kreuzberg 3.9.0__py3-none-any.whl → 3.10.0__py3-none-any.whl

kreuzberg/_extractors/_pdf.py

@@ -22,7 +22,7 @@ from kreuzberg._ocr._easyocr import EasyOCRConfig
 from kreuzberg._ocr._paddleocr import PaddleOCRConfig
 from kreuzberg._ocr._tesseract import TesseractConfig
 from kreuzberg._playa import extract_pdf_metadata, extract_pdf_metadata_sync
-from kreuzberg._types import ExtractionResult, OcrBackendType
+from kreuzberg._types import ExtractionResult, Metadata, OcrBackendType
 from kreuzberg._utils._errors import create_error_context, should_retry
 from kreuzberg._utils._pdf_lock import pypdfium_file_lock
 from kreuzberg._utils._string import normalize_spaces
@@ -33,6 +33,7 @@ from kreuzberg.exceptions import ParsingError
 
 if TYPE_CHECKING:  # pragma: no cover
     from PIL.Image import Image
+    from playa.document import Document
 
 
 class PDFExtractor(Extractor):
@@ -45,7 +46,7 @@ class PDFExtractor(Extractor):
         file_path, unlink = await create_temp_file(".pdf")
         await AsyncPath(file_path).write_bytes(content)
         try:
-            metadata = await extract_pdf_metadata(content)
+            metadata = await self._extract_metadata_with_password_attempts(content)
             result = await self.extract_path_async(file_path)
 
             result.metadata = metadata
@@ -73,7 +74,7 @@ class PDFExtractor(Extractor):
         if not result:
             result = ExtractionResult(content="", mime_type=PLAIN_TEXT_MIME_TYPE, metadata={}, chunks=[])
 
-        result.metadata = await extract_pdf_metadata(content_bytes)
+        result.metadata = await self._extract_metadata_with_password_attempts(content_bytes)
 
         if self.config.extract_tables:
             # GMFT is optional dependency
@@ -107,7 +108,7 @@ class PDFExtractor(Extractor):
 
             result = self.extract_path_sync(Path(temp_path))
 
-            metadata = extract_pdf_metadata_sync(content)
+            metadata = self._extract_metadata_with_password_attempts_sync(content)
             result.metadata = metadata
 
             return result
@@ -406,11 +407,81 @@ class PDFExtractor(Extractor):
         # Use list comprehension and join for efficient string building
         return "\n\n".join(result.content for result in results)
 
+    def _parse_with_password_attempts(self, content: bytes) -> Document:
+        """Parse PDF with password attempts."""
+        # Normalize password to list
+        if isinstance(self.config.pdf_password, str):
+            passwords = [self.config.pdf_password] if self.config.pdf_password else [""]
+        else:
+            passwords = list(self.config.pdf_password)
+
+        # Try each password in sequence
+        last_exception = None
+        for password in passwords:
+            try:
+                return parse(content, max_workers=1, password=password)
+            except Exception as e:  # noqa: PERF203, BLE001
+                last_exception = e
+                continue
+
+        # If all passwords failed, raise the last exception
+        if last_exception:
+            raise last_exception from None
+
+        # Fallback to no password
+        return parse(content, max_workers=1, password="")
+
+    def _get_passwords_to_try(self) -> list[str]:
+        """Get list of passwords to try in sequence."""
+        if isinstance(self.config.pdf_password, str):
+            return [self.config.pdf_password] if self.config.pdf_password else [""]
+        return list(self.config.pdf_password) if self.config.pdf_password else [""]
+
+    async def _extract_metadata_with_password_attempts(self, content: bytes) -> Metadata:
+        """Extract PDF metadata with password attempts."""
+        passwords = self._get_passwords_to_try()
+
+        last_exception = None
+        for password in passwords:
+            try:
+                return await extract_pdf_metadata(content, password=password)
+            except Exception as e:  # noqa: PERF203, BLE001
+                last_exception = e
+                continue
+
+        # If all passwords failed, try with empty password as fallback
+        try:
+            return await extract_pdf_metadata(content, password="")
+        except Exception:
+            if last_exception:
+                raise last_exception from None
+            raise
+
+    def _extract_metadata_with_password_attempts_sync(self, content: bytes) -> Metadata:
+        """Extract PDF metadata with password attempts (sync version)."""
+        passwords = self._get_passwords_to_try()
+
+        last_exception = None
+        for password in passwords:
+            try:
+                return extract_pdf_metadata_sync(content, password=password)
+            except Exception as e:  # noqa: PERF203, BLE001
+                last_exception = e
+                continue
+
+        # If all passwords failed, try with empty password as fallback
+        try:
+            return extract_pdf_metadata_sync(content, password="")
+        except Exception:
+            if last_exception:
+                raise last_exception from None
+            raise
+
     def _extract_with_playa_sync(self, path: Path, fallback_text: str) -> str:
         """Extract text using playa for better structure preservation."""
         with contextlib.suppress(Exception):
             content = path.read_bytes()
-            document = parse(content, max_workers=1)
+            document = self._parse_with_password_attempts(content)
 
             # Extract text while preserving structure
             pages_text = []

kreuzberg/_playa.py

@@ -24,11 +24,12 @@ FULL_DATE_LENGTH = 14
 BOM_CHAR = "\ufeff"
 
 
-async def extract_pdf_metadata(pdf_content: bytes) -> Metadata:
+async def extract_pdf_metadata(pdf_content: bytes, password: str = "") -> Metadata:
     """Extract metadata from a PDF document.
 
     Args:
         pdf_content: The bytes of the PDF document.
+        password: Password for encrypted PDF files.
 
     Raises:
         ParsingError: If the PDF metadata could not be extracted.
@@ -37,7 +38,7 @@ async def extract_pdf_metadata(pdf_content: bytes) -> Metadata:
         A dictionary of metadata extracted from the PDF.
     """
     try:
-        document = parse(pdf_content, max_workers=1)
+        document = parse(pdf_content, max_workers=1, password=password)
         metadata: Metadata = {}
 
         for raw_info in document.info:
@@ -275,13 +276,14 @@ def _extract_structure_information(document: Document, result: Metadata) -> None
             result["subtitle"] = subtitle
 
 
-def extract_pdf_metadata_sync(pdf_content: bytes) -> Metadata:
+def extract_pdf_metadata_sync(pdf_content: bytes, password: str = "") -> Metadata:
     """Synchronous version of extract_pdf_metadata.
 
     Extract metadata from a PDF document without using async/await.
 
     Args:
         pdf_content: The bytes of the PDF document.
+        password: Password for encrypted PDF files.
 
     Raises:
         ParsingError: If the PDF metadata could not be extracted.
@@ -290,7 +292,7 @@ def extract_pdf_metadata_sync(pdf_content: bytes) -> Metadata:
         A dictionary of metadata extracted from the PDF.
     """
     try:
-        document = parse(pdf_content, max_workers=1)
+        document = parse(pdf_content, max_workers=1, password=password)
         metadata: Metadata = {}
 
         for raw_info in document.info:

kreuzberg/_types.py

@@ -357,6 +357,8 @@ class ExtractionConfig:
     """The mode to use for document classification."""
     enable_quality_processing: bool = True
     """Whether to apply quality post-processing to improve extraction results."""
+    pdf_password: str | list[str] = ""
+    """Password(s) for encrypted PDF files. Can be a single password or list of passwords to try in sequence. Only used when crypto extra is installed."""
 
     def __post_init__(self) -> None:
         if self.custom_entity_patterns is not None and isinstance(self.custom_entity_patterns, dict):

{kreuzberg-3.9.0.dist-info → kreuzberg-3.10.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: kreuzberg
-Version: 3.9.0
+Version: 3.10.0
 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
@@ -29,12 +29,12 @@ Classifier: Topic :: Text Processing :: General
 Classifier: Typing :: Typed
 Requires-Python: >=3.10
 Requires-Dist: anyio>=4.9.0
-Requires-Dist: chardetng-py>=0.3.4
+Requires-Dist: chardetng-py>=0.3.5
 Requires-Dist: exceptiongroup>=1.2.2; python_version < '3.11'
-Requires-Dist: html-to-markdown[lxml]>=1.8.0
-Requires-Dist: mcp>=1.11.0
+Requires-Dist: html-to-markdown[lxml]>=1.9.0
+Requires-Dist: mcp>=1.12.2
 Requires-Dist: msgspec>=0.18.0
-Requires-Dist: playa-pdf>=0.6.1
+Requires-Dist: playa-pdf>=0.6.4
 Requires-Dist: psutil>=7.0.0
 Requires-Dist: pypdfium2==4.30.0
 Requires-Dist: python-calamine>=0.3.2
@@ -53,7 +53,8 @@ Requires-Dist: litestar[opentelemetry,standard,structlog]>=2.16.0; extra == 'all
 Requires-Dist: mailparse>=1.0.15; extra == 'all'
 Requires-Dist: paddleocr>=3.1.0; extra == 'all'
 Requires-Dist: paddlepaddle>=3.1.0; extra == 'all'
-Requires-Dist: rich>=14.0.0; extra == 'all'
+Requires-Dist: playa-pdf[crypto]>=0.6.4; extra == 'all'
+Requires-Dist: rich>=14.1.0; extra == 'all'
 Requires-Dist: semantic-text-splitter>=0.27.0; extra == 'all'
 Requires-Dist: setuptools>=80.9.0; extra == 'all'
 Requires-Dist: spacy>=3.8.7; extra == 'all'
@@ -67,8 +68,10 @@ Provides-Extra: chunking
 Requires-Dist: semantic-text-splitter>=0.27.0; extra == 'chunking'
 Provides-Extra: cli
 Requires-Dist: click>=8.2.1; extra == 'cli'
-Requires-Dist: rich>=14.0.0; extra == 'cli'
+Requires-Dist: rich>=14.1.0; extra == 'cli'
 Requires-Dist: tomli>=2.0.0; (python_version < '3.11') and extra == 'cli'
+Provides-Extra: crypto
+Requires-Dist: playa-pdf[crypto]>=0.6.4; extra == 'crypto'
 Provides-Extra: easyocr
 Requires-Dist: easyocr>=1.7.2; extra == 'easyocr'
 Provides-Extra: entity-extraction
@@ -130,14 +133,14 @@ Kreuzberg leverages established open source technologies:
 ### Extract Text with CLI
 
 ```bash
-# Extract text from any file to markdown
-uvx kreuzberg extract document.pdf > output.md
+# Extract text from any file to text format
+uvx kreuzberg extract document.pdf > output.txt
 
 # With all features (OCR, table extraction, etc.)
-uvx --from "kreuzberg[all]" kreuzberg extract invoice.pdf --ocr --format markdown
+uvx --from "kreuzberg[all]" kreuzberg extract invoice.pdf --ocr-backend tesseract --output-format text
 
 # Extract with rich metadata
-uvx kreuzberg extract report.pdf --show-metadata --format json
+uvx kreuzberg extract report.pdf --show-metadata --output-format json
 ```
 
 ### Python Usage

{kreuzberg-3.9.0.dist-info → kreuzberg-3.10.0.dist-info}/RECORD

@@ -8,9 +8,9 @@ kreuzberg/_entity_extraction.py,sha256=43VxtCPbuSrfi-XGSop-VRoZadJcNe02qRGwjGSaR
 kreuzberg/_gmft.py,sha256=JWDeBYWSDkh3tLUhwTmW5CbIBEig2A3o712BLshk7hE,25533
 kreuzberg/_language_detection.py,sha256=eEfj4tsh91SfB2_zQIdY-qD7TlPcppaFm0SqQmETS6Y,3295
 kreuzberg/_mime_types.py,sha256=2warRVqfBUNIg8JBg8yP4pRqaMPvwINosHMkJwtH_Fc,8488
-kreuzberg/_playa.py,sha256=9z4If0WHxbYQxfb8xT7T96L9Du2Fj3Ar5-rF0OHHiMM,11877
+kreuzberg/_playa.py,sha256=cJ000ZPHRhbpbP7odRuzMKn38teR6RbodoHgksbfjGE,12059
 kreuzberg/_registry.py,sha256=wGSlkS0U1zqruWQCLE95vj4a2mw1yyvf0j6rgz80sJg,3473
-kreuzberg/_types.py,sha256=Si-Kb58HgE4ckGyZnJFqbWRbCNbdyC_Y0-p75aQP838,15065
+kreuzberg/_types.py,sha256=ecT2dRg7dr06p7Dxv23YJ7Ur2m4FUCt6xGtuoS7MQaI,15259
 kreuzberg/cli.py,sha256=lmthZa0x8pz7SQfCoPPdTaUF9aQZ8W4w5FlnPcsGr9k,12438
 kreuzberg/exceptions.py,sha256=PTiAZgQwcG9hXbgYg2W7sfxksFhq5_wzOFgZGnTJAoc,2991
 kreuzberg/extraction.py,sha256=Kt1mOxdlOb35yVOdpdhiRPuTgA9BW_TTG9qwCkSxSkc,17332
@@ -23,7 +23,7 @@ kreuzberg/_extractors/_email.py,sha256=6-Mk1TRXPyy9ylWKCpgdrogyzhiFnJOTuTRld1ghO
 kreuzberg/_extractors/_html.py,sha256=lOM1Tgrrvd7vpEeFAxC1dp0Tibr6N2FEHCjgFx0FK64,1745
 kreuzberg/_extractors/_image.py,sha256=OmkqR5Cd6bTM_qHEdNHeXa5eK-3KvtLgHX-JE5oZWec,4483
 kreuzberg/_extractors/_pandoc.py,sha256=51k7XISfKaPorhapG7aIeQb94KGsfozxKyT2rwhk9Bk,26553
-kreuzberg/_extractors/_pdf.py,sha256=UlliWggWHuVwwJE-bRa7H9-_cieSa8kdrQP3x_GOxxY,17018
+kreuzberg/_extractors/_pdf.py,sha256=pn45qKYkMcmG-PzeeF5jRjrw1NwaKU3589dhpn7HvE8,19918
 kreuzberg/_extractors/_presentation.py,sha256=CUlqZl_QCdJdumsZh0BpROkFbvi9uq7yMoIt3bRTUeE,10859
 kreuzberg/_extractors/_spread_sheet.py,sha256=iagiyJsnl-89OP1eqmEv8jWl7gZBJm2x0YOyqBgLasA,13733
 kreuzberg/_extractors/_structured.py,sha256=J7op5ZZ663a_CkE2SB5eKjPQPXKHdqPst-GFCJMNGqw,5727
@@ -47,8 +47,8 @@ kreuzberg/_utils/_string.py,sha256=bCzO3UO6nXupxvtMWvHqfp1Vd9CTzEH9jmpJXQ7upAU,6
 kreuzberg/_utils/_sync.py,sha256=7LSavBmxVKQUzdjfx9fYRAI9IbJtRw8iGf_Q8B7RX9g,4923
 kreuzberg/_utils/_table.py,sha256=IomrfQBP85DZI8RmQjOVs2Siq7VP9FUTYPaZR4t3yRw,8199
 kreuzberg/_utils/_tmp.py,sha256=hVn-VVijIg2FM7EZJ899gc7wZg-TGoJZoeAcxMX-Cxg,1044
-kreuzberg-3.9.0.dist-info/METADATA,sha256=C83JYzqxhGHhrqWDUmo0eJwK_2szx9ZQt3cnkocgwBY,11876
-kreuzberg-3.9.0.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
-kreuzberg-3.9.0.dist-info/entry_points.txt,sha256=GplGhFryCP7kyAG_k-Mdahznvo2fwi73qLFg5yQfH_A,91
-kreuzberg-3.9.0.dist-info/licenses/LICENSE,sha256=-8caMvpCK8SgZ5LlRKhGCMtYDEXqTKH9X8pFEhl91_4,1066
-kreuzberg-3.9.0.dist-info/RECORD,,
+kreuzberg-3.10.0.dist-info/METADATA,sha256=4U1mSEAbT3zRir--SPZmYy09LfEfu5vUz6CUhQL8uzA,12047
+kreuzberg-3.10.0.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
+kreuzberg-3.10.0.dist-info/entry_points.txt,sha256=GplGhFryCP7kyAG_k-Mdahznvo2fwi73qLFg5yQfH_A,91
+kreuzberg-3.10.0.dist-info/licenses/LICENSE,sha256=-8caMvpCK8SgZ5LlRKhGCMtYDEXqTKH9X8pFEhl91_4,1066
+kreuzberg-3.10.0.dist-info/RECORD,,