Supercharge your AI agents with data from the web and beyond. Building the library for superintelligence. 🔥
TypeScript aiai-agentsai-crawlerai-scrapingai-searchcrawlerdata-extractionhtml-to-markdownllmmarkdownscraperscrapingweb-crawlerweb-dataweb-data-extractionweb-scraperweb-scrapingweb-searchwebscraping
Update lopdf to use load_metadata method for better performance (#2642)
Abimael Martella45a11d55203
diff --git a/apps/api/native/Cargo.toml b/apps/api/native/Cargo.tomlindex 3172bbb8f..dbce050d4 100644--- a/apps/api/native/Cargo.toml+++ b/apps/api/native/Cargo.toml@@ -12,7 +12,7 @@ crate-type = ["cdylib"]chrono = { version = "0.4", features = ["serde"] }kuchikiki = "0.8.2"lol_html = "2.6.0"-lopdf = "0.38.0"+lopdf = { git = "https://github.com/J-F-Liu/lopdf", rev = "6ee84c30f9226d6752dc8e1892298c9d46ac3eb6" }maud = "0.27.0"napi = { version = "3.0.0", features = ["serde-json", "tokio_rt"] }napi-derive = "3.0.0"diff --git a/apps/api/native/src/pdf.rs b/apps/api/native/src/pdf.rsindex 67f64de42..1678292dd 100644--- a/apps/api/native/src/pdf.rs+++ b/apps/api/native/src/pdf.rs@@ -11,37 +11,17 @@ pub struct PDFMetadata {}fn _get_pdf_metadata(path: &str) -> std::result::Result<PDFMetadata, String> {- let doc = match lopdf::Document::load(path) {- Ok(x) => x,- Err(_) => {- return Err("Failed to load PDF".to_string());+ let metadata = match lopdf::Document::load_metadata(path) {+ Ok(m) => m,+ Err(e) => {+ return Err(format!("Failed to load PDF metadata: {}", e));}};- let num_pages = doc.get_pages().len() as i32;-- let title = doc- .trailer- .get(b"Info")- .and_then(|info| {- info- .as_dict()- .and_then(|info| info.get(b"Title"))- .and_then(lopdf::decode_text_string)- })- .ok()- .or_else(|| {- doc.objects.iter().find_map(|(_i, obj)| {- obj- .as_dict()- .and_then(|obj| obj.get(b"Title"))- .and_then(lopdf::decode_text_string)- .ok()- })- })- .map(|x| x.trim().to_string());-- Ok(PDFMetadata { num_pages, title })+ Ok(PDFMetadata {+ num_pages: metadata.page_count as i32,+ title: metadata.title,+ })}/// Extract metadata from PDF file.