tablevalidator-databricks 0.1.5__tar.gz → 0.1.7__tar.gz

{tablevalidator_databricks-0.1.5 → tablevalidator_databricks-0.1.7}/PKG-INFO

@@ -1,12 +1,12 @@
 Metadata-Version: 2.4
 Name: tablevalidator-databricks
-Version: 0.1.5
+Version: 0.1.7
 Summary: Generates a widget-driven Databricks notebook UI on top of the table-validator package - no code required to run a comparison.
 License: MIT
 Requires-Python: >=3.9
 Description-Content-Type: text/markdown
 Requires-Dist: typer>=0.12
-Requires-Dist: table-validator>=0.1.19
+Requires-Dist: table-validator>=0.1.22
 Provides-Extra: dev
 Requires-Dist: pytest>=7.0; extra == "dev"
 Requires-Dist: ruff>=0.4; extra == "dev"

{tablevalidator_databricks-0.1.5 → tablevalidator_databricks-0.1.7}/pyproject.toml

@@ -11,7 +11,12 @@ requires-python = ">=3.9"
 license = { text = "MIT" }
 dependencies = [
     "typer>=0.12",
-    "table-validator>=0.1.19",
+    # 0.1.22 is the first release whose result sheets survive Arrow
+    # conversion, which is what Databricks' display() uses - older engines
+    # fail with "ArrowTypeError: Expected bytes, got a 'int' object" on a
+    # column mixing row counts with blanks. The generated notebook's
+    # %pip install floor is read from this line.
+    "table-validator>=0.1.22",
 ]
 
 [project.optional-dependencies]

{tablevalidator_databricks-0.1.5 → tablevalidator_databricks-0.1.7}/tablevalidator_databricks/cli/main.py

@@ -1,9 +1,11 @@
 """CLI entry point: `tablevalidator-databricks` console script."""
 
 from pathlib import Path
+from typing import Optional
 
 import typer
 
+from tablevalidator_databricks import __version__
 from tablevalidator_databricks.generator.generator import (
     VALID_MODES,
     _min_core_version,
@@ -22,6 +24,25 @@ app = typer.Typer(
 DEFAULT_OUTPUT_PATH = Path("TableValidator.py")
 
 
+def _version_callback(value: bool) -> None:
+    if value:
+        typer.echo(f"tablevalidator-databricks {__version__}")
+        raise typer.Exit()
+
+
+@app.callback()
+def _main(
+    version: Optional[bool] = typer.Option(
+        None,
+        "--version",
+        callback=_version_callback,
+        is_eager=True,
+        help="Show the installed version and exit.",
+    ),
+) -> None:
+    """Generate a widget-driven Databricks notebook UI for table-validator."""
+
+
 @app.command()
 def init(
     output: Path = typer.Option(
@@ -62,10 +83,16 @@ def init(
     generate_notebook(output, mode)
 
     typer.echo(f"Generated notebook: {output.resolve()}")
+    typer.echo(f"  generator version: {__version__}  (mode: {mode})")
     typer.echo(
         "Import it into Databricks (Workspace -> Import, format 'Source'), "
         "fill in the widgets, and Run All."
     )
+    typer.echo(
+        "Re-import over the old notebook if you are updating one - widgets are "
+        "baked in at generation time, so an existing notebook keeps its old "
+        "code until you replace it."
+    )
 
 
 _INFO_TEXT = """
@@ -164,6 +191,25 @@ something far too old and fails with:
 
 Re-run 'tablevalidator-databricks init' after upgrading this package -
 an already-generated notebook keeps whatever code it was written with.
+
+A WIDGET IS MISSING FROM MY NOTEBOOK
+------------------------------------
+Widgets are baked into the .py file when it is generated. Upgrading this
+package does NOT change a notebook that already exists - you have to
+generate a new one and re-import it.
+
+Check what generated the notebook you are looking at: the title cell at
+the top names the version. Compare it against your installed version:
+    tablevalidator-databricks --version
+
+Then regenerate and re-import:
+    pip install --upgrade tablevalidator-databricks
+    tablevalidator-databricks init
+
+For reference, widgets arrived in these versions:
+    0.1.1   "(manual override)" text box beside each dropdown
+    0.1.2   "(all tables in schema)" schema-wide sweep option
+    0.1.3   Yes/No dropdowns per check group (replacing a multiselect)
 """.strip("\n")
 
 

{tablevalidator_databricks-0.1.5 → tablevalidator_databricks-0.1.7}/tablevalidator_databricks/generator/generator.py

@@ -56,10 +56,13 @@ def generate_notebook(output: Path, mode: str) -> None:
         .joinpath("widget_notebook.py.tmpl")
         .read_text(encoding="utf-8")
     )
+    from tablevalidator_databricks import __version__
+
     rendered = (
         template.replace("{{MODE}}", mode)
         .replace("{{SHOW_OPTIONAL_WIDGETS}}", "True" if mode == "full" else "False")
         .replace("{{MIN_CORE_VERSION}}", _min_core_version())
+        .replace("{{GENERATOR_VERSION}}", __version__)
     )
 
     output.parent.mkdir(parents=True, exist_ok=True)

{tablevalidator_databricks-0.1.5 → tablevalidator_databricks-0.1.7}/tablevalidator_databricks/templates/widget_notebook.py.tmpl

@@ -1,12 +1,22 @@
 # Databricks notebook source
 # MAGIC %md
 # MAGIC # Table Validator
-# MAGIC Generated by `tablevalidator-databricks init --mode {{MODE}}`.
+# MAGIC Generated by `tablevalidator-databricks` **v{{GENERATOR_VERSION}}**
+# MAGIC (`init --mode {{MODE}}`).
 # MAGIC
-# MAGIC Fill in the widgets above (Source/Target Catalog, Schema, Table, and
-# MAGIC which checks to run), then **Run All**. No code needs to be written or
-# MAGIC edited - every cell below just reads the widgets and calls the
-# MAGIC `table_validator` package's `validate_tables()` API.
+# MAGIC Widgets are baked in when this file is generated - upgrading the
+# MAGIC package does NOT change a notebook that already exists. If a widget
+# MAGIC described below is missing here, this notebook was generated by an
+# MAGIC older version: run `pip install --upgrade tablevalidator-databricks`
+# MAGIC then `tablevalidator-databricks init` again, and re-import the result.
+# MAGIC
+# MAGIC **First time here?** Choose **Run All**. Widgets are created by running
+# MAGIC cells, so on a fresh notebook they only appear once the run reaches the
+# MAGIC "Create / refresh widgets" cell - that is normal, not a fault. Once they
+# MAGIC are there, fill them in and run again.
+# MAGIC
+# MAGIC No code needs to be written or edited - every cell below just reads the
+# MAGIC widgets and calls the `table_validator` package's `validate_tables()` API.
 # MAGIC
 # MAGIC **Changing Catalog or Schema after the dropdowns are already built?**
 # MAGIC Databricks widgets have no on-change event, so picking a new Catalog
@@ -43,7 +53,7 @@
 
 # COMMAND ----------
 
-# MAGIC %pip install "table-validator>={{MIN_CORE_VERSION}}"
+# MAGIC %pip install --upgrade "table-validator>={{MIN_CORE_VERSION}}"
 
 # COMMAND ----------
 
@@ -51,12 +61,41 @@ dbutils.library.restartPython()
 
 # COMMAND ----------
 
+import table_validator
 from table_validator import validate_tables
 from table_validator.config.schema import ValidationType
 
 MODE = "{{MODE}}"
 SHOW_OPTIONAL_WIDGETS = {{SHOW_OPTIONAL_WIDGETS}}
 
+# Print what is actually loaded. A %pip install without --upgrade silently
+# keeps an older already-installed engine ("Requirement already
+# satisfied"), so the version that matters is the one imported here, not
+# the one the install cell asked for.
+_MIN_CORE = "{{MIN_CORE_VERSION}}"
+print(f"table-validator engine: {table_validator.__version__}  (need >= {_MIN_CORE})")
+
+
+def _version_tuple(text):
+    parts = []
+    for chunk in str(text).split(".")[:3]:
+        digits = "".join(c for c in chunk if c.isdigit())
+        parts.append(int(digits) if digits else 0)
+    while len(parts) < 3:
+        parts.append(0)
+    return tuple(parts)
+
+
+if _version_tuple(table_validator.__version__) < _version_tuple(_MIN_CORE):
+    print(
+        f"\n  WARNING: this engine ({table_validator.__version__}) is older than "
+        f"{_MIN_CORE}, which this notebook expects.\n"
+        "  You may hit errors such as 'ArrowTypeError: Expected bytes, got a "
+        "'int' object' when displaying results.\n"
+        "  Fix: run the install cell above with --upgrade, then re-run "
+        "dbutils.library.restartPython() and start again."
+    )
+
 
 def _list_catalogs():
     return sorted(r["catalog"] for r in spark.sql("SHOW CATALOGS").collect())
@@ -176,9 +215,44 @@ def _build_cascading_widgets(prefix: str, label_prefix: str) -> None:
     dbutils.widgets.text(f"{prefix}_table_override", "", f"{label_prefix} Table (manual override)")
 
 
+def _build_check_widgets() -> None:
+    """The three check-group Yes/No dropdowns, plus (in full mode) the
+    optional filter/column text boxes.
+
+    One Yes/No dropdown per check group, rather than a single multiselect -
+    dbutils.widgets.multiselect enforces an internal "selection sequence"
+    validation against its own default that was observed to reject even a
+    same-length, same-items default on some Databricks runtimes/existing-
+    widget states (DefaultValueNotInChoicesList), with no reliable
+    workaround. Three independent dropdowns have no such cross-widget
+    validation and are the more robust primitive for this."""
+    yes_no = ["Yes", "No"]
+    # --mode schema means "schema SHAPE only" - that shape (column names,
+    # types, nullability) is validated by the COLUMN checks, so Column
+    # stays on and only the expensive Row checks (row counts + row-level
+    # data) are off by default. Turning Column off too would leave nothing
+    # producing a per-table verdict, reporting every table as SKIPPED.
+    row_default = "No" if MODE == "schema" else "Yes"
+
+    dbutils.widgets.dropdown("check_catalog_schema", "Yes", yes_no, "Run Catalog & Schema checks?")
+    dbutils.widgets.dropdown("check_column", "Yes", yes_no, "Run Column checks?")
+    dbutils.widgets.dropdown("check_row", row_default, yes_no, "Run Row checks?")
+
+    if SHOW_OPTIONAL_WIDGETS:
+        dbutils.widgets.text("primary_key", "", "Primary key column(s) (comma-separated, optional)")
+        dbutils.widgets.text("only_columns", "", "Only compare these columns (comma-separated, optional)")
+        dbutils.widgets.text("ignore_columns", "", "Skip these columns entirely (comma-separated, optional)")
+        dbutils.widgets.text("row_filter", "", "Row filter - SQL WHERE-fragment (optional)")
+
+
+# Every widget is created here, in ONE cell, so they all appear together
+# and in a predictable order - Databricks lays widgets out in creation
+# order, so splitting these across cells made them interleave oddly and
+# made later ones seem to "appear" partway through a Run All.
 catalogs = _list_catalogs()
 _build_cascading_widgets("source", "Source")
 _build_cascading_widgets("target", "Target")
+_build_check_widgets()
 
 print(
     "Widgets refreshed for the CURRENT Catalog/Schema selection.\n"
@@ -188,7 +262,7 @@ print(
 
 # COMMAND ----------
 
-# MAGIC %md ### Validation checks
+# MAGIC %md ### About the check groups
 # MAGIC "Catalog & Schema" only checks that the catalogs/schemas/tables exist
 # MAGIC and match by name - it produces no per-table PASS/FAIL on its own, so
 # MAGIC running it alone reports every table as **SKIPPED**. That's expected,
@@ -199,34 +273,6 @@ print(
 
 # COMMAND ----------
 
-# One Yes/No dropdown per check group, instead of a single multiselect -
-# dbutils.widgets.multiselect enforces an internal "selection sequence"
-# validation against its own default that was observed to reject even a
-# same-length, same-items default on some Databricks runtimes/existing-
-# widget states (DefaultValueNotInChoicesList), with no reliable
-# workaround. Three independent dropdowns have no such cross-widget
-# validation and are the more robust primitive for this.
-_YES_NO = ["Yes", "No"]
-_schema_default = "Yes"
-# --mode schema means "schema SHAPE only" - that shape (column names,
-# types, nullability) is validated by the COLUMN checks, so Column stays
-# on and only the expensive Row checks (row counts + row-level data) are
-# off by default. Turning Column off too would leave nothing producing a
-# per-table verdict, reporting every table as SKIPPED.
-_column_default = "Yes"
-_row_default = "No" if MODE == "schema" else "Yes"
-dbutils.widgets.dropdown("check_catalog_schema", _schema_default, _YES_NO, "Run Catalog & Schema checks?")
-dbutils.widgets.dropdown("check_column", _column_default, _YES_NO, "Run Column checks?")
-dbutils.widgets.dropdown("check_row", _row_default, _YES_NO, "Run Row checks?")
-
-if SHOW_OPTIONAL_WIDGETS:
-    dbutils.widgets.text("only_columns", "", "Only compare these columns (comma-separated, optional)")
-    dbutils.widgets.text("ignore_columns", "", "Skip these columns entirely (comma-separated, optional)")
-    dbutils.widgets.text("row_filter", "", "Row filter - SQL WHERE-fragment (optional)")
-    dbutils.widgets.text("primary_key", "", "Primary key column(s) (comma-separated, optional)")
-
-# COMMAND ----------
-
 # MAGIC %md ### Run validation
 
 # COMMAND ----------

{tablevalidator_databricks-0.1.5 → tablevalidator_databricks-0.1.7}/tablevalidator_databricks.egg-info/PKG-INFO

@@ -1,12 +1,12 @@
 Metadata-Version: 2.4
 Name: tablevalidator-databricks
-Version: 0.1.5
+Version: 0.1.7
 Summary: Generates a widget-driven Databricks notebook UI on top of the table-validator package - no code required to run a comparison.
 License: MIT
 Requires-Python: >=3.9
 Description-Content-Type: text/markdown
 Requires-Dist: typer>=0.12
-Requires-Dist: table-validator>=0.1.19
+Requires-Dist: table-validator>=0.1.22
 Provides-Extra: dev
 Requires-Dist: pytest>=7.0; extra == "dev"
 Requires-Dist: ruff>=0.4; extra == "dev"

{tablevalidator_databricks-0.1.5 → tablevalidator_databricks-0.1.7}/tablevalidator_databricks.egg-info/requires.txt

@@ -1,5 +1,5 @@
 typer>=0.12
-table-validator>=0.1.19
+table-validator>=0.1.22
 
 [dev]
 pytest>=7.0

{tablevalidator_databricks-0.1.5 → tablevalidator_databricks-0.1.7}/tablevalidator_databricks.egg-info/scm_version.json

@@ -1,7 +1,7 @@
 {
-  "tag": "0.1.5",
+  "tag": "0.1.7",
   "distance": 0,
-  "node": "g00d5f8a",
+  "node": "ga3dbc87",
   "dirty": false,
   "branch": "HEAD",
   "node_date": "2026-09-10"

{tablevalidator_databricks-0.1.5 → tablevalidator_databricks-0.1.7}/tests/test_cli.py

@@ -88,3 +88,36 @@ def test_info_renders_the_real_version_floor() -> None:
 
     assert "{MIN_CORE_VERSION}" not in result.output
     assert f'table-validator>={_min_core_version()}' in result.output
+
+
+def test_version_flag_reports_installed_version() -> None:
+    """`--version` exists so a user can check what they have installed
+    against what generated a notebook - the fastest way to diagnose a
+    "my widgets look different" report."""
+    from tablevalidator_databricks import __version__
+
+    result = runner.invoke(app, ["--version"])
+
+    assert result.exit_code == 0
+    assert __version__ in result.output
+
+
+def test_init_prints_the_generator_version(tmp_path: Path) -> None:
+    """init names the version it generated with, so the number is visible
+    at the moment the file is created, not just inside it."""
+    from tablevalidator_databricks import __version__
+
+    result = runner.invoke(app, ["init", "--output", str(tmp_path / "nb.py")])
+
+    assert result.exit_code == 0
+    assert __version__ in result.output
+
+
+def test_info_explains_missing_widgets() -> None:
+    """The "a widget is missing" case is a real, reported confusion -
+    info must explain that widgets are baked in at generation time."""
+    result = runner.invoke(app, ["info"])
+
+    assert "WIDGET IS MISSING" in result.output
+    assert "--version" in result.output
+    assert "0.1.3" in result.output  # the version-history table

{tablevalidator_databricks-0.1.5 → tablevalidator_databricks-0.1.7}/tests/test_generator.py

@@ -103,9 +103,11 @@ def test_generated_notebook_pins_a_core_version_floor(tmp_path: Path) -> None:
     generate_notebook(output, "full")
     content = output.read_text(encoding="utf-8")
 
-    assert 'pip install "table-validator>=' in content
+    assert 'pip install --upgrade "table-validator>=' in content
     # The floor must be new enough to actually contain validate_tables().
-    match = re.search(r'pip install "table-validator>=([0-9]+)\.([0-9]+)\.([0-9]+)"', content)
+    match = re.search(
+        r'pip install --upgrade "table-validator>=([0-9]+)\.([0-9]+)\.([0-9]+)"', content
+    )
     assert match, "no parseable version floor in the %pip install line"
     major, minor, patch = (int(g) for g in match.groups())
     assert (major, minor, patch) >= (0, 1, 9)
@@ -116,3 +118,55 @@ def test_min_core_version_reads_from_package_metadata() -> None:
     dependency, so the two can never silently drift apart."""
     assert _min_core_version() != "0.0.0"
     assert re.match(r"^[0-9]+\.[0-9]+\.[0-9]+$", _min_core_version())
+
+
+def test_generated_notebook_stamps_the_generator_version(tmp_path: Path) -> None:
+    """The notebook records which version generated it, so a user seeing
+    unexpected widgets can tell at a glance whether their file is stale."""
+    from tablevalidator_databricks import __version__
+
+    output = tmp_path / "TableValidator.py"
+    generate_notebook(output, "full")
+    content = output.read_text(encoding="utf-8")
+
+    assert __version__ in content
+    # And it explains that regenerating is what picks up new widgets.
+    assert "upgrading the" in content.lower()
+    assert "tablevalidator-databricks init" in content
+
+
+def test_pip_install_line_forces_an_upgrade(tmp_path: Path) -> None:
+    """Real user-reported failure: `%pip install "table-validator"` printed
+    "Requirement already satisfied" and kept an older engine (0.1.20),
+    which then crashed on display(). The install must force an upgrade so
+    an existing older install is actually replaced."""
+    output = tmp_path / "TableValidator.py"
+    generate_notebook(output, "full")
+    content = output.read_text(encoding="utf-8")
+
+    assert 'pip install --upgrade "table-validator>=' in content
+
+
+def test_notebook_reports_the_loaded_engine_version(tmp_path: Path) -> None:
+    """The version that matters is the one actually imported, not the one
+    the install cell asked for - the notebook must print it and warn when
+    it is below the required floor."""
+    output = tmp_path / "TableValidator.py"
+    generate_notebook(output, "full")
+    content = output.read_text(encoding="utf-8")
+
+    assert "table_validator.__version__" in content
+    assert "WARNING" in content
+    assert "ArrowTypeError" in content
+
+
+def test_core_version_floor_includes_the_arrow_fix(tmp_path: Path) -> None:
+    """The floor must be at least 0.1.22, the first release whose result
+    sheets survive Arrow conversion (what Databricks' display() uses)."""
+    output = tmp_path / "TableValidator.py"
+    generate_notebook(output, "full")
+    content = output.read_text(encoding="utf-8")
+
+    match = re.search(r'table-validator>=([0-9]+)\.([0-9]+)\.([0-9]+)"', content)
+    assert match
+    assert tuple(int(g) for g in match.groups()) >= (0, 1, 22)