kreuzberg 3.13.1__py3-none-any.whl → 3.13.3__py3-none-any.whl

kreuzberg/_api/main.py

@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+import traceback
 from functools import lru_cache
 from json import dumps, loads
 from typing import TYPE_CHECKING, Annotated, Any, Literal
@@ -69,6 +70,33 @@ def exception_handler(request: Request[Any, Any, Any], exception: KreuzbergError
     )
 
 
+def general_exception_handler(request: Request[Any, Any, Any], exception: Exception) -> Response[Any]:
+    """Temporary handler to catch ALL exceptions for debugging."""
+    error_type = type(exception).__name__
+    error_message = str(exception)
+    traceback_str = traceback.format_exc()
+
+    if request.app.logger:
+        request.app.logger.error(
+            "Unhandled exception",
+            method=request.method,
+            url=str(request.url),
+            error_type=error_type,
+            message=error_message,
+            traceback=traceback_str,
+        )
+
+    return Response(
+        content={
+            "error_type": error_type,
+            "message": error_message,
+            "traceback": traceback_str,
+            "debug": "This is a temporary debug handler",
+        },
+        status_code=HTTP_500_INTERNAL_SERVER_ERROR,
+    )
+
+
 def _convert_value_type(current_value: Any, new_value: Any) -> Any:
     if isinstance(current_value, bool):
         if isinstance(new_value, str):
@@ -121,6 +149,15 @@ def _merge_configs_cached(
     return ExtractionConfig(**config_dict)
 
 
+def _make_hashable(obj: Any) -> Any:
+    """Convert nested dicts/lists to hashable tuples."""
+    if isinstance(obj, dict):
+        return tuple(sorted((k, _make_hashable(v)) for k, v in obj.items()))
+    if isinstance(obj, list):
+        return tuple(_make_hashable(item) for item in obj)
+    return obj
+
+
 def merge_configs(
     static_config: ExtractionConfig | None,
     query_params: dict[str, Any],
@@ -128,7 +165,7 @@ def merge_configs(
 ) -> ExtractionConfig:
     """Merge configurations with precedence: header > query > static > default."""
     query_tuple = tuple(sorted(query_params.items())) if query_params else ()
-    header_tuple = tuple(sorted(header_config.items())) if header_config else None
+    header_tuple = _make_hashable(header_config) if header_config else None
 
     return _merge_configs_cached(static_config, query_tuple, header_tuple)
 
@@ -211,8 +248,9 @@ async def get_configuration() -> dict[str, Any]:
 app = Litestar(
     route_handlers=[handle_files_upload, health_check, get_configuration],
     plugins=[OpenTelemetryPlugin(OpenTelemetryConfig())],
-    logging_config=StructLoggingConfig(),
+    logging_config=StructLoggingConfig(),  # Use default config
     exception_handlers={
         KreuzbergError: exception_handler,
+        Exception: general_exception_handler,  # Catch all exceptions for debugging
     },
 )

kreuzberg/_config.py

@@ -162,6 +162,10 @@ def load_config_from_file(config_path: Path) -> dict[str, Any]:
     if config_path.name == "pyproject.toml":
         return data.get("tool", {}).get("kreuzberg", {})  # type: ignore[no-any-return]
 
+    # For any other TOML file, check if it has [tool.kreuzberg] section
+    if "tool" in data and "kreuzberg" in data["tool"]:
+        return data["tool"]["kreuzberg"]  # type: ignore[no-any-return]
+
     return data  # type: ignore[no-any-return]
 
 

kreuzberg/_extractors/_spread_sheet.py

@@ -34,8 +34,22 @@ CellValue = int | float | str | bool | time | date | datetime | timedelta
 class SpreadSheetExtractor(Extractor):
     SUPPORTED_MIME_TYPES = SPREADSHEET_MIME_TYPES
 
+    def _get_file_extension(self) -> str:
+        """Get the appropriate file extension based on MIME type."""
+        mime_to_ext = {
+            "application/vnd.ms-excel": ".xls",
+            "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet": ".xlsx",
+            "application/vnd.ms-excel.sheet.macroEnabled.12": ".xlsm",
+            "application/vnd.ms-excel.sheet.binary.macroEnabled.12": ".xlsb",
+            "application/vnd.ms-excel.addin.macroEnabled.12": ".xlam",
+            "application/vnd.ms-excel.template.macroEnabled.12": ".xltm",
+            "application/vnd.oasis.opendocument.spreadsheet": ".ods",
+        }
+        return mime_to_ext.get(self.mime_type, ".xlsx")
+
     async def extract_bytes_async(self, content: bytes) -> ExtractionResult:
-        xlsx_path, unlink = await create_temp_file(".xlsx")
+        file_extension = self._get_file_extension()
+        xlsx_path, unlink = await create_temp_file(file_extension)
         await AsyncPath(xlsx_path).write_bytes(content)
         try:
             return await self.extract_path_async(xlsx_path)
@@ -72,7 +86,8 @@ class SpreadSheetExtractor(Extractor):
             ) from e
 
     def extract_bytes_sync(self, content: bytes) -> ExtractionResult:
-        fd, temp_path = tempfile.mkstemp(suffix=".xlsx")
+        file_extension = self._get_file_extension()
+        fd, temp_path = tempfile.mkstemp(suffix=file_extension)
 
         try:
             with os.fdopen(fd, "wb") as f:

kreuzberg/_ocr/_tesseract.py

@@ -280,6 +280,9 @@ class TesseractBackend(OCRBackend[TesseractConfig]):
         }
 
     async def _execute_tesseract(self, path: Path, output_base: str, run_config: dict[str, Any]) -> None:
+        psm_value = run_config["psm"]
+        psm_str = str(psm_value.value) if hasattr(psm_value, "value") else str(psm_value)
+
         command = [
             "tesseract",
             str(path),
@@ -287,7 +290,7 @@ class TesseractBackend(OCRBackend[TesseractConfig]):
             "-l",
             run_config["language"],
             "--psm",
-            str(run_config["psm"].value),
+            psm_str,
             "--oem",
             "1",
             "--loglevel",
@@ -1089,8 +1092,16 @@ class TesseractBackend(OCRBackend[TesseractConfig]):
             }
 
     def _build_tesseract_command(
-        self, path: Path, output_base: str, language: str, psm: PSMMode, output_format: str = "text", **kwargs: Any
+        self,
+        path: Path,
+        output_base: str,
+        language: str,
+        psm: PSMMode | int,
+        output_format: str = "text",
+        **kwargs: Any,
     ) -> list[str]:
+        psm_str = str(psm.value) if hasattr(psm, "value") else str(psm)
+
         command = [
             "tesseract",
             str(path),
@@ -1098,7 +1109,7 @@ class TesseractBackend(OCRBackend[TesseractConfig]):
             "-l",
             language,
             "--psm",
-            str(psm.value),
+            psm_str,
             "--oem",
             "1",
             "--loglevel",

kreuzberg/cli.py

@@ -265,7 +265,7 @@ def cli(ctx: click.Context) -> None:
 @click.option("--easyocr-languages", help="EasyOCR language codes (comma-separated, e.g., 'en,de')")
 @click.option("--paddleocr-languages", help="PaddleOCR language codes (comma-separated, e.g., 'en,german')")
 @click.pass_context
-def extract(ctx: click.Context) -> None:
+def extract(ctx: click.Context, /, **kwargs: Any) -> None:
     """Extract text from a document.
 
     FILE can be a path to a document or '-' to read from stdin.
@@ -279,7 +279,7 @@ def extract(ctx: click.Context) -> None:
 
         extraction_config = build_extraction_config(file_config, cli_args)
 
-        result = _perform_extraction(params["file"], extraction_config, params["verbose"])
+        result = _perform_extraction(kwargs.get("file"), extraction_config, params["verbose"])
 
         _write_output(result, params["output"], params["show_metadata"], params["output_format"], params["verbose"])
 

{kreuzberg-3.13.1.dist-info → kreuzberg-3.13.3.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: kreuzberg
-Version: 3.13.1
+Version: 3.13.3
 Summary: Document intelligence framework for Python - Extract text, metadata, and structured data from diverse file formats
 Project-URL: documentation, https://kreuzberg.dev
 Project-URL: homepage, https://github.com/Goldziher/kreuzberg
@@ -31,15 +31,15 @@ Requires-Python: >=3.10
 Requires-Dist: anyio>=4.10.0
 Requires-Dist: chardetng-py>=0.3.5
 Requires-Dist: exceptiongroup>=1.2.2; python_version < '3.11'
-Requires-Dist: html-to-markdown[lxml]>=1.9.1
+Requires-Dist: html-to-markdown[lxml]>=1.10.0
 Requires-Dist: mcp>=1.13.0
 Requires-Dist: msgspec>=0.18.0
 Requires-Dist: numpy>=1.24.0
 Requires-Dist: playa-pdf>=0.7.0
-Requires-Dist: polars>=1.33.0
+Requires-Dist: polars>=1.33.1
 Requires-Dist: psutil>=7.0.0
 Requires-Dist: pypdfium2==4.30.0
-Requires-Dist: python-calamine>=0.5.2
+Requires-Dist: python-calamine>=0.5.3
 Requires-Dist: python-pptx>=1.0.2
 Requires-Dist: typing-extensions>=4.15.0; python_version < '3.12'
 Provides-Extra: additional-extensions
@@ -55,17 +55,17 @@ Requires-Dist: keybert>=0.9.0; extra == 'all'
 Requires-Dist: litestar[opentelemetry,standard,structlog]>=2.17.0; extra == 'all'
 Requires-Dist: mailparse>=1.0.15; extra == 'all'
 Requires-Dist: paddleocr>=3.2.0; extra == 'all'
-Requires-Dist: paddlepaddle>=3.1.1; extra == 'all'
+Requires-Dist: paddlepaddle>=3.2.0; extra == 'all'
 Requires-Dist: playa-pdf[crypto]>=0.7.0; extra == 'all'
 Requires-Dist: rich>=14.1.0; extra == 'all'
-Requires-Dist: semantic-text-splitter>=0.27.0; extra == 'all'
+Requires-Dist: semantic-text-splitter>=0.28.0; extra == 'all'
 Requires-Dist: setuptools>=80.9.0; extra == 'all'
 Requires-Dist: spacy>=3.8.7; extra == 'all'
 Requires-Dist: tomli>=2.0.0; (python_version < '3.11') and extra == 'all'
 Provides-Extra: api
 Requires-Dist: litestar[opentelemetry,standard,structlog]>=2.17.0; extra == 'api'
 Provides-Extra: chunking
-Requires-Dist: semantic-text-splitter>=0.27.0; extra == 'chunking'
+Requires-Dist: semantic-text-splitter>=0.28.0; extra == 'chunking'
 Provides-Extra: cli
 Requires-Dist: click>=8.2.1; extra == 'cli'
 Requires-Dist: rich>=14.1.0; extra == 'cli'
@@ -85,7 +85,7 @@ Provides-Extra: langdetect
 Requires-Dist: fast-langdetect>=0.3.2; extra == 'langdetect'
 Provides-Extra: paddleocr
 Requires-Dist: paddleocr>=3.2.0; extra == 'paddleocr'
-Requires-Dist: paddlepaddle>=3.1.1; extra == 'paddleocr'
+Requires-Dist: paddlepaddle>=3.2.0; extra == 'paddleocr'
 Requires-Dist: setuptools>=80.9.0; extra == 'paddleocr'
 Description-Content-Type: text/markdown
 

{kreuzberg-3.13.1.dist-info → kreuzberg-3.13.3.dist-info}/RECORD

@@ -1,7 +1,7 @@
 kreuzberg/__init__.py,sha256=Oh_NTp8wf0BlvD8CSBad2A493nEWH4jTE0x8v7v1Y9w,1341
 kreuzberg/__main__.py,sha256=3cIDdzTggj2kj8uKx4WShWHmCWqdZazdM3BxUGbAuSI,104
 kreuzberg/_chunker.py,sha256=tr9_KUYTSLauFois3MsB-A-0hGcTT8hTQFrqNRTii-I,1373
-kreuzberg/_config.py,sha256=Q5oiJE1XRf8ITuYcO8LZAOB3G2zNlXz2458rgPSth-U,12257
+kreuzberg/_config.py,sha256=T6ASb3N8nPQ4g5B2FxfgK82uE4pesGllezqrmZ0gSdM,12457
 kreuzberg/_constants.py,sha256=Bxc8oiN-wHwnWXT9bEiJhTUcu1ygPpra5qHirAif3b4,191
 kreuzberg/_document_classification.py,sha256=Mz_s2GJGsEl7MQ-67BPoGYCZibTy9Sw0PScUZKBjKOA,5736
 kreuzberg/_entity_extraction.py,sha256=5YpPnqoJ5aiHd_sy4bN4-Ngiq79RhCV6yaUQE8joGXo,3503
@@ -11,12 +11,12 @@ kreuzberg/_mime_types.py,sha256=kGBDSMO4XPgzUKC7iaBeChCtRQXZ9_zXq6eJydejX_k,7739
 kreuzberg/_playa.py,sha256=p4G5ymSSCbQoDeXJjH-yuVzdd4y-wKcolqDthjPtqok,11413
 kreuzberg/_registry.py,sha256=8cPpz3oZVnMwWDT2v_Q7wf-GHd5YuHmc-nkLtvPfE1I,2433
 kreuzberg/_types.py,sha256=D-2d_WG8HyByA163izGhjk7t-e4FL_N-_6UzlVso8Dg,36020
-kreuzberg/cli.py,sha256=MLeWoMcLoN6WnkbyRbOY-2dqp-vNZf7Nb-K_R5F5CoU,12730
+kreuzberg/cli.py,sha256=nPH4FDW6WkoF4gtH0s4RWmxjAveJ_-Unb6fev6x0Sko,12752
 kreuzberg/exceptions.py,sha256=PTiAZgQwcG9hXbgYg2W7sfxksFhq5_wzOFgZGnTJAoc,2991
 kreuzberg/extraction.py,sha256=jiMKiDyTf3sHyk76sMffHR-eH-_yg-DFRMuXEKufRYI,17649
 kreuzberg/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 kreuzberg/_api/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-kreuzberg/_api/main.py,sha256=JALYRD0qwyoZloWk5dNNuslBtG4GlVNc0G2oADm6cAc,7578
+kreuzberg/_api/main.py,sha256=q0ygmdAUfTkjlqAa1RdW1KxxzxQ6IX80__UTpoXipp8,8859
 kreuzberg/_extractors/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 kreuzberg/_extractors/_base.py,sha256=i2FvAhRnamEtBb4a-C7pfcdWIXnkEBw0saMQu7h1_RQ,2069
 kreuzberg/_extractors/_email.py,sha256=jn_8J4BASKJ7zFHBG0PgxNe3OT4pjmEM2tTKX8y_0AE,5887
@@ -25,7 +25,7 @@ kreuzberg/_extractors/_image.py,sha256=UqPoYfvDRX6Rd1yPhcLHJLDw6d2cUzgkqOGjh2ele
 kreuzberg/_extractors/_pandoc.py,sha256=-Ai4S1cXs7F6yeonb_7Y7_ZoWHn29E2oP1WlPtM-4HM,22505
 kreuzberg/_extractors/_pdf.py,sha256=Yv_c3xYzrGAjgTbwCGqbiQTDLjIUP_Pu7Z3GmMOqgqg,17865
 kreuzberg/_extractors/_presentation.py,sha256=ULGkt7dzeA9sYSEhpAucKZmkdv9EubzeZtOjoLP3Z2E,6994
-kreuzberg/_extractors/_spread_sheet.py,sha256=x25u2M-ufxpDd7_qrjhMEz1yFftIcOISE1qwPW09Zm0,11962
+kreuzberg/_extractors/_spread_sheet.py,sha256=UgjkLBATirc5FXUFtRN1ArLfOYhLDJxH2wFb1s9E5vA,12784
 kreuzberg/_extractors/_structured.py,sha256=PpefI_GDrdLyUgnElrbdB-MeTMKVWium4Ckxm5Zg100,5536
 kreuzberg/_mcp/__init__.py,sha256=h6DgLFO4TMUk7_wCJ2jn2Y6IkFmfzb-Z7jX-G5UCYVc,43
 kreuzberg/_mcp/server.py,sha256=iYJG6g0u7I6mWtC4R1XlxydBrPpgnp5dGJzpm9QAZig,8438
@@ -34,7 +34,7 @@ kreuzberg/_ocr/_base.py,sha256=5ef2g8JuSaZF2sDiAmoaODHbeG4MT0LtNzbtW0n9BnU,1445
 kreuzberg/_ocr/_easyocr.py,sha256=XbgpGt5tkE4xHleIGvV1cHlpOQTp43rSXBO1CyIyKTg,14599
 kreuzberg/_ocr/_paddleocr.py,sha256=58sKOHfKCHGFJNlRLrJwey8G_7xbsAAPBXB4n3hKc7k,14052
 kreuzberg/_ocr/_table_extractor.py,sha256=LhBiCX8R_xR-uK1FH3ONA_vqOmqUWANZJ2HMCBLsmNY,5513
-kreuzberg/_ocr/_tesseract.py,sha256=xGML3ygY5xMN5T3YznrKDVAH_DWfaFiteFBo_-GpjCs,48931
+kreuzberg/_ocr/_tesseract.py,sha256=H2T_iuXwa0FGCSQ_ZfXvmvqksxoOdOFAfv3uQA8E4-M,49160
 kreuzberg/_utils/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 kreuzberg/_utils/_cache.py,sha256=S6Oc4TJamiuuWeJ2ABxDFbbQh4o8w38AUyZeBEc1NN8,12767
 kreuzberg/_utils/_device.py,sha256=UxGkSTN3Up-Zn43CSyvf8CozW2xAF05Cm01LWA2FZmg,8263
@@ -50,8 +50,8 @@ kreuzberg/_utils/_string.py,sha256=wVyvEHByHBeu_6evmqJGv9Ml-NAwkyz60n8l-7L5Cw0,4
 kreuzberg/_utils/_sync.py,sha256=OWiciXPTGHIxgiGoHI2AglZ1siTNT-nU_JCgHPNzzHk,2196
 kreuzberg/_utils/_table.py,sha256=R-6owHjvcvHGhem_vDsFH7S2yMHGoUUO2PFcj-Idptk,6361
 kreuzberg/_utils/_tmp.py,sha256=wnOInBkcuQoxI1vBLvNv9NqbRCEu9Y03qfOjqQuAk3s,841
-kreuzberg-3.13.1.dist-info/METADATA,sha256=tle3wdEqeitG4inON0f59QltYcZW_xTW5iyggasDEz0,12127
-kreuzberg-3.13.1.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
-kreuzberg-3.13.1.dist-info/entry_points.txt,sha256=GplGhFryCP7kyAG_k-Mdahznvo2fwi73qLFg5yQfH_A,91
-kreuzberg-3.13.1.dist-info/licenses/LICENSE,sha256=-8caMvpCK8SgZ5LlRKhGCMtYDEXqTKH9X8pFEhl91_4,1066
-kreuzberg-3.13.1.dist-info/RECORD,,
+kreuzberg-3.13.3.dist-info/METADATA,sha256=ey7kAlKK8eTER87IiGZZpIPnYoSLwLPX2AGdOPTjj2M,12128
+kreuzberg-3.13.3.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
+kreuzberg-3.13.3.dist-info/entry_points.txt,sha256=GplGhFryCP7kyAG_k-Mdahznvo2fwi73qLFg5yQfH_A,91
+kreuzberg-3.13.3.dist-info/licenses/LICENSE,sha256=-8caMvpCK8SgZ5LlRKhGCMtYDEXqTKH9X8pFEhl91_4,1066
+kreuzberg-3.13.3.dist-info/RECORD,,