kreuzberg 3.1.3__tar.gz → 3.1.4__tar.gz

{kreuzberg-3.1.3 → kreuzberg-3.1.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: kreuzberg
-Version: 3.1.3
+Version: 3.1.4
 Summary: A text extraction library supporting PDFs, images, office documents and more
 Author-email: Na'aman Hirschfeld <nhirschfed@gmail.com>
 License: MIT
@@ -27,19 +27,19 @@ License-File: LICENSE
 Requires-Dist: anyio>=4.9.0
 Requires-Dist: charset-normalizer>=3.4.1
 Requires-Dist: exceptiongroup>=1.2.2; python_version < "3.11"
-Requires-Dist: html-to-markdown>=1.3.0
+Requires-Dist: html-to-markdown>=1.3.2
 Requires-Dist: playa-pdf>=0.4.1
 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: setuptools>=79.0.1; extra == "all"
 Provides-Extra: chunking
 Requires-Dist: semantic-text-splitter>=0.25.1; extra == "chunking"
 Provides-Extra: easyocr
@@ -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>=79.0.1; extra == "paddleocr"
 Dynamic: license-file
 
 # Kreuzberg

{kreuzberg-3.1.3 → kreuzberg-3.1.4}/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-3.1.3 → kreuzberg-3.1.4}/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-3.1.3 → kreuzberg-3.1.4}/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 → kreuzberg-3.1.4}/kreuzberg.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: kreuzberg
-Version: 3.1.3
+Version: 3.1.4
 Summary: A text extraction library supporting PDFs, images, office documents and more
 Author-email: Na'aman Hirschfeld <nhirschfed@gmail.com>
 License: MIT
@@ -27,19 +27,19 @@ License-File: LICENSE
 Requires-Dist: anyio>=4.9.0
 Requires-Dist: charset-normalizer>=3.4.1
 Requires-Dist: exceptiongroup>=1.2.2; python_version < "3.11"
-Requires-Dist: html-to-markdown>=1.3.0
+Requires-Dist: html-to-markdown>=1.3.2
 Requires-Dist: playa-pdf>=0.4.1
 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: setuptools>=79.0.1; extra == "all"
 Provides-Extra: chunking
 Requires-Dist: semantic-text-splitter>=0.25.1; extra == "chunking"
 Provides-Extra: easyocr
@@ -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>=79.0.1; extra == "paddleocr"
 Dynamic: license-file
 
 # Kreuzberg

{kreuzberg-3.1.3 → kreuzberg-3.1.4}/kreuzberg.egg-info/requires.txt

@@ -1,6 +1,6 @@
 anyio>=4.9.0
 charset-normalizer>=3.4.1
-html-to-markdown>=1.3.0
+html-to-markdown>=1.3.2
 playa-pdf>=0.4.1
 pypdfium2==4.30.0
 python-calamine>=0.3.2
@@ -10,7 +10,7 @@ python-pptx>=1.0.2
 exceptiongroup>=1.2.2
 
 [:python_version < "3.12"]
-typing-extensions>=4.13.1
+typing-extensions>=4.13.2
 
 [all]
 easyocr>=1.7.2
@@ -18,7 +18,7 @@ gmft>=0.4.1
 paddleocr>=2.10.0
 paddlepaddle>=3.0.0
 semantic-text-splitter>=0.25.1
-setuptools>=76.0.0
+setuptools>=79.0.1
 
 [chunking]
 semantic-text-splitter>=0.25.1
@@ -32,4 +32,4 @@ gmft>=0.4.1
 [paddleocr]
 paddleocr>=2.10.0
 paddlepaddle>=3.0.0
-setuptools>=76.0.0
+setuptools>=79.0.1

{kreuzberg-3.1.3 → kreuzberg-3.1.4}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "kreuzberg"
-version = "3.1.3"
+version = "3.1.4"
 description = "A text extraction library supporting PDFs, images, office documents and more"
 readme = "README.md"
 keywords = [
@@ -40,12 +40,12 @@ dependencies = [
   "anyio>=4.9.0",
   "charset-normalizer>=3.4.1",
   "exceptiongroup>=1.2.2; python_version<'3.11'",
-  "html-to-markdown>=1.3.0",
+  "html-to-markdown>=1.3.2",
   "playa-pdf>=0.4.1",
   "pypdfium2==4.30.0",                                # pinned due to bug in 4.30.1, until v5 is stable
   "python-calamine>=0.3.2",
   "python-pptx>=1.0.2",
-  "typing-extensions>=4.13.1; python_version<'3.12'",
+  "typing-extensions>=4.13.2; python_version<'3.12'",
 ]
 
 optional-dependencies.all = [
@@ -58,7 +58,7 @@ optional-dependencies.all = [
   "paddlepaddle>=3.0.0",
   # chunking
   "semantic-text-splitter>=0.25.1",
-  "setuptools>=76.0.0",
+  "setuptools>=79.0.1",
 ]
 optional-dependencies.chunking = [
   "semantic-text-splitter>=0.25.1",
@@ -72,7 +72,7 @@ optional-dependencies.gmft = [
 optional-dependencies.paddleocr = [
   "paddleocr>=2.10.0",
   "paddlepaddle>=3.0.0",
-  "setuptools>=76.0.0",
+  "setuptools>=79.0.1",
 ]
 urls.homepage = "https://github.com/Goldziher/kreuzberg"
 
@@ -85,8 +85,8 @@ dev = [
   "pytest-cov>=6.1.1",
   "pytest-mock>=3.14.0",
   "pytest-timeout>=2.3.1",
-  "ruff>=0.11.4",
-  "trio>=0.29.0",
+  "ruff>=0.11.7",
+  "trio>=0.30.0",
   "uv-bump",
 ]
 doc = [