Skip to content

Hypergraph is more powerful. Transform unstructured text into structured knowledge with LLMs. Graphs, hypergraphs, and spatio-temporal extractions — with one command.

style: fix ruff lint (F541) and format issues

yifanfeng97f690cc4e4939
diff --git a/hyperextract/cli/cli.py b/hyperextract/cli/cli.py
index 83d4f87..fd68607 100644
--- a/hyperextract/cli/cli.py
+++ b/hyperextract/cli/cli.py
@@ -237,7 +237,13 @@ def parse(
),
):
"""Extract knowledge from text to a new directory."""
- logger.info("command=parse input=%s output=%s template=%s lang=%s", input, output, template or "auto", lang or "auto")
+ logger.info(
+ "command=parse input=%s output=%s template=%s lang=%s",
+ input,
+ output,
+ template or "auto",
+ lang or "auto",
+ )
validate_config()
logger.info("stage=config_validated")
@@ -605,7 +611,12 @@ def talk(
),
):
"""Chat with Knowledge Abstract."""
- logger.info("command=talk ka_path=%s query=%s interactive=%s", ka_path, query or "loop", interactive)
+ logger.info(
+ "command=talk ka_path=%s query=%s interactive=%s",
+ ka_path,
+ query or "loop",
+ interactive,
+ )
validate_config()
path = validate_ka_with_index(ka_path)
diff --git a/hyperextract/cli/commands/config.py b/hyperextract/cli/commands/config.py
index 1bf93b4..da550b8 100644
--- a/hyperextract/cli/commands/config.py
+++ b/hyperextract/cli/commands/config.py
@@ -288,12 +288,17 @@ def init(
),
):
"""Initialize configuration interactively."""
- logger.info("command=config-init provider=%s api_key_provided=%s", provider, api_key is not None)
+ logger.info(
+ "command=config-init provider=%s api_key_provided=%s",
+ provider,
+ api_key is not None,
+ )
config = ConfigManager()
# Quick mode: provider + api_key provided
if provider and api_key:
from hyperextract.utils.client import PROVIDER_PRESETS
+
preset = PROVIDER_PRESETS.get(provider, {})
llm_model = preset.get("default_llm") or "gpt-4o-mini"
emb_model = preset.get("default_embedder") or "text-embedding-3-small"
@@ -314,7 +319,11 @@ def init(
base_url=resolved_base,
)
else:
- console.print("[yellow]Warning: Provider '{}' has no default embedder. Please configure embedder separately.[/yellow]".format(provider))
+ console.print(
+ "[yellow]Warning: Provider '{}' has no default embedder. Please configure embedder separately.[/yellow]".format(
+ provider
+ )
+ )
console.print("[bold green]Configuration saved successfully![/bold green]")
console.print()
@@ -359,10 +368,10 @@ def init(
console.print("[bold]Step 1: Choose Provider[/bold]")
providers = [
- ("openai", "OpenAI", "https://api.openai.com/v1"),
- ("bailian", "阿里云百炼", "https://dashscope.aliyuncs.com/compatible-mode/v1"),
- ("vllm", "本地 vLLM", "自定义地址"),
- ("custom", "其他 OpenAI 兼容接口", "自定义地址"),
+ ("openai", "OpenAI", "https://api.openai.com/v1"),
+ ("bailian", "阿里云百炼", "https://dashscope.aliyuncs.com/compatible-mode/v1"),
+ ("vllm", "本地 vLLM", "自定义地址"),
+ ("custom", "其他 OpenAI 兼容接口", "自定义地址"),
]
for i, (key, name, url) in enumerate(providers, 1):
console.print(f" [{i}] {name:<20} ({url})")
@@ -382,11 +391,20 @@ def init(
console.print(f"[bold]Step 2: LLM Configuration (Provider: {selected})[/bold]")
if selected == "vllm":
- llm_model = console.input(f" LLM Model: ").strip()
- llm_base_url = console.input(f" LLM Base URL (e.g. http://localhost:8000/v1): ").strip()
+ llm_model = console.input(" LLM Model: ").strip()
+ llm_base_url = console.input(
+ " LLM Base URL (e.g. http://localhost:8000/v1): "
+ ).strip()
else:
- llm_model = console.input(f" Model (default: {default_llm}): ").strip() or default_llm
- llm_base_url = console.input(f" Base URL (default: {preset_url}, press Enter to skip): ").strip() or preset_url
+ llm_model = (
+ console.input(f" Model (default: {default_llm}): ").strip() or default_llm
+ )
+ llm_base_url = (
+ console.input(
+ f" Base URL (default: {preset_url}, press Enter to skip): "
+ ).strip()
+ or preset_url
+ )
llm_api_key = None
while not llm_api_key:
@@ -396,7 +414,9 @@ def init(
llm_api_key = "dummy"
console.print(" [dim]Using 'dummy' for vLLM[/dim]")
break
- console.print(" [red]API Key is required. Please enter your API key.[/red]")
+ console.print(
+ " [red]API Key is required. Please enter your API key.[/red]"
+ )
config.set_llm(
provider=selected,
@@ -411,10 +431,19 @@ def init(
if selected == "vllm":
emb_model = console.input(" Embedder Model (e.g. bge-m3): ").strip()
- emb_base_url = console.input(" Embedder Base URL (e.g. http://localhost:8001/v1): ").strip()
+ emb_base_url = console.input(
+ " Embedder Base URL (e.g. http://localhost:8001/v1): "
+ ).strip()
else:
- emb_model = console.input(f" Model (default: {default_emb}): ").strip() or default_emb
- emb_base_url = console.input(f" Base URL (default: {preset_url}, press Enter to skip): ").strip() or preset_url
+ emb_model = (
+ console.input(f" Model (default: {default_emb}): ").strip() or default_emb
+ )
+ emb_base_url = (
+ console.input(
+ f" Base URL (default: {preset_url}, press Enter to skip): "
+ ).strip()
+ or preset_url
+ )
emb_api_key = None
while not emb_api_key:
@@ -424,7 +453,9 @@ def init(
emb_api_key = "dummy"
console.print(" [dim]Using 'dummy' for vLLM[/dim]")
break
- console.print(" [red]API Key is required. Please enter your API key.[/red]")
+ console.print(
+ " [red]API Key is required. Please enter your API key.[/red]"
+ )
config.set_embedder(
provider=selected,
diff --git a/hyperextract/cli/commands/list.py b/hyperextract/cli/commands/list.py
index ed24625..6728229 100644
--- a/hyperextract/cli/commands/list.py
+++ b/hyperextract/cli/commands/list.py
@@ -42,7 +42,9 @@ def template(
),
):
"""List all available templates with detailed information."""
- logger.info("command=list-template query=%s autotype=%s lang=%s", query, autotype, language)
+ logger.info(
+ "command=list-template query=%s autotype=%s lang=%s", query, autotype, language
+ )
target_lang = language if language else "en"
results = Gallery.list(
diff --git a/hyperextract/types/base.py b/hyperextract/types/base.py
index 2493c67..d43bc1f 100644
--- a/hyperextract/types/base.py
+++ b/hyperextract/types/base.py
@@ -247,7 +247,11 @@ class BaseAutoType(ABC, Generic[T]):
def _extract_data(self, text: str) -> T:
"""Internal: Unified extraction logic (Chunking -> LLM -> Merge)."""
- logger.debug("stage=extract_start input_chars=%d chunk_size=%d", len(text), self.chunk_size)
+ logger.debug(
+ "stage=extract_start input_chars=%d chunk_size=%d",
+ len(text),
+ self.chunk_size,
+ )
if len(text) <= self.chunk_size:
logger.debug("stage=extract_single_chunk chunk_text_preview=%s", text[:200])
@@ -263,23 +267,35 @@ class BaseAutoType(ABC, Generic[T]):
for i, chunk in enumerate(chunks):
logger.debug(
"stage=chunk_before_llm chunk_index=%d chunk_chars=%d chunk_text_preview=%s",
- i, len(chunk), chunk[:200],
+ i,
+ len(chunk),
+ chunk[:200],
)
inputs = [{"source_text": chunk} for chunk in chunks]
- logger.debug("stage=llm_batch_start max_concurrency=%d num_inputs=%d", self.max_workers, len(inputs))
+ logger.debug(
+ "stage=llm_batch_start max_concurrency=%d num_inputs=%d",
+ self.max_workers,
+ len(inputs),
+ )
extracted_data_list = self.data_extractor.batch(
inputs, config={"max_concurrency": self.max_workers}
)
- logger.debug("stage=llm_batch_complete results=%d", len(extracted_data_list))
+ logger.debug(
+ "stage=llm_batch_complete results=%d", len(extracted_data_list)
+ )
for i, result in enumerate(extracted_data_list):
logger.debug(
"stage=chunk_llm_result chunk_index=%d result_summary=%s",
- i, self._summarize_extracted(result),
+ i,
+ self._summarize_extracted(result),
)
logger.debug("stage=merge_start num_items=%d", len(extracted_data_list))
merged_data = self.merge_batch_data(extracted_data_list)
- logger.debug("stage=extract_complete merged_summary=%s", self._summarize_extracted(merged_data))
+ logger.debug(
+ "stage=extract_complete merged_summary=%s",
+ self._summarize_extracted(merged_data),
+ )
return merged_data
def _summarize_extracted(self, data: T) -> str:
diff --git a/hyperextract/types/graph.py b/hyperextract/types/graph.py
index 6f6df5f..afecdeb 100644
--- a/hyperextract/types/graph.py
+++ b/hyperextract/types/graph.py
@@ -486,7 +486,9 @@ class AutoGraph(
chunks = self.text_splitter.split_text(text)
logger.debug("stage=one_stage_split num_chunks=%d", len(chunks))
inputs = [{"source_text": chunk} for chunk in chunks]
- logger.debug("stage=one_stage_batch_start max_concurrency=%d", self.max_workers)
+ logger.debug(
+ "stage=one_stage_batch_start max_concurrency=%d", self.max_workers
+ )
graph_list = self.data_extractor.batch(
inputs, config={"max_concurrency": self.max_workers}
)
@@ -494,7 +496,11 @@ class AutoGraph(
logger.debug("stage=one_stage_merge_start")
result = self.merge_batch_data(graph_list)
- logger.debug("stage=one_stage_merge_complete nodes=%d edges=%d", len(result.nodes), len(result.edges))
+ logger.debug(
+ "stage=one_stage_merge_complete nodes=%d edges=%d",
+ len(result.nodes),
+ len(result.edges),
+ )
return result
def _extract_data_by_two_stage(
@@ -528,13 +534,21 @@ class AutoGraph(
logger.debug("stage=two_stage_node_extraction_start")
chunk_node_lists = self._extract_nodes_batch(chunks)
total_nodes = sum(len(nl.items) for nl in chunk_node_lists)
- logger.debug("stage=two_stage_node_extraction_complete chunks=%d total_nodes=%d", len(chunk_node_lists), total_nodes)
+ logger.debug(
+ "stage=two_stage_node_extraction_complete chunks=%d total_nodes=%d",
+ len(chunk_node_lists),
+ total_nodes,
+ )
# 3. Batch Extract Edges (Context-aware, returns List[EdgeListSchema])
logger.debug("stage=two_stage_edge_extraction_start")
chunk_edge_lists = self._extract_edges_batch(chunks, chunk_node_lists)
total_edges = sum(len(el.items) for el in chunk_edge_lists)
- logger.debug("stage=two_stage_edge_extraction_complete chunks=%d total_edges=%d", len(chunk_edge_lists), total_edges)
+ logger.debug(
+ "stage=two_stage_edge_extraction_complete chunks=%d total_edges=%d",
+ len(chunk_edge_lists),
+ total_edges,
+ )
# 4. Construct Partial Graphs (Tuple format for merge optimization)
partial_graphs = (
@@ -545,7 +559,11 @@ class AutoGraph(
# 5. Global Merge (passes tuples to merge_batch_data)
logger.debug("stage=two_stage_merge_start")
result = self.merge_batch_data(partial_graphs)
- logger.debug("stage=two_stage_merge_complete nodes=%d edges=%d", len(result.nodes), len(result.edges))
+ logger.debug(
+ "stage=two_stage_merge_complete nodes=%d edges=%d",
+ len(result.nodes),
+ len(result.edges),
+ )
return result
def _extract_nodes_batch(
@@ -654,7 +672,12 @@ class AutoGraph(
Returns:
Merged graph.
"""
- logger.debug("stage=merge_batch_start input_type=%s", "tuple" if not isinstance(data_list_or_tuple[0], self.graph_schema) else "list")
+ logger.debug(
+ "stage=merge_batch_start input_type=%s",
+ "tuple"
+ if not isinstance(data_list_or_tuple[0], self.graph_schema)
+ else "list",
+ )
if isinstance(data_list_or_tuple[0], self.graph_schema):
all_nodes, all_edges = [], []
@@ -680,10 +703,20 @@ class AutoGraph(
all_nodes.extend(node_list)
all_edges.extend(edge_list)
- logger.debug("stage=merge_batch_raw total_nodes=%d total_edges=%d", len(all_nodes), len(all_edges))
+ logger.debug(
+ "stage=merge_batch_raw total_nodes=%d total_edges=%d",
+ len(all_nodes),
+ len(all_edges),
+ )
merged_nodes = self.node_merger.merge(all_nodes) if all_nodes else []
merged_edges = self.edge_merger.merge(all_edges) if all_edges else []
- logger.debug("stage=merge_batch_complete merged_nodes=%d merged_edges=%d deduped_nodes=%d deduped_edges=%d", len(merged_nodes), len(merged_edges), len(all_nodes) - len(merged_nodes), len(all_edges) - len(merged_edges))
+ logger.debug(
+ "stage=merge_batch_complete merged_nodes=%d merged_edges=%d deduped_nodes=%d deduped_edges=%d",
+ len(merged_nodes),
+ len(merged_edges),
+ len(all_nodes) - len(merged_nodes),
+ len(all_edges) - len(merged_edges),
+ )
return self.graph_schema(nodes=merged_nodes, edges=merged_edges)
# ==================== Indexing & Search & Chat ====================
diff --git a/hyperextract/utils/client.py b/hyperextract/utils/client.py
index bb4c759..3b8b6e0 100644
--- a/hyperextract/utils/client.py
+++ b/hyperextract/utils/client.py
@@ -78,6 +78,7 @@ class CompatibleEmbeddings(Embeddings):
# Determine the tiktoken encoding to use for chunking
import tiktoken
+
try:
self._encoding = tiktoken.encoding_for_model(model)
except KeyError:
@@ -191,7 +192,11 @@ def _parse_client_spec(
"model": spec.get("model", ""),
"base_url": spec.get("base_url", ""),
"api_key": spec.get("api_key") or api_key,
- **{k: v for k, v in spec.items() if k not in ("provider", "model", "base_url", "api_key")},
+ **{
+ k: v
+ for k, v in spec.items()
+ if k not in ("provider", "model", "base_url", "api_key")
+ },
}
# Parse string shorthand: provider:model@url
@@ -361,13 +366,17 @@ def create_client(
embedder = provider
# Parse llm config
- llm_config = _parse_client_spec(llm or provider, api_key=api_key, default_kind="llm")
+ llm_config = _parse_client_spec(
+ llm or provider, api_key=api_key, default_kind="llm"
+ )
llm_config.update(kwargs)
# Parse embedder config
# For cloud providers, embedder defaults to llm's provider
embedder_spec = embedder or llm_config.get("provider", "")
- emb_config = _parse_client_spec(embedder_spec, api_key=api_key, default_kind="embedder")
+ emb_config = _parse_client_spec(
+ embedder_spec, api_key=api_key, default_kind="embedder"
+ )
# Build clients
llm_client = create_llm(llm_config, api_key=api_key)