kreuzberg 3.1.3__py3-none-any.whl → 3.1.5__py3-none-any.whl

kreuzberg/_extractors/_pandoc.py

@@ -227,14 +227,54 @@ class PandocExtractor(Extractor):
 
             command = ["pandoc", "--version"]
             result = await run_process(command)
+            stdout = result.stdout.decode()
 
-            version_match = re.search(r"pandoc\s+v?(\d+)\.\d+(\.\d+)?", result.stdout.decode())
-            if not version_match or int(version_match.group(1)) < MINIMAL_SUPPORTED_PANDOC_VERSION:
-                raise MissingDependencyError(
-                    "Pandoc version 2 or above is a required system dependency. Please install it on your system and make sure its available in $PATH."
-                )
+            # Try more inclusive patterns to detect the pandoc version
+            # Try common formats first
+            version_match = re.search(
+                r"pandoc(?:\.exe)?(?:\s+|\s+v|\s+version\s+)(\d+)\.(\d+)(?:\.(\d+))?", stdout, re.IGNORECASE
+            )
+
+            # Try version in parentheses format
+            if not version_match:
+                version_match = re.search(r"pandoc\s+\(version\s+(\d+)\.(\d+)(?:\.(\d+))?\)", stdout, re.IGNORECASE)
+
+            # Try hyphenated format
+            if not version_match:
+                version_match = re.search(r"pandoc-(\d+)\.(\d+)(?:\.(\d+))?", stdout)
+
+            # If still no match, check for version at the beginning of the output or any line
+            if not version_match:
+                # Match version at the start of a line (like in the test case "2.9.2.1\npandoc-types 1.20")
+                version_match = re.search(r"^(\d+)\.(\d+)(?:\.(\d+)(?:\.(\d+))?)?", stdout, re.MULTILINE)
+
+            # Try finding version-like patterns elsewhere in the text
+            if not version_match:
+                # Search for version-like patterns at the beginning of lines or after spaces
+                version_match = re.search(r"(?:^|\s)(\d+)\.(\d+)(?:\.(\d+))?(?:\s|$)", stdout)
+
+            # As a last resort, check any sequence of digits that might be a version
+            if not version_match:
+                out_lines = stdout.splitlines()
+                for line in out_lines:
+                    for token in line.split():
+                        # Match standalone version patterns like 2.11 or 2.11.4
+                        version_pattern = re.match(r"^(\d+)\.(\d+)(?:\.(\d+))?$", token)
+                        if version_pattern:
+                            version_match = version_pattern
+                            break
+                    if version_match:
+                        break
+
+            # If we found a version, check that the major version is at least the minimum required
+            if version_match and int(version_match.group(1)) >= MINIMAL_SUPPORTED_PANDOC_VERSION:
+                self._checked_version = True
+                return
 
-            self._checked_version = True
+            # If we get here, we either didn't find a version or it's too low
+            raise MissingDependencyError(
+                "Pandoc version 2 or above is a required system dependency. Please install it on your system and make sure its available in $PATH."
+            )
 
         except FileNotFoundError as e:
             raise MissingDependencyError(

kreuzberg/_ocr/_easyocr.py

@@ -127,7 +127,8 @@ class EasyOCRConfig:
     height_ths: float = 0.5
     """Maximum difference in box height for merging."""
     language: str | list[str] = "en"
-    """Language or languages to use for OCR."""
+    """Language or languages to use for OCR. Can be a single language code (e.g., 'en'),
+    a comma-separated string of language codes (e.g., 'en,ch_sim'), or a list of language codes."""
     link_threshold: float = 0.4
     """Link confidence threshold."""
     low_text: float = 0.4
@@ -354,29 +355,32 @@ class EasyOCRBackend(OCRBackend[EasyOCRConfig]):
 
     @staticmethod
     def _validate_language_code(language_codes: str | list[str]) -> list[str]:
-        """Validate and normalize a provided language code.
+        """Validate and normalize provided language codes.
 
         Args:
-            language_codes: The language code string.
+            language_codes: The language code(s), either as a string (single or comma-separated) or a list.
 
         Raises:
-            ValidationError: If the language is not supported by EasyOCR
+            ValidationError: If any of the languages are not supported by EasyOCR
 
         Returns:
-            A list with the normalized language code.
+            A list with the normalized language codes.
         """
-        if not isinstance(language_codes, list):
-            languages = [language_codes.lower()]
+        if isinstance(language_codes, str):
+            # Handle comma-separated language codes
+            languages = [lang.strip().lower() for lang in language_codes.split(",")]
         else:
+            # Handle list of language codes
             languages = [lang.lower() for lang in language_codes]
 
-        if all(lang in EASYOCR_SUPPORTED_LANGUAGE_CODES for lang in languages):
-            return languages
+        unsupported_langs = [lang for lang in languages if lang not in EASYOCR_SUPPORTED_LANGUAGE_CODES]
+        if unsupported_langs:
+            raise ValidationError(
+                "The provided language codes are not supported by EasyOCR",
+                context={
+                    "language_code": ",".join(unsupported_langs),
+                    "supported_languages": ",".join(sorted(EASYOCR_SUPPORTED_LANGUAGE_CODES)),
+                },
+            )
 
-        raise ValidationError(
-            "The provided language codes are not supported by EasyOCR",
-            context={
-                "language_code": ",".join([lang for lang in languages if lang not in EASYOCR_SUPPORTED_LANGUAGE_CODES]),
-                "supported_languages": ",".join(sorted(EASYOCR_SUPPORTED_LANGUAGE_CODES)),
-            },
-        )
+        return languages

kreuzberg/_ocr/_paddleocr.py

@@ -52,6 +52,8 @@ class PaddleOCRConfig:
     """Binarization threshold for EAST output map."""
     det_max_side_len: int = 960
     """Maximum size of image long side. Images exceeding this will be proportionally resized."""
+    det_model_dir: str | None = None
+    """Directory for detection model. If None, uses default model location."""
     drop_score: float = 0.5
     """Filter recognition results by confidence score. Results below this are discarded."""
     enable_mkldnn: bool = False
@@ -82,6 +84,8 @@ class PaddleOCRConfig:
     """Recognition algorithm."""
     rec_image_shape: str = "3,32,320"
     """Image shape for recognition algorithm in format 'channels,height,width'."""
+    rec_model_dir: str | None = None
+    """Directory for recognition model. If None, uses default model location."""
     table: bool = True
     """Whether to enable table recognition."""
     use_angle_cls: bool = True

{kreuzberg-3.1.3.dist-info → kreuzberg-3.1.5.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: kreuzberg
-Version: 3.1.3
+Version: 3.1.5
 Summary: A text extraction library supporting PDFs, images, office documents and more
 Author-email: Na'aman Hirschfeld <nhirschfed@gmail.com>
 License: MIT
@@ -25,23 +25,23 @@ Requires-Python: >=3.9
 Description-Content-Type: text/markdown
 License-File: LICENSE
 Requires-Dist: anyio>=4.9.0
-Requires-Dist: charset-normalizer>=3.4.1
+Requires-Dist: charset-normalizer>=3.4.2
 Requires-Dist: exceptiongroup>=1.2.2; python_version < "3.11"
-Requires-Dist: html-to-markdown>=1.3.0
-Requires-Dist: playa-pdf>=0.4.1
+Requires-Dist: html-to-markdown>=1.3.2
+Requires-Dist: playa-pdf>=0.4.3
 Requires-Dist: pypdfium2==4.30.0
 Requires-Dist: python-calamine>=0.3.2
 Requires-Dist: python-pptx>=1.0.2
-Requires-Dist: typing-extensions>=4.13.1; python_version < "3.12"
+Requires-Dist: typing-extensions>=4.13.2; python_version < "3.12"
 Provides-Extra: all
 Requires-Dist: easyocr>=1.7.2; extra == "all"
 Requires-Dist: gmft>=0.4.1; extra == "all"
 Requires-Dist: paddleocr>=2.10.0; extra == "all"
 Requires-Dist: paddlepaddle>=3.0.0; extra == "all"
-Requires-Dist: semantic-text-splitter>=0.25.1; extra == "all"
-Requires-Dist: setuptools>=76.0.0; extra == "all"
+Requires-Dist: semantic-text-splitter>=0.26.0; extra == "all"
+Requires-Dist: setuptools>=80.4.0; extra == "all"
 Provides-Extra: chunking
-Requires-Dist: semantic-text-splitter>=0.25.1; extra == "chunking"
+Requires-Dist: semantic-text-splitter>=0.26.0; extra == "chunking"
 Provides-Extra: easyocr
 Requires-Dist: easyocr>=1.7.2; extra == "easyocr"
 Provides-Extra: gmft
@@ -49,7 +49,7 @@ Requires-Dist: gmft>=0.4.1; extra == "gmft"
 Provides-Extra: paddleocr
 Requires-Dist: paddleocr>=2.10.0; extra == "paddleocr"
 Requires-Dist: paddlepaddle>=3.0.0; extra == "paddleocr"
-Requires-Dist: setuptools>=76.0.0; extra == "paddleocr"
+Requires-Dist: setuptools>=80.4.0; extra == "paddleocr"
 Dynamic: license-file
 
 # Kreuzberg

{kreuzberg-3.1.3.dist-info → kreuzberg-3.1.5.dist-info}/RECORD

@@ -13,21 +13,21 @@ kreuzberg/_extractors/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3h
 kreuzberg/_extractors/_base.py,sha256=YUr6A2n34LlFzbYQkiKqhXAphL9RYrvAls5SlkoQqNg,3028
 kreuzberg/_extractors/_html.py,sha256=L_vcKyv1ObI6irPaD3-WTKqxeRfZA4Rhsl3zUiAe_ws,1312
 kreuzberg/_extractors/_image.py,sha256=VQgSFSzXIMX3A52-DyvuKgfTRXUJIjYn6IX4-sQWWdg,2626
-kreuzberg/_extractors/_pandoc.py,sha256=U5CUrVilKdJqXJFUUT5xzcpy2jfJ26h7kde3p1N_N4w,20124
+kreuzberg/_extractors/_pandoc.py,sha256=OAbWvfzEx3rjim9uNMS9yBRnvkI71rYJKlgVzndsvyc,22157
 kreuzberg/_extractors/_pdf.py,sha256=eNFws_UxLgWSTC_VC_zJmVojpyQvioOXgNjSHQzBq5c,6607
 kreuzberg/_extractors/_presentation.py,sha256=K4ALrpmZ0EWyp2O-3oEmTRCS7yAET9xjinrzo13rpWo,8764
 kreuzberg/_extractors/_spread_sheet.py,sha256=1ejRZk8AE1dXS1tRIdg2S0J9Vo0wG81iKkW2IF6PjlE,4445
 kreuzberg/_ocr/__init__.py,sha256=VTqwKDlIRbjve71Y11Ztygyhv5aWG9LWTj8iX66ANxE,533
 kreuzberg/_ocr/_base.py,sha256=lNT0Tin4hzbmaamqqySxvYEwNtrJB5gGlStrANQQcyc,1637
-kreuzberg/_ocr/_easyocr.py,sha256=4K344njfpzIig7Oz82yI9D5qqFJDRrsSa6x9vDW-eT4,11140
-kreuzberg/_ocr/_paddleocr.py,sha256=NDKXiMtHjIy-Uq4hXe4qm5oUWwOrhjJaibyC708Cw5E,10422
+kreuzberg/_ocr/_easyocr.py,sha256=J8IP2Fg55dG2MH9-lpyZFounvgIgWgCrw694UkaUa9E,11491
+kreuzberg/_ocr/_paddleocr.py,sha256=FyALVb3AQFcej9NFOLy-0dkA-3uxE_ie9Mzd6ho3t68,10656
 kreuzberg/_ocr/_tesseract.py,sha256=cdnVxNpaKjxtBN4xy0Timz-uYtPA9wq9kc6kyYVeDug,9779
 kreuzberg/_utils/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 kreuzberg/_utils/_string.py,sha256=oNO0cmwjVNG0jAzaqNCjYtzvM_nxH5TW2KV-Uh3oEUU,978
 kreuzberg/_utils/_sync.py,sha256=lycobEMXk0tBMWLwkuMdOuNMStDwPKMC0V1Qgp_oi6k,4071
 kreuzberg/_utils/_tmp.py,sha256=5rqG_Nlb9xweaLqJA8Kc5csHDase9_eY_Fq93rNQGWc,1044
-kreuzberg-3.1.3.dist-info/licenses/LICENSE,sha256=-8caMvpCK8SgZ5LlRKhGCMtYDEXqTKH9X8pFEhl91_4,1066
-kreuzberg-3.1.3.dist-info/METADATA,sha256=rgfHFNmD7V_s1HPx7aFrfcYep8dMmqpHidqMcY5_MJE,6641
-kreuzberg-3.1.3.dist-info/WHEEL,sha256=CmyFI0kx5cdEMTLiONQRbGQwjIoR1aIYB7eCAQ4KPJ0,91
-kreuzberg-3.1.3.dist-info/top_level.txt,sha256=rbGkygffkZiyKhL8UN41ZOjLfem0jJPA1Whtndne0rE,10
-kreuzberg-3.1.3.dist-info/RECORD,,
+kreuzberg-3.1.5.dist-info/licenses/LICENSE,sha256=-8caMvpCK8SgZ5LlRKhGCMtYDEXqTKH9X8pFEhl91_4,1066
+kreuzberg-3.1.5.dist-info/METADATA,sha256=nvd68yES8u37CwDxpNLwwZKmRHKqsqCPLx0MqRZHFjo,6641
+kreuzberg-3.1.5.dist-info/WHEEL,sha256=DnLRTWE75wApRYVsjgc6wsVswC54sMSJhAEd4xhDpBk,91
+kreuzberg-3.1.5.dist-info/top_level.txt,sha256=rbGkygffkZiyKhL8UN41ZOjLfem0jJPA1Whtndne0rE,10
+kreuzberg-3.1.5.dist-info/RECORD,,

{kreuzberg-3.1.3.dist-info → kreuzberg-3.1.5.dist-info}/WHEEL

@@ -1,5 +1,5 @@
 Wheel-Version: 1.0
-Generator: setuptools (78.1.0)
+Generator: setuptools (80.4.0)
 Root-Is-Purelib: true
 Tag: py3-none-any