mostlyright-data 0.10.0__tar.gz → 0.10.2__tar.gz

{mostlyright_data-0.10.0 → mostlyright_data-0.10.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: mostlyright-data
-Version: 0.10.0
+Version: 0.10.2
 Summary: Mostly Right CLI for reproducible, reviewed Parquet datasets
 Project-URL: Homepage, https://mostlyright.md/
 Project-URL: Documentation, https://mostlyright.md/docs/guides/cli/
@@ -9,51 +9,88 @@ Provides-Extra: build
 Requires-Dist: hatchling==1.28.0; extra == 'build'
 Requires-Dist: uv-build==0.11.3; extra == 'build'
 Provides-Extra: dev
+Requires-Dist: antlr4-python3-runtime==4.9.3; extra == 'dev'
 Requires-Dist: astropy==8.0.1; extra == 'dev'
+Requires-Dist: cffi==2.1.0; extra == 'dev'
 Requires-Dist: cmarkgfm>=2025.10.22; extra == 'dev'
+Requires-Dist: colorlog==6.12.0; extra == 'dev'
 Requires-Dist: croniter==6.2.4; extra == 'dev'
-Requires-Dist: cryptography<51,>=50.0.0; extra == 'dev'
+Requires-Dist: cryptography==50.0.0; extra == 'dev'
 Requires-Dist: duckdb==1.5.5; extra == 'dev'
 Requires-Dist: eccodes==2.48.0; extra == 'dev'
 Requires-Dist: eccodeslib==2.48.0.26; (sys_platform != 'win32') and extra == 'dev'
 Requires-Dist: et-xmlfile==2.0.0; extra == 'dev'
+Requires-Dist: flatbuffers==25.12.19; extra == 'dev'
 Requires-Dist: h5py==3.16.0; extra == 'dev'
-Requires-Dist: idna<4,>=3.10; extra == 'dev'
+Requires-Dist: idna==3.18; extra == 'dev'
 Requires-Dist: jsonschema[format]<5,>=4.23; extra == 'dev'
+Requires-Dist: lxml==6.1.2; extra == 'dev'
+Requires-Dist: narwhals==2.25.0; extra == 'dev'
 Requires-Dist: netcdf4==1.7.4; extra == 'dev'
 Requires-Dist: numpy==2.5.1; extra == 'dev'
+Requires-Dist: omegaconf==2.3.1; extra == 'dev'
+Requires-Dist: onnxruntime==1.28.0; extra == 'dev'
+Requires-Dist: opencv-python-headless==5.0.0.93; extra == 'dev'
 Requires-Dist: openpyxl==3.1.5; extra == 'dev'
-Requires-Dist: packaging<27,>=24; extra == 'dev'
+Requires-Dist: packaging==26.3; extra == 'dev'
 Requires-Dist: pandas==3.0.3; extra == 'dev'
+Requires-Dist: pikepdf==10.10.0; extra == 'dev'
 Requires-Dist: pillow==12.3.0; extra == 'dev'
 Requires-Dist: polars-runtime-32==1.42.1; extra == 'dev'
 Requires-Dist: polars==1.42.1; extra == 'dev'
+Requires-Dist: protobuf==7.35.1; extra == 'dev'
 Requires-Dist: pyarrow==23.0.1; extra == 'dev'
+Requires-Dist: pyclipper==1.4.0; extra == 'dev'
+Requires-Dist: pycparser==3.0; extra == 'dev'
 Requires-Dist: pyhdf==0.11.7; ((sys_platform == 'linux' and platform_machine == 'x86_64') or (sys_platform == 'darwin' and platform_machine == 'arm64')) and extra == 'dev'
+Requires-Dist: pypdfium2==5.13.0; extra == 'dev'
 Requires-Dist: pypublicdecompwt==2.8.1.6; extra == 'dev'
 Requires-Dist: pyreadstat==1.3.6; extra == 'dev'
 Requires-Dist: pytest-xdist>=3.6; extra == 'dev'
 Requires-Dist: pytest>=8.3; extra == 'dev'
+Requires-Dist: python-dateutil==2.9.0.post0; extra == 'dev'
+Requires-Dist: pyyaml==6.0.3; extra == 'dev'
 Requires-Dist: rasterio==1.5.1; extra == 'dev'
 Requires-Dist: referencing<1,>=0.35; extra == 'dev'
 Requires-Dist: ruff>=0.9; extra == 'dev'
+Requires-Dist: shapely==2.1.2; extra == 'dev'
+Requires-Dist: six==1.17.0; extra == 'dev'
 Requires-Dist: zarr==3.3.0; extra == 'dev'
 Provides-Extra: hosted
 Requires-Dist: mostlyright-studio-client-v3==3.0.0; extra == 'hosted'
 Provides-Extra: local
+Requires-Dist: antlr4-python3-runtime==4.9.3; extra == 'local'
+Requires-Dist: cffi==2.1.0; extra == 'local'
+Requires-Dist: colorlog==6.12.0; extra == 'local'
 Requires-Dist: croniter==6.2.4; extra == 'local'
-Requires-Dist: cryptography<51,>=50.0.0; extra == 'local'
+Requires-Dist: cryptography==50.0.0; extra == 'local'
 Requires-Dist: duckdb==1.5.5; extra == 'local'
 Requires-Dist: et-xmlfile==2.0.0; extra == 'local'
-Requires-Dist: idna<4,>=3.10; extra == 'local'
+Requires-Dist: flatbuffers==25.12.19; extra == 'local'
+Requires-Dist: idna==3.18; extra == 'local'
+Requires-Dist: lxml==6.1.2; extra == 'local'
+Requires-Dist: narwhals==2.25.0; extra == 'local'
 Requires-Dist: numpy==2.5.1; extra == 'local'
+Requires-Dist: omegaconf==2.3.1; extra == 'local'
+Requires-Dist: onnxruntime==1.28.0; extra == 'local'
+Requires-Dist: opencv-python-headless==5.0.0.93; extra == 'local'
 Requires-Dist: openpyxl==3.1.5; extra == 'local'
-Requires-Dist: packaging<27,>=24; extra == 'local'
+Requires-Dist: packaging==26.3; extra == 'local'
 Requires-Dist: pandas==3.0.3; extra == 'local'
+Requires-Dist: pikepdf==10.10.0; extra == 'local'
+Requires-Dist: pillow==12.3.0; extra == 'local'
 Requires-Dist: polars-runtime-32==1.42.1; extra == 'local'
 Requires-Dist: polars==1.42.1; extra == 'local'
+Requires-Dist: protobuf==7.35.1; extra == 'local'
 Requires-Dist: pyarrow==23.0.1; extra == 'local'
+Requires-Dist: pyclipper==1.4.0; extra == 'local'
+Requires-Dist: pycparser==3.0; extra == 'local'
+Requires-Dist: pypdfium2==5.13.0; extra == 'local'
 Requires-Dist: pyreadstat==1.3.6; extra == 'local'
+Requires-Dist: python-dateutil==2.9.0.post0; extra == 'local'
+Requires-Dist: pyyaml==6.0.3; extra == 'local'
+Requires-Dist: shapely==2.1.2; extra == 'local'
+Requires-Dist: six==1.17.0; extra == 'local'
 Provides-Extra: local-search-neural
 Requires-Dist: anyio==4.14.2; extra == 'local-search-neural'
 Requires-Dist: certifi==2026.7.22; extra == 'local-search-neural'
@@ -76,22 +113,41 @@ Requires-Dist: tokenizers==0.23.1; extra == 'local-search-neural'
 Requires-Dist: tqdm==4.70.0; extra == 'local-search-neural'
 Requires-Dist: typing-extensions==4.16.0; extra == 'local-search-neural'
 Provides-Extra: recorder
-Requires-Dist: cryptography<51,>=50.0.0; extra == 'recorder'
+Requires-Dist: cryptography==50.0.0; extra == 'recorder'
 Provides-Extra: release
+Requires-Dist: antlr4-python3-runtime==4.9.3; extra == 'release'
+Requires-Dist: cffi==2.1.0; extra == 'release'
+Requires-Dist: colorlog==6.12.0; extra == 'release'
 Requires-Dist: croniter==6.2.4; extra == 'release'
-Requires-Dist: cryptography<51,>=50.0.0; extra == 'release'
+Requires-Dist: cryptography==50.0.0; extra == 'release'
 Requires-Dist: duckdb==1.5.5; extra == 'release'
 Requires-Dist: et-xmlfile==2.0.0; extra == 'release'
-Requires-Dist: idna<4,>=3.10; extra == 'release'
+Requires-Dist: flatbuffers==25.12.19; extra == 'release'
+Requires-Dist: idna==3.18; extra == 'release'
 Requires-Dist: jsonschema[format]<5,>=4.23; extra == 'release'
+Requires-Dist: lxml==6.1.2; extra == 'release'
+Requires-Dist: narwhals==2.25.0; extra == 'release'
 Requires-Dist: numpy==2.5.1; extra == 'release'
+Requires-Dist: omegaconf==2.3.1; extra == 'release'
+Requires-Dist: onnxruntime==1.28.0; extra == 'release'
+Requires-Dist: opencv-python-headless==5.0.0.93; extra == 'release'
 Requires-Dist: openpyxl==3.1.5; extra == 'release'
-Requires-Dist: packaging<27,>=24; extra == 'release'
+Requires-Dist: packaging==26.3; extra == 'release'
 Requires-Dist: pandas==3.0.3; extra == 'release'
+Requires-Dist: pikepdf==10.10.0; extra == 'release'
+Requires-Dist: pillow==12.3.0; extra == 'release'
 Requires-Dist: polars-runtime-32==1.42.1; extra == 'release'
 Requires-Dist: polars==1.42.1; extra == 'release'
+Requires-Dist: protobuf==7.35.1; extra == 'release'
 Requires-Dist: pyarrow==23.0.1; extra == 'release'
+Requires-Dist: pyclipper==1.4.0; extra == 'release'
+Requires-Dist: pycparser==3.0; extra == 'release'
+Requires-Dist: pypdfium2==5.13.0; extra == 'release'
 Requires-Dist: pyreadstat==1.3.6; extra == 'release'
+Requires-Dist: python-dateutil==2.9.0.post0; extra == 'release'
+Requires-Dist: pyyaml==6.0.3; extra == 'release'
+Requires-Dist: shapely==2.1.2; extra == 'release'
+Requires-Dist: six==1.17.0; extra == 'release'
 Provides-Extra: scientific
 Requires-Dist: astropy==8.0.1; extra == 'scientific'
 Requires-Dist: eccodes==2.48.0; extra == 'scientific'
@@ -129,6 +185,12 @@ python -m pip install mostlyright-data
 
 The default install is the hosted client. It works on Linux, macOS, and Windows.
 
+It also carries the `mr-data-build` agent skill, which Claude Code and Codex read to author a
+recipe without rediscovering the contract a run at a time. The first `mr-data` command you run
+places it in your agent's skill directory; there is nothing to install by hand. A skill you have
+edited yourself is never replaced, and `MOSTLYRIGHT_SKILL_AUTOINSTALL=0` switches the placement
+off.
+
 ## Sign in
 
 ```bash
@@ -179,6 +241,7 @@ says why.
 - [Hosted research sessions](https://mostlyright.md/docs/guides/hosted-research/)
 - [Credentials](https://mostlyright.md/docs/guides/credentials/)
 - [Public datasets](https://mostlyright.md/docs/guides/use-public-datasets/)
+- [Certified document extraction](docs/DOCUMENT-EXTRACTION.md)
 
 Use `mr-data --help` for the full command list and options. Commands that support
 `--json` write one JSON object.

{mostlyright_data-0.10.0 → mostlyright_data-0.10.2}/README.md

@@ -19,6 +19,12 @@ python -m pip install mostlyright-data
 
 The default install is the hosted client. It works on Linux, macOS, and Windows.
 
+It also carries the `mr-data-build` agent skill, which Claude Code and Codex read to author a
+recipe without rediscovering the contract a run at a time. The first `mr-data` command you run
+places it in your agent's skill directory; there is nothing to install by hand. A skill you have
+edited yourself is never replaced, and `MOSTLYRIGHT_SKILL_AUTOINSTALL=0` switches the placement
+off.
+
 ## Sign in
 
 ```bash
@@ -69,6 +75,7 @@ says why.
 - [Hosted research sessions](https://mostlyright.md/docs/guides/hosted-research/)
 - [Credentials](https://mostlyright.md/docs/guides/credentials/)
 - [Public datasets](https://mostlyright.md/docs/guides/use-public-datasets/)
+- [Certified document extraction](docs/DOCUMENT-EXTRACTION.md)
 
 Use `mr-data --help` for the full command list and options. Commands that support
 `--json` write one JSON object.

{mostlyright_data-0.10.0 → mostlyright_data-0.10.2}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "mostlyright-data"
-version = "0.10.0"
+version = "0.10.2"
 description = "Mostly Right CLI for reproducible, reviewed Parquet datasets"
 readme = "README.md"
 requires-python = ">=3.13,<3.14"
@@ -37,6 +37,7 @@ pypi-release = [
 local-search-neural-conversion = [
     "onnx==1.20.1",
     "onnxruntime==1.28.0",
+    "pillow==12.3.0",
 ]
 
 [tool.uv]
@@ -49,6 +50,7 @@ mostlyright-grib = [
     { path = "vendor/wheels/mostlyright_grib-0.1.0-py3-none-macosx_11_0_arm64.whl", marker = "sys_platform == 'darwin'" },
     { path = "vendor/wheels/mostlyright_grib-0.1.0-py3-none-manylinux_2_36_x86_64.whl", marker = "sys_platform == 'linux'" },
 ]
+antlr4-python3-runtime = { path = "vendor/wheels/antlr4_python3_runtime-4.9.3-py3-none-any.whl" }
 mostlyright-studio-client-v3 = { path = "vendor/wheels/mostlyright_studio_client_v3-3.0.0-py3-none-any.whl" }
 pyhdf = [
     { path = "vendor/wheels/pyhdf-0.11.7-cp313-cp313-macosx_15_0_arm64.whl", marker = "sys_platform == 'darwin' and platform_machine == 'arm64'" },
@@ -109,7 +111,14 @@ build = [
 # `mostlyright-data[local]`.
 local = [
     "croniter==6.2.4",
-    "cryptography>=50.0.0,<51",
+    # Kept character-identical to the recorder extra: keyed stream signatures must resolve the
+    # same approved cryptography coordinate whether they execute in the local closure or the
+    # independently installed recorder image.
+    "cryptography==50.0.0",
+    "cffi==2.1.0",
+    "pycparser==3.0",
+    "python-dateutil==2.9.0.post0",
+    "six==1.17.0",
     # The warm session worker's bounded-query engine, and the ONLY reason it is here rather than
     # in the base `dependencies`: duckdb ships a compiled extension, and `tests/test_thin_profile.py`
     # installs the base closure into a fresh environment and refuses any `.so`/`.pyd`/`.dylib` it
@@ -122,20 +131,40 @@ local = [
     "duckdb==1.5.5",
     # UTS46 with URL Standard validity checks. `acquisition/url_policy` requires the same
     # domain-to-ASCII behavior as the URL parser; Python's `idna` codec implements IDNA2003.
-    "idna>=3.10,<4",
+    "idna==3.18",
     "openpyxl==3.1.5",
     "et-xmlfile==2.0.0",
     "numpy==2.5.1",
-    "packaging>=24,<27",
+    "packaging==26.3",
     "pyarrow==23.0.1",
     "pandas==3.0.3",
     "polars==1.42.1",
     "polars-runtime-32==1.42.1",
+    # Certified document lane: PDFium and qpdf are native C/C++ engines.  RapidOCR 3.9.2's
+    # Apache-2.0 Python source and its three approved ONNX model bytes are vendored in this
+    # distribution, so a published wheel does not request the upstream GUI OpenCV distribution.
+    # The explicit, exact supporting packages make the resulting headless OCR coordinate a
+    # reproducible public-PyPI install graph rather than a repository-only source override.
+    "pypdfium2==5.13.0",
+    "pikepdf==10.10.0",
+    "onnxruntime==1.28.0",
+    "flatbuffers==25.12.19",
+    "protobuf==7.35.1",
+    "opencv-python-headless==5.0.0.93",
+    "antlr4-python3-runtime==4.9.3",
+    "colorlog==6.12.0",
+    "omegaconf==2.3.1",
+    "pillow==12.3.0",
+    "lxml==6.1.2",
+    "pyclipper==1.4.0",
+    "pyyaml==6.0.3",
+    "shapely==2.1.2",
     # The bounded statistical-microdata Readers use the reviewed ReadStat binding for Stata,
     # SPSS, and SAS Transport files.  It stays in the local lane: the base hosted client must
     # remain dependency-free, and the compiled binding is never imported before exact Reader
     # resolution.
     "pyreadstat==1.3.6",
+    "narwhals==2.25.0",
 ]
 # What the hosted stream recorder needs beyond the base distribution, and nothing else. The two
 # asymmetric signing algorithms in the keyed-wss connector grammar -- `rsa-pss-sha256` for Kalshi
@@ -150,7 +179,7 @@ local = [
 # `tests/test_thin_profile.py`: two specifiers that resolve to one version today and drift
 # tomorrow is exactly the failure the load-bearing-pin warning above exists for.
 recorder = [
-    "cryptography>=50.0.0,<51",
+    "cryptography==50.0.0",
 ]
 # Native scientific decoders are kept out of the ordinary local and thin-client closures. The
 # hosted Builder and networkless Courier install this extra explicitly. Every Reader still gates
@@ -243,9 +272,17 @@ build-backend = "hatchling.build"
 [tool.hatch.build.targets.wheel]
 packages = ["src/mostlyright"]
 
+# The agent skill ships INSIDE the wheel. `skills/mr-data-build` stays the one editable copy --
+# the documentation contract tests read it there -- and the same bytes are carried into the
+# package so `pip install mostlyright-data` delivers it. Without this the skill reached only
+# readers who had cloned the repository, while the documented install is pip.
+[tool.hatch.build.targets.wheel.force-include]
+"skills/mr-data-build" = "mostlyright/data_harness/skill_tree/mr-data-build"
+
 [tool.hatch.build.targets.sdist]
 only-include = [
     "scripts/hatch_build.py",
+    "skills/mr-data-build",
     "src/mostlyright",
 ]
 
@@ -273,6 +310,10 @@ tmp_path_retention_count = 96
 [tool.ruff]
 line-length = 100
 target-version = "py311"
+# Upstream RapidOCR is intentionally vendored verbatim apart from the closed-coordinate changes.
+# Its style is not reformatted as project code, but Make's explicit F/E9 pass below still makes
+# every shipped vendor module syntax- and undefined-name-safe.
+extend-exclude = ["src/mostlyright/_vendor/rapidocr"]
 
 [tool.ruff.lint]
 select = ["E", "F", "I", "UP", "B", "RUF"]