers(segment_reader);
- let current_timestamp = Utc::now().timestamp() as f64;
+ let current_timestamp = Utc::now().timestamp() as usize;
Ok(InitialSegmentScoreTweaker {
aggregator,
diff --git a/src/ranking/mod.rs b/src/ranking/mod.rs
index fa6a8100..3663c7c9 100644
--- a/src/ranking/mod.rs
+++ b/src/ranking/mod.rs
@@ -14,9 +14,10 @@
// You should have received a copy of the GNU Affero General Public License
// along with this program. If not, see <https://www.gnu.org/licenses/>.
+mod bm25;
pub mod centrality_store;
+pub mod goggles;
mod initial;
-pub mod signal_aggregator;
use std::sync::Arc;
@@ -28,7 +29,7 @@ use crate::{
webpage::region::{Region, RegionCount},
};
-use self::signal_aggregator::SignalAggregator;
+use self::goggles::SignalAggregator;
pub struct Ranker {
region_count: Arc<RegionCount>,
@@ -398,4 +399,124 @@ mod tests {
assert_eq!(result.webpages.documents.len(), 2);
assert_eq!(result.webpages.documents[0].url, "https://www.first.com");
}
+
+ #[test]
+ fn custom_signal_aggregation() {
+ let mut index = Index::temporary().expect("Unable to open index");
+
+ index
+ .insert(Webpage::new(
+ r#"
+ <html>
+ <head>
+ <title>Test website</title>
+ </head>
+ <body>
+ example
+ </body>
+ </html>
+ "#,
+ "https://www.body.com",
+ vec![],
+ 1.0,
+ 20,
+ ))
+ .expect("failed to parse webpage");
+
+ index
+ .insert(Webpage::new(
+ r#"
+ <html>
+ <head>
+ <title>Example website</title>
+ </head>
+ <body>
+ test
+ </body>
+ </html>
+ "#,
+ "https://www.title.com",
+ vec![],
+ 1.0,
+ 20,
+ ))
+ .expect("failed to parse webpage");
+
+ index
+ .insert(Webpage::new(
+ r#"
+ <html>
+ <head>
+ <title>Example website</title>
+ </head>
+ <body>
+ test
+ </body>
+ </html>
+ "#,
+ "https://www.centrality.com",
+ vec![],
+ 1.0002,
+ 500,
+ ))
+ .expect("failed to parse webpage");
+
+ index.commit().unwrap();
+
+ let searcher = Searcher::new(index, None, None);
+
+ let res = searcher
+ .search(
+ "example",
+ None,
+ Some(
+ r#"
+ @field_title = 20000000
+ @host_centrality = 0
+ "#
+ .to_string(),
+ ),
+ None,
+ )
+ .unwrap()
+ .into_websites()
+ .unwrap();
+
+ assert_eq!(res.webpages.num_docs, 3);
+ assert_eq!(&res.webpages.documents[0].url, "https://www.title.com");
+
+ let res = searcher
+ .search(
+ "example",
+ None,
+ Some(
+ r#"
+ @field_all_body= 20000000
+ @host_centrality = 0
+ "#
+ .to_string(),
+ ),
+ None,
+ )
+ .unwrap()
+ .into_websites()
+ .unwrap();
+
+ assert_eq!(res.webpages.num_docs, 3);
+ assert_eq!(&res.webpages.documents[0].url, "https://www.body.com");
+
+ let res = searcher
+ .search(
+ "example",
+ None,
+ Some("@host_centrality= 2000000".to_string()),
+ None,
+ )
+ .unwrap()
+ .into_websites()
+ .unwrap();
+
+ assert_eq!(res.webpages.num_docs, 3);
+ assert_eq!(&res.webpages.documents[0].url, "https://www.centrality.com");
+ }
}
diff --git a/src/ranking/signal_aggregator/ast.rs b/src/ranking/signal_aggregator/ast.rs
deleted file mode 100644
index cf757c38..00000000
--- a/src/ranking/signal_aggregator/ast.rs
+++ /dev/null
@@ -1,70 +0,0 @@
-// Cuely is an open source web search engine.
-// Copyright (C) 2022 Cuely ApS
-//
-// This program is free software: you can redistribute it and/or modify
-// it under the terms of the GNU Affero General Public License as
-// published by the Free Software Foundation, either version 3 of the
-// License, or (at your option) any later version.
-//
-// This program is distributed in the hope that it will be useful,
-// but WITHOUT ANY WARRANTY; without even the implied warranty of
-// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
-// GNU Affero General Public License for more details.
-//
-// You should have received a copy of the GNU Affero General Public License
-// along with this program. If not, see <https://www.gnu.org/licenses/>.
-
-use lalrpop_util::lalrpop_mod;
-
-lalrpop_mod!(pub parser, "/ranking/signal_aggregator/parser.rs");
-
-pub static PARSER: once_cell::sync::Lazy<parser::AlterationsParser> =
- once_cell::sync::Lazy::new(parser::AlterationsParser::new);
-
-#[derive(Debug, PartialEq, Eq)]
-pub enum Target {
- Signal(String),
- Field(String),
-}
-
-#[derive(Debug, PartialEq)]
-pub struct Alteration {
- pub target: Target,
- pub score: f64,
-}
-
-#[cfg(test)]
-mod tests {
- use super::*;
-
- #[test]
- fn simple() {
- let res: Vec<Alteration> = PARSER
- .parse(
- r#"
- @host_centrality = 3
- @bm25 = 100
- @field_url = 2
- "#,
- )
- .unwrap();
-
- assert_eq!(
- res,
- vec![
- Alteration {
- target: Target::Signal("host_centrality".to_string()),
- score: 3.0,
- },
- Alteration {
- target: Target::Signal("bm25".to_string()),
- score: 100.0
- },
- Alteration {
- target: Target::Field("url".to_string()),
- score: 2.0
- }
- ]
- )
- }
-}
diff --git a/src/ranking/signal_aggregator/parser.lalrpop b/src/ranking/signal_aggregator/parser.lalrpop
deleted file mode 100644
index ebdca900..00000000
--- a/src/ranking/signal_aggregator/parser.lalrpop
+++ /dev/null
@@ -1,25 +0,0 @@
-use std::str::FromStr;
-use crate::ranking::signal_aggregator::ast::*;
-
-grammar;
-
-pub Alterations: Vec<Alteration> = <Alteration*> => <>;
-
-Alteration: Alteration = {
- "@" <target:Target> "=" <score:Num> => Alteration {<>}
-}
-
-Target: Target = {
- <Name> => {
- if let Some(name) = <>.strip_prefix("field_") {
- Target::Field(name.to_string())
- } else {
- Target::Signal(<>)
- }
- },
-}
-
-Name: String = r"[a-zA-Z_]+[0-9]*" => <>.to_string();
-
-
-Num: f64 = r"[0-9]+" => f64::from_str(<>).unwrap();
\ No newline at end of file
diff --git a/src/schema.rs b/src/schema.rs
index bb0fa4d7..c4dc03d2 100644
--- a/src/schema.rs
+++ b/src/schema.rs
@@ -29,9 +29,8 @@ pub enum Field {
StemmedTitle,
StemmedCleanBody,
AllBody,
- StemmedAllBody,
Url,
- Host,
+ Site,
Domain,
DomainIfHomepage, // this field is only set if the webpage is the homepage for the site. Allows us to boost
IsHomepage,
@@ -43,17 +42,20 @@ pub enum Field {
Description,
NumTrackers,
Region,
+ NumUrlTokens,
+ NumTitleTokens,
+ NumCleanBodyTokens,
+ NumDescriptionTokens,
}
-pub static ALL_FIELDS: [Field; 19] = [
+pub static ALL_FIELDS: [Field; 22] = [
Field::Title,
Field::CleanBody,
Field::StemmedTitle,
Field::StemmedCleanBody,
Field::AllBody,
- Field::StemmedAllBody,
Field::Url,
- Field::Host,
+ Field::Site,
Field::Domain,
Field::DomainIfHomepage,
Field::IsHomepage,
@@ -65,6 +67,10 @@ pub static ALL_FIELDS: [Field; 19] = [
Field::Description,
Field::NumTrackers,
Field::Region,
+ Field::NumUrlTokens,
+ Field::NumTitleTokens,
+ Field::NumCleanBodyTokens,
+ Field::NumDescriptionTokens,
];
impl Field {
@@ -88,13 +94,9 @@ impl Field {
Field::Title => IndexingOption::Text(self.default_text_options().set_stored()),
Field::CleanBody => IndexingOption::Text(self.default_text_options()),
Field::Url => IndexingOption::Text(self.default_text_options().set_stored()),
- Field::Host => IndexingOption::Text(self.default_text_options()),
+ Field::Site => IndexingOption::Text(self.default_text_options()),
Field::Domain => IndexingOption::Text(self.default_text_options()),
- Field::AllBody => IndexingOption::Text(self.default_text_options()),
- Field::StemmedAllBody => IndexingOption::Text(
- self.default_text_options_with_tokenizer(Stemmed::as_str())
- .set_stored(),
- ),
+ Field::AllBody => IndexingOption::Text(self.default_text_options().set_stored()),
Field::DomainIfHomepage => IndexingOption::Text(self.default_text_options()),
Field::IsHomepage => IndexingOption::Numeric(
NumericOptions::default()
@@ -133,6 +135,27 @@ impl Field {
),
Field::Description => IndexingOption::Text(self.default_text_options().set_stored()),
Field::Region => IndexingOption::Numeric(
+ NumericOptions::default()
+ .set_fast(Cardinality::SingleValue)
+ .set_stored()
+ .set_indexed(),
+ ),
+ Field::NumCleanBodyTokens => IndexingOption::Numeric(
+ NumericOptions::default()
+ .set_fast(Cardinality::SingleValue)
+ .set_indexed(),
+ ),
+ Field::NumDescriptionTokens => IndexingOption::Numeric(
+ NumericOptions::default()
+ .set_fast(Cardinality::SingleValue)
+ .set_indexed(),
+ ),
+ Field::NumTitleTokens => IndexingOption::Numeric(
+ NumericOptions::default()
+ .set_fast(Cardinality::SingleValue)
+ .set_indexed(),
+ ),
+ Field::NumUrlTokens => IndexingOption::Numeric(
NumericOptions::default()
.set_fast(Cardinality::SingleValue)
.set_indexed(),
@@ -145,7 +168,7 @@ impl Field {
Field::Title => "title",
Field::CleanBody => "body",
Field::Url => "url",
- Field::Host => "host",
+ Field::Site => "site",
Field::BacklinkText => "backlink_text",
Field::Centrality => "centrality",
Field::StemmedTitle => "stemmed_title",
@@ -158,23 +181,25 @@ impl Field {
Field::LastUpdated => "last_updated",
Field::Description => "description",
Field::AllBody => "all_body",
- Field::StemmedAllBody => "stemmed_all_body",
Field::NumTrackers => "num_trackers",
Field::Region => "region",
+ Field::NumUrlTokens => "num_url_tokens",
+ Field::NumTitleTokens => "num_title_tokens",
+ Field::NumCleanBodyTokens => "num_clean_body_tokens",
+ Field::NumDescriptionTokens => "num_description_tokens",
}
}
pub fn boost(&self) -> Option<f32> {
match self {
- Field::Host => Some(6.0),
- Field::DomainIfHomepage => Some(50.0),
+ Field::Site => Some(6.0),
+ Field::DomainIfHomepage => Some(250.0),
Field::StemmedCleanBody | Field::StemmedTitle => Some(0.1),
Field::CleanBody => Some(4.0),
Field::Title => Some(10.0),
Field::Url => Some(1.0),
- Field::Domain => Some(2.0),
+ Field::Domain => Some(1.0),
Field::AllBody => Some(0.01),
- Field::StemmedAllBody => Some(0.001),
Field::BacklinkText => Some(4.0),
Field::Centrality
| Field::IsHomepage
@@ -182,13 +207,33 @@ impl Field {
| Field::FetchTimeMs
| Field::Description
| Field::NumTrackers
+ | Field::NumUrlTokens
+ | Field::NumTitleTokens
+ | Field::NumCleanBodyTokens
+ | Field::NumDescriptionTokens
| Field::Region
| Field::LastUpdated => None,
}
}
pub fn is_searchable(&self) -> bool {
- !matches!(self, Field::PrimaryImage | Field::BacklinkText)
+ !matches!(self, Field::PrimaryImage | Field::BacklinkText) && !self.is_fast()
+ }
+
+ pub fn is_fast(&self) -> bool {
+ matches!(
+ self,
+ Field::IsHomepage
+ | Field::Centrality
+ | Field::FetchTimeMs
+ | Field::LastUpdated
+ | Field::NumTrackers
+ | Field::Region
+ | Field::NumUrlTokens
+ | Field::NumTitleTokens
+ | Field::NumCleanBodyTokens
+ | Field::NumDescriptionTokens
+ )
}
pub fn from_string(name: String) -> Option<Field> {
@@ -196,7 +241,7 @@ impl Field {
"title" => Some(Field::Title),
"body" => Some(Field::CleanBody),
"url" => Some(Field::Url),
- "host" => Some(Field::Host),
+ "site" => Some(Field::Site),
"backlink_text" => Some(Field::BacklinkText),
"centrality" => Some(Field::Centrality),
"stemmed_title" => Some(Field::StemmedTitle),
@@ -209,7 +254,6 @@ impl Field {
"last_updated" => Some(Field::LastUpdated),
"description" => Some(Field::Description),
"all_body" => Some(Field::AllBody),
- "stemmed_all_body" => Some(Field::StemmedAllBody),
"num_trackers" => Some(Field::NumTrackers),
"region" => Some(Field::Region),
_ => None,
diff --git a/src/searcher.rs b/src/searcher.rs
index dbdef47e..79293ba8 100644
--- a/src/searcher.rs
+++ b/src/searcher.rs
@@ -26,7 +26,7 @@ use crate::image_store::Image;
use crate::index::Index;
use crate::inverted_index::InvertedIndexSearchResult;
use crate::query::Query;
-use crate::ranking::signal_aggregator;
+use crate::ranking::goggles::{self, SignalAggregator};
use crate::ranking::Ranker;
use crate::webpage::region::Region;
use crate::webpage::Url;
@@ -73,26 +73,30 @@ impl Searcher {
&self,
query: &str,
selected_region: Option<Region>,
- goggle_program: Option<&str>,
+ goggle_program: Option<String>,
skip_pages: Option<usize>,
) -> Result<SearchResult> {
let start = Instant::now();
let raw_query = query.to_string();
- let aggregator = goggle_program
- .and_then(|program| signal_aggregator::parse(program).ok())
- .unwrap_or_default();
+ let goggle = goggle_program.and_then(|program| goggles::parse(&program).ok());
- let query = Query::parse(
+ let mut query = Query::parse(
query,
self.index.schema(),
self.index.tokenizers(),
- &aggregator,
+ goggle
+ .as_ref()
+ .map(|goggle| &goggle.aggregator)
+ .unwrap_or(&SignalAggregator::default()),
)?;
if query.is_empty() {
return Err(Error::EmptyQuery);
}
+ if let Some(goggle) = &goggle {
+ query.set_goggle(goggle, &self.index.schema());
+ }
if let Some(bangs) = self.bangs.as_ref() {
if let Some(bang) = bangs.get(&query) {
@@ -100,7 +104,10 @@ impl Searcher {
}
}
- let mut ranker = Ranker::new(self.index.region_count.clone(), aggregator);
+ let mut ranker = Ranker::new(
+ self.index.region_count.clone(),
+ goggle.map(|goggle| goggle.aggregator).unwrap_or_default(),
+ );
if let Some(skip_pages) = skip_pages {
ranker = ranker.with_offset(NUM_RESULTS_PER_PAGE * skip_pages);
@@ -170,119 +177,6 @@ mod tests {
use super::*;
- #[test]
- fn custom_signal_aggregation() {
- let mut index = Index::temporary().expect("Unable to open index");
-
- index
- .insert(Webpage::new(
- r#"
- <html>
- <head>
- <title>Test website</title>
- </head>
- <body>
- example
- </body>
- </html>
- "#,
- "https://www.body.com",
- vec![],
- 1.0,
- 20,
- ))
- .expect("failed to parse webpage");
-
- index
- .insert(Webpage::new(
- r#"
- <html>
- <head>
- <title>Example website</title>
- </head>
- <body>
- test
- </body>
- </html>
- "#,
- "https://www.title.com",
- vec![],
- 1.0,
- 20,
- ))
- .expect("failed to parse webpage");
-
- index
- .insert(Webpage::new(
- r#"
- <html>
- <head>
- <title>Example website</title>
- </head>
- <body>
- test
- </body>
- </html>
- "#,
- "https://www.centrality.com",
- vec![],
- 1.0002,
- 500,
- ))
- .expect("failed to parse webpage");
-
- index.commit().unwrap();
-
- let searcher = Searcher::new(index, None, None);
-
- let res = searcher
- .search(
- "example",
- None,
- Some(
- r#"
- @field_title = 20000000
- @host_centrality = 0
- "#,
- ),
- None,
- )
- .unwrap()
- .into_websites()
- .unwrap();
-
- assert_eq!(res.webpages.num_docs, 3);
- assert_eq!(&res.webpages.documents[0].url, "https://www.title.com");
-
- let res = searcher
- .search(
- "example",
- None,
- Some(
- r#"
- @field_all_body= 20000000
- @host_centrality = 0
- "#,
- ),
- None,
- )
- .unwrap()
- .into_websites()
- .unwrap();
-
- assert_eq!(res.webpages.num_docs, 3);
- assert_eq!(&res.webpages.documents[0].url, "https://www.body.com");
-
- let res = searcher
- .search("example", None, Some("@host_centrality= 2000000"), None)
- .unwrap()
- .into_websites()
- .unwrap();
-
- assert_eq!(res.webpages.num_docs, 3);
- assert_eq!(&res.webpages.documents[0].url, "https://www.centrality.com");
- }
-
#[test]
fn offset_page() {
const NUM_PAGES: usize = 10;
diff --git a/src/snippet.rs b/src/snippet.rs
index 1ee8081e..8b77cd3b 100644
--- a/src/snippet.rs
+++ b/src/snippet.rs
@@ -16,6 +16,7 @@
use crate::frontend::search::html_escape;
use crate::tokenizer::Stemmed;
+use crate::webpage::region::Region;
use std::cmp::Ordering;
use std::collections::{BTreeMap, BTreeSet};
use std::ops::Range;
@@ -247,6 +248,7 @@ pub fn generate(
text: &str,
dirty_text: &str,
description: &Option<String>,
+ region: &Region,
searcher: &tantivy::Searcher,
) -> Result<String> {
let field = searcher
@@ -254,31 +256,40 @@ pub fn generate(
.get_field(Field::StemmedCleanBody.as_str())
.expect("Failed to get body field");
- let tokenizer =
- Stemmed::with_forced_language(whatlang::detect_lang(text).unwrap_or(Lang::Eng)).into();
+ let lang = match region.lang() {
+ Some(lang) => lang,
+ None => whatlang::detect_lang(text).unwrap_or(Lang::Eng),
+ };
+
+ let tokenizer = Stemmed::with_forced_language(lang).into();
let generator = SnippetGenerator::create(searcher, query, field, tokenizer)?;
let mut snippet = generator.snippet(text);
if snippet.fragment.is_empty() {
- match description {
- Some(desc) => {
- let tokenizer =
- Stemmed::with_forced_language(whatlang::detect_lang(desc).unwrap_or(Lang::Eng))
- .into();
- let generator = SnippetGenerator::create(searcher, query, field, tokenizer)?;
-
- snippet = generator.snippet(desc);
- if snippet.fragment.is_empty() {
- snippet.fragment = desc.chars().take(DEFAULT_MAX_NUM_CHARS).collect();
+ if text.is_empty() {
+ match description {
+ Some(desc) => {
+ let lang = match region.lang() {
+ Some(lang) => lang,
+ None => whatlang::detect_lang(desc).unwrap_or(Lang::Eng),
+ };
+
+ let tokenizer = Stemmed::with_forced_language(lang).into();
+ let generator = SnippetGenerator::create(searcher, query, field, tokenizer)?;
+
+ snippet = generator.snippet(desc);
+ if snippet.fragment.is_empty() {
+ snippet.fragment = desc.chars().take(DEFAULT_MAX_NUM_CHARS).collect();
+ }
}
- }
- None => {
- if text.is_empty() {
- let tokenizer = Stemmed::with_forced_language(
- whatlang::detect_lang(dirty_text).unwrap_or(Lang::Eng),
- )
- .into();
+ None => {
+ let lang = match region.lang() {
+ Some(lang) => lang,
+ None => whatlang::detect_lang(dirty_text).unwrap_or(Lang::Eng),
+ };
+
+ let tokenizer = Stemmed::with_forced_language(lang).into();
let generator = SnippetGenerator::create(searcher, query, field, tokenizer)?;
snippet = generator.snippet(dirty_text);
@@ -286,10 +297,10 @@ pub fn generate(
if snippet.fragment.is_empty() {
snippet.fragment = dirty_text.chars().take(DEFAULT_MAX_NUM_CHARS).collect();
}
- } else {
- snippet.fragment = text.chars().take(DEFAULT_MAX_NUM_CHARS).collect();
}
}
+ } else {
+ snippet.fragment = text.chars().take(DEFAULT_MAX_NUM_CHARS).collect();
}
}
diff --git a/src/webpage/just_text.rs b/src/webpage/just_text.rs
index f3b44aa4..85bc8c7f 100644
--- a/src/webpage/just_text.rs
+++ b/src/webpage/just_text.rs
@@ -60,7 +60,7 @@ macro_rules! include_stopwords {
$lang,
include_str!($file)
.lines()
- .map(|s| s.to_lowercase())
+ .map(|s| s.to_string())
.collect(),
);
)*
@@ -144,11 +144,11 @@ impl Default for JustText {
}
#[derive(Clone, Debug)]
-struct Paragraph {
+pub struct Paragraph {
is_heading: bool,
tags_count: usize,
chars_count_in_links: usize,
- text: String,
+ pub text: String,
last_was_whitespace: bool,
}
@@ -195,10 +195,12 @@ struct ClassifiedParagraph {
}
impl JustText {
- fn paragraphs(&self, root: NodeRef) -> Vec<Paragraph> {
+ pub fn paragraphs(root: NodeRef) -> Vec<Paragraph> {
let mut res = Vec::new();
- let mut preprocessor = Preprocessor::new(["script", "style", "embed", "form", "head"]);
+ let mut preprocessor = Preprocessor::new([
+ "script", "style", "embed", "form", "head", "noscript", "iframe",
+ ]);
let mut br = false;
let mut link = false;
@@ -261,7 +263,9 @@ impl JustText {
// the <br><br> is a paragraph separator and should
// not be included in the number of tags within the
// paragraph
- paragraph.tags_count -= 1;
+ if paragraph.tags_count > 0 {
+ paragraph.tags_count -= 1;
+ }
}
paragraph.is_heading = heading_count > 0;
@@ -338,26 +342,25 @@ impl JustText {
paragraph
.text
.split_whitespace()
- .filter(|word| stopwords.contains(&word.to_lowercase()))
+ .filter(|word| stopwords.contains(*word))
.count() as f64
/ paragraph.text.split_whitespace().count() as f64
}
- fn initial_classification(&self, paragraphs: Vec<Paragraph>) -> Vec<ClassifiedParagraph> {
+ fn initial_classification(
+ &self,
+ paragraphs: &[Paragraph],
+ lang: &Lang,
+ ) -> Vec<ClassifiedParagraph> {
let mut res = Vec::new();
- let stopwords = paragraphs
- .iter()
- .max_by_key(|p| p.text.len())
- .and_then(|paragraph| {
- let lang = whatlang::detect_lang(&paragraph.text).unwrap_or(Lang::Eng);
- STOPWORDS.get(&lang)
- })
+ let stopwords = STOPWORDS
+ .get(lang)
.unwrap_or_else(|| STOPWORDS.get(&Lang::Eng).unwrap());
for paragraph in paragraphs {
let classification = {
- let stopword_density: f64 = Self::calculate_stopword_density(&paragraph, stopwords);
+ let stopword_density: f64 = Self::calculate_stopword_density(paragraph, stopwords);
if paragraph.link_density() > self.max_link_density
|| paragraph.text.contains("\\xa9")
@@ -384,7 +387,7 @@ impl JustText {
};
res.push(ClassifiedParagraph {
- paragraph,
+ paragraph: paragraph.clone(),
classification: Classification::Intermediate(classification),
});
}
@@ -577,10 +580,9 @@ impl JustText {
.join(" ")
}
- pub fn extract(&self, root: NodeRef) -> String {
- let paragraphs = self.paragraphs(root);
+ pub fn extract_from_paragraphs(&self, paragraphs: &[Paragraph], lang: &Lang) -> String {
let mut classified = self
- .initial_classification(paragraphs)
+ .initial_classification(paragraphs, lang)
.into_iter()
.filter(|par| par.paragraph.text.chars().any(|c| !c.is_whitespace()))
.collect();
@@ -647,7 +649,7 @@ mod tests {
];
let just_text = JustText::default();
- let mut classifications = just_text.initial_classification(paragraphs);
+ let mut classifications = just_text.initial_classification(&paragraphs, &Lang::Eng);
assert!(classifications[0].classification.is_short());
assert!(matches!(
diff --git a/src/webpage/mod.rs b/src/webpage/mod.rs
index 57654b0d..0dd85919 100644
--- a/src/webpage/mod.rs
+++ b/src/webpage/mod.rs
@@ -22,9 +22,11 @@ use serde::{Deserialize, Serialize};
use std::{
borrow::Borrow,
collections::{HashMap, HashSet},
+ panic,
};
-use tantivy::tokenizer::Tokenizer;
+use tantivy::tokenizer::{PreTokenizedString, Tokenizer};
use uuid::Uuid;
+use whatlang::Lang;
mod just_text;
pub mod region;
@@ -33,7 +35,10 @@ mod url;
use crate::schema::{Field, ALL_FIELDS, CENTRALITY_SCALING};
pub use self::url::Url;
-use self::{just_text::JustText, region::Region};
+use self::{
+ just_text::{JustText, Paragraph},
+ region::Region,
+};
static URL_REGEX: once_cell::sync::Lazy<Regex> = once_cell::sync::Lazy::new(|| {
Regex::new(r#"(((http|ftp|https):/{2})+(([0-9a-z_-]+\.)+(aero|asia|biz|cat|com|coop|edu|gov|info|int|jobs|mil|mobi|museum|name|net|org|pro|tel|travel|ac|ad|ae|af|ag|ai|al|am|an|ao|aq|ar|as|at|au|aw|ax|az|ba|bb|bd|be|bf|bg|bh|bi|bj|bm|bn|bo|br|bs|bt|bv|bw|by|bz|ca|cc|cd|cf|cg|ch|ci|ck|cl|cm|cn|co|cr|cu|cv|cx|cy|cz|cz|de|dj|dk|dm|do|dz|ec|ee|eg|er|es|et|eu|fi|fj|fk|fm|fo|fr|ga|gb|gd|ge|gf|gg|gh|gi|gl|gm|gn|gp|gq|gr|gs|gt|gu|gw|gy|hk|hm|hn|hr|ht|hu|id|ie|il|im|in|io|iq|ir|is|it|je|jm|jo|jp|ke|kg|kh|ki|km|kn|kp|kr|kw|ky|kz|la|lb|lc|li|lk|lr|ls|lt|lu|lv|ly|ma|mc|md|me|mg|mh|mk|ml|mn|mn|mo|mp|mr|ms|mt|mu|mv|mw|mx|my|mz|na|nc|ne|nf|ng|ni|nl|no|np|nr|nu|nz|nom|pa|pe|pf|pg|ph|pk|pl|pm|pn|pr|ps|pt|pw|py|qa|re|ra|rs|ru|rw|sa|sb|sc|sd|se|sg|sh|si|sj|sj|sk|sl|sm|sn|so|sr|st|su|sv|sy|sz|tc|td|tf|tg|th|tj|tk|tl|tm|tn|to|tp|tr|tt|tv|tw|tz|ua|ug|uk|us|uy|uz|va|vc|ve|vg|vi|vn|vu|wf|ws|ye|yt|yu|za|zm|zw|arpa)(:[0-9]+)?((/([~0-9a-zA-Z\#\+%@\./_-]+))?(\?[0-9a-zA-Z\+%@/&\[\];=_-]+)?)?))\b"#).unwrap()
@@ -235,6 +240,7 @@ pub struct Html {
root: NodeRef, // this is reference counted (cheap to clone)
all_text: Option<String>,
clean_text: Option<String>,
+ lang: Option<Lang>,
}
impl Html {
@@ -245,21 +251,35 @@ impl Html {
pub fn parse_including_text(html: &str, url: &str, include_text: bool) -> Self {
let root = kuchiki::parse_html().one(html);
- let all_text = if include_text {
- Html::calculate_all_text(root.clone())
- } else {
- None
- };
- let clean_text = if include_text {
- Html::calculate_clean_text(root.clone())
- } else {
- None
- };
+ let mut all_text = None;
+ let mut clean_text = None;
+ let mut lang = None;
+
+ if include_text {
+ let paragraphs = JustText::paragraphs(root.clone());
+
+ lang = paragraphs
+ .iter()
+ .max_by_key(|paragraph| paragraph.text.len())
+ .and_then(|paragraph| {
+ whatlang::detect(&paragraph.text).and_then(|info| {
+ if info.is_reliable() && info.confidence() > 0.95 {
+ Some(info.lang())
+ } else {
+ None
+ }
+ })
+ });
+
+ all_text = Html::calculate_all_text(&paragraphs, &lang.unwrap_or(Lang::Eng));
+ clean_text = Html::calculate_clean_text(&paragraphs, &lang.unwrap_or(Lang::Eng));
+ }
Self {
root,
all_text,
clean_text,
+ lang,
url: url.to_string().into(),
}
}
@@ -267,7 +287,7 @@ impl Html {
pub fn links(&self) -> Vec<Link> {
let mut links = Vec::new();
let mut open_links = Vec::new();
- let mut preprocessor = Preprocessor::new(["script", "style", "head"]);
+ let mut preprocessor = Preprocessor::new(["script", "style", "head", "noscript"]);
for edge in self.root.traverse() {
preprocessor.update(&edge);
@@ -355,8 +375,8 @@ impl Html {
None
}
- fn calculate_clean_text(node: NodeRef) -> Option<String> {
- let text = JustText::default().extract(node);
+ fn calculate_clean_text(paragraphs: &[Paragraph], lang: &Lang) -> Option<String> {
+ let text = JustText::default().extract_from_paragraphs(paragraphs, lang);
if text.is_empty() {
None
@@ -369,7 +389,7 @@ impl Html {
self.clean_text.clone()
}
- fn calculate_all_text(node: NodeRef) -> Option<String> {
+ fn calculate_all_text(paragraphs: &[Paragraph], lang: &Lang) -> Option<String> {
let text = JustText {
max_link_density: 2.0,
length_low: 0,
@@ -378,7 +398,7 @@ impl Html {
stopwords_high: -1.0,
max_heading_distance: 10000,
}
- .extract(node);
+ .extract_from_paragraphs(paragraphs, lang);
if text.is_empty() {
None
@@ -433,34 +453,103 @@ impl Html {
metas
}
+ fn pretokenize_title(&self) -> Result<PreTokenizedString> {
+ let title = self.title();
+
+ if title.is_none() {
+ return Err(Error::EmptyField("title"));
+ }
+ let title = title.unwrap();
+
+ Ok(self.pretokenize_string(title))
+ }
+
+ fn pretokenize_all_text(&self) -> Result<PreTokenizedString> {
+ let all_text = self.all_text();
+
+ if all_text.is_none() {
+ return Err(Error::EmptyField("all body"));
+ }
+ let all_text = all_text.unwrap();
+
+ Ok(self.pretokenize_string(all_text))
+ }
+
+ fn pretokenize_clean_text(&self) -> PreTokenizedString {
+ let clean_text = self.clean_text().unwrap_or_default();
+ self.pretokenize_string(clean_text)
+ }
+
+ fn pretokenize_url(&self) -> PreTokenizedString {
+ let url = self.url().full();
+ self.pretokenize_string(url)
+ }
+
+ fn pretokenize_description(&self) -> PreTokenizedString {
+ let text = self.description().unwrap_or_default();
+
+ self.pretokenize_string(text)
+ }
+
+ fn pretokenize_string(&self, text: String) -> PreTokenizedString {
+ let mut tokens = Vec::new();
+
+ {
+ let mut stream = tokenizer::Normal::default().token_stream(&text);
+ while let Some(token) = stream.next() {
+ tokens.push(token.clone());
+ }
+ }
+
+ PreTokenizedString { text, tokens }
+ }
+
pub fn into_tantivy(self, schema: &tantivy::schema::Schema) -> Result<tantivy::Document> {
let mut doc = tantivy::Document::new();
+ let title = self.pretokenize_title()?;
+ let all_text = self.pretokenize_all_text()?;
+ let clean_text = self.pretokenize_clean_text();
+ let url = self.pretokenize_url();
+ let description = self.pretokenize_description();
+
for field in &ALL_FIELDS {
let tantivy_field = schema
.get_field(field.as_str())
.unwrap_or_else(|| panic!("Unknown field: {}", field.as_str()));
match field {
- Field::Title | Field::StemmedTitle => {
- let title = self.title();
-
- if title.is_none() {
- return Err(Error::EmptyField("title"));
- }
-
- doc.add_text(tantivy_field, title.unwrap());
+ Field::Title => doc.add_pre_tokenized_text(tantivy_field, title.clone()),
+ Field::StemmedTitle => {
+ let mut tokens = title.tokens.clone();
+ stem_tokens(&mut tokens, self.lang.unwrap_or(Lang::Eng));
+
+ doc.add_pre_tokenized_text(
+ tantivy_field,
+ PreTokenizedString {
+ text: title.text.clone(),
+ tokens,
+ },
+ );
}
- Field::CleanBody | Field::StemmedCleanBody => {
- let text = self.clean_text();
-
- doc.add_text(tantivy_field, text.unwrap_or_default());
+ Field::CleanBody => doc.add_pre_tokenized_text(tantivy_field, clean_text.clone()),
+ Field::StemmedCleanBody => {
+ let mut tokens = clean_text.tokens.clone();
+ stem_tokens(&mut tokens, self.lang.unwrap_or(Lang::Eng));
+
+ doc.add_pre_tokenized_text(
+ tantivy_field,
+ PreTokenizedString {
+ text: clean_text.text.clone(),
+ tokens,
+ },
+ );
}
Field::Description => {
- doc.add_text(tantivy_field, self.description().unwrap_or_default());
+ doc.add_pre_tokenized_text(tantivy_field, description.clone());
}
- Field::Url => doc.add_text(tantivy_field, self.url()),
- Field::Host => doc.add_text(tantivy_field, self.url().host()),
+ Field::Url => doc.add_pre_tokenized_text(tantivy_field, url.clone()),
+ Field::Site => doc.add_text(tantivy_field, self.url().site()),
Field::Domain => doc.add_text(tantivy_field, self.url().domain()),
Field::DomainIfHomepage => {
if self.url().is_homepage() {
@@ -477,16 +566,16 @@ impl Html {
self.updated_time()
.map_or(0, |time| time.timestamp().max(0) as u64),
),
- Field::StemmedAllBody | Field::AllBody => {
- let text = self.all_text();
-
- if text.is_none() {
- return Err(Error::EmptyField("all body"));
- }
-
- doc.add_text(tantivy_field, text.unwrap());
- }
+ Field::AllBody => doc.add_pre_tokenized_text(tantivy_field, all_text.clone()),
Field::NumTrackers => doc.add_u64(tantivy_field, self.trackers().len() as u64),
+ Field::NumUrlTokens => doc.add_u64(tantivy_field, url.tokens.len() as u64),
+ Field::NumTitleTokens => doc.add_u64(tantivy_field, title.tokens.len() as u64),
+ Field::NumCleanBodyTokens => {
+ doc.add_u64(tantivy_field, clean_text.tokens.len() as u64)
+ }
+ Field::NumDescriptionTokens => {
+ doc.add_u64(tantivy_field, description.tokens.len() as u64)
+ }
Field::BacklinkText
| Field::Centrality
| Field::FetchTimeMs
@@ -563,9 +652,9 @@ impl Html {
links
.into_iter()
- .filter(|link| !link.host().is_empty())
- .filter(|link| link.host() != self.url().host())
- .unique_by(|link| link.host().to_string())
+ .filter(|link| !link.site().is_empty())
+ .filter(|link| link.site() != self.url().site())
+ .unique_by(|link| link.site().to_string())
.collect()
}
@@ -688,6 +777,39 @@ impl Html {
}
}
+fn stemmer_from_lang(lang: &Lang) -> rust_stemmers::Stemmer {
+ match lang {
+ Lang::Ara => rust_stemmers::Stemmer::create(rust_stemmers::Algorithm::Arabic),
+ Lang::Dan => rust_stemmers::Stemmer::create(rust_stemmers::Algorithm::Danish),
+ Lang::Nld => rust_stemmers::Stemmer::create(rust_stemmers::Algorithm::Dutch),
+ Lang::Fin => rust_stemmers::Stemmer::create(rust_stemmers::Algorithm::Finnish),
+ Lang::Fra => rust_stemmers::Stemmer::create(rust_stemmers::Algorithm::French),
+ Lang::Deu => rust_stemmers::Stemmer::create(rust_stemmers::Algorithm::German),
+ Lang::Ell => rust_stemmers::Stemmer::create(rust_stemmers::Algorithm::Greek),
+ Lang::Hun => rust_stemmers::Stemmer::create(rust_stemmers::Algorithm::Hungarian),
+ Lang::Ita => rust_stemmers::Stemmer::create(rust_stemmers::Algorithm::Italian),
+ Lang::Por => rust_stemmers::Stemmer::create(rust_stemmers::Algorithm::Portuguese),
+ Lang::Ron => rust_stemmers::Stemmer::create(rust_stemmers::Algorithm::Romanian),
+ Lang::Rus => rust_stemmers::Stemmer::create(rust_stemmers::Algorithm::Russian),
+ Lang::Spa => rust_stemmers::Stemmer::create(rust_stemmers::Algorithm::Spanish),
+ Lang::Swe => rust_stemmers::Stemmer::create(rust_stemmers::Algorithm::Swedish),
+ Lang::Tam => rust_stemmers::Stemmer::create(rust_stemmers::Algorithm::Tamil),
+ Lang::Tur => rust_stemmers::Stemmer::create(rust_stemmers::Algorithm::Turkish),
+ _ => rust_stemmers::Stemmer::create(rust_stemmers::Algorithm::English),
+ }
+}
+
+fn stem_tokens(tokens: &mut [tantivy::tokenizer::Token], lang: Lang) {
+ let stemmer = stemmer_from_lang(&lang);
+ for token in tokens {
+ // TODO remove allocation
+ if let Ok(stemmed_str) = panic::catch_unwind(|| stemmer.stem(&token.text).into_owned()) {
+ token.text.clear();
+ token.text.push_str(&stemmed_str);
+ }
+ }
+}
+
#[derive(Debug, PartialEq, Eq)]
pub struct Link {
pub source: Url,
@@ -701,7 +823,7 @@ pub type Meta = HashMap<String, String>;
mod tests {
// TODO: make test macro to test both dom parsers
- use crate::schema_org::ImageObject;
+ use crate::{schema::create_schema, schema_org::ImageObject};
use super::*;
@@ -747,7 +869,7 @@ mod tests {
webpage.url().to_string().as_str(),
"https://www.example.com/whatever"
);
- assert_eq!(webpage.url().host(), "www.example.com");
+ assert_eq!(webpage.url().site(), "www.example.com");
assert_eq!(webpage.url().domain(), "example.com");
}
@@ -877,23 +999,22 @@ mod tests {
}
#[test]
- #[ignore = "JustText doesn't find any content on the sites. How should we split into words for Japanese, Chinese etc.?"]
fn hard_parsing() {
let webpage = Html::parse(include_str!("../../testcases/parsing/yasudaya.html"), "");
assert_eq!(
webpage.title(),
Some("パチンコ大当たり情報 - Pジューシーハニー3 大当たり詳細ページ - やすだひばりヶ丘店".to_string())
);
- assert!(webpage.clean_text().is_some());
- assert!(!webpage.clean_text().unwrap().is_empty());
+ assert!(webpage.all_text().is_some());
+ assert!(!webpage.all_text().unwrap().is_empty());
let webpage = Html::parse(include_str!("../../testcases/parsing/5390001.html"), "");
assert_eq!(
webpage.title(),
Some("特效烟机系列_山东壹线文化传播有限公司".to_string())
);
- assert!(webpage.clean_text().is_some());
- assert!(!webpage.clean_text().unwrap().is_empty());
+ assert!(webpage.all_text().is_some());
+ assert!(!webpage.all_text().unwrap().is_empty());
let webpage = Html::parse(
include_str!("../../testcases/parsing/77p2p-7.live-105.html"),
@@ -901,10 +1022,24 @@ mod tests {
);
assert_eq!(
webpage.title(),
- Some("77p2pЅu¤WЖ[¬Э - ҐDјЅ :: іnєс ".to_string())
+ Some("77p2pЅu¤WЖ[¬Э - ҐDјЅ :: іnєс".to_string())
);
- assert!(webpage.clean_text().is_some());
- assert!(!webpage.clean_text().unwrap().is_empty());
+ assert!(webpage.all_text().is_some());
+ assert!(!webpage.all_text().unwrap().is_empty());
+ }
+
+ #[test]
+ fn out_of_bounds_str() {
+ let webpage = Html::parse(
+ include_str!("../../testcases/parsing/byte_index_out_of_bounds.html"),
+ "",
+ );
+ assert_eq!(webpage.title(), Some("Test".to_string()));
+ assert!(webpage.all_text().is_some());
+ assert!(!webpage.all_text().unwrap().is_empty());
+
+ let schema = create_schema();
+ webpage.into_tantivy(&schema).unwrap();
}
#[test]
@@ -1287,7 +1422,7 @@ mod tests {
assert_eq!(
html.trackers()
.into_iter()
- .map(|url| url.host().to_string())
+ .map(|url| url.site().to_string())
.collect::<Vec<_>>(),
vec![
"cdn.segment.com".to_string(),
diff --git a/src/webpage/region.rs b/src/webpage/region.rs
index ac0ea32a..86f8e766 100644
--- a/src/webpage/region.rs
+++ b/src/webpage/region.rs
@@ -22,7 +22,7 @@ use crate::{Error, Result};
use super::Webpage;
-#[derive(PartialEq, Eq, Clone, Copy, Hash, Serialize, Deserialize)]
+#[derive(Debug, PartialEq, Eq, Clone, Copy, Hash, Serialize, Deserialize)]
pub enum Region {
All,
Denmark,
@@ -32,6 +32,12 @@ pub enum Region {
US,
}
+impl Default for Region {
+ fn default() -> Self {
+ Region::All
+ }
+}
+
pub const ALL_REGIONS: [Region; 6] = [
Region::All,
Region::Denmark,
@@ -86,28 +92,15 @@ impl Region {
}
pub fn guess_from(webpage: &Webpage) -> Result<Self> {
- match webpage
- .html
- .clean_text()
- .or_else(|| webpage.html.all_text())
- .or_else(|| webpage.html.title())
- {
- Some(text) => whatlang::detect(&text)
- .and_then(|info| {
- if info.is_reliable() && info.confidence() > 0.95 {
- match info.lang() {
- whatlang::Lang::Eng => Some(Ok(Region::US)),
- whatlang::Lang::Spa => Some(Ok(Region::Spain)),
- whatlang::Lang::Fra => Some(Ok(Region::France)),
- whatlang::Lang::Deu => Some(Ok(Region::Germany)),
- whatlang::Lang::Dan => Some(Ok(Region::Denmark)),
- _ => Some(Err(Error::UnknownRegion)),
- }
- } else {
- None
- }
- })
- .unwrap_or(Err(Error::UnknownRegion)),
+ match webpage.html.lang {
+ Some(lang) => match lang {
+ whatlang::Lang::Eng => Ok(Region::US),
+ whatlang::Lang::Spa => Ok(Region::Spain),
+ whatlang::Lang::Fra => Ok(Region::France),
+ whatlang::Lang::Deu => Ok(Region::Germany),
+ whatlang::Lang::Dan => Ok(Region::Denmark),
+ _ => Err(Error::UnknownRegion),
+ },
None => Err(Error::UnknownRegion),
}
}
@@ -115,18 +108,30 @@ impl Region {
pub fn from_id(doc: u64) -> Self {
ALL_REGIONS[doc as usize]
}
+
+ pub fn lang(&self) -> Option<whatlang::Lang> {
+ match self {
+ Region::Denmark => Some(whatlang::Lang::Dan),
+ Region::France => Some(whatlang::Lang::Fra),
+ Region::Germany => Some(whatlang::Lang::Deu),
+ Region::Spain => Some(whatlang::Lang::Spa),
+ Region::US => Some(whatlang::Lang::Eng),
+ Region::All => None,
+ }
+ }
}
#[derive(Serialize, Deserialize, Default, Clone)]
pub struct RegionCount {
map: HashMap<Region, u64>,
+ fast_count: Vec<Option<u64>>,
total_counts: u64,
path: String,
}
impl RegionCount {
pub fn open<P: AsRef<Path>>(path: P) -> Self {
- let map = if !path.as_ref().exists() {
+ let map: HashMap<Region, u64> = if !path.as_ref().exists() {
if let Some(parent) = path.as_ref().parent() {
std::fs::create_dir_all(parent).unwrap();
}
@@ -137,9 +142,22 @@ impl RegionCount {
serde_json::from_str(&json).unwrap()
};
+ let mut fast_count = Vec::new();
+
+ for (region, count) in &map {
+ let idx = region.id() as usize;
+
+ while idx >= fast_count.len() {
+ fast_count.push(None);
+ }
+
+ fast_count[idx] = Some(*count);
+ }
+
RegionCount {
total_counts: map.iter().map(|(_, count)| count).sum(),
map,
+ fast_count,
path: path.as_ref().to_str().unwrap().to_string(),
}
}
@@ -155,6 +173,20 @@ impl RegionCount {
let mut file = File::options().write(true).open(&self.path).unwrap();
file.write_all(json.as_bytes()).unwrap();
self.total_counts = self.map.iter().map(|(_, count)| count).sum();
+
+ let mut fast_count = Vec::new();
+
+ for (region, count) in &self.map {
+ let idx = region.id() as usize;
+
+ while idx >= fast_count.len() {
+ fast_count.push(None);
+ }
+
+ fast_count[idx] = Some(*count);
+ }
+
+ self.fast_count = fast_count;
}
pub fn merge(&mut self, other: Self) {
@@ -168,9 +200,12 @@ impl RegionCount {
}
pub fn score(&self, region: &Region) -> f64 {
- self.map
- .get(region)
- .map(|count| *count as f64 / self.total_counts as f64)
+ self.fast_count
+ .get(region.id() as usize)
+ .map(|count| match count {
+ Some(count) => *count as f64 / self.total_counts as f64,
+ None => 0.0,
+ })
.unwrap_or(0.0)
}
}
diff --git a/src/webpage/url.rs b/src/webpage/url.rs
index fe4e4037..02780160 100644
--- a/src/webpage/url.rs
+++ b/src/webpage/url.rs
@@ -58,40 +58,40 @@ impl Url {
&url[..start_query]
}
- pub fn host(&self) -> &str {
+ pub fn site(&self) -> &str {
let url = self.strip_protocol();
- let mut end_host = url.len();
+ let mut end_site = url.len();
if url.contains('/') {
- end_host = url.find('/').expect("The url contains atleast 1 '/'");
+ end_site = url.find('/').expect("The url contains atleast 1 '/'");
}
- &url[..end_host]
+ &url[..end_site]
}
pub fn domain(&self) -> &str {
- let host = self.host();
- let num_punctuations: usize = host.chars().map(|c| if c == '.' { 1 } else { 0 }).sum();
+ let site = self.site();
+ let num_punctuations: usize = site.chars().map(|c| if c == '.' { 1 } else { 0 }).sum();
if num_punctuations > 1 {
- let domain_index = host.rfind('.').unwrap();
- let mut start_index = host[..domain_index].rfind('.').unwrap() + 1;
+ let domain_index = site.rfind('.').unwrap();
+ let mut start_index = site[..domain_index].rfind('.').unwrap() + 1;
- if &host[start_index..] == "co.uk" {
- if let Some(new_start_index) = host[..start_index - 1].rfind('.') {
+ if &site[start_index..] == "co.uk" {
+ if let Some(new_start_index) = site[..start_index - 1].rfind('.') {
start_index = new_start_index + 1;
} else {
start_index = 0;
}
}
- &host[start_index..]
+ &site[start_index..]
} else {
- host
+ site
}
}
pub fn subdomain(&self) -> Option<&str> {
- if let Some(subdomain) = self.host().strip_suffix(self.domain()) {
+ if let Some(subdomain) = self.site().strip_suffix(self.domain()) {
if subdomain.is_empty() || subdomain == "." {
None
} else {
@@ -129,28 +129,16 @@ impl Url {
&self.0[..self.find_protocol_end()]
}
- pub fn site(&self) -> &str {
- let start = self.find_protocol_end() + 3;
- let url = &self.0[start..];
-
- let mut end_host = url.len();
- if url.contains('/') {
- end_host = url.find('/').expect("The url contains atleast 1 '/'");
- }
-
- &self.0[..end_host + start]
- }
-
pub fn is_full_path(&self) -> bool {
matches!(self.protocol(), "http" | "https" | "pdf")
}
pub fn prefix_with(&mut self, url: &Url) {
self.0 = match (url.0.ends_with('/'), self.0.starts_with('/')) {
- (true, true) => url.site().to_string() + &self.0,
- (true, false) => url.0.clone() + &self.0,
- (false, true) => url.site().to_string() + &self.0,
- (false, false) => url.0.clone() + "/" + &self.0,
+ (true, true) => url.protocol().to_string() + "://" + url.site() + &self.0,
+ (true, false) => url.full() + &self.0,
+ (false, true) => url.protocol().to_string() + "://" + url.site() + &self.0,
+ (false, false) => url.full() + "/" + &self.0,
};
}
@@ -189,10 +177,10 @@ impl Url {
if subdomain == "www" {
self.domain()
} else {
- self.host()
+ self.site()
}
} else {
- self.host()
+ self.site()
}
}
}
@@ -206,7 +194,7 @@ mod tests {
let url: Url = "//scripts.dailymail.co.uk".to_string().into();
assert_eq!(url.domain(), "dailymail.co.uk");
- assert_eq!(url.host(), "scripts.dailymail.co.uk");
+ assert_eq!(url.site(), "scripts.dailymail.co.uk");
}
#[test]
@@ -214,7 +202,7 @@ mod tests {
let url: Url = "dailymail.co.uk".to_string().into();
assert_eq!(url.domain(), "dailymail.co.uk");
- assert_eq!(url.host(), "dailymail.co.uk");
+ assert_eq!(url.site(), "dailymail.co.uk");
assert_eq!(url.full().as_str(), "https://dailymail.co.uk");
}
diff --git a/testcases/goggles/hacker_news.goggle b/testcases/goggles/hacker_news.goggle
new file mode 100644
index 00000000..55431e9e
--- /dev/null
+++ b/testcases/goggles/hacker_news.goggle
@@ -0,0 +1,6245 @@
+! name: Hacker News / 1k short
+! description: Prioritizes domains popular with the Hacker News community, minus those that would rank among the top 1000 most-viewed websites.
+! author: Goggles 101
+! public: true
+! avatar: #ff6600
+
+$discard
+$boost=4,site=blog.samaltman.com
+$boost=4,site=paulgraham.com
+$boost=4,site=blog.rust-lang.org
+$boost=4,site=blog.ycombinator.com
+$boost=4,site=ciechanow.ski
+$boost=4,site=danluu.com
+$boost=4,site=kalzumeus.com
+$boost=4,site=signal.org
+$boost=4,site=antirez.com
+$boost=4,site=jacquesmattheij.com
+$boost=4,site=chris-granger.com
+$boost=4,site=letsencrypt.org
+$boost=4,site=teslamotors.com
+$boost=4,site=lore.kernel.org
+$boost=4,site=frame.work
+$boost=4,site=righto.com
+$boost=4,site=troyhunt.com
+$boost=4,site=sec.gov
+$boost=4,site=tesla.com
+$boost=4,site=keybase.io
+$boost=4,site=blender.org
+$boost=4,site=spacex.com
+$boost=4,site=susanjfowler.com
+$boost=4,site=eff.org
+$boost=4,site=blog.golang.org
+$boost=4,site=matrix.org
+$boost=4,site=openai.com
+$boost=4,site=sublimetext.com
+$boost=4,site=mtlynch.io
+$boost=4,site=blog.pinboard.in
+$boost=4,site=whispersystems.org
+$boost=4,site=fabiensanglard.net
+$boost=4,site=drewdevault.com
+$boost=4,site=blog.interviewing.io
+$boost=4,site=github.blog
+$boost=4,site=calpaterson.com
+$boost=4,site=rachelbythebay.com
+$boost=4,site=bellard.org
+$boost=4,site=training.kalzumeus.com
+$boost=4,site=jvns.ca
+$boost=4,site=daemonology.net
+$boost=4,site=ifixit.org
+$boost=4,site=tonsky.me
+$boost=4,site=buzzfeednews.com
+$boost=4,site=deepmind.com
+$boost=4,site=postgresql.org
+$boost=4,site=krebsonsecurity.com
+$boost=4,site=overreacted.io
+$boost=4,site=stratechery.com
+$boost=4,site=brave.com
+$boost=4,site=lwn.net
+$boost=4,site=stephendiehl.com
+$boost=4,site=raganwald.posterous.com
+$boost=4,site=bugs.chromium.org
+$boost=4,site=marco.org
+$boost=4,site=words.steveklabnik.com
+$boost=4,site=backblaze.com
+$boost=4,site=robert.ocallahan.org
+$boost=4,site=blog.cryptographyengineering.com
+$boost=4,site=citizenlab.ca
+$boost=4,site=propublica.org
+$boost=4,site=blog.benjojo.co.uk
+$boost=4,site=justine.lol
+$boost=4,site=lists.zx2c4.com
+$boost=4,site=supremecourt.gov
+$boost=4,site=idlewords.com
+$boost=4,site=jepsen.io
+$boost=4,site=zachholman.com
+$boost=4,site=githubengineering.com
+$boost=4,site=tinyprojects.dev
+$boost=4,site=quantamagazine.org
+$boost=4,site=tbray.org
+$boost=4,site=sqlite.org
+$boost=4,site=sivers.org
+$boost=4,site=inkscape.org
+$boost=4,site=deepsouthventures.com
+$boost=4,site=blog.openai.com
+$boost=4,site=dcurt.is
+$boost=4,site=blog.atom.io
+$boost=4,site=ccc.de
+$boost=4,site=jeffhuang.com
+$boost=4,site=marc.info
+$boost=4,site=anandtech.com
+$boost=4,site=setosa.io
+$boost=4,site=gogameguru.com
+$boost=4,site=patrickcollison.com
+$boost=4,site=daringfireball.net
+$boost=4,site=codinghorror.com
+$boost=4,site=blog.chriszacharias.com
+$boost=4,site=blog.priceonomics.com
+$boost=4,site=petapixel.com
+$boost=4,site=aphyr.com
+$boost=4,site=blog.jupyter.org
+$boost=4,site=josephg.com
+$boost=4,site=candyjapan.com
+$boost=4,site=raspberrypi.org
+$boost=4,site=blog.triplebyte.com
+$boost=4,site=web.eecs.utk.edu
+$boost=4,site=foundationdb.org
+$boost=4,site=factorio.com
+$boost=4,site=brew.sh
+$boost=4,site=jlongster.com
+$boost=4,site=blog.dustinkirkland.com
+$boost=4,site=daniel.haxx.se
+$boost=4,site=rethinkdb.com
+$boost=4,site=andrewkelley.me
+$boost=4,site=theintercept.com
+$boost=4,site=lkml.org
+$boost=4,site=justinobeirne.com
+$boost=4,site=pine64.org
+$boost=4,site=thirtythreeforty.net
+$boost=4,site=aclu.org
+$boost=4,site=robertheaton.com
+$boost=4,site=justice.gov
+$boost=4,site=startupclass.samaltman.com
+$boost=4,site=atlasobscura.com
+$boost=4,site=wozniak.ca
+$boost=4,site=shkspr.mobi
+$boost=4,site=wikileaks.org
+$boost=4,site=lucumr.pocoo.org
+$boost=4,site=defmacro.org
+$boost=4,site=cloudplatform.googleblog.com
+$boost=4,site=gatesnotes.com
+$boost=4,site=lapcatsoftware.com
+$boost=4,site=bunniestudios.com
+$boost=4,site=puri.sm
+$boost=4,site=openculture.com
+$boost=4,site=spreadprivacy.com
+$boost=4,site=37signals.com
+$boost=4,site=thedaywefightback.org
+$boost=4,site=asia.nikkei.com
+$boost=4,site=blog.pragmaticengineer.com
+$boost=4,site=brandur.org
+$boost=4,site=worrydream.com
+$boost=4,site=gimp.org
+$boost=4,site=gabrielweinberg.com
+$boost=4,site=tyler.io
+$boost=4,site=deno.land
+$boost=4,site=sheddingbikes.com
+$boost=4,site=pud.com
+$boost=4,site=lavabit.com
+$boost=4,site=blog.bradfieldcs.com
+$boost=4,site=smithsonianmag.com
+$boost=4,site=elm-lang.org
+$boost=4,site=themacro.com
+$boost=4,site=security.googleblog.com
+$boost=4,site=voxelquest.com
+$boost=4,site=jeffgeerling.com
+$boost=4,site=stallman.org
+$boost=4,site=blog.torproject.org
+$boost=4,site=cockroachlabs.com
+$boost=4,site=earthly.dev
+$boost=4,site=nullprogram.com
+$boost=4,site=teachyourselfcs.com
+$boost=4,site=immersivemath.com
+$boost=4,site=fly.io
+$boost=4,site=blog.system76.com
+$boost=4,site=torrentfreak.com
+$boost=4,site=swift.org
+$boost=4,site=plausible.io
+$boost=4,site=blog.timescale.com
+$boost=4,site=elixir-lang.org
+$boost=4,site=einaregilsson.com
+$boost=4,site=blog.servo.org
+$boost=4,site=brendangregg.com
+$boost=4,site=blog.jessfraz.com
+$boost=4,site=gwern.net
+$boost=4,site=videolan.org
+$boost=4,site=arp242.net
+$boost=4,site=mjg59.dreamwidth.org
+$boost=4,site=stavros.io
+$boost=4,site=expeditedssl.com
+$boost=4,site=matt.might.net
+$boost=4,site=flif.info
+$boost=4,site=apenwarr.ca
+$boost=4,site=priceonomics.com
+$boost=4,site=humbledmba.com
+$boost=4,site=unrealengine.com
+$boost=4,site=people.xiph.org
+$boost=4,site=betterexplained.com
+$boost=4,site=al3x.net
+$boost=4,site=deno.com
+$boost=4,site=git.kernel.org
+$boost=4,site=djangoproject.com
+$boost=4,site=ruby-lang.org
+$boost=4,site=catonmat.net
+$boost=4,site=blog.plover.com
+$boost=4,site=steveblank.com
+$boost=4,site=blog.haschek.at
+$boost=4,site=blog.jgc.org
+$boost=4,site=utcc.utoronto.ca
+$boost=4,site=protonmail.com
+$boost=4,site=stackoverflow.blog
+$boost=4,site=feynmanlectures.caltech.edu
+$boost=4,site=thume.ca
+$boost=4,site=blog.google
+$boost=4,site=talks.golang.org
+$boost=4,site=antipope.org
+$boost=4,site=blog.coinbase.com
+$boost=4,site=freedom.press
+$boost=4,site=research.googleblog.com
+$boost=4,site=sneak.berlin
+$boost=4,site=ferd.ca
+$boost=4,site=system76.com
+$boost=4,site=jsomers.net
+$boost=4,site=nautil.us
+$boost=4,site=blog.fogus.me
+$boost=4,site=probablydance.com
+$boost=4,site=blog.stephenwolfram.com
+$boost=4,site=blog.42floors.com
+$boost=4,site=thoughtcrime.org
+$boost=4,site=winehq.org
+$boost=4,site=benkuhn.net
+$boost=4,site=dev.stephendiehl.com
+$boost=4,site=thedrive.com
+$boost=4,site=explainshell.com
+$boost=4,site=krita.org
+$boost=4,site=techdirt.com
+$boost=4,site=lists.freebsd.org
+$boost=4,site=axios.com
+$boost=4,site=triplebyte.com
+$boost=4,site=blog.filippo.io
+$boost=4,site=weblog.rubyonrails.org
+$boost=4,site=viewsourcecode.org
+$boost=4,site=tjcx.me
+$boost=4,site=scottaaronson.com
+$boost=4,site=mattkeeter.com
+$boost=4,site=edweissman.com
+$boost=4,site=citylab.com
+$boost=4,site=ziglang.org
+$boost=4,site=ffmpeg.org
+$boost=4,site=tablix.org
+$boost=4,site=opensource.googleblog.com
+$boost=4,site=lettersofnote.com
+$boost=4,site=blog.pushbullet.com
+$boost=4,site=zwischenzugs.com
+$boost=4,site=blogs.valvesoftware.com
+$boost=4,site=markosaric.com
+$boost=4,site=os.phil-opp.com
+$boost=4,site=juliareda.eu
+$boost=4,site=uspirg.org
+$boost=4,site=fasterthanli.me
+$boost=4,site=ipfs.io
+$boost=4,site=queue.acm.org
+$boost=4,site=evanmiller.org
+$boost=4,site=ejohn.org
+$boost=4,site=lcamtuf.coredump.cx
+$boost=4,site=mango.pdf.zone
+$boost=4,site=godotengine.org
+$boost=4,site=justgetflux.com
+$boost=4,site=bradfrost.com
+$boost=4,site=nim-lang.org
+$boost=4,site=simonwillison.net
+$boost=4,site=groklaw.net
+$boost=4,site=lithub.com
+$boost=4,site=dave.cheney.net
+$boost=4,site=beepb00p.xyz
+$boost=4,site=blog.fastmail.com
+$boost=4,site=seclists.org
+$boost=4,site=hillelwayne.com
+$boost=4,site=filfre.net
+$boost=4,site=oculus.com
+$boost=4,site=graydon2.dreamwidth.org
+$boost=4,site=jonathanturner.org
+$boost=4,site=jameshfisher.com
+$boost=4,site=julialang.org
+$boost=4,site=academictorrents.com
+$boost=4,site=staltz.com
+$boost=4,site=openbsd.org
+$boost=4,site=psyche.co
+$boost=4,site=blog.rongarret.info
+$boost=4,site=redblobgames.com
+$boost=4,site=dtrace.org
+$boost=4,site=boingboing.net
+$boost=4,site=schneier.com
+$boost=4,site=twobithistory.org
+$boost=4,site=tttthis.com
+$boost=4,site=windows93.net
+$boost=4,site=ascii.textfiles.com
+$boost=4,site=aaronsw.com
+$boost=4,site=blog.chromium.org
+$boost=4,site=daltoncaldwell.com
+$boost=4,site=jeffe.cs.illinois.edu
+$boost=4,site=hookrace.net
+$boost=4,site=blog.garrytan.com
+$boost=4,site=leahneukirchen.org
+$boost=4,site=journal.stuffwithstuff.com
+$boost=4,site=hanselman.com
+$boost=4,site=prog21.dadgum.com
+$boost=4,site=craigmurray.org.uk
+$boost=4,site=acko.net
+$boost=4,site=misc-stuff.terraaeon.com
+$boost=4,site=blog.robertelder.org
+$boost=4,site=patrick-breyer.de
+$boost=4,site=mainichi.jp
+$boost=4,site=notepad-plus-plus.org
+$boost=4,site=dolphin-emu.org
+$boost=4,site=adriancourreges.com
+$boost=4,site=anishathalye.com
+$boost=4,site=martin.kleppmann.com
+$boost=4,site=buy.louisck.net
+$boost=4,site=sandstorm.io
+$boost=4,site=blog.fogcreek.com
+$boost=4,site=panic.com
+$boost=4,site=svelte.dev
+$boost=4,site=justinkan.com
+$boost=4,site=ianbicking.org
+$boost=4,site=firstlook.org
+$boost=4,site=foundersatwork.posthaven.com
+$boost=4,site=acoup.blog
+$boost=4,site=jacobian.org
+$boost=4,site=hakibenita.com
+$boost=4,site=mothership.sg
+$boost=4,site=vitavonni.de
+$boost=4,site=palant.de
+$boost=4,site=benhoyt.com
+$boost=4,site=david.li
+$boost=4,site=a16z.com
+$boost=4,site=battleforthenet.com
+$boost=4,site=uncrunched.com
+$boost=4,site=nullsweep.com
+$boost=4,site=wiki.postgresql.org
+$boost=4,site=snellman.net
+$boost=4,site=craigkerstiens.com
+$boost=4,site=wikimediafoundation.org
+$boost=4,site=eater.net
+$boost=4,site=redditblog.com
+$boost=4,site=aaronkharris.com
+$boost=4,site=qz.com
+$boost=4,site=tailscale.com
+$boost=4,site=codewords.recurse.com
+$boost=4,site=bost.ocks.org
+$boost=4,site=norvig.com
+$boost=4,site=pluralistic.net
+$boost=4,site=simplethread.com
+$boost=4,site=nltimes.nl
+$boost=4,site=jitsi.org
+$boost=4,site=article.gmane.org
+$boost=4,site=qntm.org
+$boost=4,site=iterm2.com
+$boost=4,site=geoff.greer.fm
+$boost=4,site=reactjs.org
+$boost=4,site=citusdata.com
+$boost=4,site=pastebin.com
+$boost=4,site=fast.ai
+$boost=4,site=begriffs.com
+$boost=4,site=lkml.iu.edu
+$boost=4,site=eclecticlight.co
+$boost=4,site=sam.zeloof.xyz
+$boost=4,site=cfenollosa.com
+$boost=4,site=stevelosh.com
+$boost=4,site=marctenbosch.com
+$boost=4,site=thehustle.co
+$boost=4,site=brendaneich.com
+$boost=4,site=42floors.com
+$boost=4,site=blog.alexmaccaw.com
+$boost=4,site=martinfowler.com
+$boost=4,site=freedom-to-tinker.com
+$boost=4,site=iphoneincanada.ca
+$boost=4,site=ir.tesla.com
+$boost=4,site=blog.asmartbear.com
+$boost=4,site=gameprogrammingpatterns.com
+$boost=4,site=blogs.msdn.com
+$boost=4,site=meteor.com
+$boost=4,site=segment.com
+$boost=4,site=goodnewsnetwork.org
+$boost=4,site=crypto101.io
+$boost=4,site=hintjens.com
+$boost=4,site=changelog.complete.org
+$boost=4,site=webkit.org
+$boost=4,site=blog.emacsen.net
+$boost=4,site=boringtechnology.club
+$boost=4,site=swombat.com
+$boost=4,site=berthub.eu
+$boost=4,site=erikbern.com
+$boost=4,site=theregister.com
+$boost=4,site=learnpythonthehardway.org
+$boost=4,site=dynomight.net
+$boost=4,site=danshipper.com
+$boost=4,site=kennethreitz.org
+$boost=4,site=jefftk.com
+$boost=4,site=idiallo.com
+$boost=4,site=blog.burntsushi.net
+$boost=4,site=roguelazer.com
+$boost=4,site=beej.us
+$boost=4,site=destroyallsoftware.com
+$boost=4,site=eev.ee
+$boost=4,site=konklone.com
+$boost=4,site=twilio.com
+$boost=4,site=outsideonline.com
+$boost=4,site=chrisstucchio.com
+$boost=4,site=themarkup.org
+$boost=4,site=nobelprize.org
+$boost=4,site=samcurry.net
+$boost=4,site=phoboslab.org
+$boost=4,site=media.ccc.de
+$boost=4,site=tom.preston-werner.com
+$boost=4,site=iism.org
+$boost=4,site=blog.joinmastodon.org
+$boost=4,site=accc.gov.au
+$boost=4,site=ianstormtaylor.com
+$boost=4,site=imperialviolet.org
+$boost=4,site=ridiculousfish.com
+$boost=4,site=developers.googleblog.com
+$boost=4,site=theparisreview.org
+$boost=4,site=michaeleisen.org
+$boost=4,site=blog.nuclino.com
+$boost=4,site=universityofcalifornia.edu
+$boost=4,site=gravitational.com
+$boost=4,site=zserge.com
+$boost=4,site=observablehq.com
+$boost=4,site=mathoverflow.net
+$boost=4,site=joyent.com
+$boost=4,site=blog.kevmod.com
+$boost=4,site=blog.heroku.com
+$boost=4,site=strongtowns.org
+$boost=4,site=nextjs.org
+$boost=4,site=icij.org
+$boost=4,site=ifixit.com
+$boost=4,site=windytan.com
+$boost=4,site=servethehome.com
+$boost=4,site=grumpy.website
+$boost=4,site=bitwarden.com
+$boost=4,site=distill.pub
+$boost=4,site=craftinginterpreters.com
+$boost=4,site=lord.io
+$boost=4,site=eso.org
+$boost=4,site=codersnotes.com
+$boost=4,site=blog.regehr.org
+$boost=4,site=highscalability.com
+$boost=4,site=toddwschneider.com
+$boost=4,site=slatestarcodex.com
+$boost=4,site=cloudflarestatus.com
+$boost=4,site=cnx-software.com
+$boost=4,site=statnews.com
+$boost=4,site=reneweconomy.com.au
+$boost=4,site=jwz.org
+$boost=4,site=ncase.me
+$boost=4,site=metzdowd.com
+$boost=4,site=edu.mkrecny.com
+$boost=4,site=mastermind.atavist.com
+$boost=4,site=laphamsquarterly.org
+$boost=4,site=therecord.media
+$boost=4,site=salon.com
+$boost=4,site=howacarworks.com
+$boost=4,site=paulstamatiou.com
+$boost=4,site=eloquentjavascript.net
+$boost=4,site=nathanmarz.com
+$boost=4,site=explained.ai
+$boost=4,site=golang.org
+$boost=4,site=99percentinvisible.org
+$boost=4,site=restoreprivacy.com
+$boost=4,site=ycombinator.com
+$boost=4,site=joshwcomeau.com
+$boost=4,site=easydns.com
+$boost=4,site=blog.racket-lang.org
+$boost=4,site=ctrl.blog
+$boost=4,site=bjk5.com
+$boost=4,site=refactoringui.com
+$boost=4,site=reactos.org
+$boost=4,site=matthias-endler.de
+$boost=4,site=blog.usejournal.com
+$boost=4,site=deprocrastination.co
+$boost=4,site=blog.llvm.org
+$boost=4,site=blog.keras.io
+$boost=4,site=markodenic.com
+$boost=4,site=tirania.org
+$boost=4,site=copetti.org
+$boost=4,site=interviewcake.com
+$boost=4,site=newatlas.com
+$boost=4,site=blog.okcupid.com
+$boost=4,site=smallstep.com
+$boost=4,site=act.eff.org
+$boost=4,site=extremetech.com
+$boost=4,site=kottke.org
+$boost=4,site=aeon.co
+$boost=4,site=blog.printf.net
+$boost=4,site=johndcook.com
+$boost=4,site=community.letsencrypt.org
+$boost=4,site=hashicorp.com
+$boost=4,site=thewaltdisneycompany.com
+$boost=4,site=haiku-os.org
+$boost=4,site=scotthelme.co.uk
+$boost=4,site=gamasutra.com
+$boost=4,site=cdixon.org
+$boost=4,site=azeria-labs.com
+$boost=4,site=stevehanov.ca
+$boost=4,site=calebporzio.com
+$boost=4,site=blog.erlang.org
+$boost=4,site=blog.backblaze.com
+$boost=4,site=nymag.com
+$boost=4,site=blog.mailgun.com
+$boost=4,site=tokio.rs
+$boost=4,site=woodgears.ca
+$boost=4,site=zotero.org
+$boost=4,site=zeit.co
+$boost=4,site=realworldtech.com
+$boost=4,site=mattcutts.com
+$boost=4,site=thecut.com
+$boost=4,site=theunderstatement.com
+$boost=4,site=donmelton.com
+$boost=4,site=learncodethehardway.org
+$boost=4,site=element.io
+$boost=4,site=thehill.com
+$boost=4,site=wattenberger.com
+$boost=4,site=raganwald.com
+$boost=4,site=danshapiro.com
+$boost=4,site=amasci.com
+$boost=4,site=blog.instapaper.com
+$boost=4,site=pathsensitive.com
+$boost=4,site=sachagreif.com
+$boost=4,site=daedtech.com
+$boost=4,site=phoenixframework.org
+$boost=4,site=avc.com
+$boost=4,site=jasmcole.com
+$boost=4,site=lemire.me
+$boost=4,site=ycombinator.posterous.com
+$boost=4,site=crystal-lang.org
+$boost=4,site=flightradar24.com
+$boost=4,site=lofibucket.com
+$boost=4,site=agner.org
+$boost=4,site=mercurynews.com
+$boost=4,site=peteris.rocks
+$boost=4,site=haneycodes.net
+$boost=4,site=blog.royalsloth.eu
+$boost=4,site=jeremykun.com
+$boost=4,site=netmarketshare.com
+$boost=4,site=streaming.media.ccc.de
+$boost=4,site=scattered-thoughts.net
+$boost=4,site=theoutline.com
+$boost=4,site=tedunangst.com
+$boost=4,site=siberiantimes.com
+$boost=4,site=avherald.com
+$boost=4,site=hackaday.com
+$boost=4,site=hackerfactor.com
+$boost=4,site=doist.com
+$boost=4,site=ai.googleblog.com
+$boost=4,site=macwright.com
+$boost=4,site=cake.co
+$boost=4,site=blog.ptsecurity.com
+$boost=4,site=harpers.org
+$boost=4,site=research.swtch.com
+$boost=4,site=khanacademy.org
+$boost=4,site=malwaretech.com
+$boost=4,site=chronicle.com
+$boost=4,site=michaelnielsen.org
+$boost=4,site=threadreaderapp.com
+$boost=4,site=lowtechmagazine.com
+$boost=4,site=thorstenball.com
+$boost=4,site=copy.sh
+$boost=4,site=datagubbe.se
+$boost=4,site=thehftguy.com
+$boost=4,site=fishshell.com
+$boost=4,site=lists.torproject.org
+$boost=4,site=wbur.org
+$boost=4,site=blog.inkdrop.info
+$boost=4,site=retool.com
+$boost=4,site=aboutamazon.com
+$boost=4,site=randsinrepose.com
+$boost=4,site=inkandswitch.com
+$boost=4,site=bhorowitz.com
+$boost=4,site=keenen.xyz
+$boost=4,site=zachtronics.com
+$boost=4,site=blog.elementary.io
+$boost=4,site=blog.waymo.com
+$boost=4,site=ghostbin.com
+$boost=4,site=blog.cubeofm.com
+$boost=4,site=privacyinternational.org
+$boost=4,site=mondaynote.com
+$boost=4,site=swissinfo.ch
+$boost=4,site=fs.blog
+$boost=4,site=rosenzweig.io
+$boost=4,site=thehackerblog.com
+$boost=4,site=palant.info
+$boost=4,site=course.fast.ai
+$boost=4,site=niemanlab.org
+$boost=4,site=perens.com
+$boost=4,site=matuzo.at
+$boost=4,site=yosefk.com
+$boost=4,site=samsaffron.com
+$boost=4,site=cs.usfca.edu
+$boost=4,site=the-tls.co.uk
+$boost=4,site=erlang.org
+$boost=4,site=home.cern
+$boost=4,site=joelonsoftware.com
+$boost=4,site=clojure.org
+$boost=4,site=pi-hole.net
+$boost=4,site=projecteuler.net
+$boost=4,site=coreos.com
+$boost=4,site=slack.engineering
+$boost=4,site=jenkins.io
+$boost=4,site=libcello.org
+$boost=4,site=vivaldi.com
+$boost=4,site=blog.acolyer.org
+$boost=4,site=devever.net
+$boost=4,site=judiciary.house.gov
+$boost=4,site=neocities.org
+$boost=4,site=vanityfair.com
+$boost=4,site=fosdem.org
+$boost=4,site=lists.llvm.org
+$boost=4,site=forbrukerradet.no
+$boost=4,site=alexkras.com
+$boost=4,site=yehudakatz.com
+$boost=4,site=politico.eu
+$boost=4,site=fieggen.com
+$boost=4,site=fermatslibrary.com
+$boost=4,site=llvm.org
+$boost=4,site=codeofhonor.com
+$boost=4,site=sandimetz.com
+$boost=4,site=beautifulracket.com
+$boost=4,site=solar.lowtechmagazine.com
+$boost=4,site=rollingstone.com
+$boost=4,site=theglobeandmail.com
+$boost=4,site=ucsf.edu
+$boost=4,site=games.greggman.com
+$boost=4,site=robrhinehart.com
+$boost=4,site=evertpot.com
+$boost=4,site=ryancarson.com
+$boost=4,site=oculusvr.com
+$boost=4,site=reasonablypolymorphic.com
+$boost=4,site=nybooks.com
+$boost=4,site=blog.malwarebytes.com
+$boost=4,site=sourcehut.org
+$boost=4,site=antonz.org
+$boost=4,site=weavesilk.com
+$boost=4,site=githubstatus.com
+$boost=4,site=thenextweb.com
+$boost=4,site=squareup.com
+$boost=4,site=teamten.com
+$boost=4,site=minimaxir.com
+$boost=4,site=sparkfun.com
+$boost=4,site=seattletimes.com
+$boost=4,site=int10h.org
+$boost=4,site=cpbotha.net
+$boost=4,site=mosaicscience.com
+$boost=4,site=cjr.org
+$boost=4,site=lareviewofbooks.org
+$boost=4,site=theregister.co.uk
+$boost=4,site=phoronix.com
+$boost=4,site=doc.rust-lang.org
+$boost=4,site=fuchsia.googlesource.com
+$boost=4,site=tomanthony.co.uk
+$boost=4,site=blog.ethereum.org
+$boost=4,site=historytoday.com
+$boost=4,site=michaelburge.us
+$boost=4,site=hackerschool.com
+$boost=4,site=507movements.com
+$boost=4,site=backchannel.com
+$boost=4,site=openssl.org
+$boost=4,site=openwall.com
+$boost=4,site=roslyn.codeplex.com
+$boost=4,site=blog.jazzychad.net
+$boost=4,site=atrium.co
+$boost=4,site=cacm.acm.org
+$boost=4,site=firstround.com
+$boost=4,site=blog.erratasec.com
+$boost=4,site=blog.paralleluniverse.co
+$boost=4,site=boz.com
+$boost=4,site=dmitry.gr
+$boost=4,site=linuxjournal.com
+$boost=4,site=bottomupcs.com
+$boost=4,site=commoncog.com
+$boost=4,site=drmaciver.com
+$boost=4,site=lawsofux.com
+$boost=4,site=blog.wesleyac.com
+$boost=4,site=sanctum.geek.nz
+$boost=4,site=rarehistoricalphotos.com
+$boost=4,site=stephaniehurlburt.com
+$boost=4,site=ssh.com
+$boost=4,site=sahillavingia.com
+$boost=4,site=bleepingcomputer.com
+$boost=4,site=thejh.net
+$boost=4,site=codingnest.com
+$boost=4,site=motorauthority.com
+$boost=4,site=kevinlynagh.com
+$boost=4,site=home-assistant.io
+$boost=4,site=spritesmods.com
+$boost=4,site=esa.int
+$boost=4,site=obsproject.com
+$boost=4,site=1843magazine.com
+$boost=4,site=therealkatie.net
+$boost=4,site=christine.website
+$boost=4,site=aadrake.com
+$boost=4,site=zapier.com
+$boost=4,site=diegobasch.com
+$boost=4,site=lichess.org
+$boost=4,site=collectorsweekly.com
+$boost=4,site=kev.inburke.com
+$boost=4,site=matt.sh
+$boost=4,site=turkeyblocks.org
+$boost=4,site=vitalik.ca
+$boost=4,site=themarshallproject.org
+$boost=4,site=usenix.org
+$boost=4,site=blog.docker.io
+$boost=4,site=kristoff.it
+$boost=4,site=radio.garden
+$boost=4,site=tholman.com
+$boost=4,site=gregoryszorc.com
+$boost=4,site=phys.org
+$boost=4,site=prisma.io
+$boost=4,site=aprilzero.com
+$boost=4,site=blog.parse.com
+$boost=4,site=borncity.com
+$boost=4,site=grumpygamer.com
+$boost=4,site=rsms.me
+$boost=4,site=bitquabit.com
+$boost=4,site=iamnotaprogrammer.com
+$boost=4,site=chrishateswriting.com
+$boost=4,site=saagarjha.com
+$boost=4,site=supabase.io
+$boost=4,site=me.veekun.com
+$boost=4,site=250bpm.com
+$boost=4,site=blog.gregbrockman.com
+$boost=4,site=mono-project.com
+$boost=4,site=minimallyminimal.com
+$boost=4,site=9to5mac.com
+$boost=4,site=char.gd
+$boost=4,site=plumshell.com
+$boost=4,site=nixos.org
+$boost=4,site=underjord.io
+$boost=4,site=landoflisp.com
+$boost=4,site=tnhh.net
+$boost=4,site=rapgenius.com
+$boost=4,site=blog.floydhub.com
+$boost=4,site=leafletjs.com
+$boost=4,site=kqed.org
+$boost=4,site=timharford.com
+$boost=4,site=blog.innerht.ml
+$boost=4,site=blog.mecheye.net
+$boost=4,site=about.sourcegraph.com
+$boost=4,site=wi-fi.org
+$boost=4,site=4chan.org
+$boost=4,site=thedailybeast.com
+$boost=4,site=ipp.mpg.de
+$boost=4,site=arstechnica.co.uk
+$boost=4,site=news.railstutorial.org
+$boost=4,site=inputmag.com
+$boost=4,site=labs.detectify.com
+$boost=4,site=undeadly.org
+$boost=4,site=tech.channable.com
+$boost=4,site=falkvinge.net
+$boost=4,site=omniref.com
+$boost=4,site=notes.unwieldy.net
+$boost=4,site=blog.newsblur.com
+$boost=4,site=mattwarren.org
+$boost=4,site=reason.com
+$boost=4,site=ramiro.org
+$boost=4,site=nsa.gov
+$boost=4,site=japantimes.co.jp
+$boost=4,site=twitlonger.com
+$boost=4,site=humblebundle.com
+$boost=4,site=publicdomainreview.org
+$boost=4,site=blog.nuclearsecrecy.com
+$boost=4,site=qotoqot.com
+$boost=4,site=pytorch.org
+$boost=4,site=muckrock.com
+$boost=4,site=dropbox.tech
+$boost=4,site=jasondavies.com
+$boost=4,site=kite.com
+$boost=4,site=indiehackers.com
+$boost=4,site=blog.detectify.com
+$boost=4,site=sixcolors.com
+$boost=4,site=blog.qt.io
+$boost=4,site=quirksmode.org
+$boost=4,site=community.namecheap.com
+$boost=4,site=thecodist.com
+$boost=4,site=hongkongfp.com
+$boost=4,site=stephenfry.com
+$boost=4,site=coffeescript.org
+$boost=4,site=major.io
+$boost=4,site=typesetinthefuture.com
+$boost=4,site=nayuki.io
+$boost=4,site=hscott.net
+$boost=4,site=pyvideo.org
+$boost=4,site=ai.google
+$boost=4,site=michael.stapelberg.ch
+$boost=4,site=v8.dev
+$boost=4,site=digiday.com
+$boost=4,site=thewalrus.ca
+$boost=4,site=ignorethecode.net
+$boost=4,site=thread.gmane.org
+$boost=4,site=sciencealert.com
+$boost=4,site=slackhq.com
+$boost=4,site=blog.datawrapper.de
+$boost=4,site=seriouspony.com
+$boost=4,site=musicforprogramming.net
+$boost=4,site=meowni.ca
+$boost=4,site=redox-os.org
+$boost=4,site=minnie.tuhs.org
+$boost=4,site=warpspire.com
+$boost=4,site=locusmag.com
+$boost=4,site=bloombergview.com
+$boost=4,site=damninteresting.com
+$boost=4,site=joeyh.name
+$boost=4,site=newstatesman.com
+$boost=4,site=freebsd.org
+$boost=4,site=andregarzia.com
+$boost=4,site=caddyserver.com
+$boost=4,site=kukuruku.co
+$boost=4,site=jbkempf.com
+$boost=4,site=pharo.org
+$boost=4,site=netblocks.org
+$boost=4,site=chromium.googlesource.com
+$boost=4,site=phrack.org
+$boost=4,site=codelitt.com
+$boost=4,site=baekdal.com
+$boost=4,site=ianvisits.co.uk
+$boost=4,site=megaprocessor.com
+$boost=4,site=mail-archive.com
+$boost=4,site=privateinternetaccess.com
+$boost=4,site=meltingasphalt.com
+$boost=4,site=juliacomputing.com
+$boost=4,site=rein.pk
+$boost=4,site=osnews.com
+$boost=4,site=ourworldindata.org
+$boost=4,site=blog.joda.org
+$boost=4,site=blog.stephsmith.io
+$boost=4,site=linusakesson.net
+$boost=4,site=learn.shayhowe.com
+$boost=4,site=chrisfenton.com
+$boost=4,site=alexdanco.com
+$boost=4,site=bert.org
+$boost=4,site=buyforlifeproducts.com
+$boost=4,site=adventofcode.com
+$boost=4,site=wordsandbuttons.online
+$boost=4,site=blog.trailofbits.com
+$boost=4,site=saminiir.com
+$boost=4,site=buttondown.email
+$boost=4,site=polygon.com
+$boost=4,site=ryanwaggoner.com
+$boost=4,site=media.defense.gov
+$boost=4,site=wired.co.uk
+$boost=4,site=blog.apnic.net
+$boost=4,site=blog.agilebits.com
+$boost=4,site=ar.al
+$boost=4,site=tomdale.net
+$boost=4,site=doublepulsar.com
+$boost=4,site=keyvalues.com
+$boost=4,site=android-developers.googleblog.com
+$boost=4,site=startupschool.org
+$boost=4,site=android.googlesource.com
+$boost=4,site=thegradient.pub
+$boost=4,site=orgmode.org
+$boost=4,site=wsws.org
+$boost=4,site=citizenlab.org
+$boost=4,site=blog.notdot.net
+$boost=4,site=eduardo.intermeta.com.br
+$boost=4,site=joreteg.com
+$boost=4,site=hopkinsmedicine.org
+$boost=4,site=stopa.io
+$boost=4,site=rigsomelight.com
+$boost=4,site=docs.djangoproject.com
+$boost=4,site=americanaffairsjournal.org
+$boost=4,site=jsfiddle.net
+$boost=4,site=metalevel.at
+$boost=4,site=andreasgal.com
+$boost=4,site=c9x.me
+$boost=4,site=danieltenner.com
+$boost=4,site=danielcompton.net
+$boost=4,site=netflixtechblog.com
+$boost=4,site=vidarholen.net
+$boost=4,site=eli.thegreenplace.net
+$boost=4,site=cybernews.com
+$boost=4,site=blogs.smithsonianmag.com
+$boost=4,site=rambo.codes
+$boost=4,site=hackermonthly.com
+$boost=4,site=sifive.com
+$boost=4,site=plus.maths.org
+$boost=4,site=incoherency.co.uk
+$boost=4,site=circleci.com
+$boost=4,site=catern.com
+$boost=4,site=penny-arcade.com
+$boost=4,site=howistart.org
+$boost=4,site=middleeasteye.net
+$boost=4,site=unlikekinds.com
+$boost=4,site=videogameschronicle.com
+$boost=4,site=roadandtrack.com
+$boost=4,site=hyperallergic.com
+$boost=4,site=blog.level3.com
+$boost=4,site=os2museum.com
+$boost=4,site=soranews24.com
+$boost=4,site=patshaughnessy.net
+$boost=4,site=bill.harding.blog
+$boost=4,site=blog.plan99.net
+$boost=4,site=gq.com
+$boost=4,site=occrp.org
+$boost=4,site=devonzuegel.com
+$boost=4,site=ben-evans.com
+$boost=4,site=mediumdifficulty.com
+$boost=4,site=fsf.org
+$boost=4,site=learn.sparkfun.com
+$boost=4,site=dlang.org
+$boost=4,site=templeos.org
+$boost=4,site=code.blender.org
+$boost=4,site=singhkays.com
+$boost=4,site=repubblica.it
+$boost=4,site=aerofs.com
+$boost=4,site=lrb.co.uk
+$boost=4,site=eng.lyft.com
+$boost=4,site=jakearchibald.com
+$boost=4,site=blog.youtube
+$boost=4,site=blog.easydns.org
+$boost=4,site=xiph.org
+$boost=4,site=newscientist.com
+$boost=4,site=nicolasbize.com
+$boost=4,site=lineageos.org
+$boost=4,site=briancallahan.net
+$boost=4,site=blog.ltse.com
+$boost=4,site=bryanbraun.com
+$boost=4,site=blog.thea.codes
+$boost=4,site=code.mixpanel.com
+$boost=4,site=web.law.duke.edu
+$boost=4,site=technologyreview.com
+$boost=4,site=news.ncsu.edu
+$boost=4,site=acsh.org
+$boost=4,site=repl.it
+$boost=4,site=roscidus.com
+$boost=4,site=blog.ghost.org
+$boost=4,site=icann.org
+$boost=4,site=digest.bps.org.uk
+$boost=4,site=nj.com
+$boost=4,site=90percentofeverything.com
+$boost=4,site=lars.ingebrigtsen.no
+$boost=4,site=harmful.cat-v.org
+$boost=4,site=ecma-international.org
+$boost=4,site=engineering.riotgames.com
+$boost=4,site=lexaloffle.com
+$boost=4,site=en.greatfire.org
+$boost=4,site=karl-voit.at
+$boost=4,site=fair.org
+$boost=4,site=askamathematician.com
+$boost=4,site=unixism.net
+$boost=4,site=rockpapershotgun.com
+$boost=4,site=realdanlyons.com
+$boost=4,site=play0ad.com
+$boost=4,site=tedium.co
+$boost=4,site=bitcoin.org
+$boost=4,site=loup-vaillant.fr
+$boost=4,site=getpolarized.io
+$boost=4,site=thewire.in
+$boost=4,site=melmagazine.com
+$boost=4,site=sambleckley.com
+$boost=4,site=david.heinemeierhansson.com
+$boost=4,site=secondbreakfast.co
+$boost=4,site=asahi.com
+$boost=4,site=nintendo.co.jp
+$boost=4,site=annoying.technology
+$boost=4,site=kiro7.com
+$boost=4,site=markshuttleworth.com
+$boost=4,site=anders.unix.se
+$boost=4,site=blog.bitbucket.org
+$boost=4,site=blog.kubernetes.io
+$boost=4,site=devdocs.io
+$boost=4,site=justinjackson.ca
+$boost=4,site=corte.si
+$boost=4,site=kotaku.com
+$boost=4,site=cdn.ca9.uscourts.gov
+$boost=4,site=forum.dlang.org
+$boost=4,site=spacecraft.ssl.umd.edu
+$boost=4,site=neuralnetworksanddeeplearning.com
+$boost=4,site=politico.com
+$boost=4,site=notamonadtutorial.com
+$boost=4,site=duck.co
+$boost=4,site=lg.io
+$boost=4,site=vorpus.org
+$boost=4,site=objective-see.com
+$boost=4,site=erlang-solutions.com
+$boost=4,site=alexgaynor.net
+$boost=4,site=sub.garrytan.com
+$boost=4,site=xkcd.com
+$boost=4,site=blog.codinghorror.com
+$boost=4,site=wingolog.org
+$boost=4,site=pipelinedb.com
+$boost=4,site=blog.nelhage.com
+$boost=4,site=tonyarcieri.com
+$boost=4,site=geoffreylitt.com
+$boost=4,site=viniciusvacanti.com
+$boost=4,site=magicleap.com
+$boost=4,site=igvita.com
+$boost=4,site=chrisdone.com
+$boost=4,site=stackshare.io
+$boost=4,site=latacora.singles
+$boost=4,site=infoq.com
+$boost=4,site=papersplease.org
+$boost=4,site=thepiratebay.org
+$boost=4,site=calnewport.com
+$boost=4,site=krausefx.com
+$boost=4,site=blog.circleci.com
+$boost=4,site=blog.cognitect.com
+$boost=4,site=masswerk.at
+$boost=4,site=0x65.dev
+$boost=4,site=dangerousminds.net
+$boost=4,site=businessweek.com
+$boost=4,site=meta.slashdot.org
+$boost=4,site=john-millikin.com
+$boost=4,site=raptitude.com
+$boost=4,site=mersenne.org
+$boost=4,site=blog.reverberate.org
+$boost=4,site=alexisohanian.com
+$boost=4,site=electrek.co
+$boost=4,site=jeanhsu.com
+$boost=4,site=svelte.technology
+$boost=4,site=discuss.ocaml.org
+$boost=4,site=nedbatchelder.com
+$boost=4,site=haskellforall.com
+$boost=4,site=vole.wtf
+$boost=4,site=yurichev.com
+$boost=4,site=joel.is
+$boost=4,site=yager.io
+$boost=4,site=boilingsteam.com
+$boost=4,site=bigmessowires.com
+$boost=4,site=fail0verflow.com
+$boost=4,site=beagleboard.org
+$boost=4,site=floridatoday.com
+$boost=4,site=popehat.com
+$boost=4,site=blog.tutorspree.com
+$boost=4,site=daeken.com
+$boost=4,site=techstartups.com
+$boost=4,site=wikitribune.com
+$boost=4,site=sealedabstract.com
+$boost=4,site=0x46.net
+$boost=4,site=longreads.com
+$boost=4,site=dispatch.com
+$boost=4,site=lists.centos.org
+$boost=4,site=adgrok.com
+$boost=4,site=mcfunley.com
+$boost=4,site=beginners.re
+$boost=4,site=joshworth.com
+$boost=4,site=calv.info
+$boost=4,site=jdk.java.net
+$boost=4,site=keepassxc.org
+$boost=4,site=thepiratebay.se
+$boost=4,site=metatalk.metafilter.com
+$boost=4,site=gregable.com
+$boost=4,site=newrepublic.com
+$boost=4,site=cryptome.org
+$boost=4,site=smashcompany.com
+$boost=4,site=mixergy.com
+$boost=4,site=blog.bolt.io
+$boost=4,site=lesswrong.com
+$boost=4,site=recode.net
+$boost=4,site=baremetrics.com
+$boost=4,site=idea.popcount.org
+$boost=4,site=blog.nodejs.org
+$boost=4,site=dustincurtis.com
+$boost=4,site=capnproto.org
+$boost=4,site=samy.pl
+$boost=4,site=fivethirtyeight.com
+$boost=4,site=babeljs.io
+$boost=4,site=venturebeat.com
+$boost=4,site=bartoszmilewski.com
+$boost=4,site=interactivepython.org
+$boost=4,site=feross.org
+$boost=4,site=nateberkopec.com
+$boost=4,site=bnnbloomberg.ca
+$boost=4,site=sourcegraph.com
+$boost=4,site=dampfkraft.com
+$boost=4,site=sockpuppet.org
+$boost=4,site=vuejs.org
+$boost=4,site=tomshardware.com
+$boost=4,site=go.googlesource.com
+$boost=4,site=every.to
+$boost=4,site=community.signalusers.org
+$boost=4,site=features.propublica.org
+$boost=4,site=trigger.io
+$boost=4,site=blog.wolfram.com
+$boost=4,site=globalcitizen.org
+$boost=4,site=brennan.io
+$boost=4,site=typescriptlang.org
+$boost=4,site=blog.hboeck.de
+$boost=4,site=blog.siftscience.com
+$boost=4,site=blog.jetbrains.com
+$boost=4,site=libcom.org
+$boost=4,site=pudding.cool
+$boost=4,site=geek.com
+$boost=4,site=zedshaw.com
+$boost=4,site=techspot.com
+$boost=4,site=popularmechanics.com
+$boost=4,site=acesounderglass.com
+$boost=4,site=stitcher.io
+$boost=4,site=texasmonthly.com
+$boost=4,site=mattmaroon.com
+$boost=4,site=gabrielgambetta.com
+$boost=4,site=tratt.net
+$boost=4,site=blog.posterous.com
+$boost=4,site=oxide.computer
+$boost=4,site=trackchanges.postlight.com
+$boost=4,site=codahale.com
+$boost=4,site=emberjs.com
+$boost=4,site=applehelpwriter.com
+$boost=4,site=open.buffer.com
+$boost=4,site=expatsoftware.com
+$boost=4,site=edgedb.com
+$boost=4,site=etherpad.com
+$boost=4,site=firebase.com
+$boost=4,site=blog.thestartuptoolkit.com
+$boost=4,site=flynn.io
+$boost=4,site=eevblog.com
+$boost=4,site=balena.io
+$boost=4,site=bell-labs.com
+$boost=4,site=persuasion.community
+$boost=4,site=eugeneyan.com
+$boost=4,site=80000hours.org
+$boost=4,site=zerotier.com
+$boost=4,site=httptoolkit.tech
+$boost=4,site=ekioh.com
+$boost=4,site=cia.gov
+$boost=4,site=nickcraver.com
+$boost=4,site=simonsfoundation.org
+$boost=4,site=artsy.net
+$boost=4,site=tryretool.com
+$boost=4,site=bloombergquint.com
+$boost=4,site=vas3k.com
+$boost=4,site=100r.co
+$boost=4,site=motherjones.com
+$boost=4,site=nasdaq.com
+$boost=4,site=bitcointalk.org
+$boost=4,site=hackershelf.com
+$boost=4,site=slashcrypto.org
+$boost=4,site=tosdr.org
+$boost=4,site=fiercebiotech.com
+$boost=4,site=haskell.org
+$boost=4,site=tribunemag.co.uk
+$boost=4,site=alexeymk.com
+$boost=4,site=cs.vu.nl
+$boost=4,site=caltech.edu
+$boost=4,site=azarask.in
+$boost=4,site=gigaom.com
+$boost=4,site=virtualbox.org
+$boost=4,site=increment.com
+$boost=4,site=blog.janestreet.com
+$boost=4,site=textslashplain.com
+$boost=4,site=planetary.org
+$boost=4,site=allthingsd.com
+$boost=4,site=clojure.com
+$boost=4,site=git.postgresql.org
+$boost=4,site=howtogeek.com
+$boost=4,site=blog.kapeli.com
+$boost=4,site=quickdraw.withgoogle.com
+$boost=4,site=discussions.agilebits.com
+$boost=4,site=fauna.com
+$boost=4,site=blog.sandstorm.io
+$boost=4,site=openvim.com
+$boost=4,site=waymo.com
+$boost=4,site=blog.eladgil.com
+$boost=4,site=projekte.sueddeutsche.de
+$boost=4,site=whytheluckystiff.net
+$boost=4,site=the-paper-trail.org
+$boost=4,site=medium.dave-bailey.com
+$boost=4,site=kadavy.net
+$boost=4,site=symbo1ics.com
+$boost=4,site=erikrood.com
+$boost=4,site=gnome.org
+$boost=4,site=evanjones.ca
+$boost=4,site=johnsalvatier.org
+$boost=4,site=notes.eatonphil.com
+$boost=4,site=cityobservatory.org
+$boost=4,site=autocar.co.uk
+$boost=4,site=blog.alinelerner.com
+$boost=4,site=a.singlediv.com
+$boost=4,site=city-journal.org
+$boost=4,site=blog.bugsnag.com
+$boost=4,site=blog.steveklabnik.com
+$boost=4,site=jgc.org
+$boost=4,site=maraoz.com
+$boost=4,site=somethingsimilar.com
+$boost=4,site=macro.ycombinator.com
+$boost=4,site=damienkatz.net
+$boost=4,site=moserware.com
+$boost=4,site=elastic.co
+$boost=4,site=tenthousandmeters.com
+$boost=4,site=fmwconcepts.com
+$boost=4,site=recraigslist.com
+$boost=4,site=codingjohnson.com
+$boost=4,site=aftenposten.no
+$boost=4,site=blog.1password.com
+$boost=4,site=neugierig.org
+$boost=4,site=01.org
+$boost=4,site=jalopnik.com
+$boost=4,site=mnot.net
+$boost=4,site=jmoiron.net
+$boost=4,site=bjoern.brembs.net
+$boost=4,site=yieldthought.com
+$boost=4,site=news.gallup.com
+$boost=4,site=plusbryan.com
+$boost=4,site=rameznaam.com
+$boost=4,site=ij.org
+$boost=4,site=hackingdistributed.com
+$boost=4,site=aosabook.org
+$boost=4,site=fcc.gov
+$boost=4,site=multithreaded.stitchfix.com
+$boost=4,site=ybrikman.com
+$boost=4,site=neh.gov
+$boost=4,site=abc.xyz
+$boost=4,site=thecorrespondent.com
+$boost=4,site=techempower.com
+$boost=4,site=letsencrypt.status.io
+$boost=4,site=syncthing.net
+$boost=4,site=torontolife.com
+$boost=4,site=empiricalzeal.com
+$boost=4,site=brown.edu
+$boost=4,site=dustri.org
+$boost=4,site=restofworld.org
+$boost=4,site=mikeash.com
+$boost=4,site=emacsair.me
+$boost=4,site=fossil-scm.org
+$boost=4,site=rauchg.com
+$boost=4,site=oregonlive.com
+$boost=4,site=madebyevan.com
+$boost=4,site=bitbucket.org
+$boost=4,site=justinvincent.com
+$boost=4,site=shape-of-code.coding-guidelines.com
+$boost=4,site=hodinkee.com
+$boost=4,site=blog.getprismatic.com
+$boost=4,site=jeremyaboyd.com
+$boost=4,site=pcgamer.com
+$boost=4,site=tutanota.com
+$boost=4,site=review.chicagobooth.edu
+$boost=4,site=programmingisterrible.com
+$boost=4,site=zainamro.com
+$boost=4,site=waxy.org
+$boost=4,site=hisham.hm
+$boost=4,site=foundersatwork.com
+$boost=4,site=neil.computer
+$boost=4,site=youmightnotneedjquery.com
+$boost=4,site=sparktoro.com
+$boost=4,site=supporters.eff.org
+$boost=4,site=django-rest-framework.org
+$boost=4,site=fuse.wikichip.org
+$boost=4,site=alexanderell.is
+$boost=4,site=jetbrains.com
+$boost=4,site=rootsofprogress.org
+$boost=4,site=blog.jonlu.ca
+$boost=4,site=grafana.com
+$boost=4,site=joinpeertube.org
+$boost=4,site=trac.torproject.org
+$boost=4,site=kapwing.com
+$boost=4,site=jgthms.com
+$boost=4,site=lighterra.com
+$boost=4,site=bzarg.com
+$boost=4,site=itep.org
+$boost=4,site=dood.al
+$boost=4,site=seriot.ch
+$boost=4,site=go-review.googlesource.com
+$boost=4,site=benjamincongdon.me
+$boost=4,site=coderwall.com
+$boost=4,site=betabeat.com
+$boost=4,site=cryto.net
+$boost=4,site=blogs.bl.uk
+$boost=4,site=b-list.org
+$boost=4,site=mailpile.is
+$boost=4,site=chiark.greenend.org.uk
+$boost=4,site=arcfn.com
+$boost=4,site=urbit.org
+$boost=4,site=eprint.iacr.org
+$boost=4,site=dl.acm.org
+$boost=4,site=blog.sessionstack.com
+$boost=4,site=breck-mckye.com
+$boost=4,site=prospect.org
+$boost=4,site=banksimple.com
+$boost=4,site=practicaltypography.com
+$boost=4,site=blogmaverick.com
+$boost=4,site=instagram-engineering.com
+$boost=4,site=theartofmachinery.com
+$boost=4,site=cpushack.com
+$boost=4,site=guzey.com
+$boost=4,site=morrick.me
+$boost=4,site=blog.emojipedia.org
+$boost=4,site=arjunsreedharan.org
+$boost=4,site=superuser.com
+$boost=4,site=blog.fosketts.net
+$boost=4,site=darkpatterns.org
+$boost=4,site=ulisp.com
+$boost=4,site=herbsutter.com
+$boost=4,site=darkcoding.net
+$boost=4,site=latacora.micro.blog
+$boost=4,site=drdobbs.com
+$boost=4,site=hollywoodreporter.com
+$boost=4,site=thestartupfoundry.com
+$boost=4,site=blog.fastmail.fm
+$boost=4,site=wit.ai
+$boost=4,site=devttys0.com
+$boost=4,site=qubes-os.org
+$boost=4,site=robinsloan.com
+$boost=4,site=timetobleed.com
+$boost=4,site=blog.jitbit.com
+$boost=4,site=mrale.ph
+$boost=4,site=play.freeciv.org
+$boost=4,site=blackhat.com
+$boost=4,site=laforge.gnumonks.org
+$boost=4,site=sfchronicle.com
+$boost=4,site=tutorial.math.lamar.edu
+$boost=4,site=holovaty.com
+$boost=4,site=cringely.com
+$boost=4,site=folklore.org
+$boost=4,site=linuxvoice.com
+$boost=4,site=papers.ssrn.com
+$boost=4,site=brianschrader.com
+$boost=4,site=kreci.net
+$boost=4,site=nngroup.com
+$boost=4,site=jasonshen.com
+$boost=4,site=lists.freedesktop.org
+$boost=4,site=nodejs.org
+$boost=4,site=graalvm.org
+$boost=4,site=dailydot.com
+$boost=4,site=lbrandy.com
+$boost=4,site=chemistryworld.com
+$boost=4,site=solipsys.co.uk
+$boost=4,site=layerci.com
+$boost=4,site=stellar.org
+$boost=4,site=etherscan.io
+$boost=4,site=macrumors.com
+$boost=4,site=vim-adventures.com
+$boost=4,site=macleans.ca
+$boost=4,site=blog.appcanary.com
+$boost=4,site=briancarper.net
+$boost=4,site=joehewitt.com
+$boost=4,site=sfconservancy.org
+$boost=4,site=mariovittone.com
+$boost=4,site=500hats.com
+$boost=4,site=blog.tailwindcss.com
+$boost=4,site=dilbert.com
+$boost=4,site=mewo2.com
+$boost=4,site=blog.getpebble.com
+$boost=4,site=preshing.com
+$boost=4,site=medievalists.net
+$boost=4,site=lynalden.com
+$boost=4,site=hindenburgresearch.com
+$boost=4,site=joindiaspora.com
+$boost=4,site=chadfowler.com
+$boost=4,site=git.sr.ht
+$boost=4,site=science.org
+$boost=4,site=esquire.com
+$boost=4,site=augustl.com
+$boost=4,site=paulrouget.com
+$boost=4,site=joelgrus.com
+$boost=4,site=electronjs.org
+$boost=4,site=fastmail.blog
+$boost=4,site=trixter.oldskool.org
+$boost=4,site=palladiummag.com
+$boost=4,site=dwmkerr.com
+$boost=4,site=blueorigin.com
+$boost=4,site=blogs.perl.org
+$boost=4,site=peter.bourgon.org
+$boost=4,site=boats.gitlab.io
+$boost=4,site=news.usc.edu
+$boost=4,site=media.uzh.ch
+$boost=4,site=rubini.us
+$boost=4,site=ognjen.io
+$boost=4,site=dragonflybsd.org
+$boost=4,site=vagrantup.com
+$boost=4,site=reclaimthenet.org
+$boost=4,site=blogs.lse.ac.uk
+$boost=4,site=threader.app
+$boost=4,site=pgdash.io
+$boost=4,site=namecheap.com
+$boost=4,site=mtgox.com
+$boost=4,site=blog.filepicker.io
+$boost=4,site=mullvad.net
+$boost=4,site=m.signalvnoise.com
+$boost=4,site=gamerant.com
+$boost=4,site=computerhistory.org
+$boost=4,site=bennettnotes.com
+$boost=4,site=amnesty.org
+$boost=4,site=quip.com
+$boost=4,site=balsamiq.com
+$boost=4,site=conwaylife.com
+$boost=4,site=quoteinvestigator.com
+$boost=4,site=stephanboyer.com
+$boost=4,site=kkovacs.eu
+$boost=4,site=cs.virginia.edu
+$boost=4,site=blog.heapanalytics.com
+$boost=4,site=khronos.org
+$boost=4,site=infotrope.net
+$boost=4,site=r6.ca
+$boost=4,site=standardnotes.org
+$boost=4,site=thanassis.space
+$boost=4,site=essential.com
+$boost=4,site=firebase.googleblog.com
+$boost=4,site=theamericanscholar.org
+$boost=4,site=mxb.dev
+$boost=4,site=moderncrypto.org
+$boost=4,site=gresearch.co.uk
+$boost=4,site=news.rapgenius.com
+$boost=4,site=duartes.org
+$boost=4,site=robinwieruch.de
+$boost=4,site=blog.viktomas.com
+$boost=4,site=thenation.com
+$boost=4,site=evilmartians.com
+$boost=4,site=unixsheikh.com
+$boost=4,site=lanyrd.com
+$boost=4,site=slowboring.com
+$boost=4,site=seldo.com
+$boost=4,site=elischiff.com
+$boost=4,site=arrl.org
+$boost=4,site=spacetelescope.org
+$boost=4,site=writing.markchristian.org
+$boost=4,site=elaineou.com
+$boost=4,site=blog.lastpass.com
+$boost=4,site=netzpolitik.org
+$boost=4,site=ghuntley.com
+$boost=4,site=knightcolumbia.org
+$boost=4,site=comparitech.com
+$boost=4,site=scanlime.org
+$boost=4,site=eklitzke.org
+$boost=4,site=scala-lang.org
+$boost=4,site=obscuritory.com
+$boost=4,site=habitatchronicles.com
+$boost=4,site=giftrocket.com
+$boost=4,site=debugbear.com
+$boost=4,site=makezine.com
+$boost=4,site=timkadlec.com
+$boost=4,site=momentjs.com
+$boost=4,site=euractiv.com
+$boost=4,site=daniellemorrill.com
+$boost=4,site=writings.stephenwolfram.com
+$boost=4,site=gkogan.co
+$boost=4,site=needwant.com
+$boost=4,site=medialens.org
+$boost=4,site=blog.wepay.com
+$boost=4,site=dustmop.io
+$boost=4,site=comp.nus.edu.sg
+$boost=4,site=remotebase.io
+$boost=4,site=smashingmagazine.com
+$boost=4,site=kylehalladay.com
+$boost=4,site=fosterelli.co
+$boost=4,site=braveclojure.com
+$boost=4,site=nomasters.io
+$boost=4,site=reasonstobecheerful.world
+$boost=4,site=x264dev.multimedia.cx
+$boost=4,site=androidpolice.com
+$boost=4,site=ribbonfarm.com
+$boost=4,site=jcs.org
+$boost=4,site=codusoperandi.com
+$boost=4,site=peternixey.com
+$boost=4,site=theplayerstribune.com
+$boost=4,site=basho.com
+$boost=4,site=curiousgnu.com
+$boost=4,site=gafferongames.com
+$boost=4,site=idontplaydarts.com
+$boost=4,site=forrestthewoods.com
+$boost=4,site=earthquake.usgs.gov
+$boost=4,site=blog.bitpay.com
+$boost=4,site=brookings.edu
+$boost=4,site=18f.gsa.gov
+$boost=4,site=minecraftforum.net
+$boost=4,site=lyft.com
+$boost=4,site=hhvm.com
+$boost=4,site=blog.wolfire.com
+$boost=4,site=nbviewer.ipython.org
+$boost=4,site=eenewseurope.com
+$boost=4,site=smh.com.au
+$boost=4,site=yubico.com
+$boost=4,site=larrysanger.org
+$boost=4,site=qualys.com
+$boost=4,site=charlieharrington.com
+$boost=4,site=casual-effects.com
+$boost=4,site=blog.macromates.com
+$boost=4,site=beza1e1.tuxen.de
+$boost=4,site=dragan.rocks
+$boost=4,site=kevq.uk
+$boost=4,site=blog.ploeh.dk
+$boost=4,site=spoon-tamago.com
+$boost=4,site=maizure.org
+$boost=4,site=californiaglobe.com
+$boost=4,site=yarchive.net
+$boost=4,site=gail.com
+$boost=4,site=snikt.net
+$boost=4,site=blog.plaid.com
+$boost=4,site=ebb.org
+$boost=4,site=insomnia.rest
+$boost=4,site=biorxiv.org
+$boost=4,site=cmdrtaco.net
+$boost=4,site=tom7.org
+$boost=4,site=taler.net
+$boost=4,site=docs.rs
+$boost=4,site=ioccc.org
+$boost=4,site=domainnamewire.com
+$boost=4,site=dafoster.net
+$boost=4,site=boringcompany.com
+$boost=4,site=bankofengland.co.uk
+$boost=4,site=blog.ntpsec.org
+$boost=4,site=huyenchip.com
+$boost=4,site=pganalyze.com
+$boost=4,site=bl.ocks.org
+$boost=4,site=codeamigo.dev
+$boost=4,site=blog.x.company
+$boost=4,site=inews.co.uk
+$boost=4,site=talyarkoni.org
+$boost=4,site=pijul.org
+$boost=4,site=thegatesnotes.com
+$boost=4,site=bartongeorge.io
+$boost=4,site=arslan.io
+$boost=4,site=bytecodealliance.org
+$boost=4,site=journal.paul.querna.org
+$boost=4,site=jasonlbaptiste.com
+$boost=4,site=sharelatex.com
+$boost=4,site=jessesquires.com
+$boost=4,site=iquilezles.org
+$boost=4,site=yeokhengmeng.com
+$boost=4,site=software.rajivprab.com
+$boost=4,site=fbi.gov
+$boost=4,site=sirupsen.com
+$boost=4,site=studyfinds.org
+$boost=4,site=rdegges.com
+$boost=4,site=cemerick.com
+$boost=4,site=cr.openjdk.java.net
+$boost=4,site=mapbox.com
+$boost=4,site=lihaoyi.com
+$boost=4,site=ams.org
+$boost=4,site=oilshell.org
+$boost=4,site=darktable.org
+$boost=4,site=madinamerica.com
+$boost=4,site=blog.scrapinghub.com
+$boost=4,site=sheshbabu.com
+$boost=4,site=diveintomark.org
+$boost=4,site=modernfarmer.com
+$boost=4,site=news.ucsc.edu
+$boost=4,site=rtl-sdr.com
+$boost=4,site=byuu.org
+$boost=4,site=bugzilla.gnome.org
+$boost=4,site=sfwa.org
+$boost=4,site=spinellis.gr
+$boost=4,site=0pointer.net
+$boost=4,site=learnxinyminutes.com
+$boost=4,site=9to5google.com
+$boost=4,site=fourmilab.ch
+$boost=4,site=hynek.me
+$boost=4,site=sarahmei.com
+$boost=4,site=blog.piekniewski.info
+$boost=4,site=cb.vu
+$boost=4,site=news.bloomberglaw.com
+$boost=4,site=blog.darknedgy.net
+$boost=4,site=telegra.ph
+$boost=4,site=lists.swift.org
+$boost=4,site=adit.io
+$boost=4,site=thurrott.com
+$boost=4,site=socket.io
+$boost=4,site=maryrosecook.com
+$boost=4,site=tomayko.com
+$boost=4,site=netmeister.org
+$boost=4,site=transition.fcc.gov
+$boost=4,site=playground.tensorflow.org
+$boost=4,site=macworld.com
+$boost=4,site=mahdiyusuf.com
+$boost=4,site=nplusonemag.com
+$boost=4,site=kunle.app
+$boost=4,site=cheatsheetseries.owasp.org
+$boost=4,site=blogs.janestreet.com
+$boost=4,site=chromium.org
+$boost=4,site=aht.seriouseats.com
+$boost=4,site=arcadeblogger.com
+$boost=4,site=lock.cmpxchg8b.com
+$boost=4,site=codewithoutrules.com
+$boost=4,site=blog.roboflow.ai
+$boost=4,site=jolla.com
+$boost=4,site=textfiles.com
+$boost=4,site=landshark.io
+$boost=4,site=syfy.com
+$boost=4,site=a1k0n.net
+$boost=4,site=thirtybees.com
+$boost=4,site=wp.josh.com
+$boost=4,site=community.rapid7.com
+$boost=4,site=variety.com
+$boost=4,site=petekeen.net
+$boost=4,site=pragprog.com
+$boost=4,site=blog.abevoelker.com
+$boost=4,site=uxdesign.cc
+$boost=4,site=tomblomfield.com
+$boost=4,site=platformer.news
+$boost=4,site=airbnb.design
+$boost=4,site=forum.bitcoin.org
+$boost=4,site=victorzhou.com
+$boost=4,site=newcriterion.com
+$boost=4,site=blog.sagemath.com
+$boost=4,site=secret.club
+$boost=4,site=codebutler.com
+$boost=4,site=lambda-the-ultimate.org
+$boost=4,site=universetoday.com
+$boost=4,site=vc.gg
+$boost=4,site=bostonglobe.com
+$boost=4,site=slides.com
+$boost=4,site=openid.net
+$boost=4,site=sizovs.net
+$boost=4,site=minimalmac.com
+$boost=4,site=pandodaily.com
+$boost=4,site=blogs.technet.com
+$boost=4,site=tynan.com
+$boost=4,site=antipolygraph.org
+$boost=4,site=dev.lemmy.ml
+$boost=4,site=spectator.co.uk
+$boost=4,site=cognitivemedium.com
+$boost=4,site=cyanogenmod.org
+$boost=4,site=codepen.io
+$boost=4,site=ianthehenry.com
+$boost=4,site=speakerdeck.com
+$boost=4,site=en.lichess.org
+$boost=4,site=datagenetics.com
+$boost=4,site=people.kernel.org
+$boost=4,site=ma.tt
+$boost=4,site=people.freebsd.org
+$boost=4,site=measuringmeasures.com
+$boost=4,site=catb.org
+$boost=4,site=defn.io
+$boost=4,site=fastly.com
+$boost=4,site=eetimes.com
+$boost=4,site=chicagomaroon.com
+$boost=4,site=blog.openbloc.fr
+$boost=4,site=blog.moertel.com
+$boost=4,site=nathanbarry.com
+$boost=4,site=bombich.com
+$boost=4,site=mojang.com
+$boost=4,site=pgbovine.net
+$boost=4,site=bicycling.com
+$boost=4,site=perell.com
+$boost=4,site=mjtsai.com
+$boost=4,site=softwarebyrob.com
+$boost=4,site=fusion.net
+$boost=4,site=beta.observablehq.com
+$boost=4,site=lists.gnupg.org
+$boost=4,site=gleam.run
+$boost=4,site=thestar.com
+$boost=4,site=militarytimes.com
+$boost=4,site=blog.memsql.com
+$boost=4,site=cr.yp.to
+$boost=4,site=thepointsguy.com
+$boost=4,site=dinosaurpictures.org
+$boost=4,site=quillette.com
+$boost=4,site=neil.fraser.name
+$boost=4,site=www3.nhk.or.jp
+$boost=4,site=muppetlabs.com
+$boost=4,site=blogs.gnome.org
+$boost=4,site=readonlymemory.vg
+$boost=4,site=news.osu.edu
+$boost=4,site=cgl.uwaterloo.ca
+$boost=4,site=sebastianmarshall.com
+$boost=4,site=physicstoday.scitation.org
+$boost=4,site=edge.org
+$boost=4,site=blog.bugreplay.com
+$boost=4,site=medrxiv.org
+$boost=4,site=permalink.gmane.org
+$boost=4,site=notechmagazine.com
+$boost=4,site=blog.ipfs.io
+$boost=4,site=ninjasandrobots.com
+$boost=4,site=blog.johnnovak.net
+$boost=4,site=tldr.sh
+$boost=4,site=internals.rust-lang.org
+$boost=4,site=formidable.com
+$boost=4,site=thenounproject.com
+$boost=4,site=blog.earbits.com
+$boost=4,site=gamehistory.org
+$boost=4,site=influxdata.com
+$boost=4,site=pythonspeed.com
+$boost=4,site=observer.com
+$boost=4,site=mastodon.social
+$boost=4,site=whatever.scalzi.com
+$boost=4,site=rietta.com
+$boost=4,site=cushionapp.com
+$boost=4,site=cantorsparadise.com
+$boost=4,site=intelligence.senate.gov
+$boost=4,site=blog.nullspace.io
+$boost=4,site=keleshev.com
+$boost=4,site=freethink.com
+$boost=4,site=singularityhub.com
+$boost=4,site=greenteapress.com
+$boost=4,site=hazlitt.net
+$boost=4,site=zainrizvi.io
+$boost=4,site=welcometothejungle.co
+$boost=4,site=joeduffyblog.com
+$boost=4,site=tbbuck.com
+$boost=4,site=blog.fawny.org
+$boost=4,site=cyclingtips.com
+$boost=4,site=f-secure.com
+$boost=4,site=tweag.io
+$boost=4,site=edri.org
+$boost=4,site=brooker.co.za
+$boost=4,site=learnyouahaskell.com
+$boost=4,site=dmitrysoshnikov.com
+$boost=4,site=frankgroeneveld.nl
+$boost=4,site=inconvergent.net
+$boost=4,site=editorconfig.org
+$boost=4,site=atom.io
+$boost=4,site=mattgemmell.com
+$boost=4,site=fastcoexist.com
+$boost=4,site=blog.koehntopp.info
+$boost=4,site=htmx.org
+$boost=4,site=markozivanovic.com
+$boost=4,site=number-none.com
+$boost=4,site=sequoia-pgp.org
+$boost=4,site=tanelpoder.com
+$boost=4,site=prospectmagazine.co.uk
+$boost=4,site=bouk.co
+$boost=4,site=courtlistener.com
+$boost=4,site=lav.io
+$boost=4,site=landave.io
+$boost=4,site=blog.phusion.nl
+$boost=4,site=nickjanetakis.com
+$boost=4,site=stateofjs.com
+$boost=4,site=hakim.se
+$boost=4,site=blog.datomic.com
+$boost=4,site=docs.racket-lang.org
+$boost=4,site=jamie-wong.com
+$boost=4,site=mocko.org.uk
+$boost=4,site=matasano.com
+$boost=4,site=okigiveup.net
+$boost=4,site=logicmag.io
+$boost=4,site=andrewchen.co
+$boost=4,site=airpair.com
+$boost=4,site=cowlark.com
+$boost=4,site=semianalysis.com
+$boost=4,site=kk.org
+$boost=4,site=log.nadim.cc
+$boost=4,site=collaborativefund.com
+$boost=4,site=ventusky.com
+$boost=4,site=nintendo.com
+$boost=4,site=worldofindie.co.uk
+$boost=4,site=jeroenjanssens.com
+$boost=4,site=libreboot.org
+$boost=4,site=blog.roboflow.com
+$boost=4,site=pagetable.com
+$boost=4,site=eejournal.com
+$boost=4,site=blog.sanctum.geek.nz
+$boost=4,site=blog.elcomsoft.com
+$boost=4,site=judegomila.com
+$boost=4,site=cirw.in
+$boost=4,site=corner.squareup.com
+$boost=4,site=flightfox.com
+$boost=4,site=tip.golang.org
+$boost=4,site=nzherald.co.nz
+$boost=4,site=wezm.net
+$boost=4,site=crontab.guru
+$boost=4,site=degoes.net
+$boost=4,site=investors.modernatx.com
+$boost=4,site=linustechtips.com
+$boost=4,site=blog.tanyakhovanova.com
+$boost=4,site=amandapeyton.com
+$boost=4,site=forumserver.twoplustwo.com
+$boost=4,site=belkadan.com
+$boost=4,site=nicolasgallagher.com
+$boost=4,site=leowid.com
+$boost=4,site=blog.izs.me
+$boost=4,site=electrospaces.net
+$boost=4,site=granta.com
+$boost=4,site=blog.darksky.net
+$boost=4,site=shipilev.net
+$boost=4,site=nanex.net
+$boost=4,site=leetcode.com
+$boost=4,site=numfocus.org
+$boost=4,site=tomscott.com
+$boost=4,site=trmm.net
+$boost=4,site=marketplace.org
+$boost=4,site=codecademy.com
+$boost=4,site=bit-player.org
+$boost=4,site=postmake.io
+$boost=4,site=journal.standardnotes.org
+$boost=4,site=semantic-ui.com
+$boost=4,site=blog.dshr.org
+$boost=4,site=rogerebert.com
+$boost=4,site=web.dev
+$boost=4,site=bravenewgeek.com
+$boost=4,site=benjiweber.co.uk
+$boost=4,site=theinformation.com
+$boost=4,site=jmmv.dev
+$boost=4,site=nycresistor.com
+$boost=4,site=glench.com
+$boost=4,site=blog.kicksend.com
+$boost=4,site=neverworkintheory.org
+$boost=4,site=thebillfold.com
+$boost=4,site=superhighway98.com
+$boost=4,site=tylerneylon.com
+$boost=4,site=thetruesize.com
+$boost=4,site=zipcpu.com
+$boost=4,site=cs.uni.edu
+$boost=4,site=bradleytaunt.com
+$boost=4,site=scripting.com
+$boost=4,site=cybertec-postgresql.com
+$boost=4,site=wcvb.com
+$boost=4,site=dslreports.com
+$boost=4,site=justi.cz
+$boost=4,site=blog.data.gov.sg
+$boost=4,site=gutenberg.org
+$boost=4,site=sendwithus.com
+$boost=4,site=chrismm.com
+$boost=4,site=stevecorona.com
+$boost=4,site=eager.io
+$boost=4,site=insidesources.com
+$boost=4,site=chrisseaton.com
+$boost=4,site=behind-the-enemy-lines.com
+$boost=4,site=commondreams.org
+$boost=4,site=tech.marksblogg.com
+$boost=4,site=blog.bitquabit.com
+$boost=4,site=utkusen.com
+$boost=4,site=lab.hakim.se
+$boost=4,site=calacanis.com
+$boost=4,site=blog.metaobject.com
+$boost=4,site=blogs.discovermagazine.com
+$boost=4,site=flutter.io
+$boost=4,site=chriskrycho.com
+$boost=4,site=washingtonmonthly.com
+$boost=4,site=nora.codes
+$boost=4,site=blog.professorbeekums.com
+$boost=4,site=yro.slashdot.org
+$boost=4,site=noncombatant.org
+$boost=4,site=irrlicht3d.org
+$boost=4,site=analog.com
+$boost=4,site=byuu.net
+$boost=4,site=learntemail.sam.today
+$boost=4,site=discuss.joelonsoftware.com
+$boost=4,site=hamberg.no
+$boost=4,site=culture.pl
+$boost=4,site=godbolt.org
+$boost=4,site=usgamer.net
+$boost=4,site=openmymind.net
+$boost=4,site=pgcon.org
+$boost=4,site=blog.insightdatascience.com
+$boost=4,site=blog.mapbox.com
+$boost=4,site=heap.io
+$boost=4,site=openttd.org
+$boost=4,site=gazehawk.com
+$boost=4,site=blog.headius.com
+$boost=4,site=jquerymobile.com
+$boost=4,site=strlen.com
+$boost=4,site=designfestival.com
+$boost=4,site=blog.kowalczyk.info
+$boost=4,site=natewienert.com
+$boost=4,site=blog.iron.io
+$boost=4,site=optaplanner.org
+$boost=4,site=blog.directededge.com
+$boost=4,site=acm.org
+$boost=4,site=postgres.heroku.com
+$boost=4,site=upguard.com
+$boost=4,site=panix.com
+$boost=4,site=arangodb.com
+$boost=4,site=ctvnews.ca
+$boost=4,site=tips.ariyh.com
+$boost=4,site=henrikwarne.com
+$boost=4,site=sorbet.org
+$boost=4,site=whattofix.com
+$boost=4,site=lukeplant.me.uk
+$boost=4,site=areallybadidea.com
+$boost=4,site=rednuht.org
+$boost=4,site=cappuccino.org
+$boost=4,site=blog.wallaroolabs.com
+$boost=4,site=story.californiasunday.com
+$boost=4,site=xato.net
+$boost=4,site=thinkfaster.co
+$boost=4,site=blog.algorithmia.com
+$boost=4,site=n-o-d-e.net
+$boost=4,site=talkingquickly.co.uk
+$boost=4,site=blamcast.net
+$boost=4,site=opendemocracy.net
+$boost=4,site=pixelbeat.org
+$boost=4,site=blog.2ndquadrant.com
+$boost=4,site=klinger.io
+$boost=4,site=invisibleup.com
+$boost=4,site=mattgreer.org
+$boost=4,site=georgehart.com
+$boost=4,site=gsuiteupdates.googleblog.com
+$boost=4,site=programming.guide
+$boost=4,site=revealnews.org
+$boost=4,site=renderman.pixar.com
+$boost=4,site=swtch.com
+$boost=4,site=akkartik.name
+$boost=4,site=ma.ttias.be
+$boost=4,site=blog.freecodecamp.com
+$boost=4,site=cognitiontoday.com
+$boost=4,site=data.triplebyte.com
+$boost=4,site=unix4lyfe.org
+$boost=4,site=coconut-lang.org
+$boost=4,site=discourse.nixos.org
+$boost=4,site=macwright.org
+$boost=4,site=varnish-cache.org
+$boost=4,site=blog.lyft.com
+$boost=4,site=martinrue.com
+$boost=4,site=bittersoutherner.com
+$boost=4,site=somebits.com
+$boost=4,site=certsimple.com
+$boost=4,site=calvertjournal.com
+$boost=4,site=pentestpartners.com
+$boost=4,site=wiki.gentoo.org
+$boost=4,site=00f.net
+$boost=4,site=geekwire.com
+$boost=4,site=georgestocker.com
+$boost=4,site=yacoset.com
+$boost=4,site=spencerfry.com
+$boost=4,site=thenewatlantis.com
+$boost=4,site=j2kun.svbtle.com
+$boost=4,site=bothsidesofthetable.com
+$boost=4,site=thetimes.co.uk
+$boost=4,site=corecursive.com
+$boost=4,site=wildml.com
+$boost=4,site=smallcultfollowing.com
+$boost=4,site=theartnewspaper.com
+$boost=4,site=cdixon.posterous.com
+$boost=4,site=readwriteweb.com
+$boost=4,site=planscope.io
+$boost=4,site=allthingsdistributed.com
+$boost=4,site=boston.com
+$boost=4,site=blog.kentcdodds.com
+$boost=4,site=brianstorti.com
+$boost=4,site=austinvernon.eth.link
+$boost=4,site=residentadvisor.net
+$boost=4,site=documentcloud.org
+$boost=4,site=howardism.org
+$boost=4,site=fsharpforfunandprofit.com
+$boost=4,site=semiaccurate.com
+$boost=4,site=hetzner.com
+$boost=4,site=aiweirdness.com
+$boost=4,site=bit-101.com
+$boost=4,site=41latitude.com
+$boost=4,site=amnh.org
+$boost=4,site=stories.californiasunday.com
+$boost=4,site=pxlnv.com
+$boost=4,site=batsov.com
+$boost=4,site=gamedev.net
+$boost=4,site=gatesfoundation.org
+$boost=4,site=meta.ath0.com
+$boost=4,site=radicle.xyz
+$boost=4,site=gigamonkeys.com
+$boost=4,site=thetyee.ca
+$boost=4,site=ucdavis.edu
+$boost=4,site=teenage.engineering
+$boost=4,site=approximatelycorrect.com
+$boost=4,site=console.dev
+$boost=4,site=qemu.org
+$boost=4,site=chadaustin.me
+$boost=4,site=javascript.info
+$boost=4,site=crawshaw.io
+$boost=4,site=minireference.com
+$boost=4,site=0x00sec.org
+$boost=4,site=jtaby.com
+$boost=4,site=hackernoon.com
+$boost=4,site=aem1k.com
+$boost=4,site=theatlanticwire.com
+$boost=4,site=tennessean.com
+$boost=4,site=graphicallinearalgebra.net
+$boost=4,site=jpmens.net
+$boost=4,site=eugenewei.com
+$boost=4,site=blog.packagecloud.io
+$boost=4,site=zef.me
+$boost=4,site=blog.makezine.com
+$boost=4,site=yannesposito.com
+$boost=4,site=apolloinrealtime.org
+$boost=4,site=nicole.express
+$boost=4,site=docs.browserless.io
+$boost=4,site=joshondesign.com
+$boost=4,site=aviationweek.com
+$boost=4,site=philip.greenspun.com
+$boost=4,site=ariadne.space
+$boost=4,site=blog.sia.tech
+$boost=4,site=cato.org
+$boost=4,site=blog.hipchat.com
+$boost=4,site=ruben.verborgh.org
+$boost=4,site=straightdope.com
+$boost=4,site=joshbarczak.com
+$boost=4,site=aarongreenspan.com
+$boost=4,site=changelog.com
+$boost=4,site=jdlm.info
+$boost=4,site=blog.cleancoder.com
+$boost=4,site=kyrobeshay.com
+$boost=4,site=js1k.com
+$boost=4,site=socketsite.com
+$boost=4,site=spinics.net
+$boost=4,site=h-online.com
+$boost=4,site=humanwhocodes.com
+$boost=4,site=angular.io
+$boost=4,site=digitalcameraworld.com
+$boost=4,site=amd.com
+$boost=4,site=scienceblogs.com
+$boost=4,site=jonathan-cook.net
+$boost=4,site=randomwire.com
+$boost=4,site=blog.crunchydata.com
+$boost=4,site=timeshighereducation.co.uk
+$boost=4,site=mihaileric.com
+$boost=4,site=seatgeek.com
+$boost=4,site=fanfiction.net
+$boost=4,site=thecounter.org
+$boost=4,site=haproxy.com
+$boost=4,site=movingfulcrum.com
+$boost=4,site=onstartups.com
+$boost=4,site=noyb.eu
+$boost=4,site=millionshort.com
+$boost=4,site=blog.sketchapp.com
+$boost=4,site=tillett.info
+$boost=4,site=community.ui.com
+$boost=4,site=adnauseam.io
+$boost=4,site=intoli.com
+$boost=4,site=jsoftware.com
+$boost=4,site=chicagotribune.com
+$boost=4,site=developer.squareup.com
+$boost=4,site=pixelambacht.nl
+$boost=4,site=lambdaschool.com
+$boost=4,site=consumerist.com
+$boost=4,site=reaktor.com
+$boost=4,site=blog.rescuetime.com
+$boost=4,site=dustycloud.org
+$boost=4,site=the.exa.website
+$boost=4,site=adventures.michaelfbryan.com
+$boost=4,site=blog.xamarin.com
+$boost=4,site=micahlerner.com
+$boost=4,site=canonical.org
+$boost=4,site=scmp.com
+$boost=4,site=grantland.com
+$boost=4,site=nextplatform.com
+$boost=4,site=blog.netlab.360.com
+$boost=4,site=m.nautil.us
+$boost=4,site=cell.com
+$boost=4,site=cheapskatesguide.org
+$boost=3,site=w3techs.com
+$boost=3,site=literaryreview.co.uk
+$boost=3,site=explorabl.es
+$boost=3,site=blog.felipe.rs
+$boost=3,site=unherd.com
+$boost=3,site=garry.posterous.com
+$boost=3,site=worthdoingbadly.com
+$boost=3,site=rob.conery.io
+$boost=3,site=io9.com
+$boost=3,site=nolanlawson.com
+$boost=3,site=soundguys.com
+$boost=3,site=starterstory.com
+$boost=3,site=blog.jse.li
+$boost=3,site=knowingless.com
+$boost=3,site=decrypt.co
+$boost=3,site=euronews.com
+$boost=3,site=nirsoft.net
+$boost=3,site=exploringjs.com
+$boost=3,site=recurse.com
+$boost=3,site=legacy.com
+$boost=3,site=docs.fcc.gov
+$boost=3,site=channel9.msdn.com
+$boost=3,site=blog.ksplice.com
+$boost=3,site=valvesoftware.com
+$boost=3,site=blog.netbsd.org
+$boost=3,site=hyper.sh
+$boost=3,site=blog.waleedkhan.name
+$boost=3,site=sugru.com
+$boost=3,site=feefighters.com
+$boost=3,site=dejanseo.com.au
+$boost=3,site=ianww.com
+$boost=3,site=blog.cr.yp.to
+$boost=3,site=kentnguyen.com
+$boost=3,site=jonathanturley.org
+$boost=3,site=snyk.io
+$boost=3,site=pluto.jhuapl.edu
+$boost=3,site=calendar.perfplanet.com
+$boost=3,site=tinyapps.org
+$boost=3,site=blog.evjang.com
+$boost=3,site=lightbluetouchpaper.org
+$boost=3,site=blog.arduino.cc
+$boost=3,site=psychnewsdaily.com
+$boost=3,site=hypercritical.co
+$boost=3,site=0a.io
+$boost=3,site=erowid.org
+$boost=3,site=inklestudios.com
+$boost=3,site=blog.soylent.com
+$boost=3,site=fpcomplete.com
+$boost=3,site=dmitryfrank.com
+$boost=3,site=red-lang.org
+$boost=3,site=parallella.org
+$boost=3,site=habr.com
+$boost=3,site=musictech.net
+$boost=3,site=josephwalla.com
+$boost=3,site=hsivonen.fi
+$boost=3,site=thoughtcatalog.com
+$boost=3,site=scylladb.com
+$boost=3,site=stochasticlifestyle.com
+$boost=3,site=vaibhavsagar.com
+$boost=3,site=searchengineland.com
+$boost=3,site=nrk.no
+$boost=3,site=blog.app.net
+$boost=3,site=mail.openjdk.java.net
+$boost=3,site=sciencebulletin.org
+$boost=3,site=paperplanes.de
+$boost=3,site=jessenoller.com
+$boost=3,site=blog.hirelite.com
+$boost=3,site=bemmu.com
+$boost=3,site=adguard.com
+$boost=3,site=protocol.com
+$boost=3,site=redislabs.com
+$boost=3,site=coding2learn.org
+$boost=3,site=tonywright.com
+$boost=3,site=writing.kemitchell.com
+$boost=3,site=voxeu.org
+$boost=3,site=math.ucr.edu
+$boost=3,site=gekk.info
+$boost=3,site=dwitter.net
+$boost=3,site=tomdispatch.com
+$boost=3,site=erichgrunewald.com
+$boost=3,site=everydayastronaut.com
+$boost=3,site=dashbit.co
+$boost=3,site=kroah.com
+$boost=3,site=terrybisson.com
+$boost=3,site=ploum.net
+$boost=3,site=explosion.ai
+$boost=3,site=blog.repl.it
+$boost=3,site=neurosciencenews.com
+$boost=3,site=supermemo.com
+$boost=3,site=diyphotography.net
+$boost=3,site=kcsrk.info
+$boost=3,site=blog.rjmetrics.com
+$boost=3,site=kakoune.org
+$boost=3,site=saurik.com
+$boost=3,site=thetoolsweneed.com
+$boost=3,site=atthis.link
+$boost=3,site=softwarepreservation.org
+$boost=3,site=kleinbottle.com
+$boost=3,site=interrupt.memfault.com
+$boost=3,site=philipwalton.com
+$boost=3,site=foxbusiness.com
+$boost=3,site=jrsinclair.com
+$boost=3,site=space.com
+$boost=3,site=sachin.posterous.com
+$boost=3,site=khn.org
+$boost=3,site=pigdog.org
+$boost=3,site=zackkanter.com
+$boost=3,site=informationdiet.com
+$boost=3,site=giantrobotlasers.com
+$boost=3,site=incolumitas.com
+$boost=3,site=jeffkreeftmeijer.com
+$boost=3,site=bostonreview.net
+$boost=3,site=techreport.com
+$boost=3,site=baldurbjarnason.com
+$boost=3,site=sicpers.info
+$boost=3,site=write.as
+$boost=3,site=everything2.com
+$boost=3,site=typicalprogrammer.com
+$boost=3,site=blog.scaleway.com
+$boost=3,site=rethrick.com
+$boost=3,site=inference-review.com
+$boost=3,site=blog.23andme.com
+$boost=3,site=mikealche.com
+$boost=3,site=hashcat.net
+$boost=3,site=madebyloren.com
+$boost=3,site=blockchain.com
+$boost=3,site=computerworld.com
+$boost=3,site=lauren.vortex.com
+$boost=3,site=commoncrawl.org
+$boost=3,site=svds.com
+$boost=3,site=blog.goodstuff.im
+$boost=3,site=lhup.edu
+$boost=3,site=jasonevanish.com
+$boost=3,site=nextcity.org
+$boost=3,site=nuitka.net
+$boost=3,site=dataswamp.org
+$boost=3,site=limpet.net
+$boost=3,site=nextjournal.com
+$boost=3,site=sinews.siam.org
+$boost=3,site=blog.zactownsend.com
+$boost=3,site=m.guardiannews.com
+$boost=3,site=mazzo.li
+$boost=3,site=spideroak.com
+$boost=3,site=startribune.com
+$boost=3,site=gitbook.com
+$boost=3,site=brusselstimes.com
+$boost=3,site=benedikt-bitterli.me
+$boost=3,site=fedoramagazine.org
+$boost=3,site=rte.ie
+$boost=3,site=f-droid.org
+$boost=3,site=blog.dwolla.com
+$boost=3,site=news.trust.org
+$boost=3,site=about.bnef.com
+$boost=3,site=txti.es
+$boost=3,site=samizdat.mines.edu
+$boost=3,site=murilopereira.com
+$boost=3,site=bitcoincharts.com
+$boost=3,site=material.io
+$boost=3,site=blog.rapportive.com
+$boost=3,site=m-cacm.acm.org
+$boost=3,site=blog.halide.cam
+$boost=3,site=billthelizard.com
+$boost=3,site=edmunds.com
+$boost=3,site=wiki.c2.com
+$boost=3,site=thinkvitamin.com
+$boost=3,site=ipython.org
+$boost=3,site=panelbear.com
+$boost=3,site=stuartsierra.com
+$boost=3,site=clojurescript.org
+$boost=3,site=maps.org
+$boost=3,site=dextronet.com
+$boost=3,site=happyassassin.net
+$boost=3,site=blockchain.info
+$boost=3,site=realm.io
+$boost=3,site=til.simonwillison.net
+$boost=3,site=fingerprintjs.com
+$boost=3,site=cs-blog.khanacademy.org
+$boost=3,site=bay12games.com
+$boost=3,site=ginzametrics.com
+$boost=3,site=blog.simpleanalytics.com
+$boost=3,site=viewfromthewing.com
+$boost=3,site=softwaremaniacs.org
+$boost=3,site=blog.mariadb.org
+$boost=3,site=levels.io
+$boost=3,site=craigmod.com
+$boost=3,site=overcomingbias.com
+$boost=3,site=users.rust-lang.org
+$boost=3,site=scrimba.com
+$boost=3,site=sbnation.com
+$boost=3,site=benswann.com
+$boost=3,site=thelocal.dk
+$boost=3,site=pitchfork.com
+$boost=3,site=wekeroad.com
+$boost=3,site=animalnewyork.com
+$boost=3,site=beets.io
+$boost=3,site=ibiblio.org
+$boost=3,site=kernelnewbies.org
+$boost=3,site=fullstackreact.com
+$boost=3,site=ricostacruz.com
+$boost=3,site=pointersgonewild.com
+$boost=3,site=512pixels.net
+$boost=3,site=blog.wilshipley.com
+$boost=3,site=devlog.hexops.com
+$boost=3,site=sacbee.com
+$boost=3,site=publicbooks.org
+$boost=3,site=hforsten.com
+$boost=3,site=okmij.org
+$boost=3,site=dataquest.io
+$boost=3,site=nocsdegree.com
+$boost=3,site=pandas.pydata.org
+$boost=3,site=blog.ezyang.com
+$boost=3,site=cs.gmu.edu
+$boost=3,site=weblog.raganwald.com
+$boost=3,site=pedestrianobservations.com
+$boost=3,site=dangrover.com
+$boost=3,site=nostarch.com
+$boost=3,site=foreignpolicy.com
+$boost=3,site=benchmarksgame-team.pages.debian.net
+$boost=3,site=scummvm.org
+$boost=3,site=theweek.com
+$boost=3,site=status.heroku.com
+$boost=3,site=mensjournal.com
+$boost=3,site=braythwayt.com
+$boost=3,site=clayallsopp.com
+$boost=3,site=blog.japaric.io
+$boost=3,site=linuxuprising.com
+$boost=3,site=pando.com
+$boost=3,site=nextcloud.com
+$boost=3,site=blog.luap.info
+$boost=3,site=preetamnath.com
+$boost=3,site=loper-os.org
+$boost=3,site=mining.com
+$boost=3,site=thestartuptoolkit.com
+$boost=3,site=exple.tive.org
+$boost=3,site=percona.com
+$boost=3,site=tug.org
+$boost=3,site=badassjs.com
+$boost=3,site=florentcrivello.com
+$boost=3,site=cntraveler.com
+$boost=3,site=postmasculine.com
+$boost=3,site=blade.nagaokaut.ac.jp
+$boost=3,site=blog.demofox.org
+$boost=3,site=zeit.de
+$boost=3,site=louwrentius.com
+$boost=3,site=jeffwofford.com
+$boost=3,site=lists.canonical.org
+$boost=3,site=blog.aboutamazon.com
+$boost=3,site=discourse.haskell.org
+$boost=3,site=marginalrevolution.com
+$boost=3,site=southampton.ac.uk
+$boost=3,site=wccftech.com
+$boost=3,site=blog.jcoglan.com
+$boost=3,site=today.duke.edu
+$boost=3,site=forums.swift.org
+$boost=3,site=sebastianraschka.com
+$boost=3,site=michaelfogleman.com
+$boost=3,site=confluent.io
+$boost=3,site=maa.org
+$boost=3,site=serverfault.com
+$boost=3,site=fitzgeraldnick.com
+$boost=3,site=2600.com
+$boost=3,site=rustyshelf.org
+$boost=3,site=svnweb.freebsd.org
+$boost=3,site=argmin.net
+$boost=3,site=theawl.com
+$boost=3,site=keygen.sh
+$boost=3,site=freightwaves.com
+$boost=3,site=randomhacks.net
+$boost=3,site=susam.in
+$boost=3,site=exercism.io
+$boost=3,site=trekhleb.dev
+$boost=3,site=securelist.com
+$boost=3,site=techrights.org
+$boost=3,site=franciskim.co
+$boost=3,site=ergoemacs.org
+$boost=3,site=currentaffairs.org
+$boost=3,site=arc.aiaa.org
+$boost=3,site=youtube.googleblog.com
+$boost=3,site=pulitzer.org
+$boost=3,site=computer.rip
+$boost=3,site=blog.dgraph.io
+$boost=3,site=blog.fastforwardlabs.com
+$boost=3,site=bridgewater.com
+$boost=3,site=mycroft.ai
+$boost=3,site=open.salon.com
+$boost=3,site=blog.longreads.com
+$boost=3,site=nvie.com
+$boost=3,site=adamdrake.com
+$boost=3,site=cordcuttersnews.com
+$boost=3,site=kayak.com
+$boost=3,site=aiexperiments.withgoogle.com
+$boost=3,site=blog.statuspage.io
+$boost=3,site=hyperpolyglot.org
+$boost=3,site=scaleway.com
+$boost=3,site=bennettfeely.com
+$boost=3,site=ktvu.com
+$boost=3,site=tomsguide.com
+$boost=3,site=jitbit.com
+$boost=3,site=time.is
+$boost=3,site=tinyclouds.org
+$boost=3,site=gurjeet.singh.im
+$boost=3,site=blog.mailgun.net
+$boost=3,site=harry.me
+$boost=3,site=archpaper.com
+$boost=3,site=news.ycombinator.com
+$boost=3,site=durmonski.com
+$boost=3,site=blog.zorinaq.com
+$boost=3,site=blog.peepcode.com
+$boost=3,site=kivy.org
+$boost=3,site=crowdtilt.com
+$boost=3,site=wyden.senate.gov
+$boost=3,site=watsi.org
+$boost=3,site=wolframalpha.com
+$boost=3,site=nethack.org
+$boost=3,site=tidbits.com
+$boost=3,site=theappendix.net
+$boost=3,site=fossjobs.net
+$boost=3,site=blog.robinhood.com
+$boost=3,site=realtimerendering.com
+$boost=3,site=blog.theincredibleholk.org
+$boost=3,site=ee.ryerson.ca
+$boost=3,site=databricks.com
+$boost=3,site=defensenews.com
+$boost=3,site=rcoh.me
+$boost=3,site=mijailovic.net
+$boost=3,site=gallium.inria.fr
+$boost=3,site=blog.replit.com
+$boost=3,site=learn.adafruit.com
+$boost=3,site=joelrunyon.com
+$boost=3,site=blog.frizk.net
+$boost=3,site=alarmingdevelopment.org
+$boost=3,site=robots.thoughtbot.com
+$boost=3,site=includeos.org
+$boost=3,site=mrcoles.com
+$boost=3,site=c2.com
+$boost=3,site=whonix.org
+$boost=3,site=fivebooks.com
+$boost=3,site=wilfred.me.uk
+$boost=3,site=whydoeseverythingsuck.com
+$boost=3,site=pcworld.com
+$boost=3,site=bamsoftware.com
+$boost=3,site=cs.unm.edu
+$boost=3,site=idris-lang.org
+$boost=3,site=raspberrypi.com
+$boost=3,site=effectiviology.com
+$boost=3,site=asciiflow.com
+$boost=3,site=philosophically.com
+$boost=3,site=alexa.com
+$boost=3,site=algorithmicbotany.org
+$boost=3,site=yasoob.me
+$boost=3,site=varun.ca
+$boost=3,site=medievalbooks.nl
+$boost=3,site=kgw.com
+$boost=3,site=junglecoder.com
+$boost=3,site=insideintercom.io
+$boost=3,site=html5rocks.com
+$boost=3,site=arxiv-vanity.com
+$boost=3,site=micahflee.com
+$boost=3,site=nccgroup.trust
+$boost=3,site=ashtonkemerling.com
+$boost=3,site=amusingplanet.com
+$boost=3,site=lea.verou.me
+$boost=3,site=nationalinterest.org
+$boost=3,site=lists.webkit.org
+$boost=3,site=opkode.com
+$boost=3,site=michaelgeist.ca
+$boost=3,site=exstreamist.com
+$boost=3,site=0xax.gitbooks.io
+$boost=3,site=jasoncrawford.org
+$boost=3,site=lego.com
+$boost=3,site=discourse.julialang.org
+$boost=3,site=open-std.org
+$boost=3,site=indiatoday.in
+$boost=3,site=winestockwebdesign.com
+$boost=3,site=mobiforge.com
+$boost=3,site=regulations.gov
+$boost=3,site=trackinghappiness.com
+$boost=3,site=blog.infinitenegativeutility.com
+$boost=3,site=blogs.balsamiq.com
+$boost=3,site=treyhunner.com
+$boost=3,site=beyondloom.com
+$boost=3,site=lennyrachitsky.com
+$boost=3,site=raymii.org
+$boost=3,site=blog.amirkhella.com
+$boost=3,site=mooreds.com
+$boost=3,site=hudsandguis.com
+$boost=3,site=flask.pocoo.org
+$boost=3,site=dol.gov
+$boost=3,site=wololo.net
+$boost=3,site=dashes.com
+$boost=3,site=mortoray.com
+$boost=3,site=mediawiki.org
+$boost=3,site=banterly.net
+$boost=3,site=justin.tv
+$boost=3,site=koreaherald.com
+$boost=3,site=two-wrongs.com
+$boost=3,site=3ders.org
+$boost=3,site=millcomputing.com
+$boost=3,site=remysharp.com
+$boost=3,site=magenta.tensorflow.org
+$boost=3,site=techradar.com
+$boost=3,site=pushing-pixels.org
+$boost=3,site=mymodernmet.com
+$boost=3,site=designyoutrust.com
+$boost=3,site=boards.4chan.org
+$boost=3,site=blog.framebase.io
+$boost=3,site=openwrt.org
+$boost=3,site=blog.codecombat.com
+$boost=3,site=informit.com
+$boost=3,site=evilsocket.net
+$boost=3,site=governing.com
+$boost=3,site=masteringemacs.org
+$boost=3,site=tomassetti.me
+$boost=3,site=gun.io
+$boost=3,site=netlify.com
+$boost=3,site=ntsb.gov
+$boost=3,site=cs.kent.ac.uk
+$boost=3,site=kernel.org
+$boost=3,site=flyertalk.com
+$boost=3,site=unlimitednovelty.com
+$boost=3,site=math.andrej.com
+$boost=3,site=healeycodes.com
+$boost=3,site=blog.jmdawson.co.uk
+$boost=3,site=metaoptimize.com
+$boost=3,site=mynorthwest.com
+$boost=3,site=asciinema.org
+$boost=3,site=pydanny.com
+$boost=3,site=warisboring.com
+$boost=3,site=blog.serverfault.com
+$boost=3,site=narratively.com
+$boost=3,site=freakonomics.com
+$boost=3,site=caixinglobal.com
+$boost=3,site=freelists.org
+$boost=3,site=devblog.bu.mp
+$boost=3,site=mislav.uniqpath.com
+$boost=3,site=heydonworks.com
+$boost=3,site=glyph.twistedmatrix.com
+$boost=3,site=paulirish.com
+$boost=3,site=vldb.org
+$boost=3,site=squeakland.org
+$boost=3,site=epi.org
+$boost=3,site=tryolabs.com
+$boost=3,site=willcrichton.net
+$boost=3,site=psmag.com
+$boost=3,site=meiert.com
+$boost=3,site=ghacks.net
+$boost=3,site=shamusyoung.com
+$boost=3,site=zeptobars.com
+$boost=3,site=cm.bell-labs.com
+$boost=3,site=fastcolabs.com
+$boost=3,site=aging.nautil.us
+$boost=3,site=hackerone.com
+$boost=3,site=storage.courtlistener.com
+$boost=3,site=danlebrero.com
+$boost=3,site=mapzen.com
+$boost=3,site=bernsteinbear.com
+$boost=3,site=austinmann.com
+$boost=3,site=vinaysahni.com
+$boost=3,site=burakkanber.com
+$boost=3,site=mollyrocket.com
+$boost=3,site=feeding.cloud.geek.nz
+$boost=3,site=cliffle.com
+$boost=3,site=mikeperham.com
+$boost=3,site=sequoiacap.com
+$boost=3,site=thestranger.com
+$boost=3,site=datacenterdynamics.com
+$boost=3,site=nongnu.org
+$boost=3,site=blog.pkh.me
+$boost=3,site=debarghyadas.com
+$boost=3,site=jborichevskiy.com
+$boost=3,site=saturdayeveningpost.com
+$boost=3,site=engineyard.com
+$boost=3,site=jeremymikkola.com
+$boost=3,site=the-scientist.com
+$boost=3,site=heterodoxacademy.org
+$boost=3,site=daveceddia.com
+$boost=3,site=eighty-twenty.org
+$boost=3,site=itvision.altervista.org
+$boost=3,site=citypages.com
+$boost=3,site=houstonchronicle.com
+$boost=3,site=perl.com
+$boost=3,site=mcclatchydc.com
+$boost=3,site=over-yonder.net
+$boost=3,site=donatstudios.com
+$boost=3,site=johnmyleswhite.com
+$boost=3,site=samsclass.info
+$boost=3,site=markmaunder.com
+$boost=3,site=tauday.com
+$boost=3,site=riscv.org
+$boost=3,site=blog.blazingdb.com
+$boost=3,site=garrett.damore.org
+$boost=3,site=ilikebigbits.com
+$boost=3,site=mediagoblin.org
+$boost=3,site=keithclark.co.uk
+$boost=3,site=reviews.llvm.org
+$boost=3,site=census.gov
+$boost=3,site=carbonbrief.org
+$boost=3,site=1000memories.com
+$boost=3,site=tommorris.org
+$boost=3,site=csswizardry.com
+$boost=3,site=t3x.org
+$boost=3,site=alpinelinux.org
+$boost=3,site=news.blizzard.com
+$boost=3,site=lowendmac.com
+$boost=3,site=jakubmarian.com
+$boost=3,site=smackdown.blogsblogsblogs.com
+$boost=3,site=gamefromscratch.com
+$boost=3,site=history.com
+$boost=3,site=aps.org
+$boost=3,site=bitbashing.io
+$boost=3,site=thecreativeindependent.com
+$boost=3,site=filecoin.io
+$boost=3,site=archdaily.com
+$boost=3,site=critter.blog
+$boost=3,site=blog.darklang.com
+$boost=3,site=youteam.io
+$boost=3,site=developer.oculus.com
+$boost=3,site=caitlingreen.org
+$boost=3,site=stuffin.space
+$boost=3,site=logicmatters.net
+$boost=3,site=coindesk.com
+$boost=3,site=blog.checklyhq.com
+$boost=3,site=reifyworks.com
+$boost=3,site=tensorflow.org
+$boost=3,site=medicalxpress.com
+$boost=3,site=cabinetmagazine.org
+$boost=3,site=vez.mrsk.me
+$boost=3,site=tenderlovemaking.com
+$boost=3,site=timdettmers.com
+$boost=3,site=timesonline.co.uk
+$boost=3,site=pdfs.semanticscholar.org
+$boost=3,site=trustwave.com
+$boost=3,site=ambrevar.xyz
+$boost=3,site=theroot.com
+$boost=3,site=tylertringas.com
+$boost=3,site=globenewswire.com
+$boost=3,site=applevis.com
+$boost=3,site=metafilter.com
+$boost=3,site=erickhun.com
+$boost=3,site=blog.wearewizards.io
+$boost=3,site=cpj.org
+$boost=3,site=boyter.org
+$boost=3,site=profgalloway.com
+$boost=3,site=mentallandscape.com
+$boost=3,site=hackertarget.com
+$boost=3,site=pyret.org
+$boost=3,site=obdev.at
+$boost=3,site=mercurial-scm.org
+$boost=3,site=datanitro.com
+$boost=3,site=huffingtonpost.in
+$boost=3,site=fangpenlin.com
+$boost=3,site=discovermagazine.com
+$boost=3,site=blog.documentfoundation.org
+$boost=3,site=dotsies.org
+$boost=3,site=tails.boum.org
+$boost=3,site=randfishkin.com
+$boost=3,site=tomasp.net
+$boost=3,site=taiwannews.com.tw
+$boost=3,site=modern-sql.com
+$boost=3,site=carlos.bueno.org
+$boost=3,site=kansascity.com
+$boost=3,site=cdm.link
+$boost=3,site=nighthacks.com
+$boost=3,site=blog.vrypan.net
+$boost=3,site=nyxt.atlas.engineer
+$boost=3,site=pjreddie.com
+$boost=3,site=fury.com
+$boost=3,site=blog.skylight.io
+$boost=3,site=kickstarter.engineering
+$boost=3,site=darpa.mil
+$boost=3,site=psyarxiv.com
+$boost=3,site=spottedsun.com
+$boost=3,site=matt.aimonetti.net
+$boost=3,site=news.dartlang.org
+$boost=3,site=musescore.org
+$boost=3,site=flightglobal.com
+$boost=3,site=theatlanticcities.com
+$boost=3,site=search.cpan.org
+$boost=3,site=p01.org
+$boost=3,site=blog.dantup.com
+$boost=3,site=thedailywtf.com
+$boost=3,site=e.foundation
+$boost=3,site=toroid.org
+$boost=3,site=forum.sublimetext.com
+$boost=3,site=praxis.fortelabs.co
+$boost=3,site=leostartsup.com
+$boost=3,site=web.cecs.pdx.edu
+$boost=3,site=ofbrooklyn.com
+$boost=3,site=revk.uk
+$boost=3,site=rawkes.com
+$boost=3,site=furbo.org
+$boost=3,site=nickoneill.com
+$boost=3,site=compilers.iecc.com
+$boost=3,site=climatechangenews.com
+$boost=3,site=avichal.com
+$boost=3,site=healthcaredive.com
+$boost=3,site=lucasfcosta.com
+$boost=3,site=stevecoast.com
+$boost=3,site=dariusforoux.com
+$boost=3,site=nzz.ch
+$boost=3,site=trac.webkit.org
+$boost=3,site=waze.com
+$boost=3,site=liip.ch
+$boost=3,site=ericlippert.com
+$boost=3,site=matpalm.com
+$boost=3,site=roshfu.com
+$boost=3,site=factmag.com
+$boost=3,site=grafana.org
+$boost=3,site=divio.com
+$boost=3,site=cs.huji.ac.il
+$boost=3,site=gothamist.com
+$boost=3,site=ajc.com
+$boost=3,site=mathpix.com
+$boost=3,site=vulture.com
+$boost=3,site=picolisp.com
+$boost=3,site=goldsborough.me
+$boost=3,site=arkadiyt.com
+$boost=3,site=flowingdata.com
+$boost=3,site=grist.org
+$boost=3,site=clusterhq.com
+$boost=3,site=postmarketos.org
+$boost=3,site=monkey.org
+$boost=3,site=coderholic.com
+$boost=3,site=blog.silktide.com
+$boost=3,site=docs.house.gov
+$boost=3,site=nhs.uk
+$boost=3,site=insideclimatenews.org
+$boost=3,site=holloway.com
+$boost=3,site=airspacemag.com
+$boost=3,site=grsecurity.net
+$boost=3,site=kevinrooke.com
+$boost=3,site=engineering.gusto.com
+$boost=3,site=blog.christianperone.com
+$boost=3,site=badscience.net
+$boost=3,site=jerf.org
+$boost=3,site=vfoley.xyz
+$boost=3,site=dangrossman.info
+$boost=3,site=blog.timac.org
+$boost=3,site=nme.com
+$boost=3,site=withblue.ink
+$boost=3,site=modernperlbooks.com
+$boost=3,site=vasilishynkarenka.com
+$boost=3,site=startupboy.com
+$boost=3,site=thesoundingline.com
+$boost=3,site=blog.kotowicz.net
+$boost=3,site=ubuntuvibes.com
+$boost=3,site=tindie.com
+$boost=3,site=falstad.com
+$boost=3,site=alisdair.mcdiarmid.org
+$boost=3,site=brennancenter.org
+$boost=3,site=theanarchistlibrary.org
+$boost=3,site=cftc.gov
+$boost=3,site=lisp-journey.gitlab.io
+$boost=3,site=namepros.com
+$boost=3,site=sahandsaba.com
+$boost=3,site=blog.rstudio.org
+$boost=3,site=neilkakkar.com
+$boost=3,site=news.nationalpost.com
+$boost=3,site=snopes.com
+$boost=3,site=brainpickings.org
+$boost=3,site=mitmproxy.org
+$boost=3,site=blog.balancedpayments.com
+$boost=3,site=breakingdefense.com
+$boost=3,site=jerrygamblin.com
+$boost=3,site=helloerik.com
+$boost=3,site=diogomonica.com
+$boost=3,site=oversight.house.gov
+$boost=3,site=blog.yossarian.net
+$boost=3,site=tim.blog
+$boost=3,site=blog.marcocantu.com
+$boost=3,site=stilldrinking.org
+$boost=3,site=troubles.md
+$boost=3,site=8bitworkshop.com
+$boost=3,site=blog.watsi.org
+$boost=3,site=vatican.va
+$boost=3,site=gridsagegames.com
+$boost=3,site=blog.wakemate.com
+$boost=3,site=eastbaytimes.com
+$boost=3,site=blog.eyas.sh
+$boost=3,site=agwa.name
+$boost=3,site=lexfridman.com
+$boost=3,site=chargen.matasano.com
+$boost=3,site=blogs.esa.int
+$boost=3,site=hackage.haskell.org
+$boost=3,site=askthepilot.com
+$boost=3,site=hackertourism.com
+$boost=3,site=varianceexplained.org
+$boost=3,site=blog.spark.io
+$boost=3,site=danielmiessler.com
+$boost=3,site=blog.cloudant.com
+$boost=3,site=thechangelog.com
+$boost=3,site=mrdoob.com
+$boost=3,site=saveur.com
+$boost=3,site=moderntreasury.com
+$boost=3,site=blogs.suntimes.com
+$boost=3,site=ironicsans.com
+$boost=3,site=blog.capitalandgrowth.org
+$boost=3,site=thisiscolossal.com
+$boost=3,site=benchling.engineering
+$boost=3,site=symmetrymagazine.org
+$boost=3,site=feld.com
+$boost=3,site=everynoise.com
+$boost=3,site=input.fontbureau.com
+$boost=3,site=factordaily.com
+$boost=3,site=addyosmani.com
+$boost=3,site=reallifemag.com
+$boost=3,site=niedzejkob.p4.team
+$boost=3,site=vercel.com
+$boost=3,site=blog.sesse.net
+$boost=3,site=mynoise.net
+$boost=3,site=justingarrison.com
+$boost=3,site=cse.buffalo.edu
+$boost=3,site=jamesgolick.com
+$boost=3,site=gnunet.org
+$boost=3,site=tampabay.com
+$boost=3,site=jide.com
+$boost=3,site=producthabits.com
+$boost=3,site=fourhourworkweek.com
+$boost=3,site=refer.ly
+$boost=3,site=aei.org
+$boost=3,site=mentalfloss.com
+$boost=3,site=izbicki.me
+$boost=3,site=mathiasbynens.be
+$boost=3,site=pcjs.org
+$boost=3,site=arcan-fe.com
+$boost=3,site=americamagazine.org
+$boost=3,site=mailman.nanog.org
+$boost=3,site=mysliderule.com
+$boost=3,site=leonardofed.io
+$boost=3,site=sourcetrail.com
+$boost=3,site=linuxfromscratch.org
+$boost=3,site=moot.it
+$boost=3,site=popula.com
+$boost=3,site=press.web.cern.ch
+$boost=3,site=ownyourbits.com
+$boost=3,site=getstream.io
+$boost=3,site=dcgross.com
+$boost=3,site=unsupervisedmethods.com
+$boost=3,site=willmcgugan.com
+$boost=3,site=wiki.tcl-lang.org
+$boost=3,site=blog.ircmaxell.com
+$boost=3,site=mux.com
+$boost=3,site=irishtimes.com
+$boost=3,site=dpc.pw
+$boost=3,site=dusted.codes
+$boost=3,site=quanticdev.com
+$boost=3,site=neowin.net
+$boost=3,site=chrislord.net
+$boost=3,site=blog.otoro.net
+$boost=3,site=thenational.scot
+$boost=3,site=inquirer.com
+$boost=3,site=modelviewculture.com
+$boost=3,site=info.iet.unipi.it
+$boost=3,site=storehouse.co
+$boost=3,site=anfractuosity.com
+$boost=3,site=danwang.co
+$boost=3,site=doxsey.net
+$boost=3,site=bugs.ruby-lang.org
+$boost=3,site=whyy.org
+$boost=3,site=itch.io
+$boost=3,site=amigalove.com
+$boost=3,site=medium.freecodecamp.com
+$boost=3,site=blog.doist.com
+$boost=3,site=swyx.io
+$boost=3,site=eidel.io
+$boost=3,site=senrigan.io
+$boost=3,site=pvk.ca
+$boost=3,site=ccs.neu.edu
+$boost=3,site=lysator.liu.se
+$boost=3,site=philcalcado.com
+$boost=3,site=lists.dragonflybsd.org
+$boost=3,site=theringer.com
+$boost=3,site=bzg.fr
+$boost=3,site=crypto.com
+$boost=3,site=fibery.io
+$boost=3,site=kimonote.com
+$boost=3,site=news.softpedia.com
+$boost=3,site=successfulsoftware.net
+$boost=3,site=blog.intercom.com
+$boost=3,site=embeddedrelated.com
+$boost=3,site=haveibeenpwned.com
+$boost=3,site=blog.ryotak.me
+$boost=3,site=blog.piston.rs
+$boost=3,site=joshmitteldorf.scienceblog.com
+$boost=3,site=reviews.com
+$boost=3,site=svpow.com
+$boost=3,site=blog.nahurst.com
+$boost=3,site=engineering.kablamo.com.au
+$boost=3,site=blog.adafruit.com
+$boost=3,site=earth.nullschool.net
+$boost=3,site=inventwithpython.com
+$boost=3,site=chalkdustmagazine.com
+$boost=3,site=blog.objectmentor.com
+$boost=3,site=aholme.co.uk
+$boost=3,site=dkriesel.com
+$boost=3,site=pipeline.corante.com
+$boost=3,site=action.eff.org
+$boost=3,site=nand2tetris.org
+$boost=3,site=gpfault.net
+$boost=3,site=meshedinsights.com
+$boost=3,site=missiveapp.com
+$boost=3,site=blog.sendgrid.com
+$boost=3,site=mitadmissions.org
+$boost=3,site=info.meteor.com
+$boost=3,site=thewirecutter.com
+$boost=3,site=aboveavalon.com
+$boost=3,site=haxe.org
+$boost=3,site=snarky.ca
+$boost=3,site=blog.mapillary.com
+$boost=3,site=alternet.org
+$boost=3,site=martinheinz.dev
+$boost=3,site=stevebenjamins.com
+$boost=3,site=kuro5hin.org
+$boost=3,site=archlinux.org
+$boost=3,site=bennjordan.com
+$boost=3,site=tromey.com
+$boost=3,site=medlifestyle.news
+$boost=3,site=hackpad.com
+$boost=3,site=zinascii.com
+$boost=3,site=thegenesisblock.com
+$boost=3,site=rkoutnik.com
+$boost=3,site=blog.angularjs.org
+$boost=3,site=learnyousomeerlang.com
+$boost=3,site=blog.rubymotion.com
+$boost=3,site=berkshirehathaway.com
+$boost=3,site=bigstory.ap.org
+$boost=3,site=dji.com
+$boost=3,site=momjian.us
+$boost=3,site=anderspitman.net
+$boost=3,site=ubiquity.acm.org
+$boost=3,site=archive.vector.org.uk
+$boost=3,site=bitcoinfoundation.org
+$boost=3,site=esr.ibiblio.org
+$boost=3,site=caranddriver.com
+$boost=3,site=bristol.ac.uk
+$boost=3,site=keacher.com
+$boost=3,site=planet.com
+$boost=3,site=braintreepayments.com
+$boost=3,site=topic.com
+$boost=3,site=lyx.org
+$boost=3,site=hal.inria.fr
+$boost=3,site=blog.powerdns.com
+$boost=3,site=discovermeteor.com
+$boost=3,site=diveintohtml5.org
+$boost=3,site=danwin.com
+$boost=3,site=thingist.com
+$boost=3,site=sf.curbed.com
+$boost=3,site=artima.com
+$boost=3,site=gregegan.net
+$boost=3,site=veed.io
+$boost=3,site=kguttag.com
+$boost=3,site=passwordstore.org
+$boost=3,site=patft.uspto.gov
+$boost=3,site=luckypeach.com
+$boost=3,site=256kilobytes.com
+$boost=3,site=kamalmarhubi.com
+$boost=3,site=levels.fyi
+$boost=3,site=grubstreet.com
+$boost=3,site=kubernetes.io
+$boost=3,site=privpapers.ssrn.com
+$boost=3,site=finledger.com
+$boost=3,site=ircv3.net
+$boost=3,site=webtorrent.io
+$boost=3,site=animagraffs.com
+$boost=3,site=coinbase.com
+$boost=3,site=lifehacker.com
+$boost=3,site=blog.mongodb.org
+$boost=3,site=lobe.ai
+$boost=3,site=blogs.ams.org
+$boost=3,site=psychologicalscience.org
+$boost=3,site=autonews.com
+$boost=3,site=google-styleguide.googlecode.com
+$boost=3,site=venomousporridge.com
+$boost=3,site=bennorthrop.com
+$boost=3,site=blog.lusis.org
+$boost=3,site=defcon.org
+$boost=3,site=railstips.org
+$boost=3,site=johnwdefeo.com
+$boost=3,site=metabase.com
+$boost=3,site=jordansmith.io
+$boost=3,site=blog.brokep.com
+$boost=3,site=davnicwil.com
+$boost=3,site=daverupert.com
+$boost=3,site=internalpointers.com
+$boost=3,site=blog.scottnonnenberg.com
+$boost=3,site=webmasters.googleblog.com
+$boost=3,site=inference.vc
+$boost=3,site=white.ucc.asn.au
+$boost=3,site=dis.4chan.org
+$boost=3,site=mta.openssl.org
+$boost=3,site=blog.sourcerer.io
+$boost=3,site=techtransparencyproject.org
+$boost=3,site=agfundernews.com
+$boost=3,site=centauri-dreams.org
+$boost=3,site=portswigger.net
+$boost=3,site=tagide.com
+$boost=3,site=flak.tedunangst.com
+$boost=3,site=ferrous-systems.com
+$boost=3,site=shapecatcher.com
+$boost=3,site=solovyov.net
+$boost=3,site=blog.pmarca.com
+$boost=3,site=benfrederickson.com
+$boost=3,site=maps.stamen.com
+$boost=3,site=blog.pyston.org
+$boost=3,site=blog.freshdesk.com
+$boost=3,site=radionz.co.nz
+$boost=3,site=stephenmann.io
+$boost=3,site=berkeleyside.com
+$boost=3,site=engineering.zenpayroll.com
+$boost=3,site=nadiaeghbal.com
+$boost=3,site=chartio.com
+$boost=3,site=slidetocode.com
+$boost=3,site=mic.com
+$boost=3,site=ayende.com
+$boost=3,site=sixrevisions.com
+$boost=3,site=ligo.caltech.edu
+$boost=3,site=insider.com
+$boost=3,site=grahamc.com
+$boost=3,site=sbcl.org
+$boost=3,site=codingvc.com
+$boost=3,site=blog.buoyant.io
+$boost=3,site=thisamericanlife.org
+$boost=3,site=livescience.com
+$boost=3,site=blog.udacity.com
+$boost=3,site=mint.com
+$boost=3,site=quantstart.com
+$boost=3,site=i3wm.org
+$boost=3,site=lugassy.net
+$boost=3,site=renato.athaydes.com
+$boost=3,site=john.ankarstrom.se
+$boost=3,site=deis.com
+$boost=3,site=blog.travis-ci.com
+$boost=3,site=akshat.posterous.com
+$boost=3,site=blog.danieljanus.pl
+$boost=3,site=lambdassociates.org
+$boost=3,site=paypal-engineering.com
+$boost=3,site=macdaddy.io
+$boost=3,site=homepages.inf.ed.ac.uk
+$boost=3,site=isocpp.org
+$boost=3,site=secjuice.com
+$boost=3,site=martinvigo.com
+$boost=3,site=osblog.stephenmarz.com
+$boost=3,site=business.financialpost.com
+$boost=3,site=thetech.com
+$boost=3,site=harvardlawreview.org
+$boost=3,site=leafo.net
+$boost=3,site=ckeditor.com
+$boost=3,site=topgear.com
+$boost=3,site=thebaffler.com
+$boost=3,site=wilwheaton.net
+$boost=3,site=andyadams.org
+$boost=3,site=applidium.com
+$boost=3,site=techspecs.blog
+$boost=3,site=cafbit.com
+$boost=3,site=jacobinmag.com
+$boost=3,site=hueniverse.com
+$boost=3,site=tent.io
+$boost=3,site=blog.caranddriver.com
+$boost=3,site=avweb.com
+$boost=3,site=kieranhealy.org
+$boost=3,site=webrtchacks.com
+$boost=3,site=penzba.co.uk
+$boost=3,site=onbeing.org
+$boost=3,site=makeartwithpython.com
+$boost=3,site=storagemojo.com
+$boost=3,site=chrisnorstrom.com
+$boost=3,site=mir.aculo.us
+$boost=3,site=automatetheboringstuff.com
+$boost=3,site=doc.cat-v.org
+$boost=3,site=azulsystems.com
+$boost=3,site=shopify.engineering
+$boost=3,site=windy.com
+$boost=3,site=mygengo.com
+$boost=3,site=microship.com
+$boost=3,site=isro.gov.in
+$boost=3,site=amitness.com
+$boost=3,site=journal.neilgaiman.com
+$boost=3,site=blog.longnow.org
+$boost=3,site=developers.lyst.com
+$boost=3,site=threads2.scripting.com
+$boost=3,site=intego.com
+$boost=3,site=abc7news.com
+$boost=3,site=ac.els-cdn.com
+$boost=3,site=ooni.torproject.org
+$boost=3,site=rodneybrooks.com
+$boost=3,site=elie.net
+$boost=3,site=metrics.torproject.org
+$boost=3,site=scss.tcd.ie
+$boost=3,site=zdziarski.com
+$boost=3,site=blog.lizzie.io
+$boost=3,site=nntp.perl.org
+$boost=3,site=skorks.com
+$boost=3,site=lastweekinaws.com
+$boost=3,site=projectgus.com
+$boost=3,site=thedailymuse.com
+$boost=3,site=kernelmag.dailydot.com
+$boost=3,site=audubon.org
+$boost=3,site=juretriglav.si
+$boost=3,site=ias.edu
+$boost=3,site=startupsopensourced.com
+$boost=3,site=scott.a16z.com
+$boost=3,site=webdistortion.com
+$boost=3,site=my.matterport.com
+$boost=3,site=world.hey.com
+$boost=3,site=freenetproject.org
+$boost=3,site=cryptologie.net
+$boost=3,site=ewherry.com
+$boost=3,site=aaronklein.com
+$boost=3,site=metachris.com
+$boost=3,site=info.ucl.ac.be
+$boost=3,site=maritime-executive.com
+$boost=3,site=nationalpost.com
+$boost=3,site=stu.mp
+$boost=3,site=lensrentals.com
+$boost=3,site=programmingzen.com
+$boost=3,site=applied-data.science
+$boost=3,site=2uo.de
+$boost=3,site=americanscientist.org
+$boost=3,site=vault.fbi.gov
+$boost=3,site=stripes.com
+$boost=3,site=aralbalkan.com
+$boost=3,site=washingtonian.com
+$boost=3,site=jamey.thesharps.us
+$boost=3,site=visualwebsiteoptimizer.com
+$boost=3,site=blogs.uw.edu
+$boost=3,site=blog.sqlizer.io
+$boost=3,site=workshop.chromeexperiments.com
+$boost=3,site=jamieweb.net
+$boost=3,site=royal.pingdom.com
+$boost=3,site=askamanager.org
+$boost=3,site=briarproject.org
+$boost=3,site=everythingsysadmin.com
+$boost=3,site=serokell.io
+$boost=3,site=ruslanspivak.com
+$boost=3,site=wesmckinney.com
+$boost=3,site=blog.mattbierner.com
+$boost=3,site=asimovinstitute.org
+$boost=3,site=common-lisp.net
+$boost=3,site=u-tokyo.ac.jp
+$boost=3,site=dailykos.com
+$boost=3,site=pagerduty.com
+$boost=3,site=austinkleon.com
+$boost=3,site=blog.last.fm
+$boost=3,site=blog.clearbit.com
+$boost=3,site=blog.thinkst.com
+$boost=3,site=hivelogic.com
+$boost=3,site=hechingerreport.org
+$boost=3,site=fortressofdoors.com
+$boost=3,site=kushagragour.in
+$boost=3,site=ultimatehackingkeyboard.com
+$boost=3,site=rushter.com
+$boost=3,site=tarsnap.com
+$boost=3,site=wtsp.com
+$boost=3,site=elementary.io
+$boost=3,site=destevez.net
+$boost=3,site=gentoo.org
+$boost=3,site=nbviewer.jupyter.org
+$boost=3,site=blog.angular.io
+$boost=3,site=techrepublic.com
+$boost=3,site=virgin.com
+$boost=3,site=blog.getspool.com
+$boost=3,site=lucidchart.com
+$boost=3,site=blog.readability.com
+$boost=3,site=narrative.ly
+$boost=3,site=davidgerard.co.uk
+$boost=3,site=tldp.org
+$boost=3,site=chillingeffects.org
+$boost=3,site=mattbruenig.com
+$boost=3,site=engineering.khanacademy.org
+$boost=3,site=aurellem.org
+$boost=3,site=mrmoneymustache.com
+$boost=3,site=themetricsystem.rjmetrics.com
+$boost=3,site=popey.com
+$boost=3,site=datapointed.net
+$boost=3,site=net.in.tum.de
+$boost=3,site=smbc-comics.com
+$boost=3,site=blog.gittip.com
+$boost=3,site=rmurphey.com
+$boost=3,site=tuhs.org
+$boost=3,site=istartedsomething.com
+$boost=3,site=ilyabirman.net
+$boost=3,site=amix.dk
+$boost=3,site=insidehighered.com
+$boost=3,site=actu.epfl.ch
+$boost=3,site=buyforlife.com
+$boost=3,site=staysaasy.com
+$boost=3,site=okayhq.com
+$boost=3,site=engineering.mixpanel.com
+$boost=3,site=netresec.com
+$boost=3,site=jott.live
+$boost=3,site=plot.ly
+$boost=3,site=edwardtufte.com
+$boost=3,site=grack.com
+$boost=3,site=aeonmagazine.com
+$boost=3,site=lowlevelbits.org
+$boost=3,site=blog.dotcloud.com
+$boost=3,site=robwalling.com
+$boost=3,site=luajit.org
+$boost=3,site=mccormick.cx
+$boost=3,site=usehaystack.io
+$boost=3,site=supermemo.guru
+$boost=3,site=mgba.io
+$boost=3,site=forum.nim-lang.org
+$boost=3,site=sympa.inria.fr
+$boost=3,site=lists.nongnu.org
+$boost=3,site=ds9a.nl
+$boost=3,site=thinkpol.ca
+$boost=3,site=nrc.nl
+$boost=3,site=faa.gov
+$boost=3,site=smerity.com
+$boost=3,site=omaha.com
+$boost=3,site=wweek.com
+$boost=3,site=stroustrup.com
+$boost=3,site=heraldsun.com.au
+$boost=3,site=wiki.alopex.li
+$boost=3,site=dartlang.org
+$boost=3,site=cyclingnews.com
+$boost=3,site=minix3.org
+$boost=3,site=cryptomuseum.com
+$boost=3,site=blog.sendhub.com
+$boost=3,site=web.williams.edu
+$boost=3,site=sourceware.org
+$boost=3,site=blog.revolutionanalytics.com
+$boost=3,site=blog.feabhas.com
+$boost=3,site=rubenerd.com
+$boost=3,site=longnow.org
+$boost=3,site=kevinslin.com
+$boost=3,site=testing.googleblog.com
+$boost=3,site=merbist.com
+$boost=3,site=romancortes.com
+$boost=3,site=ottverse.com
+$boost=3,site=panopticlick.eff.org
+$boost=3,site=presslabs.com
+$boost=3,site=archaeology.org
+$boost=3,site=qt.io
+$boost=3,site=rust-lang.org
+$boost=3,site=komonews.com
+$boost=3,site=redecentralize.org
+$boost=3,site=spencerdailey.com
+$boost=3,site=ryanbigg.com
+$boost=3,site=codeascraft.com
+$boost=3,site=davidsimon.com
+$boost=3,site=nodered.org
+$boost=3,site=releases.llvm.org
+$boost=3,site=blog.npmjs.org
+$boost=3,site=nothings.org
+$boost=3,site=minecraft.net
+$boost=3,site=nber.org
+$boost=3,site=docs.keydb.dev
+$boost=3,site=startuplessonslearned.com
+$boost=3,site=rtpg.co
+$boost=3,site=c9.io
+$boost=3,site=cs.tufts.edu
+$boost=3,site=carlschwan.eu
+$boost=3,site=disneyresearch.com
+$boost=3,site=kilianvalkhof.com
+$boost=3,site=easylang.online
+$boost=3,site=simonschreibt.de
+$boost=3,site=answers.onstartups.com
+$boost=3,site=casetext.com
+$boost=3,site=thekidshouldseethis.com
+$boost=3,site=blog.bittorrent.com
+$boost=3,site=blog.deepgram.com
+$boost=3,site=grooveshark.com
+$boost=3,site=matthewrayfield.com
+$boost=3,site=blog.merovius.de
+$boost=3,site=michael-noll.com
+$boost=3,site=detroitnews.com
+$boost=3,site=chromestatus.com
+$boost=3,site=spacy.io
+$boost=3,site=blog.theoldreader.com
+$boost=3,site=larrysummers.com
+$boost=3,site=devo.ps
+$boost=3,site=jackg.org
+$boost=3,site=leancrew.com
+$boost=3,site=blog.smartdec.net
+$boost=3,site=aaronparecki.com
+$boost=3,site=eandt.theiet.org
+$boost=3,site=lennysnewsletter.com
+$boost=3,site=blog.factual.com
+$boost=3,site=archlinux.me
+$boost=3,site=cglab.ca
+$boost=3,site=blog.jpalardy.com
+$boost=3,site=en.yna.co.kr
+$boost=3,site=linux.slashdot.org
+$boost=3,site=trevormckendrick.com
+$boost=3,site=cmcenroe.me
+$boost=3,site=robertovitillo.com
+$boost=3,site=metacademy.org
+$boost=3,site=lethain.com
+$boost=3,site=severe-weather.eu
+$boost=3,site=federalreserve.gov
+$boost=3,site=passcode.csmonitor.com
+$boost=3,site=chrismorgan.info
+$boost=3,site=kurzweilai.net
+$boost=3,site=federalregister.gov
+$boost=3,site=carolchen.me
+$boost=3,site=limi.net
+$boost=3,site=sijmen.ruwhof.net
+$boost=3,site=winaero.com
+$boost=3,site=techquila.co.in
+$boost=3,site=theamericanconservative.com
+$boost=3,site=marijnhaverbeke.nl
+$boost=3,site=swizec.com
+$boost=3,site=thebestpageintheuniverse.net
+$boost=3,site=epsilontheory.com
+$boost=3,site=exurbe.com
+$boost=3,site=motortrend.com
+$boost=3,site=blog.seattlepi.com
+$boost=3,site=enlightenment.org
+$boost=3,site=seriouseats.com
+$boost=3,site=appleinsider.com
+$boost=3,site=holistics.io
+$boost=3,site=dannyguo.com
+$boost=3,site=letsworkshop.com
+$boost=3,site=androidauthority.com
+$boost=3,site=visalist.io
+$boost=3,site=makelinux.net
+$boost=3,site=avikdas.com
+$boost=3,site=wiki.freebsd.org
+$boost=3,site=pouet.net
+$boost=3,site=blocksandfiles.com
+$boost=3,site=docracy.com
+$boost=3,site=useit.com
+$boost=3,site=sapiens.org
+$boost=3,site=ccn.com
+$boost=3,site=mattblodgett.com
+$boost=3,site=france24.com
+$boost=3,site=zacharyburt.com
+$boost=3,site=congress.gov
+$boost=3,site=newsroom.mastercard.com
+$boost=3,site=youexec.com
+$boost=3,site=pragmaticpineapple.com
+$boost=3,site=ruder.io
+$boost=3,site=cpomagazine.com
+$boost=3,site=fontsinuse.com
+$boost=3,site=uscis.gov
+$boost=3,site=movio.co
+$boost=3,site=justwatch.com
+$boost=3,site=seeedstudio.com
+$boost=3,site=journal.dedasys.com
+$boost=3,site=css-tricks.com
+$boost=3,site=radian.org
+$boost=3,site=futurism.com
+$boost=3,site=brandon.invergo.net
+$boost=3,site=madebymike.com.au
+$boost=3,site=seomoz.org
+$boost=3,site=blog.invisiblethings.org
+$boost=3,site=greghendershott.com
+$boost=3,site=frankchimero.com
+$boost=3,site=dodgycoder.net
+$boost=3,site=txt.fyi
+$boost=3,site=cisa.gov
+$boost=3,site=failory.com
+$boost=3,site=jamanetwork.com
+$boost=3,site=objc.io
+$boost=3,site=duo.com
+$boost=3,site=futurice.com
+$boost=3,site=sifted.eu
+$boost=3,site=laravel-news.com
+$boost=3,site=archagon.net
+$boost=3,site=tinyletter.com
+$boost=3,site=blog.scoutapp.com
+$boost=3,site=tomorrowcorporation.com
+$boost=3,site=blog.scottlogic.com
+$boost=3,site=vincit.fi
+$boost=3,site=hiringlab.org
+$boost=3,site=adioso.com
+$boost=3,site=intelligence.org
+$boost=3,site=musicmachinery.com
+$boost=3,site=spaceref.com
+$boost=3,site=fantasticanachronism.com
+$boost=3,site=calculist.org
+$boost=3,site=lizthedeveloper.com
+$boost=3,site=elliotjaystocks.com
+$boost=3,site=julian.com
+$boost=3,site=garbagecollected.org
+$boost=3,site=good.is
+$boost=3,site=elidedbranches.com
+$boost=3,site=people.gnome.org
+$boost=3,site=goteleport.com
+$boost=3,site=rome.ro
+$boost=3,site=molily.de
+$boost=3,site=douglas-self.com
+$boost=3,site=rubymotion.com
+$boost=3,site=shubhro.com
+$boost=3,site=blog.jim-nielsen.com
+$boost=3,site=deseret.com
+$boost=3,site=dallasnews.com
+$boost=3,site=purecss.io
+$boost=3,site=lowercasecapital.com
+$boost=3,site=blog.getpaint.net
+$boost=3,site=futureofcoding.org
+$boost=3,site=ocregister.com
+$boost=3,site=materialize.com
+$boost=3,site=teddziuba.com
+$boost=3,site=thelocal.de
+$boost=3,site=webaim.org
+$boost=3,site=firstthings.com
+$boost=3,site=yodaiken.com
+$boost=3,site=goodcheapandfast.com
+$boost=3,site=shirky.com
+$boost=3,site=baltimoresun.com
+$boost=3,site=greaterwrong.com
+$boost=3,site=bastian.rieck.me
+$boost=3,site=toastytech.com
+$boost=3,site=raw.githubusercontent.com
+$boost=3,site=wealthsimple.com
+$boost=3,site=sanfrancisco.cbslocal.com
+$boost=3,site=blog.palantir.com
+$boost=3,site=learngitbranching.js.org
+$boost=3,site=arr.am
+$boost=3,site=i.blackhat.com
+$boost=3,site=lolware.net
+$boost=3,site=sherylcanter.com
+$boost=3,site=andrewgelman.com
+$boost=3,site=pouchdb.com
+$boost=3,site=compose.io
+$boost=3,site=work.qz.com
+$boost=3,site=news.itsfoss.com
+$boost=3,site=wiki.nikitavoloboev.xyz
+$boost=3,site=antigonejournal.com
+$boost=3,site=spectator.us
+$boost=3,site=stressgrid.com
+$boost=3,site=forums.tigsource.com
+$boost=3,site=thenewstack.io
+$boost=3,site=blog.adamsmith.cc
+$boost=3,site=gallup.com
+$boost=3,site=institutionalinvestor.com
+$boost=3,site=lowrisc.org
+$boost=3,site=jshakespeare.com
+$boost=3,site=nasaspaceflight.com
+$boost=3,site=emacsconf.org
+$boost=3,site=scottporad.com
+$boost=3,site=cam.ly
+$boost=3,site=en.wikibooks.org
+$boost=3,site=boredpanda.com
+$boost=3,site=adam.heroku.com
+$boost=3,site=transportevolved.com
+$boost=3,site=jamesfriend.com.au
+$boost=3,site=kennethreitz.com
+$boost=3,site=blog.chewxy.com
+$boost=3,site=sfexaminer.com
+$boost=3,site=hitchdev.com
+$boost=3,site=airlineratings.com
+$boost=3,site=blog.dustincurtis.com
+$boost=3,site=andymatuschak.org
+$boost=3,site=cs.bham.ac.uk
+$boost=3,site=robertputt.co.uk
+$boost=3,site=miamiherald.com
+$boost=3,site=blog.fitbit.com
+$boost=3,site=blog.codeship.com
+$boost=3,site=attn.com
+$boost=3,site=paleotronic.com
+$boost=3,site=cuddly-octo-palm-tree.com
+$boost=3,site=flownet.com
+$boost=3,site=sacrideo.us
+$boost=3,site=marmelab.com
+$boost=3,site=browser.geekbench.com
+$boost=3,site=blairreeves.me
+$boost=3,site=thesquareplanet.com
+$boost=3,site=thesephist.com
+$boost=3,site=freenas.org
+$boost=3,site=kwokchain.com
+$boost=3,site=bostinnovation.com
+$boost=3,site=love2d.org
+$boost=3,site=en.people.cn
+$boost=3,site=blog.sucuri.net
+$boost=3,site=tink.uk
+$boost=3,site=developer.palm.com
+$boost=3,site=inessential.com
+$boost=3,site=blog.qt.digia.com
+$boost=3,site=cssdeck.com
+$boost=3,site=medium.freecodecamp.org
+$boost=3,site=philly.com
+$boost=3,site=nat.org
+$boost=3,site=architecturaldigest.com
+$boost=3,site=adactio.com
+$boost=3,site=anvil.works
+$boost=3,site=arduino.cc
+$boost=3,site=blog.embed.ly
+$boost=3,site=anyvite.com
+$boost=3,site=eater.com
+$boost=3,site=moxie.org
+$boost=3,site=jacquescorbytuech.com
+$boost=3,site=n8n.io
+$boost=3,site=commonsware.com
+$boost=3,site=snoyman.com
+$boost=3,site=blog.patricktriest.com
+$boost=3,site=leaf.dragonflybsd.org
+$boost=3,site=gamedeveloper.com
+$boost=3,site=decisionsciencenews.com
+$boost=3,site=marinetraffic.com
+$boost=3,site=dallasobserver.com
+$boost=3,site=consumeraffairs.com
+$boost=3,site=english.aljazeera.net
+$boost=3,site=eng.paxos.com
+$boost=3,site=scobleizer.com
+$boost=3,site=endler.dev
+$boost=3,site=assemblyai.com
+$boost=3,site=spaceweatherarchive.com
+$boost=3,site=todayifoundout.com
+$boost=3,site=itprotoday.com
+$boost=3,site=lifespan.io
+$boost=3,site=pspdfkit.com
+$boost=3,site=slackware.com
+$boost=3,site=imba.io
+$boost=3,site=blog.wizsec.jp
+$boost=3,site=neal.fun
+$boost=3,site=shift.newco.co
+$boost=3,site=c.learncodethehardway.org
+$boost=3,site=hosted.ap.org
+$boost=3,site=onebigfluke.com
+$boost=3,site=getmagicnow.com
+$boost=3,site=projects.thestar.com
+$boost=3,site=ralfj.de
+$boost=3,site=vantage.sh
+$boost=3,site=linuxfoundation.org
+$boost=3,site=usefyi.com
+$boost=3,site=bingocardcreator.com
+$boost=3,site=blog.robenkleene.com
+$boost=3,site=mako.cc
+$boost=3,site=blog.frontapp.com
+$boost=3,site=virtuallyfun.com
+$boost=3,site=csmonitor.com
+$boost=3,site=nintendo.co.uk
+$boost=3,site=codon.com
+$boost=3,site=sixthtone.com
+$boost=3,site=selectedintelligence.com
+$boost=3,site=elasticsearch.org
+$boost=3,site=dwheeler.com
+$boost=3,site=kim.com
+$boost=3,site=nbcsandiego.com
+$boost=3,site=paragonie.com
+$boost=3,site=bostonherald.com
+$boost=3,site=mathwithbaddrawings.com
+$boost=3,site=openshot.org
+$boost=3,site=blog.aerofs.com
+$boost=3,site=chicagomag.com
+$boost=3,site=blogs.embarcadero.com
+$boost=3,site=ispreview.co.uk
+$boost=3,site=twistlock.com
+$boost=3,site=lukew.com
+$boost=3,site=news.efinancialcareers.com
+$boost=3,site=travelandleisure.com
+$boost=3,site=blog.okturtles.com
+$boost=3,site=ilsr.org
+$boost=3,site=usb.org
+$boost=3,site=dish.andrewsullivan.com
+$boost=3,site=likewise.am
+$boost=3,site=blog.idonethis.com
+$boost=3,site=blog.talosintelligence.com
+$boost=3,site=garrysub.posterous.com
+$boost=3,site=insinuator.net
+$boost=3,site=hypothes.is
+$boost=3,site=cdt.org
+$boost=3,site=herman.bearblog.dev
+$boost=3,site=archiveteam.org
+$boost=3,site=seths.blog
+$boost=3,site=blog.internetcases.com
+$boost=3,site=olifro.st
+$boost=3,site=commerce.gov
+$boost=3,site=cosmosmagazine.com
+$boost=3,site=onemileatatime.com
+$boost=3,site=blog.cardiogr.am
+$boost=3,site=nsf.gov
+$boost=3,site=michaelfeathers.silvrback.com
+$boost=3,site=cardiff.ac.uk
+$boost=3,site=stuartkhall.com
+$boost=3,site=ablogtowatch.com
+$boost=3,site=submarinecablemap.com
+$boost=3,site=blog.pandora.com
+$boost=3,site=undark.org
+$boost=3,site=indiestatik.com
+$boost=3,site=ldeming.com
+$boost=3,site=shadertoy.com
+$boost=3,site=blog.forecast.io
+$boost=3,site=thumbtack.com
+$boost=3,site=blog.yoshuawuyts.com
+$boost=3,site=meaningness.com
+$boost=3,site=climatecentral.org
+$boost=3,site=solarianprogrammer.com
+$boost=3,site=allenpike.com
+$boost=3,site=prometheus.io
+$boost=3,site=zocalopublicsquare.org
+$boost=3,site=jsonresume.org
+$boost=3,site=ts.data61.csiro.au
+$boost=3,site=downloadsquad.switched.com
+$boost=3,site=chris-said.io
+$boost=3,site=eurekalert.org
+$boost=3,site=nav.al
+$boost=3,site=contextis.com
+$boost=3,site=alpha.app.net
+$boost=3,site=simon.shimmerproject.org
+$boost=3,site=blog.andyet.com
+$boost=3,site=harvardmagazine.com
+$boost=3,site=montrealgazette.com
+$boost=3,site=ossblog.org
+$boost=3,site=snazz.xyz
+$boost=3,site=securethoughts.com
+$boost=3,site=dcrainmaker.com
+$boost=3,site=cyberscoop.com
+$boost=3,site=engineering.imvu.com
+$boost=3,site=zenodo.org
+$boost=3,site=nczonline.net
+$boost=3,site=awealthofcommonsense.com
+$boost=3,site=jcp.org
+$boost=3,site=tabletmag.com
+$boost=3,site=spritzinc.com
+$boost=3,site=pl-enthusiast.net
+$boost=3,site=blog.rtwilson.com
+$boost=3,site=parislemon.com
+$boost=3,site=perimeterx.com
+$boost=3,site=blog.ndepend.com
+$boost=3,site=gregreda.com
+$boost=3,site=cooper.com
+$boost=3,site=core77.com
+$boost=3,site=paraschopra.com
+$boost=3,site=openspace.sfmoma.org
+$boost=3,site=blog.stephenmarz.com
+$boost=3,site=blog.workshape.io
+$boost=3,site=ngrok.com
+$boost=3,site=code.flickr.net
+$boost=3,site=superpowered.com
+$boost=3,site=cs.tau.ac.il
+$boost=3,site=excamera.com
+$boost=3,site=bjorn.tipling.com
+$boost=3,site=people.idsia.ch
+$boost=3,site=olark.com
+$boost=3,site=blog.sigplan.org
+$boost=3,site=volokh.com
+$boost=3,site=betterprogramming.pub
+$boost=3,site=consumerreports.org
+$boost=3,site=spin.atomicobject.com
+$boost=3,site=opencv.org
+$boost=3,site=blog.imgtec.com
+$boost=3,site=pauladamsmith.com
+$boost=3,site=neosmart.net
+$boost=3,site=reprints.longform.org
+$boost=3,site=xda-developers.com
+$boost=3,site=reefpoints.dockyard.com
+$boost=3,site=benguild.com
+$boost=3,site=naughtycomputer.uk
+$boost=3,site=kb.cert.org
+$boost=3,site=insidescience.org
+$boost=3,site=op-co.de
+$boost=3,site=roadsandkingdoms.com
+$boost=3,site=netbsd.org
+$boost=3,site=lite.cnn.io
+$boost=3,site=redeye.firstround.com
+$boost=3,site=forum.xda-developers.com
+$boost=3,site=posthog.com
+$boost=3,site=ftp.iza.org
+$boost=3,site=craphound.com
+$boost=3,site=rijksmuseum.nl
+$boost=3,site=hipmunk.com
+$boost=3,site=fuchsia.dev
+$boost=3,site=infrequently.org
+$boost=3,site=nationalreview.com
+$boost=3,site=futilitycloset.com
+$boost=3,site=hayabusa2.jaxa.jp
+$boost=3,site=malisper.me
+$boost=3,site=blog.jwhitham.org
+$boost=3,site=bemmu.posterous.com
+$boost=3,site=gizmodo.com.au
+$boost=3,site=meyerweb.com
+$boost=3,site=novaspivack.com
+$boost=3,site=pjrc.com
+$boost=3,site=get.slack.help
+$boost=3,site=mwi.usma.edu
+$boost=3,site=aaronstannard.com
+$boost=3,site=messynessychic.com
+$boost=3,site=defectivebydesign.org
+$boost=3,site=bash.org
+$boost=3,site=venturehacks.com
+$boost=3,site=thepublicdiscourse.com
+$boost=3,site=browser.engineering
+$boost=3,site=peachesnstink.com
+$boost=3,site=cs.utah.edu
+$boost=3,site=inkstonenews.com
+$boost=3,site=privacytools.io
+$boost=3,site=brianknapp.me
+$boost=3,site=mediaite.com
+$boost=3,site=theinquirer.net
+$boost=3,site=dhruvonmath.com
+$boost=3,site=lucb1e.com
+$boost=3,site=hetzner.de
+$boost=3,site=davemart.in
+$boost=3,site=cryptopals.com
+$boost=3,site=paperswithcode.com
+$boost=3,site=blog.hipmunk.com
+$boost=3,site=citizen428.net
+$boost=3,site=shitpost.plover.com
+$boost=3,site=lispcast.com
+$boost=3,site=webofstories.com
+$boost=3,site=news.northeastern.edu
+$boost=3,site=digitalmars.com
+$boost=3,site=mysk.blog
+$boost=3,site=janestreet.com
+$boost=3,site=francescodilorenzo.com
+$boost=3,site=binghamton.edu
+$boost=3,site=blog.businessofsoftware.org
+$boost=3,site=dougengelbart.org
+$boost=3,site=tkainrad.dev
+$boost=3,site=machinelearningplus.com
+$boost=3,site=200ok.ch
+$boost=3,site=singlelunch.com
+$boost=3,site=chairnerd.seatgeek.com
+$boost=3,site=govtech.com
+$boost=3,site=news.mongabay.com
+$boost=3,site=erikmcclure.com
+$boost=3,site=pyimagesearch.com
+$boost=3,site=thejournal.ie
+$boost=3,site=christophermeiklejohn.com
+$boost=3,site=csiro.au
+$boost=3,site=maxmasnick.com
+$boost=3,site=bitmatica.com
+$boost=3,site=bbs.archlinux.org
+$boost=3,site=milapneupane.com.np
+$boost=3,site=lists.randombit.net
+$boost=3,site=tyrrrz.me
+$boost=3,site=avdi.org
+$boost=3,site=tuxphones.com
+$boost=3,site=wolfstreet.com
+$boost=3,site=klarasystems.com
+$boost=3,site=blog.leahhanson.us
+$boost=3,site=lua-users.org
+$boost=3,site=blog.tjll.net
+$boost=3,site=massivelyop.com
+$boost=3,site=kuow.org
+$boost=3,site=chinafile.com
+$boost=3,site=allaboutcircuits.com
+$boost=3,site=physorg.com
+$boost=3,site=altdevblogaday.com
+$boost=3,site=codeahoy.com
+$boost=3,site=perl.plover.com
+$boost=3,site=wpri.com
+$boost=3,site=brianchristner.io
+$boost=3,site=yeoman.io
+$boost=3,site=structuredprocrastination.com
+$boost=3,site=galactanet.com
+$boost=3,site=alexmuir.com
+$boost=3,site=gapingvoid.com
+$boost=3,site=doisinkidney.com
+$boost=3,site=jeffknupp.com
+$boost=3,site=weblogs.asp.net
+$boost=3,site=eclipse.org
+$boost=3,site=gmo.com
+$boost=3,site=h14s.p5r.org
+$boost=3,site=privatdozent.co
+$boost=3,site=conorpp.com
+$boost=3,site=dev-blog.apollodata.com
+$boost=3,site=bof.nl
+$boost=3,site=potaroo.net
+$boost=3,site=disneyanimation.com
+$boost=3,site=perspectives.mvdirona.com
+$boost=3,site=digiconomist.net
+$boost=3,site=futhark-lang.org
+$boost=3,site=blog.stephanbehnke.com
+$boost=3,site=themillions.com
+$boost=3,site=charlespetzold.com
+$boost=3,site=contrast.ie
+$boost=3,site=threatpost.com
+$boost=3,site=diginomica.com
+$boost=3,site=iro.umontreal.ca
+$boost=3,site=goodui.org
+$boost=3,site=kevinmahoney.co.uk
+$boost=3,site=den.dev
+$boost=3,site=blog.danlew.net
+$boost=3,site=insidermag.net
+$boost=3,site=eviltrout.com
+$boost=3,site=bartwronski.com
+$boost=3,site=laughtonelectronics.com
+$boost=3,site=toomas.net
+$boost=3,site=peoplespolicyproject.org
+$boost=3,site=kpcb.com
+$boost=3,site=mcsweeneys.net
+$boost=3,site=weeklystandard.com
+$boost=3,site=therideshareguy.com
+$boost=3,site=evhead.com
+$boost=3,site=praeclarum.org
+$boost=3,site=rs-online.com
+$boost=3,site=blog.linuxmint.com
+$boost=3,site=blog.soylent.me
+$boost=3,site=cabel.me
+$boost=3,site=wildbit.com
+$boost=3,site=zenhabits.net
+$boost=3,site=devblog.avdi.org
+$boost=3,site=blog.phpfog.com
+$boost=3,site=thetransportpolitic.com
+$boost=3,site=forestry.io
+$boost=3,site=dabeaz.com
+$boost=3,site=brightball.com
+$boost=3,site=style.org
+$boost=3,site=elezea.com
+$boost=3,site=aarp.org
+$boost=3,site=ottawacitizen.com
+$boost=3,site=micropython.org
+$boost=3,site=metamorphblog.com
+$boost=3,site=blog.jez.io
+$boost=3,site=aatishb.com
+$boost=3,site=celinehh.com
+$boost=3,site=davidyat.es
+$boost=3,site=code.joejag.com
+$boost=3,site=washingtonexaminer.com
+$boost=3,site=resetera.com
+$boost=3,site=quantum.country
+$boost=3,site=ahungry.com
+$boost=3,site=elementaryos.org
+$boost=3,site=debuggex.com
+$boost=3,site=planetizen.com
+$boost=3,site=eweek.com
+$boost=3,site=tatiyants.com
+$boost=3,site=discover.dc.nihr.ac.uk
+$boost=3,site=web.cs.ucdavis.edu
+$boost=3,site=xmpp.org
+$boost=3,site=electricitymap.org
+$boost=3,site=projects.propublica.org
+$boost=3,site=e-estonia.com
+$boost=3,site=blog.basho.com
+$boost=3,site=ungleich.ch
+$boost=3,site=cassandracapital.net
+$boost=3,site=aiyprojects.withgoogle.com
+$boost=3,site=jwt.io
+$boost=3,site=esri.com
+$boost=3,site=gaborcselle.com
+$boost=3,site=venam.nixers.net
+$boost=3,site=7pace.com
+$boost=3,site=joshmcguigan.com
+$boost=3,site=promarket.org
+$boost=3,site=tav.espians.com
+$boost=3,site=datastax.com
+$boost=3,site=cmyr.net
+$boost=3,site=hisutton.com
+$boost=3,site=lsub.org
+$boost=3,site=alpblog.heroku.com
+$boost=3,site=gpuopen.com
+$boost=3,site=hakaimagazine.com
+$boost=3,site=marcan.st
+$boost=3,site=benjaminreinhardt.com
+$boost=3,site=adityaravishankar.com
+$boost=3,site=alistapart.com
+$boost=3,site=msgpack.org
+$boost=3,site=keionline.org
+$boost=3,site=logicalclocks.com
+$boost=3,site=svedic.org
+$boost=3,site=newsroom.unsw.edu.au
+$boost=3,site=commandlinefu.com
+$boost=3,site=clockworkpi.com
+$boost=3,site=typography.com
+$boost=3,site=sigarch.org
+$boost=3,site=countbayesie.com
+$boost=3,site=blog.echen.me
+$boost=3,site=groovehq.com
+$boost=3,site=calteches.library.caltech.edu
+$boost=3,site=rochester.edu
+$boost=3,site=blog.bjrn.se
+$boost=3,site=lawandcrime.com
+$boost=3,site=betashop.com
+$boost=3,site=ericsink.com
+$boost=3,site=cleantechnica.com
+$boost=3,site=amiunique.org
+$boost=3,site=bvckup2.com
+$boost=3,site=ieftimov.com
+$boost=3,site=ewanvalentine.io
+$boost=3,site=secalerts.co
+$boost=3,site=patrick-wied.at
+$boost=3,site=pitt.edu
+$boost=3,site=hothardware.com
+$boost=3,site=gridstylesheets.org
+$boost=3,site=sciencex.com
+$boost=3,site=developers.slashdot.org
+$boost=3,site=blog.confluent.io
+$boost=3,site=glandium.org
+$boost=3,site=arandomurl.com
+$boost=3,site=blog.nugget.one
+$boost=3,site=rd.com
+$boost=3,site=slifty.com
+$boost=3,site=lighttable.com
+$boost=3,site=lurklurk.org
+$boost=3,site=derstandard.at
+$boost=3,site=lvh.io
+$boost=3,site=rs.io
+$boost=3,site=eslint.org
+$boost=3,site=cucumbertown.com
+$boost=3,site=z.cash
+$boost=3,site=nanonets.com
+$boost=3,site=crunchgear.com
+$boost=3,site=mbuffett.com
+$boost=3,site=cybergibbons.com
+$boost=3,site=lua.org
+$boost=3,site=monodraw.helftone.com
+$boost=3,site=davidgcohen.com
+$boost=3,site=jfrog.com
+$boost=3,site=johnkary.net
+$boost=3,site=jakobgreenfeld.com
+$boost=3,site=encosia.com
+$boost=3,site=martinkool.com
+$boost=3,site=use-the-index-luke.com
+$boost=3,site=siteanalytics.compete.com
+$boost=3,site=madewithlove.be
+$boost=3,site=wildlyinaccurate.com
+$boost=3,site=news.agu.org
+$boost=3,site=hnsearch.com
+$boost=3,site=thevaluable.dev
+$boost=3,site=cve.mitre.org
+$boost=3,site=r2d3.us
+$boost=3,site=wsbtv.com
+$boost=3,site=regex101.com
+$boost=3,site=fastcodesign.com
+$boost=3,site=waitbutwhy.com
+$boost=3,site=wrapbootstrap.com
+$boost=3,site=angrystaffofficer.com
+$boost=3,site=pythonprogramming.net
+$boost=3,site=codemirror.net
+$boost=3,site=erthalion.info
+$boost=3,site=news.wsu.edu
+$boost=3,site=jmtd.net
+$boost=3,site=businessinsider.com.au
+$boost=3,site=keithschwarz.com
+$boost=3,site=blog.testdouble.com
+$boost=3,site=rationalconspiracy.com
+$boost=3,site=blog.superfeedr.com
+$boost=3,site=hn-books.com
+$boost=3,site=techsolidarity.org
+$boost=3,site=annualletter.gatesfoundation.org
+$boost=3,site=ardanlabs.com
+$boost=3,site=protonvpn.com
+$boost=3,site=winrumors.com
+$boost=3,site=arcticstartup.com
+$boost=3,site=blogs.akamai.com
+$boost=3,site=howitactuallyworks.com
+$boost=3,site=theage.com.au
+$boost=3,site=blogs.computerworld.com
+$boost=3,site=news.artnet.com
+$boost=3,site=yongfook.com
+$boost=3,site=ask.metafilter.com
+$boost=3,site=blog.zencoder.com
+$boost=3,site=ashfurrow.com
+$boost=3,site=rachelandrew.co.uk
+$boost=3,site=natureofcode.com
+$boost=3,site=jnd.org
+$boost=3,site=cryptoslate.com
+$boost=3,site=electricliterature.com
+$boost=3,site=cs.toronto.edu
+$boost=3,site=laughingsquid.com
+$boost=3,site=news.netcraft.com
+$boost=3,site=mattfriz.com
+$boost=3,site=blogs.computerworlduk.com
+$boost=3,site=weblog.jamisbuck.org
+$boost=3,site=particletree.com
+$boost=3,site=loopinsight.com
+$boost=3,site=blog.qualys.com
+$boost=3,site=moz.com
+$boost=3,site=chalmers.se
+$boost=3,site=homepage.ntlworld.com
+$boost=3,site=jpost.com
+$boost=3,site=teenvogue.com
+$boost=3,site=blog.bufferapp.com
+$boost=3,site=pharr.org
+$boost=3,site=cforcoding.com
+$boost=3,site=mixpanel.com
+$boost=3,site=thevintagenews.com
+$boost=3,site=bdickason.com
+$boost=3,site=immunant.com
+$boost=3,site=madhadron.com
+$boost=3,site=vim.org
+$boost=3,site=blog.prototypr.io
+$boost=3,site=webpolicy.org
+$boost=3,site=coding.smashingmagazine.com
+$boost=3,site=wizardzines.com
+$boost=3,site=itnews.com.au
+$boost=3,site=codethinked.com
+$boost=3,site=technologynetworks.com
+$boost=3,site=blog.wolframalpha.com
+$boost=3,site=blog.juxt.pro
+$boost=3,site=defector.com
+$boost=3,site=jamesyu.org
+$boost=3,site=dailytech.com
+$boost=3,site=linear.app
+$boost=3,site=clerk.house.gov
+$boost=3,site=keepachangelog.com
+$boost=3,site=eng.localytics.com
+$boost=3,site=emacsrocks.com
+$boost=3,site=dassur.ma
+$boost=3,site=sightline.org
+$boost=3,site=gamesradar.com
+$boost=3,site=sysdig.com
+$boost=3,site=cloudkick.com
+$boost=3,site=devopsu.com
+$boost=3,site=believermag.com
+$boost=3,site=unhandledexpression.com
+$boost=3,site=efficiencyiseverything.com
+$boost=3,site=therobotreport.com
+$boost=3,site=howtographql.com
+$boost=3,site=archive.today
+$boost=3,site=progrium.com
+$boost=3,site=redfin.engineering
+$boost=3,site=nexedi.com
+$boost=3,site=home.treasury.gov
+$boost=3,site=opensourceconnections.com
+$boost=3,site=aras-p.info
+$boost=3,site=battlepenguin.com
+$boost=3,site=solokeys.com
+$boost=3,site=alexkyte.me
+$boost=3,site=digitalmusicnews.com
+$boost=3,site=kenneth.io
+$boost=3,site=mywiki.wooledge.org
+$boost=3,site=epic.org
+$boost=3,site=behavioralscientist.org
+$boost=3,site=jack.minardi.org
+$boost=3,site=zeptobars.ru
+$boost=3,site=bramcohen.com
+$boost=3,site=frankforce.com
+$boost=3,site=erniemiller.org
+$boost=3,site=robservatory.com
+$boost=3,site=artofmanliness.com
+$boost=3,site=world-nuclear-news.org
+$boost=3,site=openmirage.org
+$boost=3,site=prism-break.org
+$boost=3,site=slab.com
+$boost=3,site=monkeylearn.com
+$boost=3,site=metaredux.com
+$boost=3,site=eurogamer.net
+$boost=3,site=wisdomination.com
+$boost=3,site=src.chromium.org
+$boost=3,site=tudelft.nl
+$boost=3,site=blog.project-insanity.org
+$boost=3,site=hadihariri.com
+$boost=3,site=blog.xot.nl
+$boost=3,site=andrewchenblog.com
+$boost=3,site=jordan-wright.com
+$boost=3,site=journal.burningman.org
+$boost=3,site=blog.tensorflow.org
+$boost=3,site=standblog.org
+$boost=3,site=ant.design
+$boost=3,site=overapi.com
+$boost=3,site=suntimes.com
+$boost=3,site=thecable.foreignpolicy.com
+$boost=3,site=bestinclass.dk
+$boost=3,site=hn.algolia.com
+$boost=3,site=gwillem.gitlab.io
+$boost=3,site=elbenshira.com
+$boost=3,site=bankunderground.co.uk
+$boost=3,site=crowdsupply.com
+$boost=3,site=lists.linuxfoundation.org
+$boost=3,site=prideout.net
+$boost=3,site=andrewoneverything.com
+$boost=3,site=peterlyons.com
+$boost=3,site=vogue.com
+$boost=3,site=future.a16z.com
+$boost=3,site=thedotpost.com
+$boost=3,site=davidwalsh.name
+$boost=3,site=heise.de
+$boost=3,site=regex.info
+$boost=3,site=blog.jayway.com
+$boost=3,site=mattmazur.com
+$boost=3,site=0x80.pl
+$boost=3,site=pine64.com
+$boost=3,site=gog.com
+$boost=3,site=jdh.hamkins.org
+$boost=3,site=openjdk.java.net
+$boost=3,site=code.djangoproject.com
+$boost=3,site=airbnb.io
+$boost=3,site=neatorama.com
+$boost=3,site=intridea.com
+$boost=3,site=fastmailstatus.com
+$boost=3,site=nas.sr
+$boost=3,site=danielbmarkham.com
+$boost=3,site=blogs.kde.org
+$boost=3,site=wordfence.com
+$boost=3,site=quariety.com
+$boost=3,site=ritholtz.com
+$boost=3,site=jasonadriaan.com
+$boost=3,site=joelcalifa.com
+$boost=3,site=cs.umd.edu
+$boost=3,site=zacharyvoase.com
+$boost=3,site=decisionproblem.com
+$boost=3,site=steamdb.info
+$boost=3,site=blog.chef.io
+$boost=3,site=techraptor.net
+$boost=3,site=nanog.org
+$boost=3,site=igoro.com
+$boost=3,site=math.toronto.edu
+$boost=3,site=tether.to
+$boost=3,site=adrianmejia.com
+$boost=3,site=adblockplus.org
+$boost=3,site=techxplore.com
+$boost=3,site=professional-troublemaker.com
+$boost=3,site=idealab.talkingpointsmemo.com
+$boost=3,site=nesslabs.com
+$boost=3,site=nationalgeographic.co.uk
+$boost=3,site=iridakos.com
+$boost=3,site=theagilepanda.com
+$boost=3,site=blog.skyliner.io
+$boost=3,site=django-cms.org
+$boost=3,site=blog.path.com
+$boost=3,site=villagevoice.com
+$boost=3,site=somehowmanage.com
+$boost=3,site=mnn.com
+$boost=3,site=abovethelaw.com
+$boost=3,site=blog.logrocket.com
+$boost=3,site=apress.com
+$boost=3,site=curbed.com
+$boost=3,site=digitizor.com
+$boost=3,site=blog.lareviewofbooks.org
+$boost=3,site=cambus.net
+$boost=3,site=blog.rize.io
+$boost=3,site=2ality.com
+$boost=3,site=thedomains.com
+$boost=3,site=examine.com
+$boost=3,site=speirs.org
+$boost=3,site=cs.brown.edu
+$boost=3,site=dnalounge.com
+$boost=3,site=nbcbayarea.com
+$boost=3,site=macperformanceguide.com
+$boost=3,site=gocardless.com
+$boost=3,site=driverless.id
+$boost=3,site=blog.l0cal.com
+$boost=3,site=labs.ripe.net
+$boost=3,site=britishlibrary.typepad.co.uk
+$boost=3,site=trustedsec.com
+$boost=3,site=nicholaswindsorhoward.com
+$boost=3,site=wefunder.com
+$boost=3,site=grpc.io
+$boost=3,site=innovationaus.com
+$boost=3,site=algorithmwatch.org
+$boost=3,site=kylewritescode.com
+$boost=3,site=doc.ic.ac.uk
+$boost=3,site=r-bloggers.com
+$boost=3,site=kxan.com
+$boost=3,site=blog.shodan.io
+$boost=3,site=blog.couchdb.org
+$boost=3,site=blogs.igalia.com
+$boost=3,site=adambard.com
+$boost=3,site=xania.org
+$boost=3,site=pinboard.in
+$boost=3,site=lobste.rs
+$boost=3,site=interestingengineering.com
+$boost=3,site=jcdav.is
+$boost=3,site=zwbetz.com
+$boost=3,site=compellingsciencefiction.com
+$boost=3,site=headmelted.com
+$boost=3,site=blog.itlater.com
+$boost=3,site=joystiq.com
+$boost=3,site=blog.loadzero.com
+$boost=3,site=exploringbinary.com
+$boost=3,site=towardsdatascience.com
+$boost=3,site=ostif.org
+$boost=3,site=cs.ucr.edu
+$boost=3,site=transportenvironment.org
+$boost=3,site=sciencenews.org
+$boost=3,site=blog.portswigger.net
+$boost=3,site=dshipper.posterous.com
+$boost=3,site=abopen.com
+$boost=3,site=realclearscience.com
+$boost=3,site=llnl.gov
+$boost=3,site=brandmark.io
+$boost=3,site=eflorenzano.com
+$boost=3,site=ethanzuckerman.com
+$boost=3,site=montulli.org
+$boost=3,site=dvorak.org
+$boost=3,site=octopart.com
+$boost=3,site=ithare.com
+$boost=3,site=isc.sans.edu
+$boost=3,site=starling-software.com
+$boost=3,site=sedimental.org
+$boost=3,site=nitrokey.com
+$boost=3,site=fool.com
+$boost=3,site=ntietz.com
+$boost=3,site=thepointmag.com
+$boost=3,site=symantec.com
+$boost=3,site=blog.adacore.com
+$boost=3,site=blog.davidtate.org
+$boost=3,site=blog.gitlab.org
+$boost=3,site=blog.dijit.sh
+$boost=3,site=boredzo.org
+$boost=3,site=bishopfox.com
+$boost=3,site=blog.sproutcore.com
+$boost=3,site=nakedcapitalism.com
+$boost=3,site=hackaday.io
+$boost=3,site=nccgroup.com
+$boost=3,site=kansas.com
+$boost=3,site=myfonts.com
+$boost=3,site=databreaches.net
+$boost=3,site=design.google
+$boost=3,site=bvp.com
+$boost=3,site=eriwen.com
+$boost=3,site=flightaware.com
+$boost=3,site=us2.campaign-archive1.com
+$boost=3,site=engineering.skroutz.gr
+$boost=3,site=arcentry.com
+$boost=3,site=redmonk.com
+$boost=3,site=jesseduffield.com
+$boost=3,site=foodandwine.com
+$boost=3,site=vandenoever.info
+$boost=3,site=tubetime.us
+$boost=3,site=xenbits.xen.org
+$boost=3,site=nofilmschool.com
+$boost=3,site=joi.ito.com
+$boost=3,site=blog.garlicsim.org
+$boost=3,site=akr.am
+$boost=3,site=nypl.org
+$boost=3,site=shmuplations.com
+$boost=3,site=blog.statwing.com
+$boost=3,site=boston.conman.org
+$boost=3,site=lilium.com
+$boost=3,site=phk.freebsd.dk
+$boost=3,site=underthehood.meltwater.com
+$boost=3,site=airliners.net
+$boost=2,site=gobankingrates.com
+$boost=2,site=maxplanck.nautil.us
+$boost=2,site=leandromoreira.com.br
+$boost=2,site=24ways.org
+$boost=2,site=federaltimes.com
+$boost=2,site=slash7.com
+$boost=2,site=projectrho.com
+$boost=2,site=pencil.evolus.vn
+$boost=2,site=newfoodeconomy.org
+$boost=2,site=geekosystem.com
+$boost=2,site=bengtan.com
+$boost=2,site=jasonlefkowitz.net
+$boost=2,site=teamtreehouse.com
+$boost=2,site=blog.quarkslab.com
+$boost=2,site=artnome.com
+$boost=2,site=blog.chron.com
+$boost=2,site=blog.close.io
+$boost=2,site=lonesysadmin.net
+$boost=2,site=lemonde.fr
+$boost=2,site=guernicamag.com
+$boost=2,site=blog.oleganza.com
+$boost=2,site=poetryfoundation.org
+$boost=2,site=archive.computerhistory.org
+$boost=2,site=philzimmermann.com
+$boost=2,site=sqlalchemy.org
+$boost=2,site=rubyinside.com
+$boost=2,site=rjbs.manxome.org
+$boost=2,site=lab.simurai.com
+$boost=2,site=shortcutfoo.com
+$boost=2,site=asymco.com
+$boost=2,site=johnmathews.eu
+$boost=2,site=blog.vivekhaldar.com
+$boost=2,site=t3.org
+$boost=2,site=coloradosun.com
+$boost=2,site=gamesindustry.biz
+$boost=2,site=benjoffe.com
+$boost=2,site=codeblog.jonskeet.uk
+$boost=2,site=mattdesl.svbtle.com
+$boost=2,site=wilshipley.com
+$boost=2,site=lunduke.com
+$boost=2,site=blog.documentcloud.org
+$boost=2,site=pratt.duke.edu
+$boost=2,site=forum.palemoon.org
+$boost=2,site=ring.cx
+$boost=2,site=healthimaging.com
+$boost=2,site=blog.lawnstarter.com
+$boost=2,site=drorpoleg.com
+$boost=2,site=abajournal.com
+$boost=2,site=akitasoftware.com
+$boost=2,site=blog.algolia.com
+$boost=2,site=cgit.freedesktop.org
+$boost=2,site=creativegood.com
+$boost=2,site=krypt.co
+$boost=2,site=david-peter.de
+$boost=2,site=markdotto.com
+$boost=2,site=tuaw.com
+$boost=2,site=jessepollak.me
+$boost=2,site=evasi0n.com
+$boost=2,site=continuum.io
+$boost=2,site=inconshreveable.com
+$boost=2,site=go350.com
+$boost=2,site=newhumanist.org.uk
+$boost=2,site=thelocal.se
+$boost=2,site=ldpreload.com
+$boost=2,site=blog.valerieaurora.org
+$boost=2,site=tympanus.net
+$boost=2,site=gcaptain.com
+$boost=2,site=fireeye.com
+$boost=2,site=mkdev.me
+$boost=2,site=thoughtmaybe.com
+$boost=2,site=blog.lukaszolejnik.com
+$boost=2,site=getmondo.co.uk
+$boost=2,site=open.bufferapp.com
+$boost=2,site=forums.macrumors.com
+$boost=2,site=scottberkun.com
+$boost=2,site=ontwik.com
+$boost=2,site=git-scm.com
+$boost=2,site=adafruit.com
+$boost=2,site=mathjax.org
+$boost=2,site=chinalawblog.com
+$boost=2,site=dw.de
+$boost=2,site=udacity.com
+$boost=2,site=dianeravitch.net
+$boost=2,site=narrator.ai
+$boost=2,site=broadbandnow.com
+$boost=2,site=theaircurrent.com
+$boost=2,site=bytecellar.com
+$boost=2,site=rifters.com
+$boost=2,site=mikewest.org
+$boost=2,site=loyaltylobby.com
+$boost=2,site=blog.hut8labs.com
+$boost=2,site=metmuseum.org
+$boost=2,site=undo.io
+$boost=2,site=news.discovery.com
+$boost=2,site=hulry.com
+$boost=2,site=greenheartgames.com
+$boost=2,site=thehindu.com
+$boost=2,site=ably.io
+$boost=2,site=chrispenner.ca
+$boost=2,site=logodesignlove.com
+$boost=2,site=neocomputer.org
+$boost=2,site=m.motherjones.com
+$boost=2,site=ro-che.info
+$boost=2,site=ofnumbers.com
+$boost=2,site=rdist.root.org
+$boost=2,site=blog.alexellis.io
+$boost=2,site=notebookcheck.net
+$boost=2,site=new.livestream.com
+$boost=2,site=sonyaellenmann.com
+$boost=2,site=builtin.com
+$boost=2,site=thecritic.co.uk
+$boost=2,site=bitworking.org
+$boost=2,site=iwantmyname.com
+$boost=2,site=circleid.com
+$boost=2,site=woboq.com
+$boost=2,site=instructables.com
+$boost=2,site=markmail.org
+$boost=2,site=inverse.com
+$boost=2,site=robinhood.com
+$boost=2,site=blog.tojicode.com
+$boost=2,site=usa.streetsblog.org
+$boost=2,site=sanderknape.com
+$boost=2,site=technical.ly
+$boost=2,site=trek10.com
+$boost=2,site=theprint.in
+$boost=2,site=000fff.org
+$boost=2,site=networkworld.com
+$boost=2,site=apps.bostonglobe.com
+$boost=2,site=pfrazee.hashbase.io
+$boost=2,site=commentarymagazine.com
+$boost=2,site=petewarden.com
+$boost=2,site=exeter.ac.uk
+$boost=2,site=statwing.com
+$boost=2,site=unenvironment.org
+$boost=2,site=paintcodeapp.com
+$boost=2,site=andreschweighofer.com
+$boost=2,site=forums.unrealengine.com
+$boost=2,site=wikidata.org
+$boost=2,site=draculatheme.com
+$boost=2,site=blog.flightcaster.com
+$boost=2,site=vancouversun.com
+$boost=2,site=cc.gatech.edu
+$boost=2,site=madebymany.com
+$boost=2,site=afp.com
+$boost=2,site=huyng.com
+$boost=2,site=jeremymorgan.com
+$boost=2,site=thereformedbroker.com
+$boost=2,site=bigocheatsheet.com
+$boost=2,site=matduggan.com
+$boost=2,site=deis.io
+$boost=2,site=poolp.org
+$boost=2,site=alteredqualia.com
+$boost=2,site=erlang-factory.com
+$boost=2,site=paranoia.dubfire.net
+$boost=2,site=cato-at-liberty.org
+$boost=2,site=netflixprize.com
+$boost=2,site=utilitydive.com
+$boost=2,site=olioapps.com
+$boost=2,site=rcrowley.org
+$boost=2,site=hastebin.com
+$boost=2,site=dotnetfoundation.org
+$boost=2,site=platformonomics.com
+$boost=2,site=alf.nu
+$boost=2,site=antoinevastel.com
+$boost=2,site=liveweave.com
+$boost=2,site=causal.app
+$boost=2,site=liw.fi
+$boost=2,site=redis.io
+$boost=2,site=collectiveidea.com
+$boost=2,site=felleisen.org
+$boost=2,site=blog.nobugware.com
+$boost=2,site=windowscentral.com
+$boost=2,site=xobni.com
+$boost=2,site=outage.report
+$boost=2,site=justsecurity.org
+$boost=2,site=blog.rfox.eu
+$boost=2,site=way-cooler.org
+$boost=2,site=adamierymenko.com
+$boost=2,site=mesosphere.com
+$boost=2,site=meatfighter.com
+$boost=2,site=dsc.discovery.com
+$boost=2,site=sciencenordic.com
+$boost=2,site=uxmovement.com
+$boost=2,site=death.andgravity.com
+$boost=2,site=staffeng.com
+$boost=2,site=godottutorials.pro
+$boost=2,site=gladwell.com
+$boost=2,site=cgpgrey.com
+$boost=2,site=source.opennews.org
+$boost=2,site=unsharpen.com
+$boost=2,site=haseebq.com
+$boost=2,site=blog.diasporafoundation.org
+$boost=2,site=pugetsystems.com
+$boost=2,site=roadmap.sh
+$boost=2,site=rawstory.com
+$boost=2,site=nylas.com
+$boost=2,site=tunetheweb.com
+$boost=2,site=subtraction.com
+$boost=2,site=adam.blog.heroku.com
+$boost=2,site=gdcvault.com
+$boost=2,site=n26.com
+$boost=2,site=passportindex.org
+$boost=2,site=sfweekly.com
+$boost=2,site=wiki.csswg.org
+$boost=2,site=remoteok.io
+$boost=2,site=guru3d.com
+$boost=2,site=wisdom.weizmann.ac.il
+$boost=2,site=fred.stlouisfed.org
+$boost=2,site=quantcast.com
+$boost=2,site=dspguide.com
+$boost=2,site=monitor.firefox.com
+$boost=2,site=manchester.ac.uk
+$boost=2,site=ecowatch.com
+$boost=2,site=exponents.co
+$boost=2,site=ubere.ng
+$boost=2,site=lists.cs.uiuc.edu
+$boost=2,site=erasmatazz.com
+$boost=2,site=prestonbyrne.com
+$boost=2,site=andyet.net
+$boost=2,site=blog.getfirebug.com
+$boost=2,site=joshtronic.com
+$boost=2,site=cat-bus.com
+$boost=2,site=oos.moxiecode.com
+$boost=2,site=dlo.me
+$boost=2,site=openbazaar.org
+$boost=2,site=tumbledry.org
+$boost=2,site=ssd.eff.org
+$boost=2,site=worldometers.info
+$boost=2,site=denverpost.com
+$boost=2,site=blog.teambox.com
+$boost=2,site=staff.science.uu.nl
+$boost=2,site=chromium-review.googlesource.com
+$boost=2,site=antoarts.com
+$boost=2,site=redcode.nl
+$boost=2,site=morningstar.com
+$boost=2,site=nomachetejuggling.com
+$boost=2,site=ericholscher.com
+$boost=2,site=nfx.com
+$boost=2,site=kara.allthingsd.com
+$boost=2,site=taptaptap.com
+$boost=2,site=rhizome.org
+$boost=2,site=c-faq.com
+$boost=2,site=ocweekly.com
+$boost=2,site=yesmagazine.org
+$boost=2,site=boston.cbslocal.com
+$boost=2,site=sethvargo.com
+$boost=2,site=teamliquid.net
+$boost=2,site=kn100.me
+$boost=2,site=carsonified.com
+$boost=2,site=makecode.com
+$boost=2,site=leejamesrobinson.com
+$boost=2,site=journals.lww.com
+$boost=2,site=philfreo.com
+$boost=2,site=fredwu.me
+$boost=2,site=desertsun.com
+$boost=2,site=blog.noctua-software.com
+$boost=2,site=elazzabi.com
+$boost=2,site=camendesign.com
+$boost=2,site=spaceflightnow.com
+$boost=2,site=cs.uic.edu
+$boost=2,site=lpalmieri.com
+$boost=2,site=call-with-current-continuation.org
+$boost=2,site=pub.gajendra.net
+$boost=2,site=write.privacytools.io
+$boost=2,site=blog.gitprime.com
+$boost=2,site=deepstreamhub.com
+$boost=2,site=fosspost.org
+$boost=2,site=staradvertiser.com
+$boost=2,site=daggle.com
+$boost=2,site=text.sourcegraph.com
+$boost=2,site=pulumi.com
+$boost=2,site=deepstream.io
+$boost=2,site=rebol.com
+$boost=2,site=oilprice.com
+$boost=2,site=itsnicethat.com
+$boost=2,site=theblackvault.com
+$boost=2,site=jeena.net
+$boost=2,site=mixnode.com
+$boost=2,site=extremelearning.com.au
+$boost=2,site=news.crunchbase.com
+$boost=2,site=the-japan-news.com
+$boost=2,site=tor.com
+$boost=2,site=graphics.pixar.com
+$boost=2,site=deekit.com
+$boost=2,site=gkbrk.com
+$boost=2,site=jayhuang.org
+$boost=2,site=lunchbag.ca
+$boost=2,site=vanschneider.com
+$boost=2,site=seobook.com
+$boost=2,site=lfe.io
+$boost=2,site=unitedstatescourts.org
+$boost=2,site=listennotes.com
+$boost=2,site=techblog.livingsocial.com
+$boost=2,site=blogoscoped.com
+$boost=2,site=density.io
+$boost=2,site=dbasch.posterous.com
+$boost=2,site=martiancraft.com
+$boost=2,site=blog.meldium.com
+$boost=2,site=blog.codeclimate.com
+$boost=2,site=abtinforouzandeh.com
+$boost=2,site=rachelnabors.com
+$boost=2,site=zencoder.com
+$boost=2,site=itworld.com
+$boost=2,site=blog.ipinfo.io
+$boost=2,site=popsci.com
+$boost=2,site=autoweek.com
+$boost=2,site=bbot.org
+$boost=2,site=crookedtimber.org
+$boost=2,site=cultofmac.com
+$boost=2,site=en.wikisource.org
+$boost=2,site=kaggle.com
+$boost=2,site=librepcb.org
+$boost=2,site=nationalobserver.com
+$boost=2,site=skitch.com
+$boost=2,site=bfilipek.com
+$boost=2,site=fee.org
+$boost=2,site=doomworld.com
+$boost=2,site=librivox.org
+$boost=2,site=crispymtn.com
+$boost=2,site=nydailynews.com
+$boost=2,site=brilliantmaps.com
+$boost=2,site=tiamat.tsotech.com
+$boost=2,site=npf.io
+$boost=2,site=jonathan.bergknoff.com
+$boost=2,site=belitsoft.com
+$boost=2,site=blog.thecybershadow.net
+$boost=2,site=blog.privacytools.io
+$boost=2,site=elifesciences.org
+$boost=2,site=nemil.com
+$boost=2,site=jsonfeed.org
+$boost=2,site=opendatastructures.org
+$boost=2,site=yifan.lu
+$boost=2,site=sierraclub.org
+$boost=2,site=lbatalha.com
+$boost=2,site=libmill.org
+$boost=2,site=tricycle.org
+$boost=2,site=seattletimes.nwsource.com
+$boost=2,site=killscreen.com
+$boost=2,site=pixeltrue.com
+$boost=2,site=randalolson.com
+$boost=2,site=blog.thijsalkema.de
+$boost=2,site=zephoria.org
+$boost=2,site=paulromer.net
+$boost=2,site=blog.gtk.org
+$boost=2,site=blog.flattr.net
+$boost=2,site=seedmagazine.com
+$boost=2,site=technosophos.com
+$boost=2,site=monzo.com
+$boost=2,site=imperial.ac.uk
+$boost=2,site=secretgeek.net
+$boost=2,site=facebook.design
+$boost=2,site=sachachua.com
+$boost=2,site=codesections.com
+$boost=2,site=davidwong.fr
+$boost=2,site=root.cern.ch
+$boost=2,site=indieweb.org
+$boost=2,site=blog.reemer.com
+$boost=2,site=mrspeaker.net
+$boost=2,site=sun-sentinel.com
+$boost=2,site=gorban.org
+$boost=2,site=bgr.com
+$boost=2,site=0day.work
+$boost=2,site=blog.realkinetic.com
+$boost=2,site=earther.com
+$boost=2,site=positech.co.uk
+$boost=2,site=onthemedia.org
+$boost=2,site=cleveralgorithms.com
+$boost=2,site=pixelapse.com
+$boost=2,site=berniesanders.com
+$boost=2,site=a.wholelottanothing.org
+$boost=2,site=events.ccc.de
+$boost=2,site=chicagobusiness.com
+$boost=2,site=nsl.com
+$boost=2,site=srcco.de
+$boost=2,site=testpilot.firefox.com
+$boost=2,site=noahveltman.com
+$boost=2,site=jamiequint.com
+$boost=2,site=reverb.com
+$boost=2,site=pri.org
+$boost=2,site=trustnodes.com
+$boost=2,site=blogs.mathworks.com
+$boost=2,site=cncf.io
+$boost=2,site=opensource.com
+$boost=2,site=news.brown.edu
+$boost=2,site=forth.com
+$boost=2,site=blockstream.com
+$boost=2,site=getaether.net
+$boost=2,site=hooktheory.com
+$boost=2,site=irssi.org
+$boost=2,site=jemma.dev
+$boost=2,site=gittup.org
+$boost=2,site=dropboxforum.com
+$boost=2,site=blog.neocities.org
+$boost=2,site=aopa.org
+$boost=2,site=math-blog.com
+$boost=2,site=meetup.hackernewsdc.org
+$boost=2,site=ibtimes.sg
+$boost=2,site=curiousrat.com
+$boost=2,site=ponylang.io
+$boost=2,site=noisebridge.net
+$boost=2,site=apps.fcc.gov
+$boost=2,site=eenews.net
+$boost=2,site=dhs.gov
+$boost=2,site=uu.se
+$boost=2,site=blog.streak.com
+$boost=2,site=bluishcoder.co.nz
+$boost=2,site=prolost.com
+$boost=2,site=corgibytes.com
+$boost=2,site=grouptalent.com
+$boost=2,site=dankaminsky.com
+$boost=2,site=blog.rubbingalcoholic.com
+$boost=2,site=paul.copplest.one
+$boost=2,site=andrewhy.de
+$boost=2,site=techfounder.net
+$boost=2,site=spinroot.com
+$boost=2,site=garethrees.org
+$boost=2,site=blog.gedrap.me
+$boost=2,site=oceans.nautil.us
+$boost=2,site=autonlab.org
+$boost=2,site=kallithea-scm.org
+$boost=2,site=getrevue.co
+$boost=2,site=whatsonweibo.com
+$boost=2,site=tass.com
+$boost=2,site=paste.lisp.org
+$boost=2,site=rollcall.com
+$boost=2,site=reference.wolfram.com
+$boost=2,site=kevin.burke.dev
+$boost=2,site=decisiondata.org
+$boost=2,site=benbernardblog.com
+$boost=2,site=usefathom.com
+$boost=2,site=pcpro.co.uk
+$boost=2,site=progit.org
+$boost=2,site=ffii.org
+$boost=2,site=byrslf.co
+$boost=2,site=sigops.org
+$boost=2,site=blog.whoishiring.io
+$boost=2,site=mcvuk.com
+$boost=2,site=google.pl
+$boost=2,site=blog.untogether.co.uk
+$boost=2,site=uh.edu
+$boost=2,site=fable.io
+$boost=2,site=hizook.com
+$boost=2,site=ovh.com
+$boost=2,site=tedinski.com
+$boost=2,site=entrepreneur.venturebeat.com
+$boost=2,site=c-spanvideo.org
+$boost=2,site=historyextra.com
+$boost=2,site=defenseone.com
+$boost=2,site=volument.com
+$boost=2,site=imaging-resource.com
+$boost=2,site=popupcity.net
+$boost=2,site=skyandtelescope.com
+$boost=2,site=viva64.com
+$boost=2,site=earnestcapital.com
+$boost=2,site=jmarbach.com
+$boost=2,site=chrisharrison.net
+$boost=2,site=jsonline.com
+$boost=2,site=tesla-cdn.thron.com
+$boost=2,site=neuroscientistnews.com
+$boost=2,site=wondersofphysics.com
+$boost=2,site=lanl.gov
+$boost=2,site=rfc-editor.org
+$boost=2,site=industrialempathy.com
+$boost=2,site=performancejs.com
+$boost=2,site=infiniteundo.com
+$boost=2,site=heroku.com
+$boost=2,site=blog.lse.epita.fr
+$boost=2,site=lacker.io
+$boost=2,site=timeline.com
+$boost=2,site=wiki.blender.org
+$boost=2,site=blog.fugue.co
+$boost=2,site=golem.de
+$boost=2,site=pingcap.com
+$boost=2,site=john.onolan.org
+$boost=2,site=blog.mocality.co.ke
+$boost=2,site=status.coinbase.com
+$boost=2,site=attejuvonen.fi
+$boost=2,site=serversforhackers.com
+$boost=2,site=work.caltech.edu
+$boost=2,site=fuelyourcoding.com
+$boost=2,site=dyn.com
+$boost=2,site=blog.tabini.ca
+$boost=2,site=israel21c.org
+$boost=2,site=rubyonrails.org
+$boost=2,site=designweek.co.uk
+$boost=2,site=collison.ie
+$boost=2,site=evilmadscientist.com
+$boost=2,site=muddylemon.com
+$boost=2,site=agateau.com
+$boost=2,site=blog.miguelgrinberg.com
+$boost=2,site=nmap.org
+$boost=2,site=theblockcrypto.com
+$boost=2,site=haaretz.com
+$boost=2,site=codecapsule.com
+$boost=2,site=ethz.ch
+$boost=2,site=blog.bloomca.me
+$boost=2,site=blog.snapeda.com
+$boost=2,site=evonomics.com
+$boost=2,site=www2.macleans.ca
+$boost=2,site=composition.al
+$boost=2,site=js13kgames.com
+$boost=2,site=lists.basho.com
+$boost=2,site=sciencenorway.no
+$boost=2,site=speaking.io
+$boost=2,site=exceptionnotfound.net
+$boost=2,site=thisismynext.com
+$boost=2,site=unchartedterritories.tomaspueyo.com
+$boost=2,site=skilled.io
+$boost=2,site=thebreakfastpost.com
+$boost=2,site=blog.bumblebeelabs.com
+$boost=2,site=materialize.io
+$boost=2,site=haystacksoftware.com
+$boost=2,site=blog.modernmechanix.com
+$boost=2,site=math3ma.com
+$boost=2,site=retractionwatch.com
+$boost=2,site=appleoutsider.com
+$boost=2,site=dreamsongs.com
+$boost=2,site=gohugo.io
+$boost=2,site=nicolodavis.com
+$boost=2,site=andreykurenkov.com
+$boost=2,site=conversionxl.com
+$boost=2,site=qemu-advent-calendar.org
+$boost=2,site=enfranchisedmind.com
+$boost=2,site=ahrefs.com
+$boost=2,site=spf13.com
+$boost=2,site=adamschwartz.co
+$boost=2,site=scikit-learn.org
+$boost=2,site=lessonsoffailure.com
+$boost=2,site=charlesleifer.com
+$boost=2,site=zeux.io
+$boost=2,site=iximiuz.com
+$boost=2,site=kerkour.com
+$boost=2,site=racked.com
+$boost=2,site=news.morningstar.com
+$boost=2,site=hfet.org
+$boost=2,site=enterprisecraftsmanship.com
+$boost=2,site=dailynous.com
+$boost=2,site=sandiegouniontribune.com
+$boost=2,site=playframework.com
+$boost=2,site=betterhumans.coach.me
+$boost=2,site=thomaspark.co
+$boost=2,site=mimimatthews.com
+$boost=2,site=chris.beams.io
+$boost=2,site=whoismcafee.com
+$boost=2,site=stlouisfed.org
+$boost=2,site=fredrikdeboer.com
+$boost=2,site=blog.cerebralab.com
+$boost=2,site=weworkremotely.com
+$boost=2,site=english.yonhapnews.co.kr
+$boost=2,site=blog.ffwll.ch
+$boost=2,site=taoofmac.com
+$boost=2,site=nenadseo.com
+$boost=2,site=our.status.im
+$boost=2,site=justinhileman.info
+$boost=2,site=financemagnates.com
+$boost=2,site=srlabs.de
+$boost=2,site=emacsredux.com
+$boost=2,site=colorforth.com
+$boost=2,site=tech.trivago.com
+$boost=2,site=rusty.ozlabs.org
+$boost=2,site=unionleader.com
+$boost=2,site=huckmag.com
+$boost=2,site=3quarksdaily.com
+$boost=2,site=basicattentiontoken.org
+$boost=2,site=stackfield.com
+$boost=2,site=laktek.com
+$boost=2,site=elasticsearch.com
+$boost=2,site=thomashunter.name
+$boost=2,site=tapbots.com
+$boost=2,site=shime.sh
+$boost=2,site=news.consumerreports.org
+$boost=2,site=alwaysrightinstitute.com
+$boost=2,site=plover.net
+$boost=2,site=oleb.net
+$boost=2,site=mymoneyblog.com
+$boost=2,site=blog.clarifai.com
+$boost=2,site=labs.earthpeople.se
+$boost=2,site=blog.greweb.fr
+$boost=2,site=truthout.org
+$boost=2,site=rowetel.com
+$boost=2,site=acha.ninja
+$boost=2,site=blog.carlmjohnson.net
+$boost=2,site=programming-journal.org
+$boost=2,site=news18.com
+$boost=2,site=sam.bluwiki.com
+$boost=2,site=monbiot.com
+$boost=2,site=rawgit.com
+$boost=2,site=foxla.com
+$boost=2,site=dvt.name
+$boost=2,site=developer.marvel.com
+$boost=2,site=lowendtalk.com
+$boost=2,site=docs.celeryproject.org
+$boost=2,site=poststatus.com
+$boost=2,site=blockgeeks.com
+$boost=2,site=itnext.io
+$boost=2,site=blog.existentialize.com
+$boost=2,site=semiengineering.com
+$boost=2,site=digibarn.com
+$boost=2,site=mikemcquaid.com
+$boost=2,site=hackflow.com
+$boost=2,site=lists.opensuse.org
+$boost=2,site=weblog.rogueamoeba.com
+$boost=2,site=webdesignmuseum.org
+$boost=2,site=garron.blog
+$boost=2,site=instadeq.com
+$boost=2,site=brisbanetimes.com.au
+$boost=2,site=leanpub.com
+$boost=2,site=fendrich.se
+$boost=2,site=edn.com
+$boost=2,site=synthtopia.com
+$boost=2,site=michaelbromley.co.uk
+$boost=2,site=refsmmat.com
+$boost=2,site=jonbischke.com
+$boost=2,site=blogs.kqed.org
+$boost=2,site=xconomy.com
+$boost=2,site=timeslive.co.za
+$boost=2,site=keithp.com
+$boost=2,site=blog.online.net
+$boost=2,site=atariarchives.org
+$boost=2,site=wearethemighty.com
+$boost=2,site=heatst.com
+$boost=2,site=blog.wekeroad.com
+$boost=2,site=dnainfo.com
+$boost=2,site=wufoo.com
+$boost=2,site=zompist.com
+$boost=2,site=nbr.co.nz
+$boost=2,site=tomtunguz.com
+$boost=2,site=blog.twilio.com
+$boost=2,site=nhplace.com
+$boost=2,site=the-digital-reader.com
+$boost=2,site=vpri.org
+$boost=2,site=bailis.org
+$boost=2,site=maximise.dk
+$boost=2,site=99designs.com
+$boost=2,site=crummy.com
+$boost=2,site=ibtimes.co.uk
+$boost=2,site=6sqft.com
+$boost=2,site=symkat.com
+$boost=2,site=survata.com
+$boost=2,site=greenwichtime.com
+$boost=2,site=leandomainsearch.com
+$boost=2,site=penguindreams.org
+$boost=2,site=windowsteamblog.com
+$boost=2,site=more-magic.net
+$boost=2,site=wundergraph.com
+$boost=2,site=stereogum.com
+$boost=2,site=media.defcon.org
+$boost=2,site=physics.aps.org
+$boost=2,site=addedbytes.com
+$boost=2,site=theconversation.edu.au
+$boost=2,site=carwoo.com
+$boost=2,site=swiftbysundell.com
+$boost=2,site=bigthink.com
+$boost=2,site=nltk.org
+$boost=2,site=atlas.cern
+$boost=2,site=nyse.com
+$boost=2,site=icyphox.sh
+$boost=2,site=hypothesis.works
+$boost=2,site=jacobobryant.com
+$boost=2,site=blogs.zappos.com
+$boost=2,site=blog.joellehman.com
+$boost=2,site=conal.net
+$boost=2,site=dadhacker.com
+$boost=2,site=stars.chromeexperiments.com
+$boost=2,site=blog.thousandeyes.com
+$boost=2,site=drbunsen.org
+$boost=2,site=blog.tinfoilsecurity.com
+$boost=2,site=landley.net
+$boost=2,site=jjbeshara.com
+$boost=2,site=searchenginejournal.com
+$boost=2,site=samgentle.com
+$boost=2,site=mywinamp.com
+$boost=2,site=acm.uiuc.edu
+$boost=2,site=blogs.sfweekly.com
+$boost=2,site=theaviationist.com
+$boost=2,site=mikealrogers.com
+$boost=2,site=unofficialgoogledatascience.com
+$boost=2,site=wildbunny.co.uk
+$boost=2,site=vimcasts.org
+$boost=2,site=home.web.cern.ch
+$boost=2,site=news.virginia.edu
+$boost=2,site=blog.zawodny.com
+$boost=2,site=blog.mousereeve.com
+$boost=2,site=hackernewsbooks.com
+$boost=2,site=detexify.kirelabs.org
+$boost=2,site=blip.tv
+$boost=2,site=goodstuff.im
+$boost=2,site=opensignal.com
+$boost=2,site=the-pastry-box-project.net
+$boost=2,site=parkerhiggins.net
+$boost=2,site=roughtype.com
+$boost=2,site=ecosia.org
+$boost=2,site=dougmccune.com
+$boost=2,site=lambdaisland.com
+$boost=2,site=blog.8thlight.com
+$boost=2,site=akaptur.com
+$boost=2,site=appneta.com
+$boost=2,site=eos.org
+$boost=2,site=eia.gov
+$boost=2,site=freesoftwaremagazine.com
+$boost=2,site=blog.isquaredsoftware.com
+$boost=2,site=effectgames.com
+$boost=2,site=danbirken.com
+$boost=2,site=georgesaines.com
+$boost=2,site=cilium.io
+$boost=2,site=luisant.ca
+$boost=2,site=kde.org
+$boost=2,site=gengo.ai
+$boost=2,site=ask.slashdot.org
+$boost=2,site=inf.ethz.ch
+$boost=2,site=mv-voice.com
+$boost=2,site=visual6502.org
+$boost=2,site=codeofrob.com
+$boost=2,site=ludumdare.com
+$boost=2,site=bradford-delong.com
+$boost=2,site=sidsite.com
+$boost=2,site=persquaremile.com
+$boost=2,site=pathdependent.com
+$boost=2,site=marketplacepulse.com
+$boost=2,site=blog.steve.fi
+$boost=2,site=stefankendall.com
+$boost=2,site=machinelearningmastery.com
+$boost=2,site=tomcritchlow.com
+$boost=2,site=vqronline.org
+$boost=2,site=pmarchive.com
+$boost=2,site=blog.exodusintel.com
+$boost=2,site=icelandmonitor.mbl.is
+$boost=2,site=learnui.design
+$boost=2,site=wisdomandwonder.com
+$boost=2,site=gnupg.org
+$boost=2,site=otherhand.org
+$boost=2,site=andyswan.com
+$boost=2,site=unikernel.org
+$boost=2,site=theappeal.org
+$boost=2,site=multivax.com
+$boost=2,site=hilarymason.com
+$boost=2,site=metro.co.uk
+$boost=2,site=fxguide.com
+$boost=2,site=mail.gnome.org
+$boost=2,site=ripe.net
+$boost=2,site=polygraph.cool
+$boost=2,site=oonwoye.com
+$boost=2,site=centernetworks.com
+$boost=2,site=blog.kenperlin.com
+$boost=2,site=ericasadun.com
+$boost=2,site=scrapingbee.com
+$boost=2,site=mail-index.netbsd.org
+$boost=2,site=mikeindustries.com
+$boost=2,site=ritsumei.ac.jp
+$boost=2,site=typing.io
+$boost=2,site=blog.britishmuseum.org
+$boost=2,site=mailbox.org
+$boost=2,site=hackety.org
+$boost=2,site=invisible-island.net
+$boost=2,site=blog.minio.io
+$boost=2,site=etymonline.com
+$boost=2,site=founderdating.com
+$boost=2,site=users.softlab.ece.ntua.gr
+$boost=2,site=read.acloud.guru
+$boost=2,site=julian.digital
+$boost=2,site=ben.balter.com
+$boost=2,site=osc.universityofcalifornia.edu
+$boost=2,site=futureoflife.org
+$boost=2,site=multicians.org
+$boost=2,site=allthatsinteresting.com
+$boost=2,site=machinethink.net
+$boost=2,site=blogs.denverpost.com
+$boost=2,site=mishadoff.com
+$boost=2,site=getfirebug.com
+$boost=2,site=dxomark.com
+$boost=2,site=mycpu.org
+$boost=2,site=raphkoster.com
+$boost=2,site=blog.doordash.com
+$boost=2,site=paul.kinlan.me
+$boost=2,site=sliderulemuseum.com
+$boost=2,site=hpcwire.com
+$boost=2,site=flaviocopes.com
+$boost=2,site=nocomplexity.com
+$boost=2,site=what-if.xkcd.com
+$boost=2,site=healthcareitnews.com
+$boost=2,site=turingpi.com
+$boost=2,site=taskwarrior.org
+$boost=2,site=grepular.com
+$boost=2,site=nuclearsecrecy.com
+$boost=2,site=maxkle.in
+$boost=2,site=csd.uwo.ca
+$boost=2,site=alanhollis.com
+$boost=2,site=splitgraph.com
+$boost=2,site=instapainting.com
+$boost=2,site=taxfoundation.org
+$boost=2,site=delta.chat
+$boost=2,site=appft1.uspto.gov
+$boost=2,site=leerob.io
+$boost=2,site=chrisshort.net
+$boost=2,site=blog.venanti.us
+$boost=2,site=news.cnrs.fr
+$boost=2,site=500px.com
+$boost=2,site=fearlesssalarynegotiation.com
+$boost=2,site=frankmcsherry.org
+$boost=2,site=nts.strzibny.name
+$boost=2,site=developers.memsql.com
+$boost=2,site=sr.ht
+$boost=2,site=ebiquity.umbc.edu
+$boost=2,site=news.ubc.ca
+$boost=2,site=dailyherald.com
+$boost=2,site=felixge.de
+$boost=2,site=blog.ayoungprogrammer.com
+$boost=2,site=inhabitat.com
+$boost=2,site=lifehacker.com.au
+$boost=2,site=science.howstuffworks.com
+$boost=2,site=cancer.nautil.us
+$boost=2,site=appcubby.com
+$boost=2,site=math.utah.edu
+$boost=2,site=tooslowexception.com
+$boost=2,site=patrickcraig.co.uk
+$boost=2,site=perfectionkills.com
+$boost=2,site=ncona.com
+$boost=2,site=cnil.fr
+$boost=2,site=littlebitofcode.com
+$boost=2,site=zerodha.tech
+$boost=2,site=medicine.wustl.edu
+$boost=2,site=blog.socialcops.com
+$boost=2,site=ifttt.com
+$boost=2,site=blog.scalyr.com
+$boost=2,site=psypost.org
+$boost=2,site=abcactionnews.com
+$boost=2,site=blog.marksweep.com
+$boost=2,site=basecamp.com
+$boost=2,site=michael.otacoo.com
+$boost=2,site=toucharcade.com
+$boost=2,site=adage.com
+$boost=2,site=blog.klipse.tech
+$boost=2,site=dev.to
+$boost=2,site=deadline.com
+$boost=2,site=expressvpn.com
+$boost=2,site=zverovich.net
+$boost=2,site=bnl.gov
+$boost=2,site=discuss.linuxcontainers.org
+$boost=2,site=codesketch.com
+$boost=2,site=leveragedsellout.com
+$boost=2,site=imore.com
+$boost=2,site=people.inf.ethz.ch
+$boost=2,site=blog.smartbear.com
+$boost=2,site=ryanflorence.com
+$boost=2,site=createdigitalmusic.com
+$boost=2,site=nhm.ac.uk
+$boost=2,site=blog.hartleybrody.com
+$boost=2,site=futureboy.us
+$boost=2,site=windowslatest.com
+$boost=2,site=ewontfix.com
+$boost=2,site=jng.imagine27.com
+$boost=2,site=gogs.io
+$boost=2,site=geoffboeing.com
+$boost=2,site=phillymag.com
+$boost=2,site=blog.authy.com
+$boost=2,site=adamwathan.me
+$boost=2,site=distributionhacks.com
+$boost=2,site=thinkster.io
+$boost=2,site=terminusdb.com
+$boost=2,site=technoblogy.com
+$boost=2,site=30hourjobs.com
+$boost=2,site=ftrain.com
+$boost=2,site=foundation.zurb.com
+$boost=2,site=blog.viraptor.info
+$boost=2,site=bennesvig.com
+$boost=2,site=expressionsofchange.org
+$boost=2,site=blog.urfix.com
+$boost=2,site=lists.x.org
+$boost=2,site=huggingface.co
+$boost=2,site=zeuxcg.org
+$boost=2,site=vmssoftware.com
+$boost=2,site=jimplush.com
+$boost=2,site=thereboot.com
+$boost=2,site=antimatter15.com
+$boost=2,site=blog.davidecoppola.com
+$boost=2,site=kotaku.com.au
+$boost=2,site=3perf.com
+$boost=2,site=ryxcommar.com
+$boost=2,site=scitechdaily.com
+$boost=2,site=rjzaworski.com
+$boost=2,site=dolthub.com
+$boost=2,site=jperla.com
+$boost=2,site=roberto.open-lab.com
+$boost=2,site=tofugu.com
+$boost=2,site=preprints.org
+$boost=2,site=blog.kaggle.com
+$boost=2,site=getcarina.com
+$boost=2,site=thebarentsobserver.com
+$boost=2,site=improbable.com
+$boost=2,site=type.method.ac
+$boost=2,site=aleph.se
+$boost=2,site=lists.macosforge.org
+$boost=2,site=zguide.zeromq.org
+$boost=2,site=onethingwell.org
+$boost=2,site=gittip.com
+$boost=2,site=newsobserver.com
+$boost=2,site=voat.co
+$boost=2,site=viemu.com
+$boost=2,site=embeddeduse.com
+$boost=2,site=ideolalia.com
+$boost=2,site=andrewdumont.me
+$boost=2,site=sriramk.com
+$boost=2,site=chenhuijing.com
+$boost=2,site=yegor256.com
+$boost=2,site=cs.nott.ac.uk
+$boost=2,site=trustedreviews.com
+$boost=2,site=mitxela.com
+$boost=2,site=mail.haskell.org
+$boost=2,site=belfasttelegraph.co.uk
+$boost=2,site=zackshapiro.com
+$boost=2,site=collectspace.com
+$boost=2,site=alchemistowl.org
+$boost=2,site=interfluidity.com
+$boost=2,site=mithril.js.org
+$boost=2,site=sunboxlabs.com
+$boost=2,site=mattermark.com
+$boost=2,site=ballotpedia.org
+$boost=2,site=r2c.dev
+$boost=2,site=goaccess.io
+$boost=2,site=michaelwest.com.au
+$boost=2,site=redsymbol.net
+$boost=2,site=journals.aps.org
+$boost=2,site=blog.gingerlime.com
+$boost=2,site=bugsplat.info
+$boost=2,site=datastuff.tech
+$boost=2,site=cliffc.org
+$boost=2,site=leapmotion.com
+$boost=2,site=dom.as
+$boost=2,site=rickcarlino.com
+$boost=2,site=electronicsweekly.com
+$boost=2,site=blog.mythic-beasts.com
+$boost=2,site=blog.codiqa.com
+$boost=2,site=urbandictionary.com
+$boost=2,site=blog.ericgoldman.org
+$boost=2,site=thatcomputerscientist.com
+$boost=2,site=shankman.com
+$boost=2,site=blog.uservoice.com
+$boost=2,site=shafyy.com
+$boost=2,site=23andme.com
+$boost=2,site=thebulletin.org
+$boost=2,site=sekao.net
+$boost=2,site=om.co
+$boost=2,site=reacttraining.com
+$boost=2,site=phili.pe
+$boost=2,site=kamerontanseli.ghost.io
+$boost=2,site=deccanherald.com
+$boost=2,site=recipes.hypotheses.org
+$boost=2,site=scotthyoung.com
+$boost=2,site=blog.kraken.com
+$boost=2,site=ello.co
+$boost=2,site=asiatimes.com
+$boost=2,site=blog.tindie.com
+$boost=2,site=plagiarismtoday.com
+$boost=2,site=the74million.org
+$boost=2,site=analytics.usa.gov
+$boost=2,site=blog.jpl-consulting.com
+$boost=2,site=gosquared.com
+$boost=2,site=unite.ai
+$boost=2,site=pastie.org
+$boost=2,site=bluetooth.com
+$boost=2,site=heartbeat.fritz.ai
+$boost=2,site=blog.liw.fi
+$boost=2,site=placesjournal.org
+$boost=2,site=interfacelab.com
+$boost=2,site=programming-motherfucker.com
+$boost=2,site=thespacereview.com
+$boost=2,site=zippia.com
+$boost=2,site=alumnit.ca
+$boost=2,site=blog.scienceexchange.com
+$boost=2,site=danielallendeutsch.com
+$boost=2,site=learningclojure.com
+$boost=2,site=neo4j.com
+$boost=2,site=insentricity.com
+$boost=2,site=johnhawks.net
+$boost=2,site=gist.io
+$boost=2,site=irishexaminer.com
+$boost=2,site=blog.runnable.com
+$boost=2,site=technology.stitchfix.com
+$boost=2,site=anyasq.com
+$boost=2,site=eigenfoo.xyz
+$boost=2,site=sunlightfoundation.com
+$boost=2,site=makegameswith.us
+$boost=2,site=wesleytansey.com
+$boost=2,site=startupljackson.com
+$boost=2,site=xprize.org
+$boost=2,site=mongodb.org
+$boost=2,site=drive.com.au
+$boost=2,site=crate.io
+$boost=2,site=voice.kadira.io
+$boost=2,site=blog.drewolson.org
+$boost=2,site=slog.thestranger.com
+$boost=2,site=remarkedly.com
+$boost=2,site=artem.krylysov.com
+$boost=2,site=collabora.com
+$boost=2,site=handelsblatt.com
+$boost=2,site=dailycamera.com
+$boost=2,site=jozefg.bitbucket.org
+$boost=2,site=baus.net
+$boost=2,site=davisr.me
+$boost=2,site=ghc.haskell.org
+$boost=2,site=akkshaya.blog
+$boost=2,site=boristhebrave.com
+$boost=2,site=ootbcomp.com
+$boost=2,site=puzzlescript.net
+$boost=2,site=mattjibson.com
+$boost=2,site=simpleflying.com
+$boost=2,site=houseoflawandorder.com
+$boost=2,site=abstractions.nautil.us
+$boost=2,site=scalescale.com
+$boost=2,site=ostechnix.com
+$boost=2,site=londonreconnections.com
+$boost=2,site=sandro.volpee.de
+$boost=2,site=reghardware.com
+$boost=2,site=shellcheck.net
+$boost=2,site=freep.com
+$boost=2,site=cs.hmc.edu
+$boost=2,site=post-gazette.com
+$boost=2,site=englishrussia.com
+$boost=2,site=2ton.com.au
+$boost=2,site=news.err.ee
+$boost=2,site=academictimes.com
+$boost=2,site=beyondgrep.com
+$boost=2,site=lesspenguiny.com
+$boost=2,site=tinyhack.com
+$boost=2,site=mailgun.com
+$boost=2,site=homepage.divms.uiowa.edu
+$boost=2,site=grimoire.ca
+$boost=2,site=unicornfree.com
+$boost=2,site=afr.com
+$boost=2,site=routefifty.com
+$boost=2,site=nintendolife.com
+$boost=2,site=wolfram.com
+$boost=2,site=rfi.fr
+$boost=2,site=talkweb.eu
+$boost=2,site=elidourado.com
+$boost=2,site=math.uwaterloo.ca
+$boost=2,site=industryweek.com
+$boost=2,site=prerender.cloud
+$boost=2,site=thezdi.com
+$boost=2,site=ableton.com
+$boost=2,site=kylerush.net
+$boost=2,site=hbswk.hbs.edu
+$boost=2,site=hopesandfears.com
+$boost=2,site=mappingignorance.org
+$boost=2,site=code.jsoftware.com
+$boost=2,site=thefrailestthing.com
+$boost=2,site=blog.gopheracademy.com
+$boost=2,site=threedots.tech
+$boost=2,site=robnapier.net
+$boost=2,site=blog.indutny.com
+$boost=2,site=cloudonaut.io
+$boost=2,site=techprowd.com
+$boost=2,site=smalldata.tech
+$boost=2,site=notebook.drmaciver.com
+$boost=2,site=getpostman.com
+$boost=2,site=nationalaffairs.com
+$boost=2,site=roarmag.org
+$boost=2,site=blog.ezliu.com
+$boost=2,site=engineering.heroku.com
+$boost=2,site=automobilemag.com
+$boost=2,site=lihautan.com
+$boost=2,site=aschmann.net
+$boost=2,site=ionicframework.com
+$boost=2,site=blog.davidsingleton.org
+$boost=2,site=blog.jessitron.com
+$boost=2,site=orau.org
+$boost=2,site=digesttime.com
+$boost=2,site=kswizz.com
+$boost=2,site=wmpoweruser.com
+$boost=2,site=nimbleindustries.io
+$boost=2,site=bookforum.com
+$boost=2,site=genode.org
+$boost=2,site=decorrespondent.nl
+$boost=2,site=theorangeduck.com
+$boost=2,site=blog.redfin.com
+$boost=2,site=threema.ch
+$boost=2,site=blog.gaborcselle.com
+$boost=2,site=eastbayexpress.com
+$boost=2,site=nilsjuenemann.de
+$boost=2,site=ausgamers.com
+$boost=2,site=nickkolenda.com
+$boost=2,site=datadome.co
+$boost=2,site=matusiak.eu
+$boost=2,site=foaas.com
+$boost=2,site=quiverquant.com
+$boost=2,site=pennmedicine.org
+$boost=2,site=panamapapers.icij.org
+$boost=2,site=rare-technologies.com
+$boost=2,site=wusa9.com
+$boost=2,site=exupero.org
+$boost=2,site=nodesource.com
+$boost=2,site=tryapl.org
+$boost=2,site=adamcaudill.com
+$boost=2,site=flexport.engineering
+$boost=2,site=theneurosphere.com
+$boost=2,site=news.rice.edu
+$boost=2,site=blog.ninlabs.com
+$boost=2,site=twistedoakstudios.com
+$boost=2,site=qvault.io
+$boost=2,site=ourmachinery.com
+$boost=2,site=linuxatemyram.com
+$boost=2,site=blog.kraih.com
+$boost=2,site=blog.nyaruka.com
+$boost=2,site=jeffdechambeau.com
+$boost=2,site=opensoul.org
+$boost=2,site=cortes.design
+$boost=2,site=pawfal.org
+$boost=2,site=blog.typesafe.com
+$boost=2,site=blog.andrewcantino.com
+$boost=2,site=boeing.com
+$boost=2,site=blog.custora.com
+$boost=2,site=naftaliharris.com
+$boost=2,site=javascriptkicks.com
+$boost=2,site=the-toast.net
+$boost=2,site=cs.rit.edu
+$boost=2,site=m.economictimes.com
+$boost=2,site=vessenes.com
+$boost=2,site=ineteconomics.org
+$boost=2,site=0fps.net
+$boost=2,site=justcramer.com
+$boost=2,site=qualiacomputing.com
+$boost=2,site=improbable.io
+$boost=2,site=blog.dilbert.com
+$boost=2,site=standpointmag.co.uk
+$boost=2,site=pixelmator.com
+$boost=2,site=heraldscotland.com
+$boost=2,site=yuzu-emu.org
+$boost=2,site=physicscentral.com
+$boost=2,site=overclock3d.net
+$boost=2,site=crainsnewyork.com
+$boost=2,site=livefreeordichotomize.com
+$boost=2,site=lostechies.com
+$boost=2,site=weblog.bocoup.com
+$boost=2,site=rechargenews.com
+$boost=2,site=weatherspark.com
+$boost=2,site=ceylon-lang.org
+$boost=2,site=androidrookies.com
+$boost=2,site=nwn.blogs.com
+$boost=2,site=nathany.com
+$boost=2,site=deepdotweb.com
+$boost=2,site=pitheringabout.com
+$boost=2,site=steve.yegge.googlepages.com
+$boost=2,site=useronboard.com
+$boost=2,site=kjmagnetics.com
+$boost=2,site=ftp.openbsd.org
+$boost=2,site=bogleheads.org
+$boost=2,site=v5.chriskrycho.com
+$boost=2,site=jordanmechner.com
+$boost=2,site=fare.tunes.org
+$boost=2,site=dangermouse.net
+$boost=2,site=pdf.yt
+$boost=2,site=robbreport.com
+$boost=2,site=insideevs.com
+$boost=2,site=logolounge.com
+$boost=2,site=ftp.arl.army.mil
+$boost=2,site=linux.com
+$boost=2,site=csoonline.com
+$boost=2,site=catapult.co
+$boost=2,site=blog.webkid.io
+$boost=2,site=uwgb.edu
+$boost=2,site=girinstud.io
+$boost=2,site=adilmoujahid.com
+$boost=2,site=rntz.net
+$boost=2,site=smore.com
+$boost=2,site=news.rthk.hk
+$boost=2,site=c-span.org
+$boost=2,site=m.stopa.io
+$boost=2,site=opb.org
+$boost=2,site=opendylan.org
+$boost=2,site=tailwindcss.com
+$boost=2,site=youtube-creators.googleblog.com
+$boost=2,site=plosone.org
+$boost=2,site=la.curbed.com
+$boost=2,site=gfycat.com
+$boost=2,site=pressfreedomfoundation.org
+$boost=2,site=tested.com
+$boost=2,site=blog.gruntwork.io
+$boost=2,site=innoq.com
+$boost=2,site=granolashotgun.com
+$boost=2,site=bu.edu
+$boost=2,site=mozillalabs.com
+$boost=2,site=thinkgrowth.org
+$boost=2,site=soledadpenades.com
+$boost=2,site=snips.ai
+$boost=2,site=insightdataengineering.com
+$boost=2,site=bugs.webkit.org
+$boost=2,site=serpentine.com
+$boost=2,site=crimereads.com
+$boost=2,site=muut.com
+$boost=2,site=deque.blog
+$boost=2,site=nearform.com
+$boost=2,site=us-cert.gov
+$boost=2,site=regexcrossword.com
+$boost=2,site=chris-lamb.co.uk
+$boost=2,site=unretrofied.com
+$boost=2,site=rsaconference.com
+$boost=2,site=nhregister.com
+$boost=2,site=cryoshon.co
+$boost=2,site=salon.thefamily.co
+$boost=2,site=leakedsource.com
+$boost=2,site=clicktorelease.com
+$boost=2,site=chipworks.com
+$boost=2,site=hardkernel.com
+$boost=2,site=news.slashdot.org
+$boost=2,site=top500.org
+$boost=2,site=technicshistory.com
+$boost=2,site=segment.io
+$boost=2,site=thebuild.com
+$boost=2,site=thistooshallgrow.com
+$boost=2,site=alliance.nautil.us
+$boost=2,site=slashgeek.net
+$boost=2,site=zeroequalsfalse.press
+$boost=2,site=blog.zarfhome.com
+$boost=2,site=ryanpickren.com
+$boost=2,site=theengineer.co.uk
+$boost=2,site=the-magazine.org
+$boost=2,site=dennybritz.com
+$boost=2,site=ben.casnocha.com
+$boost=2,site=alexirpan.com
+$boost=2,site=siliconvalleywatcher.com
+$boost=2,site=chill.com
+$boost=2,site=kanoki.org
+$boost=2,site=knowablemagazine.org
+$boost=2,site=fun-mooc.fr
+$boost=2,site=longbets.org
+$boost=2,site=papers.nips.cc
+$boost=2,site=dyske.com
+$boost=2,site=kth.se
+$boost=2,site=tawheedkader.com
+$boost=2,site=plasmasturm.org
+$boost=2,site=blog.backtype.com
+$boost=2,site=feaneron.com
+$boost=2,site=tech.backtype.com
+$boost=2,site=annashipman.co.uk
+$boost=2,site=sive.rs
+$boost=2,site=keen.io
+$boost=2,site=www1.informatik.uni-erlangen.de
+$boost=2,site=ewanleith.com
+$boost=2,site=tech.puredanger.com
+$boost=2,site=freshman.tech
+$boost=2,site=weakdh.org
+$boost=2,site=boundary.com
+$boost=2,site=pcmonk.me
+$boost=2,site=wfla.com
+$boost=2,site=blog.adamnash.com
+$boost=2,site=english.martinvarsavsky.net
+$boost=2,site=usv.com
+$boost=2,site=federalnewsnetwork.com
+$boost=2,site=theatlas.com
+$boost=2,site=riknieu.com
+$boost=2,site=itwire.com
+$boost=2,site=standalone-sysadmin.com
+$boost=2,site=autoblog.com
+$boost=2,site=pv-magazine-usa.com
+$boost=2,site=blog.andrewvc.com
+$boost=2,site=mappingthejourney.com
+$boost=2,site=blogs.s-osg.org
+$boost=2,site=erica.biz
+$boost=2,site=humantransit.org
+$boost=2,site=anaxi.com
+$boost=2,site=bedrocklinux.org
+$boost=2,site=sina.is
+$boost=2,site=joachim-breitner.de
+$boost=2,site=o.canada.com
+$boost=2,site=michaelrbernste.in
+$boost=2,site=computerworlduk.com
+$boost=2,site=tech.newstatesman.com
+$boost=2,site=abc7chicago.com
+$boost=2,site=bytes.yingw787.com
+$boost=2,site=mikekarnj.com
+$boost=2,site=designinstruct.com
+$boost=2,site=fsecurify.com
+$boost=2,site=deepsource.io
+$boost=2,site=n-gate.com
+$boost=2,site=25hoursaday.com
+$boost=2,site=global.handelsblatt.com
+$boost=2,site=pragli.com
+$boost=2,site=manuel.bernhardt.io
+$boost=2,site=blog.jooq.org
+$boost=2,site=portfolio.com
+$boost=2,site=harpersbazaar.com
+$boost=2,site=snapframework.com
+$boost=2,site=hussmanfunds.com
+$boost=2,site=benlog.com
+$boost=2,site=fluentcpp.com
+$boost=2,site=toptal.com
+$boost=2,site=openssh.com
+$boost=2,site=timesofisrael.com
+$boost=2,site=tomdalling.com
+$boost=2,site=stuff.co.nz
+$boost=2,site=heather.cs.ucdavis.edu
+$boost=2,site=theincidentaleconomist.com
+$boost=2,site=radiapp.com
+$boost=2,site=blog.ramen.is
+$boost=2,site=geti2p.net
+$boost=2,site=hey.com
+$boost=2,site=33bits.org
+$boost=2,site=pressherald.com
+$boost=2,site=staycaffeinated.com
+$boost=2,site=igorkromin.net
+$boost=2,site=news.northwestern.edu
+$boost=2,site=canny.io
+$boost=2,site=modernhealthcare.com
+$boost=2,site=about.google
+$boost=2,site=ifweassume.com
+$boost=2,site=blog.readme.io
+$boost=2,site=bschaatsbergen.com
+$boost=2,site=bendyworks.com
+$boost=2,site=randywestergren.com
+$boost=2,site=crn.com.au
+$boost=2,site=substack.net
+$boost=2,site=toothycat.net
+$boost=2,site=pixelscommander.com
+$boost=2,site=thinq.co.uk
+$boost=2,site=huffingtonpost.com.au
+$boost=2,site=depesz.com
+$boost=2,site=w2lessons.com
+$boost=2,site=mostlymaths.net
+$boost=2,site=putanumonit.com
+$boost=2,site=executeprogram.com
+$boost=2,site=computing.co.uk
+$boost=2,site=eblong.com
+$boost=2,site=asylum.madhouse-project.org
+$boost=2,site=valerieaurora.org
+$boost=2,site=owasp.org
+$boost=2,site=petition.parliament.uk
+$boost=2,site=fastml.com
+$boost=2,site=nerderati.com
+$boost=2,site=nutsvolts.com
+$boost=2,site=vine.co
+$boost=2,site=grobmeier.de
+$boost=2,site=coursetro.com
+$boost=2,site=rleonardi.com
+$boost=2,site=iq.opengenus.org
+$boost=2,site=tech.slashdot.org
+$boost=2,site=stanforddaily.com
+$boost=2,site=mapd.com
+$boost=2,site=theblaze.com
+$boost=2,site=codingkaiser.blog
+$boost=2,site=techblog.badoo.com
+$boost=2,site=lawfareblog.com
+$boost=2,site=taskandpurpose.com
+$boost=2,site=programmingdesignsystems.com
+$boost=2,site=us.pycon.org
+$boost=2,site=blog.kirigin.com
+$boost=2,site=blog.perfectaudience.com
+$boost=2,site=mreinhold.org
+$boost=2,site=thoughtworks.com
+$boost=2,site=blog.frankel.ch
+$boost=2,site=deathandtaxesmag.com
+$boost=2,site=7news.com.au
+$boost=2,site=keminglabs.com
+$boost=2,site=research.securitum.com
+$boost=2,site=blog.nanonets.com
+$boost=2,site=roguelynn.com
+$boost=2,site=zenhack.net
+$boost=2,site=thelongandshort.org
+$boost=2,site=howtosplitanatom.com
+$boost=2,site=losangeles.cbslocal.com
+$boost=2,site=freebsdfoundation.org
+$boost=2,site=pqina.nl
+$boost=2,site=ramblemaps.com
+$boost=2,site=blog.httpwatch.com
+$boost=2,site=texastribune.org
+$boost=2,site=vagosec.org
+$boost=2,site=blog.michielborkent.nl
+$boost=2,site=desmogblog.com
+$boost=2,site=informationarchitects.net
+$boost=2,site=mpg.de
+$boost=2,site=devcenter.heroku.com
+$boost=2,site=blog.hackerrank.com
+$boost=2,site=mailboxapp.com
+$boost=2,site=libregraphicsworld.org
+$boost=2,site=honeycomb.io
+$boost=2,site=davidhauser.com
+$boost=2,site=philanthropy.com
+$boost=2,site=armstrongeconomics.com
+$boost=2,site=techpowerup.com
+$boost=2,site=fourweekmba.com
+$boost=2,site=runwayml.com
+$boost=2,site=cs.uwaterloo.ca
+$boost=2,site=jeffarchibald.ca
+$boost=2,site=monograph.io
+$boost=2,site=jaredforsyth.com
+$boost=2,site=rework.withgoogle.com
+$boost=2,site=butwhatfor.com
+$boost=2,site=webglfundamentals.org
+$boost=2,site=skepchick.org
+$boost=2,site=t37.net
+$boost=2,site=blockstack.org
+$boost=2,site=ticalc.org
+$boost=2,site=konghq.com
+$boost=2,site=orlandosentinel.com
+$boost=2,site=ansuz.sooke.bc.ca
+$boost=2,site=platform.uno
+$boost=2,site=i-kh.net
+$boost=2,site=thinkingagriculture.io
+$boost=2,site=colorado.edu
+$boost=2,site=philipkiely.com
+$boost=2,site=theultralinx.com
+$boost=2,site=the-witness.net
+$boost=2,site=woot.com
+$boost=2,site=blog.crowdtilt.com
+$boost=2,site=docs.python-guide.org
+$boost=2,site=blog.indextank.com
+$boost=2,site=theirrelevantinvestor.com
+$boost=2,site=herman.asia
+$boost=2,site=quartzy.qz.com
+$boost=2,site=peterbe.com
+$boost=2,site=shakycode.com
+$boost=2,site=spectrum.chat
+$boost=2,site=alugy.com
+$boost=2,site=modern.ie
+$boost=2,site=ilovetypography.com
+$boost=2,site=parametric.press
+$boost=2,site=blog.staffjoy.com
+$boost=2,site=grahamcluley.com
+$boost=2,site=benediktmeurer.de
+$boost=2,site=loggly.com
+$boost=2,site=typophile.com
+$boost=2,site=themorningnews.com
+$boost=2,site=blog.estimote.com
+$boost=2,site=nola.com
+$boost=2,site=blog.fanout.io
+$boost=2,site=theoreticalminimum.com
+$boost=2,site=serhack.me
+$boost=2,site=digital.nhs.uk
+$boost=2,site=investor.yahoo.net
+$boost=2,site=gamingonlinux.com
+$boost=2,site=interactions.acm.org
+$boost=2,site=codeberg.org
+$boost=2,site=arm.com
+$boost=2,site=fedscoop.com
+$boost=2,site=eyeondesign.aiga.org
+$boost=2,site=vihart.com
+$boost=2,site=webmonkey.com
+$boost=2,site=pacifict.com
+$boost=2,site=anaconda.com
+$boost=2,site=jamesclear.com
+$boost=2,site=images.guide
+$boost=2,site=sandofsky.com
+$boost=2,site=tessrinearson.com
+$boost=2,site=mysqlserverteam.com
+$boost=2,site=virology.ws
+$boost=2,site=b.lesseverything.com
+$boost=2,site=forklog.media
+$boost=2,site=kevinrose.com
+$boost=2,site=sankalpjonna.com
+$boost=2,site=mathbabe.org
+$boost=2,site=nickmchardy.com
+$boost=2,site=spectrumnews.org
+$boost=2,site=fictiv.com
+$boost=2,site=riskbasedsecurity.com
+$boost=2,site=juliagalef.com
+$boost=2,site=medium.learningbyshipping.com
+$boost=2,site=developer.yahoo.net
+$boost=2,site=lostbookofsales.com
+$boost=2,site=blog.rainforestqa.com
+$boost=2,site=zeldman.com
+$boost=2,site=shogan.co.uk
+$boost=2,site=blog.friendfeed.com
+$boost=2,site=tech.instacart.com
+$boost=2,site=well-typed.com
+$boost=2,site=madebynathan.com
+$boost=2,site=dam.brown.edu
+$boost=2,site=electrive.com
+$boost=2,site=open.bekk.no
+$boost=2,site=launch.is
+$boost=2,site=mindyourdecisions.com
+$boost=2,site=avc.blogs.com
+$boost=2,site=underscore.io
+$boost=2,site=beakerbrowser.com
+$boost=2,site=tibleiz.net
+$boost=2,site=learnlayout.com
+$boost=2,site=statescoop.com
+$boost=2,site=mkaz.blog
+$boost=2,site=khalilstemmler.com
+$boost=2,site=pepijndevos.nl
+$boost=2,site=en.arguman.org
+$boost=2,site=dougseven.com
+$boost=2,site=hsj.co.uk
+$boost=2,site=washingtoncitypaper.com
+$boost=2,site=stefanjudis.com
+$boost=2,site=remotehabits.com
+$boost=2,site=pawelurbanek.com
+$boost=2,site=mattrogish.com
+$boost=2,site=dataorienteddesign.com
+$boost=2,site=adapteva.com
+$boost=2,site=waldo.jaquith.org
+$boost=2,site=sentient.ai
+$boost=2,site=cheeaun.com
+$boost=2,site=adam.chlipala.net
+$boost=2,site=fastmail.com
+$boost=2,site=law.duke.edu
+$boost=2,site=uglyduck.ca
+$boost=2,site=labs.sentinelone.com
+$boost=2,site=whoisnnamdi.com
+$boost=2,site=eagain.net
+$boost=2,site=docs.fedoraproject.org
+$boost=2,site=branchfree.org
+$boost=2,site=howmanypeopleareinspacerightnow.com
+$boost=2,site=devhints.io
+$boost=2,site=http.cat
+$boost=2,site=aspi.org.au
+$boost=2,site=lalo.li
+$boost=2,site=cyanogenmod.com
+$boost=2,site=math.brown.edu
+$boost=2,site=embracethered.com
+$boost=2,site=codemanship.co.uk
+$boost=2,site=bombthrower.com
+$boost=2,site=open.dgraph.io
+$boost=2,site=dlewis.net
+$boost=2,site=infoworld.com
+$boost=2,site=blog.evanweaver.com
+$boost=2,site=webdocs.cs.ualberta.ca
+$boost=2,site=runnable.com
+$boost=2,site=scienceblog.com
+$boost=2,site=caterina.net
+$boost=2,site=cockeyed.com
+$boost=2,site=garry.tv
+$boost=2,site=tech.mattmillman.com
+$boost=2,site=pavelfatin.com
+$boost=2,site=blog.testmunk.com
+$boost=2,site=playnice.ly
+$boost=2,site=thedriven.io
+$boost=2,site=rr-project.org
+$boost=2,site=vis4.net
+$boost=2,site=purelyfunctional.tv
+$boost=2,site=cachestocaches.com
+$boost=2,site=novelr.com
+$boost=2,site=myasmine.com
+$boost=2,site=hanshq.net
+$boost=2,site=powazek.com
+$boost=2,site=mjt.me.uk
+$boost=2,site=businessinsider.fr
+$boost=2,site=benmetcalfe.com
+$boost=2,site=cloudant.com
+$boost=2,site=buddycloud.com
+$boost=2,site=devinhelton.com
+$boost=2,site=blog.onedrive.com
+$boost=2,site=hackoftheday.securitytube.net
+$boost=2,site=news.rpi.edu
+$boost=2,site=kyso.io
+$boost=2,site=blog.cryptoaustralia.org.au
+$boost=2,site=pixelmonkey.org
+$boost=2,site=pandoralive.info
+$boost=2,site=lessmilk.com
+$boost=2,site=blog.everyblock.com
+$boost=2,site=blog.keep.network
+$boost=2,site=dpreview.com
+$boost=2,site=blog.calbucci.com
+$boost=2,site=blogs.discovery.com
+$boost=2,site=flic.kr
+$boost=2,site=bostonmagazine.com
+$boost=2,site=haydenplanetarium.org
+$boost=2,site=lukesmith.xyz
+$boost=2,site=blog.djy.io
+$boost=2,site=treasury.gov
+$boost=2,site=abovethecrowd.com
+$boost=2,site=yogthos.net
+$boost=2,site=adn.com
+$boost=2,site=eveonline.com
+$boost=2,site=blog.metamorphium.com
+$boost=2,site=dannyvankooten.com
+$boost=2,site=creativereview.co.uk
+$boost=2,site=blog.hasura.io
+$boost=2,site=dosync.posterous.com
+$boost=2,site=openreview.net
+$boost=2,site=blog.serverdensity.com
+$boost=2,site=manybutfinite.com
+$boost=2,site=rbth.com
+$boost=2,site=fefe.de
+$boost=2,site=mcla.ug
+$boost=2,site=chem.tufts.edu
+$boost=2,site=tantek.com
+$boost=2,site=greig.cc
+$boost=2,site=xaprb.com
+$boost=2,site=badssl.com
+$boost=2,site=bbirnbaum.com
+$boost=2,site=ruudvanasseldonk.com
+$boost=2,site=softwareatscale.dev
+$boost=2,site=heap.engineering
+$boost=2,site=norswap.com
+$boost=2,site=reactionwheel.net
+$boost=2,site=rasterize.io
+$boost=2,site=adlervermillion.com
+$boost=2,site=timeless.judofyr.net
+$boost=2,site=githut.info
+$boost=2,site=gadgette.com
+$boost=2,site=frantic.im
+$boost=2,site=creativeapplications.net
+$boost=2,site=wecode.wepay.com
+$boost=2,site=cs.ucdavis.edu
+$boost=2,site=data.heapanalytics.com
+$boost=2,site=blog.mixergy.com
+$boost=2,site=sudo.ws
+$boost=2,site=simplystatistics.org
+$boost=2,site=pbpython.com
+$boost=2,site=techinasia.com
+$boost=2,site=scirra.com
+$boost=2,site=mrmacintosh.com
+$boost=2,site=omarabid.com
+$boost=2,site=reactkungfu.com
+$boost=2,site=paulbridger.com
+$boost=2,site=globalpost.com
+$boost=2,site=david-smith.org
+$boost=2,site=philipotoole.com
+$boost=2,site=20bits.com
+$boost=2,site=kickscondor.com
+$boost=2,site=blog.zdsmith.com
+$boost=2,site=iclarified.com
+$boost=2,site=x-plane.com
+$boost=2,site=mirage.io
+$boost=2,site=blog.expensify.com
+$boost=2,site=hedgehogreview.com
+$boost=2,site=simple.com
+$boost=2,site=blog.laptopmag.com
+$boost=2,site=kubuszok.com
+$boost=2,site=whatblag.com
+$boost=2,site=insight.kellogg.northwestern.edu
+$boost=2,site=skerritt.blog
+$boost=2,site=shacknews.com
+$boost=2,site=hcn.org
+$boost=2,site=desmoinesregister.com
+$boost=2,site=intelligentliving.co
+$boost=2,site=nextbigfuture.com
+$boost=2,site=arneswinnen.net
+$boost=2,site=bellingcat.com
+$boost=2,site=blog.delicious.com
+$boost=2,site=hackthesystem.com
+$boost=2,site=blog.comae.io
+$boost=2,site=altdevblogaday.org
+$boost=2,site=w2.eff.org
+$boost=2,site=blog.dominodatalab.com
+$boost=2,site=burlingtonfreepress.com
+$boost=2,site=betterwebtype.com
+$boost=2,site=tech.yipit.com
+$boost=2,site=capitolhillseattle.com
+$boost=2,site=flyosity.com
+$boost=2,site=juffalow.com
+$boost=2,site=alexstjohn.com
+$boost=2,site=awardwinningfjords.com
+$boost=2,site=humanbioscience.org
+$boost=2,site=gao.gov
+$boost=2,site=sci-news.com
+$boost=2,site=teslarati.com
+$boost=2,site=creativedeletion.com
+$boost=2,site=hushkit.net
+$boost=2,site=sketch.com
+$boost=2,site=governmentattic.org
+$boost=2,site=iflscience.com
+$boost=2,site=bmwblog.com
+$boost=2,site=papercups.io
+$boost=2,site=deeplearning4j.org
+$boost=2,site=kaiostech.com
+$boost=2,site=kdab.com
+$boost=2,site=sci-hub.tw
+$boost=2,site=mobilecrunch.com
+$boost=2,site=blog.liangzan.net
+$boost=2,site=paulcavallaro.com
+$boost=2,site=freebsdnews.net
+$boost=2,site=cs.dartmouth.edu
+$boost=2,site=slashdot.org
+$boost=2,site=beta.theglobeandmail.com
+$boost=2,site=franzoni.eu
+$boost=2,site=unz.com
+$boost=2,site=launch.gust.com
+$boost=2,site=mattlakeman.org
+$boost=2,site=us.battle.net
+$boost=2,site=vitess.io
+$boost=2,site=parsonsmatt.org
+$boost=2,site=sciam.com
+$boost=2,site=blog.inklingmarkets.com
+$boost=2,site=darknetdiaries.com
+$boost=2,site=blog.mixpanel.com
+$boost=2,site=blog.kippt.com
+$boost=2,site=blog.jayfields.com
+$boost=2,site=bionic.fullstory.com
+$boost=2,site=blog.reybango.com
+$boost=2,site=tasvideos.org
+$boost=2,site=nathancunn.com
+$boost=2,site=bls.gov
+$boost=2,site=slashgear.com
+$boost=2,site=get21stnight.com
+$boost=2,site=wadefoster.net
+$boost=2,site=blog.mongohq.com
+$boost=2,site=blog.samwhited.com
+$boost=2,site=lukethomas.com
+$boost=2,site=johnmurray.io
+$boost=2,site=scpr.org
+$boost=2,site=serverless.com
+$boost=2,site=mintpressnews.com
+$boost=2,site=uxdesign.smashingmagazine.com
+$boost=2,site=dcscience.net
+$boost=2,site=badlogicgames.com
+$boost=2,site=kevinwilliampang.com
+$boost=2,site=benjaminmayo.co.uk
+$boost=2,site=tonymacx86.com
+$boost=2,site=news.firedoglake.com
+$boost=2,site=gitorious.org
+$boost=2,site=indystar.com
+$boost=2,site=nextbigwhat.com
+$boost=2,site=vectorized.io
+$boost=2,site=fakesteve.net
+$boost=2,site=theness.com
+$boost=2,site=blog.zenkit.com
+$boost=2,site=consequenceofsound.net
+$boost=2,site=raccoon.onyxbits.de
+$boost=2,site=linaro.org
+$boost=2,site=insights.hpe.com
+$boost=2,site=webflow.com
+$boost=2,site=hhs.gov
+$boost=2,site=blog.folyo.me
+$boost=2,site=blog.playcanvas.com
+$boost=2,site=oldblog.antirez.com
+$boost=2,site=beekeeperstudio.io
+$boost=2,site=capiche.com
+$boost=2,site=cbp.gov
+$boost=2,site=kapeli.com
+$boost=2,site=blog.sigfpe.com
+$boost=2,site=nianticlabs.com
+$boost=2,site=econlog.econlib.org
+$boost=2,site=bitsofco.de
+$boost=2,site=blog.gyrosco.pe
+$boost=2,site=blog.n01se.net
+$boost=2,site=iamwil.posterous.com
+$boost=2,site=minyanville.com
+$boost=2,site=jcjc-dev.com
+$boost=2,site=decryptmedia.com
+$boost=2,site=alexvermeer.com
+$boost=2,site=reddragdiva.dreamwidth.org
+$boost=2,site=anarchogeek.com
+$boost=2,site=truthonthemarket.com
+$boost=2,site=numenta.com
+$boost=2,site=perceptualedge.com
+$boost=2,site=garykessler.net
+$boost=2,site=loopj.com
+$boost=2,site=newteevee.com
+$boost=2,site=morningconsult.com
+$boost=2,site=mrmrs.io
+$boost=2,site=springboard.com
+$boost=2,site=computinghistory.org.uk
+$boost=2,site=ownlocal.com
+$boost=2,site=wellcomecollection.org
+$boost=2,site=linuxreviews.org
+$boost=2,site=monadical.com
+$boost=2,site=gamesfromwithin.com
+$boost=2,site=universal-sci.com
+$boost=2,site=esoteric.codes
+$boost=2,site=petermichaud.com
+$boost=2,site=laptopmag.com
+$boost=2,site=3news.co.nz
+$boost=2,site=sel4.systems
+$boost=2,site=accountingtoday.com
+$boost=2,site=host.madison.com
+$boost=2,site=kdenlive.org
+$boost=2,site=bitstamp.net
+$boost=2,site=stonemaiergames.com
+$boost=2,site=ceronman.com
+$boost=2,site=airmail.news
+$boost=2,site=nickwignall.com
+$boost=2,site=develop-online.net
+$boost=2,site=brunocalza.me
+$boost=2,site=gadgetopia.com
+$boost=2,site=seekingalpha.com
+$boost=2,site=jofreeman.com
+$boost=2,site=muckandbrass.com
+$boost=2,site=vg247.com
+$boost=2,site=chetansurpur.com
+$boost=2,site=deislabs.io
+$boost=2,site=philpapers.org
+$boost=2,site=rubygems.org
+$boost=2,site=laquadrature.net
+$boost=2,site=dheera.net
+$boost=2,site=use.perl.org
+$boost=2,site=dannorth.net
+$boost=2,site=orbitalindex.com
+$boost=2,site=historyofyesterday.com
+$boost=2,site=citymetric.com
+$boost=2,site=codecommit.com
+$boost=2,site=danlec.com
+$boost=2,site=endlessparentheses.com
+$boost=2,site=vrfinal.com
+$boost=2,site=the-american-interest.com
+$boost=2,site=elinux.org
+$boost=2,site=immadsnewworld.com
+$boost=2,site=kylecordes.com
+$boost=2,site=ndss-symposium.org
+$boost=2,site=accu.org
+$boost=2,site=americans.org
+$boost=2,site=languagengine.co
+$boost=2,site=al-monitor.com
+$boost=2,site=marksimonson.com
+$boost=2,site=danilocampos.com
+$boost=2,site=mathpages.com
+$boost=2,site=nakamotoinstitute.org
+$boost=2,site=blog.ionelmc.ro
+$boost=2,site=strawpoll.me
+$boost=2,site=theweek.co.uk
+$boost=2,site=morningbrew.com
+$boost=2,site=ozar.me
+$boost=2,site=9fans.net
+$boost=2,site=pythonhosted.org
+$boost=2,site=fark.com
+$boost=2,site=techendo.co
+$boost=2,site=downlode.org
+$boost=2,site=upstart.me
+$boost=2,site=extendslogic.com
+$boost=2,site=theundefeated.com
+$boost=2,site=blog.appsignal.com
+$boost=2,site=informationarchitects.jp
+$boost=2,site=intrade.com
+$boost=2,site=lukasz.langa.pl
+$boost=2,site=energy.gov
+$boost=2,site=talkpython.fm
+$boost=2,site=yourbasic.org
+$boost=2,site=sachinagarwal.com
+$boost=2,site=tech.adroll.com
+$boost=2,site=bowery.io
+$boost=2,site=blog.sourcing.io
+$boost=2,site=bostinno.com
+$boost=2,site=jatan.space
+$boost=2,site=evrone.com
+$boost=2,site=labs.domipheus.com
+$boost=2,site=thrillist.com
+$boost=2,site=gizmag.com
+$boost=2,site=danweinreb.org
+$boost=2,site=tyk.io
+$boost=2,site=mymicrocontroller.com
+$boost=2,site=samharris.org
+$boost=2,site=blog.tildeloop.com
+$boost=2,site=discuss.emberjs.com
+$boost=2,site=blog.superuser.com
+$boost=2,site=oliveremberton.com
+$boost=2,site=vpnpro.com
+$boost=2,site=lastwordonnothing.com
+$boost=2,site=stevenf.com
+$boost=2,site=cornellsun.com
+$boost=2,site=letoverlambda.com
+$boost=2,site=snapcraft.io
+$boost=2,site=fastcocreate.com
+$boost=2,site=mixmax.com
+$boost=2,site=blog.goodaudience.com
+$boost=2,site=wikitravel.org
+$boost=2,site=elzr.com
+$boost=2,site=gerger.co
+$boost=2,site=developer.amazonwebservices.com
+$boost=2,site=blog.postmarkapp.com
+$boost=2,site=gamepro.com
+$boost=2,site=sencha.com
+$boost=2,site=christianheilmann.com
+$boost=2,site=mg.co.za
+$boost=2,site=warhistoryonline.com
+$boost=2,site=sakurity.com
+$boost=2,site=blog.yalabot.com
+$boost=2,site=securityweek.com
+$boost=2,site=myce.com
+$boost=2,site=progressiveapp.store
+$boost=2,site=teamblog.supportbee.com
+$boost=2,site=strangelyconsistent.org
+$boost=2,site=themarysue.com
+$boost=2,site=attendly.com
+$boost=2,site=snell-pym.org.uk
+$boost=2,site=influxdb.com
+$boost=2,site=worksinprogress.co
+$boost=2,site=scripps.edu
+$boost=2,site=dissenter.firedoglake.com
+$boost=2,site=crackstation.net
+$boost=2,site=afar.com
+$boost=2,site=hasura.io
+$boost=2,site=mtholyoke.edu
+$boost=2,site=blog.startifact.com
+$boost=2,site=colourlovers.com
+$boost=2,site=gettingmoreawesome.com
+$boost=2,site=themorningnews.org
+$boost=2,site=hardcoresoftware.learningbyshipping.com
+$boost=2,site=bramlambrecht.com
+$boost=2,site=spacesafetymagazine.com
+$boost=2,site=suckless.org
+$boost=2,site=staticgen.com
+$boost=2,site=thefader.com
+$boost=2,site=independent.ie
+$boost=2,site=pivotallabs.com
+$boost=2,site=opsian.com
+$boost=2,site=taigeair.com
+$boost=2,site=talkingpointsmemo.com
+$boost=2,site=williampross.com
+$boost=2,site=popmotion.io
+$boost=2,site=robb.weblaws.org
+$boost=2,site=globaldev.co.uk
+$boost=2,site=pencerw.com
+$boost=2,site=michelenasti.com
+$boost=2,site=sourcesort.com
+$boost=2,site=monash.edu
+$boost=2,site=haydenjames.io
+$boost=2,site=blog.webb.page
+$boost=2,site=tcfast.com
+$boost=2,site=blog.fikesfarm.com
+$boost=2,site=telecompaper.com
+$boost=2,site=mikegerwitz.com
+$boost=2,site=catless.ncl.ac.uk
+$boost=2,site=techflash.com
+$boost=2,site=omis.me
+$boost=2,site=anildash.com
+$boost=2,site=authy.com
+$boost=2,site=bitterwinter.org
+$boost=2,site=news.illinois.edu
+$boost=2,site=trilema.com
+$boost=2,site=nshipster.com
+$boost=2,site=uppy.io
+$boost=2,site=probmods.org
+$boost=2,site=labs.umbrella.com
+$boost=2,site=lawyersgunsmoneyblog.com
+$boost=2,site=tech.okcupid.com
+$boost=2,site=jseliger.com
+$boost=2,site=openrightsgroup.org
+$boost=2,site=zendev.com
+$boost=2,site=deepsense.ai
+$boost=2,site=reading.supply
+$boost=2,site=fullstackpython.com
+$boost=2,site=cloudrail.com
+$boost=2,site=crt.sh
+$boost=2,site=charity.wtf
+$boost=2,site=gettingreal.37signals.com
+$boost=2,site=blog.spiderlabs.com
+$boost=2,site=javascript.christmas
+$boost=2,site=ruby-forum.com
+$boost=2,site=crypt.codemancers.com
+$boost=2,site=aomedia.org
+$boost=2,site=physics-astronomy.org
+$boost=2,site=av8n.com
+$boost=2,site=signalsandthreads.com
+$boost=2,site=apptorium.com
+$boost=2,site=smarterware.org
+$boost=2,site=toys.lerdorf.com
+$boost=2,site=blog.handcraft.com
+$boost=2,site=alex-ionescu.com
+$boost=2,site=team.adzerk.com
+$boost=2,site=trottercashion.com
+$boost=2,site=dylanpaulus.com
+$boost=2,site=shiroyasha.io
+$boost=2,site=sunlightlabs.com
+$boost=2,site=johnbender.us
+$boost=2,site=orionmagazine.org
+$boost=2,site=500ish.com
+$boost=2,site=eclogiselle.com
+$boost=2,site=j00ru.vexillium.org
+$boost=2,site=partow.net
+$boost=2,site=zpub.com
+$boost=2,site=news.usni.org
+$boost=2,site=charemza.name
+$boost=2,site=gondor.io
+$boost=2,site=starcraft.com
+$boost=2,site=sorentwo.com
+$boost=2,site=thebrowser.com
+$boost=2,site=sslmate.com
+$boost=2,site=blog.nearlyfreespeech.net
+$boost=2,site=blog.peer5.com
+$boost=2,site=understandlegacycode.com
+$boost=2,site=www3.imperial.ac.uk
+$boost=2,site=marketingland.com
+$boost=2,site=plainsite.org
+$boost=2,site=fabcapo.com
+$boost=2,site=osohq.com
+$boost=2,site=marc.merlins.org
+$boost=2,site=satyajit.ranjeev.in
+$boost=2,site=transcend.io
+$boost=2,site=elweb.co
+$boost=2,site=arqbackup.com
+$boost=2,site=turnkeylinux.org
+$boost=2,site=preciousplastic.com
+$boost=2,site=stonetemple.com
+$boost=2,site=oeis.org
+$boost=2,site=bath.ac.uk
+$boost=2,site=theodi.org
+$boost=2,site=saucelabs.com
+$boost=2,site=educative.io
+$boost=2,site=sway.com
+$boost=2,site=datacolada.org
+$boost=2,site=convox.com
+$boost=2,site=tokbox.com
+$boost=2,site=review.firstround.com
+$boost=2,site=netmagazine.com
+$boost=2,site=stanfordalumni.org
+$boost=2,site=fox5ny.com
+$boost=2,site=phx.corporate-ir.net
+$boost=2,site=gist.githubusercontent.com
+$boost=2,site=welivesecurity.com
+$boost=2,site=linfo.org
+$boost=2,site=vikingcodeschool.com
+$boost=2,site=talsraviv.com
+$boost=2,site=brainspl.at
+$boost=2,site=sefsar.com
+$boost=2,site=anmolsarma.in
+$boost=2,site=mikekohn.net
+$boost=2,site=doc-ok.org
+$boost=2,site=duncandavidson.com
+$boost=2,site=redbullstratos.com
+$boost=2,site=lodev.org
+$boost=2,site=sqreen.io
+$boost=2,site=weburbanist.com
+$boost=2,site=homelandsecuritynewswire.com
+$boost=2,site=yuiblog.com
+$boost=2,site=mgalligan.com
+$boost=2,site=startupsanonymous.com
+$boost=2,site=commentary.org
+$boost=2,site=paidcontent.org
+$boost=2,site=news.voyage.auto
+$boost=2,site=blog.safia.rocks
+$boost=2,site=pagely.com
+$boost=2,site=huckmagazine.com
+$boost=2,site=amplitude.com
+$boost=2,site=academicvc.com
+$boost=2,site=zurb.com
+$boost=2,site=fogcreek.com
+$boost=2,site=pubs.aeaweb.org
+$boost=2,site=ochronus.online
+$boost=2,site=mukulrathi.co.uk
+$boost=2,site=tinfoilsecurity.com
+$boost=2,site=blog.intercom.io
+$boost=2,site=counterpointresearch.com
+$boost=2,site=caines.ca
+$boost=2,site=bitmessage.org
+$boost=2,site=sameroom.io
+$boost=2,site=grapheneos.org
+$boost=2,site=ryanhoover.me
+$boost=2,site=ghost.org
+$boost=2,site=ably.com
+$boost=2,site=dgraph.io
+$boost=2,site=pcg-random.org
+$boost=2,site=electronicdesign.com
+$boost=2,site=blog.benjamin-encz.de
+$boost=2,site=blog.apollographql.com
+$boost=2,site=interaktiv.morgenpost.de
+$boost=2,site=neogaf.com
+$boost=2,site=jasongullickson.posterous.com
+$boost=2,site=lvguowei.me
+$boost=2,site=theapeiron.co.uk
+$boost=2,site=insideairbnb.com
+$boost=2,site=business-standard.com
+$boost=2,site=blog.chiariglione.org
+$boost=2,site=news.byu.edu
+$boost=2,site=ico.org.uk
+$boost=2,site=asktog.com
+$boost=2,site=thelightphone.com
+$boost=2,site=weblog.muledesign.com
+$boost=2,site=functional.christmas
+$boost=2,site=dendory.net
+$boost=2,site=css-irl.info
+$boost=2,site=blog.rubybestpractices.com
+$boost=2,site=buzzsumo.com
+$boost=2,site=edinburghhacklab.com
+$boost=2,site=uc.edu
+$boost=2,site=deprogrammaticaipsum.com
+$boost=2,site=blog.teamtreehouse.com
+$boost=2,site=poynter.org
+$boost=2,site=blog.thirdyearmba.com
+$boost=2,site=hackingwithswift.com
+$boost=2,site=43folders.com
+$boost=2,site=blog.siliconpublishing.com
+$boost=2,site=psiloveyou.xyz
+$boost=2,site=thenewinquiry.com
+$boost=2,site=grc.com
+$boost=2,site=yle.fi
+$boost=2,site=buildzoom.com
+$boost=2,site=kmjn.org
+$boost=2,site=miaminewtimes.com
+$boost=2,site=pipedream.com
+$boost=2,site=xoofx.com
+$boost=2,site=9elements.com
+$boost=2,site=safetyresearch.net
+$boost=2,site=memcached.org
+$boost=2,site=letterstoanewdeveloper.com
+$boost=2,site=viewfromthewing.boardingarea.com
+$boost=2,site=nanxiao.me
+$boost=2,site=ferrucc.io
+$boost=2,site=blog.mrmeyer.com
+$boost=2,site=dailycal.org
+$boost=2,site=nbcphiladelphia.com
+$boost=2,site=ww2.kqed.org
+$boost=2,site=nitter.net
+$boost=2,site=mathieupassenaud.fr
+$boost=2,site=uploadvr.com
+$boost=2,site=lincolnloop.com
+$boost=2,site=gaming.icrontic.com
+$boost=2,site=techvibes.com
+$boost=2,site=news.gatech.edu
+$boost=2,site=leastauthority.com
+$boost=2,site=find.xyz
+$boost=2,site=rhodesmill.org
+$boost=2,site=authzed.com
+$boost=2,site=agweb.com
+$boost=2,site=allegro.tech
+$boost=2,site=nex-3.com
+$boost=2,site=macaw.co
+$boost=2,site=nic.ferrier.me.uk
+$boost=2,site=planetscale.com
+$boost=2,site=arshadchowdhury.com
+$boost=2,site=adbusters.org
+$boost=2,site=noteflakes.com
+$boost=2,site=blog.ellenchisa.com
+$boost=2,site=ajaxian.com
+$boost=2,site=getchip.com
+$boost=2,site=dutchnews.nl
+$boost=2,site=efavdb.com
+$boost=2,site=mabez.dev
+$boost=2,site=azul.com
+$boost=2,site=bonappetit.com
+$boost=2,site=microconf.com
+$boost=2,site=mxb.at
+$boost=2,site=rancher.com
+$boost=2,site=antoniocangiano.com
+$boost=2,site=modernretail.co
+$boost=2,site=blog.veitheller.de
+$boost=2,site=thehistoryblog.com
+$boost=2,site=javacodegeeks.com
+$boost=2,site=blog.stevensanderson.com
+$boost=2,site=foundingfuel.com
+$boost=2,site=simpleprogrammer.com
+$boost=2,site=sigusr2.net
diff --git a/testcases/goggles/quickstart.goggle b/testcases/goggles/quickstart.goggle
new file mode 100644
index 00000000..2b23842e
--- /dev/null
+++ b/testcases/goggles/quickstart.goggle
@@ -0,0 +1,111 @@
+! name: Quickstart
+! description: A quick tour of Goggles
+! public: false
+! author: Goggles 101
+
+! Goggles are simple, self-contained text files which can be hosted in Github or
+! Gitlab. These files contain instructions allowing you to tell the search engine how
+! you'd like your results to be ranked. You can target specific URL patterns
+! (and, soon, website titles and other aspects of Web pages) and indicate how
+! their ranking should be altered (e.g. boosted, downranked, or completely
+! discarded from the results).
+
+! A Goggle source file consists of instructions, one per line. Empty lines, or
+! comments (starting with an exclamation mark: '!') are ignored.
+
+! You probably already noticed the header of this file which contains some
+! metadata about your Goggle, such as: name, description, public and author.
+! These are *mandatory*.
+
+! Additionally, you can specify the following optional metadata attributes:
+! * homepage — specifies a homepage URL displayed on your Goggle's profile.
+! * issues — specifies a URL where users can report issues for your Goggle.
+! * transferred_to — Allows to transfer ownership of a Goggle.
+! * avatar — specifies a *valid* HEX color code for your Goggle.
+! * license — specifies the license of a Goggle's instructions.
+
+! The simplest instruction is a plain-text pattern which can be found in URLs.
+! The following would match any search result whose URL contains the pattern
+! as-is:
+/this/is/a/pattern
+
+! It is also possible to use some limited "globbing" capabilities such as the
+! '*' character which will match zero, one or more characters (note: the number
+! of wildcards allowed in a given instruction is limited):
+/this/is/*/pattern
+
+! By default, a pattern can match anywhere in the URL, but there are specific
+! characters which can be used to indicate prefix or suffix matches: we call them
+! "anchors".
+!
+! The '|' character can be used at the beginning or end of an instruction to
+! indicate anchoring. The following instruction will match a prefix of the URL:
+|https://en.
+
+! The following will match a suffix of the URL:
+/some/path.html|
+
+! Both can be used at the same time:
+|https://example.com|
+
+! Additionally, each instruction can specify a list of options, following the
+! '$' character and separated by commas (','). Options can be used to more finely
+! target specific search results, or to indicate how a matched result's ranking
+! should be altered.
+
+! The most basic option is 'site=', which can be used to limit a instruction to
+! a specific website, based on its domain. Options can be specified on their own
+! (e.g. if you want to target any page of a site) or in conjunction with a
+! pattern:
+$site=example.com
+/blog/$site=example.com
+
+! Another set of options can be used to indicate what you want your instruction
+! to target. By default any instruction will apply to a URL, but we will add the
+! ability to match other aspects of a page too, in the future:
+!
+! web3$inurl
+! web3$intitle
+! web3$indescription
+! web3$incontent
+
+! Finally, you can specify an 'action', which indicates how the ranking of a
+! matched result should be changed by your instruction. This is the mechanism
+! through which you can customize the ranking of results to your liking. You can
+! use one of three possible actions in your instructions, and by default, any
+! instruction without an action will be considered as 'boost':
+/r/rust/
+/r/rust/$boost
+/r/rust/$boost=2
+/r/rust/$boost=3
+
+! The value associated with the option indicate the 'strength' with which you
+! want to alter the ranking (note: that it is currently limited to a maximum
+! value of 10). It can be used to boost results differently, even inside of the
+! same Goggle (e.g. some results should be favored more than others).
+!
+! You can also downrank results:
+/r/google/$downrank
+/r/google/$downrank=2
+/r/google/$downrank=3
+
+! Tou can also discard results completely:
+$discard,site=idontwanttobepartoftheresults.com
+/this/is/spam/$discard
+
+! Last but not least, you can control exactly how the signals gets prioritized for
+! the final ranking. We countinously add more signals used during ranking,
+! but you can e.g. heavily prioritize webpages with a high centrality by specifying
+@host_centrality = 2048
+
+! Or you can boost matches with a specific field with the 'field' prefix e.g
+@field_title = 50
+
+! Individually, each instruction can either target a very narrow set of pages (or
+! even a single page), or a wider range of them, to apply reranking to a bigger
+! set of results. In combination, hundreds or more instructions can allow you to
+! express complex reranking functions.
+
+! Although the Goggles language could express instructions to search through a
+! small set of websites or act as a blocklist, Goggles really shine when used to
+! express boosting and downranking across many domains and pages.
\ No newline at end of file
diff --git a/testcases/parsing/byte_index_out_of_bounds.html b/testcases/parsing/byte_index_out_of_bounds.html
new file mode 100644
index 00000000..fb99648e
--- /dev/null
+++ b/testcases/parsing/byte_index_out_of_bounds.html
@@ -0,0 +1,167 @@
+<html>
+
+<head>
+ <title>Test</title>
+</head>
+
+<body>
+ <div>
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ παγαινετ
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ This is a test website
+ </div>
+</body>
+
+</html>
\ No newline at end of file