feat: model max_length_unit on SchemaDiscoveryColumn

.github/workflows/release.yml

@@ -60,3 +60,27 @@ jobs:
 
       - name: Publish to PyPI
         uses: pypa/gh-action-pypi-publish@release/v1
+
+  github-release:
+    name: Publish GitHub Release
+    needs: publish
+    runs-on: ubuntu-latest
+    permissions:
+      contents: write
+    steps:
+      - name: Download distributions
+        uses: actions/download-artifact@v4
+        with:
+          name: dist
+          path: dist/
+
+      - name: Create GitHub Release
+        env:
+          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        run: |
+          gh release create "${GITHUB_REF_NAME}" \
+            --repo "${GITHUB_REPOSITORY}" \
+            --title "${GITHUB_REF_NAME}" \
+            --verify-tag \
+            --notes "See [HISTORY.rst](https://github.com/${GITHUB_REPOSITORY}/blob/${GITHUB_REF_NAME}/HISTORY.rst) for changelog." \
+            dist/*

HISTORY.rst

@@ -2,6 +2,15 @@
 History
 =======
 
+1.3.1 (2026-09-09)
+------------------
+
+* Added ``max_length_unit`` to ``SchemaDiscoveryColumn``, typed by the new ``LengthUnit`` enum
+  (``chars`` or ``bytes``), reporting the unit a column's ``max_length`` is counted in.
+  It is ``None`` for results from older servers that do not report a unit.
+
+Requires server version 3.26.17
+
 1.3.0 (2026-09-04)
 ------------------
 

RELEASING.rst

@@ -68,31 +68,10 @@ Releasing to PyPI
    You can verify that the release appears at https://pypi.org/project/datamasque-python/,
    and Read the Docs has built the new tag
    at https://datamasque-python.readthedocs.io/.
-5. Create the GitHub Release (see below).
-
-Creating the GitHub Release
-===========================
-
-Do this after the PyPI publish has succeeded,
-so the release never points at a version nobody can install.
-
-1. On the repository home page,
-   find **Releases** in the right-hand sidebar under the *About* box,
-   and click it,
-   then **Draft a new release**.
-2. **Choose a tag** — pick the existing tag you pushed as part of releasing to PyPI.
-   Don't let GitHub create a new one; the tag should already exist from the release flow.
-3. Set the target to ``main``.
-4. Title it the same as the tag, e.g. ``v1.2.2``.
-5. Paste the ``HISTORY.rst`` entry for this version into the body, converted to Markdown.
-   **Generate release notes** is a reasonable starting point for the commit list,
-   but the hand-written changelog is what users read.
-6. Leave *Set as a pre-release* unticked for a normal release.
-7. Click **Publish release**.
-
-Creating releases can also be done from the ``gh`` CLI::
-
-    gh release create v1.2.2 --title v1.2.2 --notes-file notes.md
+5. The GitHub Release is created automatically once the PyPI publish succeeds.
+   ``release.yml`` runs a ``github-release`` job that creates the release for the pushed tag,
+   attaches the built distributions, and links to ``HISTORY.rst`` at that tag for the changelog.
+   Nothing to do by hand; just confirm it appears under **Releases** on the repository home page.
 
 Releasing a dev build to TestPyPI
 =================================

datamasque/client/__init__.py

@@ -73,6 +73,7 @@
     ForeignKeyRef,
     InDataDiscoveryConfig,
     InDataDiscoveryRule,
+    LengthUnit,
     ReferencingForeignKey,
     RulesetGenerationRequest,
     RulesetGenerationWithRGConfigRequest,
@@ -239,6 +240,7 @@
     "InvalidRulesetError",
     "JsonPath",
     "LengthEntry",
+    "LengthUnit",
     "LengthsStatistics",
     "LicenseInfo",
     "Locator",

datamasque/client/models/discovery.py

@@ -367,13 +367,22 @@ class ReferencingForeignKey(BaseModel):
     referencing_column: str  # Dotted path: "schema.table.column".
 
 
+class LengthUnit(Enum):
+    """The unit a column's declared width (`max_length`) is counted in."""
+
+    chars = "chars"
+    bytes = "bytes"
+
+
 class SchemaDiscoveryColumn(BaseModel):
     """Column-level data in a schema discovery result."""
 
     model_config = ConfigDict(extra="allow")
 
     data_type: Optional[str] = None
     max_length: Optional[int] = None
+    # None when the server does not report a unit (older runs count chars).
+    max_length_unit: Optional[LengthUnit] = None
     foreign_keys: list[ForeignKeyRef]
     discovery_matches: list[DiscoveryMatch]
     numeric_precision: Optional[int] = None

pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "datamasque-python"
-version = "1.3.0"
+version = "1.3.1"
 description = "Official Python client for the DataMasque data-masking API."
 authors = [
     { name = "DataMasque Ltd" },

tests/test_discovery.py

@@ -21,6 +21,7 @@
     FileRulesetGenerationRequest,
     FileRulesetGenerationWithRGConfigRequest,
     InDataDiscoveryConfig,
+    LengthUnit,
     RGConfig,
     RGConfigId,
     RulesetGenerationRequest,
@@ -795,6 +796,22 @@ def test_schema_discovery_result_parses_safe_data_preview():
     assert preview.statistics_common.count_distinct == 988
 
 
+def test_schema_discovery_result_parses_max_length_unit() -> None:
+    row = _schema_discovery_row(1, "name")
+    row["data"]["max_length"] = 10
+    row["data"]["max_length_unit"] = "bytes"
+    result = SchemaDiscoveryResult.model_validate(row)
+    assert result.data.max_length == 10
+    assert result.data.max_length_unit is LengthUnit.bytes
+
+
+def test_schema_discovery_result_without_max_length_unit_is_none() -> None:
+    row = _schema_discovery_row(1, "name")
+    assert "max_length_unit" not in row["data"]
+    result = SchemaDiscoveryResult.model_validate(row)
+    assert result.data.max_length_unit is None
+
+
 def test_list_schema_discovery_results_follows_pagination(client):
     run_id = RunId(42)
     page1 = {