release: prepare ExcelAlchemy 2.2.4

CHANGELOG.md

@@ -20,19 +20,42 @@ the Pydantic adapter layer.
 - Added a regression test for unsupported annotated declarations to prevent
   native Python annotations from slipping through the workbook schema path
 
+### Changed
+
+- Clarified `FieldMetaInfo` as a compatibility facade over layered metadata
+  objects instead of treating it as the primary internal metadata model
+- Moved more core consumers and built-in codecs onto the layered metadata
+  objects (`declared`, `runtime`, `presentation`, and `constraints`)
+- Continued reducing the effective responsibility carried by the flat
+  `FieldMetaInfo` compatibility surface in the 2.x implementation
+- Concentrated necessary dynamic typing boundaries into explicit aliases in the
+  codec and metadata layers instead of leaving ad hoc `Any` usage scattered
+  across the codebase
+- Replaced a number of remaining loose `Any` annotations in the runtime path
+  with more explicit `object` or workbook-boundary aliases where the behavior
+  was already concrete
+- Added smoke coverage for the repository examples so the annotated schema and
+  custom storage examples are exercised directly in tests
+
 ### Compatibility Notes
 
 - No public import or export workflow API was removed in this release
 - Valid `ExcelFieldCodec` and `CompositeExcelFieldCodec` declarations continue
   to work unchanged
 - Unsupported native annotations with `ExcelMeta(...)` now fail early with the
   intended `ProgrammaticError`
+- `FieldMeta(...)` and `ExcelMeta(...)` remain the stable public metadata entry
+  points while internal metadata continues to consolidate behind them
 
 ### Release Summary
 
 - unsupported annotated declarations now fail with the intended error again
 - codec resolution is stricter and easier to reason about
 - the validation fix is protected by an explicit integration regression test
+- metadata internals continue to move toward layered objects rather than a flat
+  central record
+- runtime typing boundaries are more explicit without changing the public API
+- repository examples now have direct smoke coverage in the test suite
 
 ## [2.2.2] - 2026-04-03
 

MIGRATIONS.md

@@ -124,6 +124,19 @@ Additional top-level module guidance:
 - `excelalchemy.header_models` is internal and should not be imported in application code
 - `docs/public-api.md` summarizes stable public modules, compatibility modules, and internal modules
 
+## Import Inspection Names
+
+The 2.2 line also clarifies the recommended names for inspecting import-run
+state from the facade:
+
+- prefer `worksheet_table` over `df`
+- prefer `header_table` over `header_df`
+- prefer `cell_error_map` over `cell_errors`
+- prefer `row_error_map` over `row_errors`
+
+The old names still work as compatibility aliases in the 2.x line, but new
+code should use the clearer names above.
+
 ## Recommended Upgrade Checklist
 
 1. Upgrade your Python runtime to 3.12+.

README.md

@@ -178,15 +178,44 @@ pip install "ExcelAlchemy[minio]"
 Practical examples live in the repository:
 
 - [`examples/annotated_schema.py`](https://github.com/RayCarterLab/ExcelAlchemy/blob/main/examples/annotated_schema.py)
+- [`examples/employee_import_workflow.py`](https://github.com/RayCarterLab/ExcelAlchemy/blob/main/examples/employee_import_workflow.py)
+- [`examples/create_or_update_import.py`](https://github.com/RayCarterLab/ExcelAlchemy/blob/main/examples/create_or_update_import.py)
+- [`examples/date_and_range_fields.py`](https://github.com/RayCarterLab/ExcelAlchemy/blob/main/examples/date_and_range_fields.py)
+- [`examples/selection_fields.py`](https://github.com/RayCarterLab/ExcelAlchemy/blob/main/examples/selection_fields.py)
 - [`examples/custom_storage.py`](https://github.com/RayCarterLab/ExcelAlchemy/blob/main/examples/custom_storage.py)
+- [`examples/export_workflow.py`](https://github.com/RayCarterLab/ExcelAlchemy/blob/main/examples/export_workflow.py)
+- [`examples/minio_storage.py`](https://github.com/RayCarterLab/ExcelAlchemy/blob/main/examples/minio_storage.py)
 - [`examples/fastapi_upload.py`](https://github.com/RayCarterLab/ExcelAlchemy/blob/main/examples/fastapi_upload.py)
+- [`examples/README.md`](https://github.com/RayCarterLab/ExcelAlchemy/blob/main/examples/README.md)
+
+If you want the recommended reading order, start with
+[`examples/README.md`](https://github.com/RayCarterLab/ExcelAlchemy/blob/main/examples/README.md).
 
 ## Public API Boundaries
 
 If you want to know which modules are stable public entry points versus
 compatibility shims or internal modules, see
 [`docs/public-api.md`](https://github.com/RayCarterLab/ExcelAlchemy/blob/main/docs/public-api.md).
 
+## Import Inspection Names
+
+When you inspect import-run state from the facade, prefer the clearer 2.2 names:
+
+- `alchemy.worksheet_table`
+- `alchemy.header_table`
+- `alchemy.cell_error_map`
+- `alchemy.row_error_map`
+
+The older aliases:
+
+- `alchemy.df`
+- `alchemy.header_df`
+- `alchemy.cell_errors`
+- `alchemy.row_errors`
+
+still work in the 2.x line as compatibility paths, but new application code
+should use the clearer names above.
+
 ## Locale-Aware Workbook Output
 
 `locale` affects workbook-facing display text such as:
@@ -316,7 +345,7 @@ The short version:
 | Topic | v1-style risk | Current v2 design |
 | --- | --- | --- |
 | Field access | Tight coupling to `__fields__` / `ModelField` | Adapter over `model_fields` |
-| Metadata ownership | Excel metadata mixed with validation internals | `FieldMetaInfo` owns Excel metadata |
+| Metadata ownership | Excel metadata mixed with validation internals | `FieldMetaInfo` is a compatibility facade over layered Excel metadata |
 | Validation integration | Deep reliance on internals | Adapter + explicit runtime validation |
 | Upgrade path | Brittle | Layered |
 

README_cn.md

@@ -102,6 +102,24 @@ pip install ExcelAlchemy
 pip install "ExcelAlchemy[minio]"
 ```
 
+## 示例
+
+仓库里有一组更贴近实际接入的示例：
+
+- [`examples/annotated_schema.py`](https://github.com/RayCarterLab/ExcelAlchemy/blob/main/examples/annotated_schema.py)
+- [`examples/employee_import_workflow.py`](https://github.com/RayCarterLab/ExcelAlchemy/blob/main/examples/employee_import_workflow.py)
+- [`examples/create_or_update_import.py`](https://github.com/RayCarterLab/ExcelAlchemy/blob/main/examples/create_or_update_import.py)
+- [`examples/date_and_range_fields.py`](https://github.com/RayCarterLab/ExcelAlchemy/blob/main/examples/date_and_range_fields.py)
+- [`examples/selection_fields.py`](https://github.com/RayCarterLab/ExcelAlchemy/blob/main/examples/selection_fields.py)
+- [`examples/custom_storage.py`](https://github.com/RayCarterLab/ExcelAlchemy/blob/main/examples/custom_storage.py)
+- [`examples/export_workflow.py`](https://github.com/RayCarterLab/ExcelAlchemy/blob/main/examples/export_workflow.py)
+- [`examples/minio_storage.py`](https://github.com/RayCarterLab/ExcelAlchemy/blob/main/examples/minio_storage.py)
+- [`examples/fastapi_upload.py`](https://github.com/RayCarterLab/ExcelAlchemy/blob/main/examples/fastapi_upload.py)
+- [`examples/README.md`](https://github.com/RayCarterLab/ExcelAlchemy/blob/main/examples/README.md)
+
+如果你想按推荐顺序来阅读，建议先看
+[`examples/README.md`](https://github.com/RayCarterLab/ExcelAlchemy/blob/main/examples/README.md)。
+
 ## 快速开始
 
 ```python
@@ -192,6 +210,24 @@ alchemy = ExcelAlchemy(ExporterConfig(Importer, storage=InMemoryExcelStorage()))
 
 如果你希望使用内置 Minio 实现，推荐显式传入 `storage=MinioStorageGateway(...)`，而不是再把 Minio 配置散落到门面层。
 
+## 导入结果状态查看命名
+
+如果你需要从 facade 上查看一次导入后的中间状态，推荐使用 2.2 这套更清晰的命名：
+
+- `alchemy.worksheet_table`
+- `alchemy.header_table`
+- `alchemy.cell_error_map`
+- `alchemy.row_error_map`
+
+旧别名：
+
+- `alchemy.df`
+- `alchemy.header_df`
+- `alchemy.cell_errors`
+- `alchemy.row_errors`
+
+在 2.x 里仍然可用，用于兼容旧代码；但新代码建议统一使用前面这组更明确的名字。
+
 ## 为什么这样设计
 
 ### 为什么去掉 pandas
@@ -210,7 +246,7 @@ alchemy = ExcelAlchemy(ExporterConfig(Importer, storage=InMemoryExcelStorage()))
 Excel 元数据不应该深绑到 Pydantic 内部结构上。
 所以现在的分层是：
 
-- `FieldMetaInfo` 负责 Excel 元数据
+- `FieldMetaInfo` 是对外兼容 façade，内部再组合声明层、运行时绑定层、展示层和导入约束层
 - `helper/pydantic.py` 只做适配
 - 真正的业务校验仍然由 ExcelAlchemy 控制
 

docs/architecture.md

@@ -99,8 +99,10 @@ flowchart LR
 
 `src/excelalchemy/metadata.py`
 
-- owns Excel field metadata
-- exposes workbook comment fragments
+- exposes `FieldMeta(...)` / `ExcelMeta(...)` as the stable public entry points
+- keeps `FieldMetaInfo` as a compatibility facade for the 2.x line
+- splits the real metadata state into declaration, runtime binding,
+  workbook presentation, and import-constraint layers
 - keeps runtime metadata separate from validation backend internals
 
 ### Pydantic Integration

docs/public-api.md

@@ -34,6 +34,9 @@ These modules are the recommended import paths for application code:
   The recommended backend configuration pattern in the 2.x line.
 - `ExcelArtifact`
   The recommended return shape when you need bytes, base64, or data URLs.
+- import inspection names:
+  Prefer `worksheet_table`, `header_table`, `cell_error_map`, and
+  `row_error_map` when reading import-run state from the facade.
 
 ## Compatibility Modules In 2.x
 
@@ -92,3 +95,10 @@ direction is:
   and `excelalchemy.codecs`
 - backend integration through `ExcelStorage`
 - internal orchestration and helper modules treated as implementation details
+
+For import-run state naming, the long-term direction is also:
+
+- clear facade inspection names such as `worksheet_table`, `header_table`,
+  `cell_error_map`, and `row_error_map`
+- older aliases such as `df`, `header_df`, `cell_errors`, and `row_errors`
+  retained only as 2.x compatibility paths

docs/releases/2.2.3.md

@@ -9,6 +9,8 @@ line.
 - present `2.2.3` as a focused validation-correctness release
 - restore the intended failure mode for unsupported annotated declarations
 - ship regression coverage for the declaration guard in the Pydantic adapter
+- continue consolidating the remaining runtime typing gray areas
+- add direct smoke coverage for repository examples
 
 ## Release Positioning
 
@@ -19,6 +21,8 @@ line.
 - unsupported native annotations with `ExcelMeta(...)` now fail early with the
   intended `ProgrammaticError`
 - the validation path is safer and easier to reason about
+- runtime typing boundaries are more explicit in core codec and metadata paths
+- repository examples now have direct smoke coverage in tests
 
 ## Before Tagging
 
@@ -28,6 +32,8 @@ line.
    `src/excelalchemy/helper/pydantic.py`.
 4. Confirm the regression test in
    `tests/integration/test_excelalchemy_workflows.py`.
+5. Confirm the example smoke tests in
+   `tests/integration/test_examples_smoke.py`.
 
 ## Local Verification
 
@@ -36,9 +42,11 @@ Run these commands from the repository root:
 ```bash
 uv sync --extra development
 uv run ruff check src/excelalchemy/helper/pydantic.py \
-  tests/integration/test_excelalchemy_workflows.py
+  tests/integration/test_excelalchemy_workflows.py \
+  tests/integration/test_examples_smoke.py
 uv run pyright
-uv run pytest tests/integration/test_excelalchemy_workflows.py -q
+uv run pytest tests/integration/test_excelalchemy_workflows.py \
+  tests/integration/test_examples_smoke.py -q
 rm -rf dist
 uv build
 uvx twine check dist/*
@@ -60,6 +68,8 @@ themes clearly:
 - unsupported annotated declarations now fail with the intended error again
 - codec resolution in the Pydantic adapter is stricter and more explicit
 - the fix is protected by a regression test
+- runtime typing boundaries are more explicit and less ad hoc
+- repository examples are exercised directly in the test suite
 
 ## Recommended Release Messaging
 
@@ -69,6 +79,8 @@ Prefer wording that emphasizes stability and correctness:
 - "restores the intended ProgrammaticError path"
 - "tightens codec resolution in the Pydantic adapter"
 - "adds regression coverage for unsupported annotated declarations"
+- "continues tightening runtime typing boundaries without breaking the public API"
+- "adds smoke coverage for repository examples"
 
 ## Done When
 

examples/README.md

@@ -0,0 +1,89 @@
+# Examples
+
+These examples are organized as a recommended learning path rather than a flat list.
+
+## Recommended Reading Order
+
+1. `annotated_schema.py`
+   - Start here if you want to learn the declaration style first.
+   - Shows the modern `Annotated[..., Field(...), ExcelMeta(...)]` pattern.
+2. `employee_import_workflow.py`
+   - Read this next if you want to understand the core import story.
+   - Shows template generation, workbook upload, import execution, and result reporting.
+3. `create_or_update_import.py`
+   - Read this after the basic import flow.
+   - Shows `ImporterConfig.for_create_or_update(...)` with `is_data_exist`, `creator`, and `updater`.
+4. `export_workflow.py`
+   - Read this once the import flow is clear.
+   - Shows artifact generation, export uploads, and a custom storage-backed export task.
+5. `custom_storage.py`
+   - Read this when you want to implement your own `ExcelStorage`.
+   - Keeps the example minimal and focused on the protocol boundary.
+6. `date_and_range_fields.py`
+   - Read this if you want to understand workbook-friendly date, date range, number range, and money fields.
+7. `selection_fields.py`
+   - Read this if your domain uses approval forms, assignments, ownership trees, or selection-heavy templates.
+8. `minio_storage.py`
+   - Read this if you need the built-in Minio path in the current 2.x line.
+   - This reflects the current 2.x compatibility-based Minio path rather than a future 3.x-only storage story.
+9. `fastapi_upload.py`
+   - Read this last as an integration sketch.
+   - It is useful once the import and storage examples already make sense.
+
+## By Goal
+
+- Learn the declaration style:
+  - `annotated_schema.py`
+- Learn the core import flow:
+  - `employee_import_workflow.py`
+  - `create_or_update_import.py`
+- Learn export and storage integration:
+  - `export_workflow.py`
+  - `custom_storage.py`
+  - `minio_storage.py`
+- Learn field families:
+  - `date_and_range_fields.py`
+  - `selection_fields.py`
+- Learn web integration:
+  - `fastapi_upload.py`
+
+## Storage and Backend Integration
+
+- `custom_storage.py`
+  - Shows a minimal custom `ExcelStorage` implementation for export uploads.
+- `export_workflow.py`
+  - Shows a realistic export flow with artifact generation and upload.
+- `minio_storage.py`
+  - Shows the built-in Minio-backed storage path currently available in the 2.x line.
+- `fastapi_upload.py`
+  - Shows a FastAPI integration sketch for template download and workbook import.
+
+## How To Run
+
+Run examples from the repository root:
+
+```bash
+uv run python examples/annotated_schema.py
+uv run python examples/employee_import_workflow.py
+uv run python examples/create_or_update_import.py
+uv run python examples/date_and_range_fields.py
+uv run python examples/selection_fields.py
+uv run python examples/custom_storage.py
+uv run python examples/export_workflow.py
+uv run python examples/minio_storage.py
+```
+
+If you want to try the FastAPI sketch, install FastAPI first and then run your
+preferred ASGI server against `examples.fastapi_upload:app`.
+
+## Notes
+
+- The examples intentionally use in-memory storage so they stay self-contained.
+- They are meant to show the recommended public API shape for the stable 2.x
+  line.
+- If you want a production backend, prefer `storage=...` with
+  `MinioStorageGateway` or your own `ExcelStorage` implementation.
+- The built-in `minio_storage.py` example reflects the current 2.x Minio path,
+  which still uses the compatibility configuration fields under the hood.
+- The smoke tests in `tests/integration/test_examples_smoke.py` cover the main
+  example entry points directly.

examples/annotated_schema.py

@@ -1,16 +1,18 @@
 """Minimal example that uses Annotated + ExcelMeta declarations."""
 
-from __future__ import annotations
-
 from typing import Annotated
 
 from pydantic import BaseModel, Field
 
-from excelalchemy import Email, ExcelAlchemy, ExcelMeta, ImporterConfig, Number
+from excelalchemy import Email, ExcelAlchemy, ExcelMeta, ImporterConfig, Number, String
 
 
 class EmployeeImporter(BaseModel):
-    full_name: Annotated[str, Field(min_length=2), ExcelMeta(label='Full name', order=1, hint='Use the legal name')]
+    full_name: Annotated[
+        String,
+        Field(min_length=2),
+        ExcelMeta(label='Full name', order=1, hint='Use the legal name'),
+    ]
     age: Annotated[Number, Field(ge=18), ExcelMeta(label='Age', order=2)]
     work_email: Annotated[
         Email,