extract-python 0.9.1__tar.gz → 0.10.1__tar.gz

{extract_python-0.9.1 → extract_python-0.10.1}/PKG-INFO

@@ -1,13 +1,13 @@
 Metadata-Version: 2.5
 Name: extract-python
-Version: 0.9.1
+Version: 0.10.1
 Summary: Structured content extraction
 Project-URL: Homepage, https://github.com/ICIJ/extract-python
 Project-URL: Repository, https://github.com/ICIJ/extract-python
 Project-URL: Issues, https://github.com/ICIJ/extract-python/issues
 Author-email: Clément Doumouro <cdoumouro@icij.org>
 Requires-Python: <3.15,>=3.13
-Requires-Dist: extract-core~=0.9.0
+Requires-Dist: extract-core~=0.10.4
 Requires-Dist: icij-common~=0.8.2
 Requires-Dist: pympler~=1.1
 Provides-Extra: benches

{extract_python-0.9.1 → extract_python-0.10.1}/extract_python/docling_.py

@@ -182,8 +182,6 @@ def _to_markdown_doc(
     #  nested in the tree structured
     md_dir_name = path_to_artifacts_dirname(input_path)
     md_dir = output_path / md_dir_name
-    if md_dir.exists():
-        raise FileExistsError(f"directory {md_dir} already exists")
     # Let's avoid issue of duplicated input file names flattened top level
     md_filename = md_dir_name + OutputFormat.MARKDOWN
     with tempfile.TemporaryDirectory(ignore_cleanup_errors=True) as td:

{extract_python-0.9.1 → extract_python-0.10.1}/pyproject.toml

@@ -8,7 +8,7 @@ authors = [
 readme = "README.md"
 requires-python = ">=3.13,<3.15"
 dependencies = [
-  "extract-core~=0.9.0",
+  "extract-core~=0.10.4",
   "icij-common~=0.8.2",
   "pympler~=1.1",
 ]

{extract_python-0.9.1 → extract_python-0.10.1}/uv.lock

@@ -927,7 +927,7 @@ provides-extras = ["cli"]
 
 [package.metadata.requires-dev]
 dev = [
-    { name = "docling", specifier = "~=2.127" },
+    { name = "docling", specifier = "~=2.132" },
     { name = "docling-slim", extras = ["models-vlm-inline"], specifier = "~=2.127" },
     { name = "mpmath", specifier = "~=1.3.0" },
     { name = "nest-asyncio", specifier = ">=1.6.0" },
@@ -988,7 +988,7 @@ requires-dist = [
     { name = "icij-common", specifier = "~=0.8.2" },
     { name = "markdown2", marker = "extra == 'benches'", specifier = ">=2.5.4" },
     { name = "marker-pdf", marker = "extra == 'marker'", specifier = "~=1.10" },
-    { name = "mineru", extras = ["torch", "vlm"], marker = "extra == 'mineru'", specifier = "~=4.0" },
+    { name = "mineru", extras = ["torch"], marker = "extra == 'mineru'", specifier = "~=4.0" },
     { name = "notebook", marker = "extra == 'benches'", specifier = ">=7.4.5" },
     { name = "pydantic-extra-types", extras = ["pycountry"], marker = "extra == 'mineru'", specifier = "~=2.11" },
     { name = "pympler", specifier = "~=1.1" },