kreuzberg 3.11.1__tar.gz → 3.11.2__tar.gz

{kreuzberg-3.11.1 → kreuzberg-3.11.2}/.pre-commit-config.yaml

@@ -5,13 +5,11 @@ repos:
       - id: commitlint
         stages: [commit-msg]
         additional_dependencies: ["@commitlint/config-conventional"]
-  # Temporarily disabled - ai-rulez Go build failing in CI
-  # TODO: Re-enable once ai-rulez v1.4.4+ Python migration is stable
-  # - repo: https://github.com/Goldziher/ai-rulez
-  #   rev: v1.4.3
-  #   hooks:
-  #     - id: ai-rulez-validate
-  #     - id: ai-rulez-generate
+  - repo: https://github.com/Goldziher/ai-rulez
+    rev: v1.5.1
+    hooks:
+      - id: ai-rulez-validate
+      - id: ai-rulez-generate
   - repo: https://github.com/pre-commit/pre-commit-hooks
     rev: v6.0.0
     hooks:
@@ -55,7 +53,7 @@ repos:
     hooks:
       - id: pyproject-fmt
   - repo: https://github.com/astral-sh/ruff-pre-commit
-    rev: v0.12.8
+    rev: v0.12.9
     hooks:
       - id: ruff
         args: ["--fix", "--unsafe-fixes"]
@@ -68,7 +66,7 @@ repos:
         additional_dependencies:
           - tomli
   - repo: https://github.com/jsh9/pydoclint
-    rev: 0.6.7
+    rev: 0.6.10
     hooks:
       - id: pydoclint
         args:

{kreuzberg-3.11.1 → kreuzberg-3.11.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: kreuzberg
-Version: 3.11.1
+Version: 3.11.2
 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
@@ -32,7 +32,7 @@ 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]>=1.9.0
-Requires-Dist: mcp>=1.12.4
+Requires-Dist: mcp>=1.13.0
 Requires-Dist: msgspec>=0.18.0
 Requires-Dist: playa-pdf>=0.7.0
 Requires-Dist: psutil>=7.0.0
@@ -52,7 +52,7 @@ Requires-Dist: gmft>=0.4.2; extra == 'all'
 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>=3.1.0; extra == 'all'
+Requires-Dist: paddleocr>=3.1.1; extra == 'all'
 Requires-Dist: paddlepaddle>=3.1.0; extra == 'all'
 Requires-Dist: pandas>=2.3.1; extra == 'all'
 Requires-Dist: playa-pdf[crypto]>=0.7.0; extra == 'all'
@@ -84,7 +84,7 @@ Requires-Dist: gmft>=0.4.2; extra == 'gmft'
 Provides-Extra: langdetect
 Requires-Dist: fast-langdetect>=0.3.2; extra == 'langdetect'
 Provides-Extra: paddleocr
-Requires-Dist: paddleocr>=3.1.0; extra == 'paddleocr'
+Requires-Dist: paddleocr>=3.1.1; extra == 'paddleocr'
 Requires-Dist: paddlepaddle>=3.1.0; extra == 'paddleocr'
 Requires-Dist: setuptools>=80.9.0; extra == 'paddleocr'
 Description-Content-Type: text/markdown

{kreuzberg-3.11.1 → kreuzberg-3.11.2}/kreuzberg/_gmft.py

@@ -444,14 +444,26 @@ def _extract_tables_in_process(
                 cropped_image.save(img_bytes, format="PNG")
                 img_bytes.seek(0)
 
-                results.append(
-                    {
-                        "cropped_image_bytes": img_bytes.getvalue(),
-                        "page_number": cropped_table.page.page_number,
-                        "text": data_frame.to_markdown(),
-                        "df_csv": data_frame.to_csv(index=False),
-                    }
-                )
+                if data_frame.empty:
+                    results.append(
+                        {
+                            "cropped_image_bytes": img_bytes.getvalue(),
+                            "page_number": cropped_table.page.page_number,
+                            "text": data_frame.to_markdown(),
+                            "df_columns": data_frame.columns.tolist(),
+                            "df_csv": None,
+                        }
+                    )
+                else:
+                    results.append(
+                        {
+                            "cropped_image_bytes": img_bytes.getvalue(),
+                            "page_number": cropped_table.page.page_number,
+                            "text": data_frame.to_markdown(),
+                            "df_columns": None,
+                            "df_csv": data_frame.to_csv(index=False),
+                        }
+                    )
 
             result_queue.put((True, results))
 
@@ -532,7 +544,10 @@ def _extract_tables_isolated(
                 img = Image.open(io.BytesIO(table_dict["cropped_image_bytes"]))
                 import pandas as pd  # noqa: PLC0415
 
-                df = pd.read_csv(StringIO(table_dict["df_csv"]))
+                if table_dict["df_csv"] is None:
+                    df = pd.DataFrame(columns=table_dict["df_columns"])
+                else:
+                    df = pd.read_csv(StringIO(table_dict["df_csv"]))
 
                 tables.append(
                     TableData(
@@ -638,7 +653,10 @@ async def _extract_tables_isolated_async(
                 img = Image.open(io.BytesIO(table_dict["cropped_image_bytes"]))
                 import pandas as pd  # noqa: PLC0415
 
-                df = pd.read_csv(StringIO(table_dict["df_csv"]))
+                if table_dict["df_csv"] is None:
+                    df = pd.DataFrame(columns=table_dict["df_columns"])
+                else:
+                    df = pd.read_csv(StringIO(table_dict["df_csv"]))
 
                 tables.append(
                     TableData(

{kreuzberg-3.11.1 → kreuzberg-3.11.2}/pyproject.toml

@@ -5,7 +5,7 @@ requires = [ "hatchling" ]
 
 [project]
 name = "kreuzberg"
-version = "3.11.1"
+version = "3.11.2"
 description = "Document intelligence framework for Python - Extract text, metadata, and structured data from diverse file formats"
 readme = "README.md"
 keywords = [
@@ -61,7 +61,7 @@ dependencies = [
   "chardetng-py>=0.3.5",
   "exceptiongroup>=1.2.2; python_version<'3.11'",
   "html-to-markdown[lxml]>=1.9.0",
-  "mcp>=1.12.4",
+  "mcp>=1.13.0",
   "msgspec>=0.18.0",
   "playa-pdf>=0.7.0",                                 # pinned due to breaking changes in 0.5.0
   "psutil>=7.0.0",
@@ -97,7 +97,7 @@ optional-dependencies.entity-extraction = [ "keybert>=0.9.0", "spacy>=3.8.7" ]
 optional-dependencies.gmft = [ "gmft>=0.4.2" ]
 optional-dependencies.langdetect = [ "fast-langdetect>=0.3.2" ]
 optional-dependencies.paddleocr = [
-  "paddleocr>=3.1.0",
+  "paddleocr>=3.1.1",
   "paddlepaddle>=3.1.0",
   "setuptools>=80.9.0",
 ]
@@ -117,7 +117,7 @@ dev = [
   "pytest-mock>=3.14.0",
   "pytest-rerunfailures>=15.1",
   "pytest-timeout>=2.4.0",
-  "ruff>=0.12.8",
+  "ruff>=0.12.9",
   "trio>=0.30.0",
   "uv-bump",
 ]

{kreuzberg-3.11.1 → kreuzberg-3.11.2}/tests/gmft_test.py

@@ -669,6 +669,125 @@ class TestGMFTInlineExtractionEdgeCases:
             pytest.skip("GMFT dependency not available for inline testing")
 
 
+class TestGMFTWithoutTables:
+    """Test GMFT behavior with PDFs that have no tables - issue #104."""
+
+    @pytest.mark.anyio
+    async def test_extract_tables_pdf_without_tables_async(self) -> None:
+        """Test that extract_tables handles PDFs without tables gracefully (async)."""
+        pdf_path = Path("tests/test_source_files/searchable.pdf")
+
+        try:
+            tables = await extract_tables(pdf_path)
+
+            assert isinstance(tables, list)
+
+            for table in tables:
+                assert "page_number" in table
+                assert "text" in table
+                assert "df" in table
+                assert "cropped_image" in table
+
+                import pandas as pd
+
+                assert isinstance(table["df"], pd.DataFrame)
+        except MissingDependencyError:
+            pytest.skip("GMFT dependency not installed")
+
+    def test_extract_tables_pdf_without_tables_sync(self) -> None:
+        """Test that extract_tables_sync handles PDFs without tables gracefully (sync)."""
+        # Using searchable.pdf which is a simple text PDF without tables
+        pdf_path = Path("tests/test_source_files/searchable.pdf")
+
+        try:
+            tables = extract_tables_sync(pdf_path)
+
+            assert isinstance(tables, list)
+
+            for table in tables:
+                assert "page_number" in table
+                assert "text" in table
+                assert "df" in table
+                assert "cropped_image" in table
+
+                import pandas as pd
+
+                assert isinstance(table["df"], pd.DataFrame)
+        except MissingDependencyError:
+            pytest.skip("GMFT dependency not installed")
+
+    @pytest.mark.anyio
+    async def test_extract_file_with_gmft_pdf_without_tables(self) -> None:
+        """Test that extract_file with extract_tables=True handles PDFs without tables gracefully."""
+        pdf_path = Path("tests/test_source_files/searchable.pdf")
+
+        config = ExtractionConfig(
+            extract_tables=True,
+            gmft_config=GMFTConfig(
+                detector_base_threshold=0.85,
+                remove_null_rows=True,
+                enable_multi_header=True,
+            ),
+        )
+
+        try:
+            result = await extract_file(pdf_path, config=config)
+
+            assert result.content
+            assert "Sample PDF" in result.content
+
+            assert hasattr(result, "tables")
+            assert isinstance(result.tables, list)
+
+            for table in result.tables:
+                assert "page_number" in table
+                assert "text" in table
+                assert "df" in table
+                assert "cropped_image" in table
+
+                import pandas as pd
+
+                assert isinstance(table["df"], pd.DataFrame)
+        except MissingDependencyError:
+            pytest.skip("GMFT dependency not installed")
+
+    def test_extract_file_sync_with_gmft_pdf_without_tables(self) -> None:
+        """Test that extract_file_sync with extract_tables=True handles PDFs without tables gracefully."""
+        pdf_path = Path("tests/test_source_files/searchable.pdf")
+
+        from kreuzberg.extraction import extract_file_sync
+
+        config = ExtractionConfig(
+            extract_tables=True,
+            gmft_config=GMFTConfig(
+                detector_base_threshold=0.85,
+                remove_null_rows=True,
+                enable_multi_header=True,
+            ),
+        )
+
+        try:
+            result = extract_file_sync(pdf_path, config=config)
+
+            assert result.content
+            assert "Sample PDF" in result.content
+
+            assert hasattr(result, "tables")
+            assert isinstance(result.tables, list)
+
+            for table in result.tables:
+                assert "page_number" in table
+                assert "text" in table
+                assert "df" in table
+                assert "cropped_image" in table
+
+                import pandas as pd
+
+                assert isinstance(table["df"], pd.DataFrame)
+        except MissingDependencyError:
+            pytest.skip("GMFT dependency not installed")
+
+
 class TestGMFTConfigSerialization:
     """Test GMFTConfig serialization for multiprocessing."""