Merged
Conversation
- Remove table module and all functionality - Remove lexicon module and DSL parsers - Deletes filter and select PEG grammars - Removes parser implementations - Update project configuration - Modifies pyproject.toml dependencies - Updates uv.lock with fewer dependencies - Strip README documentation - Clean up package exports in __init__.py
- Add KEEP wrapper to preserve empty values
- Prevents removal of {}, [], '', None
- Implements __init__, __repr__, __eq__
- Export KEEP from main package
- Document KEEP usage in README
- Add process_output function - Combines DROP, KEEP, and empty removal - Accepts remove_empty parameter (default True) - Update README examples to use process_output - Replace process_drops with process_output - Add remove_empty option documentation - Export process_output in __init__.py
- Add @Mapper decorator for data transformations - Auto-processes DROP, KEEP, and empty values - Supports remove_empty parameter - Update README with decorator-based examples - Replace direct process_output calls - Show composable mapping functions - Export mapper from chidian package
- Add context.py with mapping_context context manager using contextvars - Update grab() to check strict mode from context - In strict mode, missing keys raise KeyError instead of returning None - Distinguishes between 'key not found' and 'key exists with None value' - Update README with strict mode documentation
- Remove mapper.py (old Mapper class, replaced by @Mapper decorator) - Remove partials.py (old FunctionChain/ChainableFunction utilities) - Remove lib/data_mapping_helpers.py (old helper functions)
- Add test_grab.py for grab() function (renamed from get) - Add test_mapper_new.py for @Mapper, DROP, KEEP, mapping_context - Update test_lib.py to use grab - Remove old test files: test_get.py, test_put.py, test_mapper.py, test_partials.py, test_data_mapping.py, test_property_based.py, test_types.py, structstest.py All 46 tests passing
- Fix tests path link - Document actual error type (KeyError) in strict mode
- Add Ok/Err result types as frozen dataclasses - Add CheckFn, Path, ValidationError type aliases
- Add V dataclass with __and__/__or__ composition operators - Add DictV for nested dict validation with field validators - Add ListV for list validation with item validators - Add to_validator() dispatch function for type coercion
- Required, Optional for presence validation - IsType for type checking - InRange, MinLength, MaxLength for length constraints - InSet for enum-like validation - Matches for regex patterns - Predicate for custom functions - Eq, Gt, Gte, Lt, Lte, Between for comparisons
- V: base validator with check function, composition via & and | - DictV: nested dict validation with field validators - ListV: list validation with item validator and length constraints - to_validator(): type coercion dispatch function
- Required/Optional: presence modifiers - IsType: type instance checking - InRange/MinLength/MaxLength: length constraints - InSet: enum-like validation - Matches: regex pattern validation - Predicate: custom predicate functions - Eq/Gt/Gte/Lt/Lte/Between: comparison validators
- validate(): validate dict data against schema, returns Ok/Err - to_pydantic(): compile schema to Pydantic BaseModel subclass
Exports: Ok, Err, V, DictV, ListV, to_validator, Required, Optional, IsType, InRange, MinLength, MaxLength, InSet, Matches, Predicate, Eq, Gt, Gte, Lt, Lte, Between, validate, to_pydantic
![cubic-dev-ai[bot]](https://avatars.githubusercontent.com/in/1082092?s=60&v=4){kind=small}
There was a problem hiding this comment.
13 issues found across 42 files
Prompt for AI agents (all 13 issues)
Check if these issues are valid — if so, understand the root cause of each and fix them.
<file name="chidian/decorator.py">
<violation number="1" location="chidian/decorator.py:11">
The mapper signature uses the PEP 604 `Callable | None` union without a `from __future__ import annotations`, which raises a SyntaxError on the project’s supported Python 3.8/3.9 versions. Use `Optional[Callable]` (and import Optional) or add the future import so the file can be parsed on supported interpreters.</violation>
</file>
<file name="chidian/process.py">
<violation number="1" location="chidian/process.py:61">
KEEP currently bypasses DROP processing entirely, so DROP sentinels inside KEEP-wrapped containers leak through instead of being applied.</violation>
</file>
<file name="chidian/validation/schema.py">
<violation number="1" location="chidian/validation/schema.py:81">
`match` statements require Python 3.10+, but this module is supposed to run on Python 3.8+, so it will not even import on the supported runtimes.</violation>
<violation number="2" location="chidian/validation/schema.py:88">
Returning `dict[str, Any]` / `dict[str, Any] | None` at runtime requires Python 3.9/3.10 features, so `_extract_pydantic_field` will crash on Python 3.8 even though that version is supported.</violation>
<violation number="3" location="chidian/validation/schema.py:93">
Using `list[item_type]` / `list[item_type] | None` at runtime requires Python 3.9/3.10 features, so optional/required list fields cannot be generated under the supported Python 3.8 runtime.</violation>
</file>
<file name="chidian/validation/types.py">
<violation number="1" location="chidian/validation/types.py:16">
`dataclass(slots=True)` requires Python 3.10+, but the project promises Python ≥3.8, so this module will crash during import on 3.8/3.9.</violation>
<violation number="2" location="chidian/validation/types.py:44">
Using `tuple[str | int, ...]` relies on Python 3.9/3.10-only syntax, so the module cannot even be parsed on Python 3.8 despite the declared support range.</violation>
<violation number="3" location="chidian/validation/types.py:45">
`tuple[Path, str]` depends on PEP 585 built-in generics, which do not exist on Python 3.8, so this assignment crashes at import on the supported interpreter versions.</violation>
<violation number="4" location="chidian/validation/types.py:46">
`list[ValidationError]` relies on Python 3.9’s PEP 585 generics, so the module cannot be imported on Python 3.8, contradicting the stated support range.</violation>
</file>
<file name="README.md">
<violation number="1" location="README.md:70">
`patient_summary` cannot be chained with `normalize_user` as written because the latter does not output the `data.*` structure `patient_summary` expects; the example will raise a KeyError.</violation>
</file>
<file name="chidian/core.py">
<violation number="1" location="chidian/core.py:31">
Docstring incorrectly states that strict mode raises ValueError for missing paths even though traverse_path actually propagates KeyError/IndexError/TypeError, which misleads callers about which exceptions to handle.</violation>
</file>
<file name="chidian/drop.py">
<violation number="1" location="chidian/drop.py:98">
DROP.PARENT used directly in a list removes only the list instead of its parent container, contradicting the documented semantics.</violation>
<violation number="2" location="chidian/drop.py:121">
DROP.GRANDPARENT and higher applied directly inside a list only remove ancestors one level too shallow because the propagation subtracts two levels.</violation>
</file>
Reply to cubic to teach it or ask questions. Re-run a review with @cubic-dev-ai review this PR
| from .process import process_output | ||
|
|
||
|
|
||
| def mapper(_func: Callable | None = None, *, remove_empty: bool = True) -> Callable: |
There was a problem hiding this comment.
The mapper signature uses the PEP 604 Callable | None union without a from __future__ import annotations, which raises a SyntaxError on the project’s supported Python 3.8/3.9 versions. Use Optional[Callable] (and import Optional) or add the future import so the file can be parsed on supported interpreters.
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. At chidian/decorator.py, line 11:
<comment>The mapper signature uses the PEP 604 `Callable | None` union without a `from __future__ import annotations`, which raises a SyntaxError on the project’s supported Python 3.8/3.9 versions. Use `Optional[Callable]` (and import Optional) or add the future import so the file can be parsed on supported interpreters.</comment>
<file context>
@@ -0,0 +1,52 @@
+from .process import process_output
+
+
+def mapper(_func: Callable | None = None, *, remove_empty: bool = True) -> Callable:
+ """
+ Decorator that transforms a mapping function into a callable mapper.
</file context>
| raise _DropSignal(data.value) | ||
|
|
||
| # Handle KEEP wrapper - unwrap and mark as preserved | ||
| if isinstance(data, KEEP): |
There was a problem hiding this comment.
KEEP currently bypasses DROP processing entirely, so DROP sentinels inside KEEP-wrapped containers leak through instead of being applied.
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. At chidian/process.py, line 61:
<comment>KEEP currently bypasses DROP processing entirely, so DROP sentinels inside KEEP-wrapped containers leak through instead of being applied.</comment>
<file context>
@@ -0,0 +1,153 @@
+ raise _DropSignal(data.value)
+
+ # Handle KEEP wrapper - unwrap and mark as preserved
+ if isinstance(data, KEEP):
+ return data.value # Return the wrapped value as-is, skip empty check
+
</file context>
✅ Addressed in 694acea
| case ListV(required=req, items=items): | ||
| item_type, _ = _extract_pydantic_field(items) | ||
| if req: | ||
| return (list[item_type], ...) # type: ignore[valid-type] |
There was a problem hiding this comment.
Using list[item_type] / list[item_type] | None at runtime requires Python 3.9/3.10 features, so optional/required list fields cannot be generated under the supported Python 3.8 runtime.
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. At chidian/validation/schema.py, line 93:
<comment>Using `list[item_type]` / `list[item_type] | None` at runtime requires Python 3.9/3.10 features, so optional/required list fields cannot be generated under the supported Python 3.8 runtime.</comment>
<file context>
@@ -0,0 +1,96 @@
+ case ListV(required=req, items=items):
+ item_type, _ = _extract_pydantic_field(items)
+ if req:
+ return (list[item_type], ...) # type: ignore[valid-type]
+ return (list[item_type] | None, None) # type: ignore[valid-type]
+
</file context>
| return (TypingOptional[t or Any], None) | ||
| case DictV(required=req): | ||
| if req: | ||
| return (dict[str, Any], ...) |
There was a problem hiding this comment.
Returning dict[str, Any] / dict[str, Any] | None at runtime requires Python 3.9/3.10 features, so _extract_pydantic_field will crash on Python 3.8 even though that version is supported.
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. At chidian/validation/schema.py, line 88:
<comment>Returning `dict[str, Any]` / `dict[str, Any] | None` at runtime requires Python 3.9/3.10 features, so `_extract_pydantic_field` will crash on Python 3.8 even though that version is supported.</comment>
<file context>
@@ -0,0 +1,96 @@
+ return (TypingOptional[t or Any], None)
+ case DictV(required=req):
+ if req:
+ return (dict[str, Any], ...)
+ return (dict[str, Any] | None, None)
+ case ListV(required=req, items=items):
</file context>
|
|
||
| def _extract_pydantic_field(v: V | DictV | ListV) -> tuple[Any, Any]: | ||
| """Extract Pydantic field type and default from validator.""" | ||
| match v: |
There was a problem hiding this comment.
match statements require Python 3.10+, but this module is supposed to run on Python 3.8+, so it will not even import on the supported runtimes.
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. At chidian/validation/schema.py, line 81:
<comment>`match` statements require Python 3.10+, but this module is supposed to run on Python 3.8+, so it will not even import on the supported runtimes.</comment>
<file context>
@@ -0,0 +1,96 @@
+
+def _extract_pydantic_field(v: V | DictV | ListV) -> tuple[Any, Any]:
+ """Extract Pydantic field type and default from validator."""
+ match v:
+ case V(required=True, type_hint=t):
+ return (t or Any, ...)
</file context>
|
|
||
| # Type aliases | ||
| CheckFn = Callable[[Any], bool] | ||
| Path = tuple[str | int, ...] |
There was a problem hiding this comment.
Using tuple[str | int, ...] relies on Python 3.9/3.10-only syntax, so the module cannot even be parsed on Python 3.8 despite the declared support range.
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. At chidian/validation/types.py, line 44:
<comment>Using `tuple[str | int, ...]` relies on Python 3.9/3.10-only syntax, so the module cannot even be parsed on Python 3.8 despite the declared support range.</comment>
<file context>
@@ -0,0 +1,46 @@
+
+# Type aliases
+CheckFn = Callable[[Any], bool]
+Path = tuple[str | int, ...]
+ValidationError = tuple[Path, str]
+ValidationErrors = list[ValidationError]
</file context>
README.md
Outdated
| from myproject.mappings import normalize_user, patient_summary | ||
|
|
||
| # Chain mappings | ||
| result = patient_summary(normalize_user(raw_data)) |
There was a problem hiding this comment.
patient_summary cannot be chained with normalize_user as written because the latter does not output the data.* structure patient_summary expects; the example will raise a KeyError.
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. At README.md, line 70:
<comment>`patient_summary` cannot be chained with `normalize_user` as written because the latter does not output the `data.*` structure `patient_summary` expects; the example will raise a KeyError.</comment>
<file context>
@@ -1,129 +1,289 @@
+from myproject.mappings import normalize_user, patient_summary
+
+# Chain mappings
+result = patient_summary(normalize_user(raw_data))
+```
+
</file context>
✅ Addressed in 694acea
chidian/core.py
Outdated
| Value at path or default if not found | ||
|
|
||
| Raises: | ||
| ValueError: In strict mode (via mapping_context), if path not found |
There was a problem hiding this comment.
Docstring incorrectly states that strict mode raises ValueError for missing paths even though traverse_path actually propagates KeyError/IndexError/TypeError, which misleads callers about which exceptions to handle.
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. At chidian/core.py, line 31:
<comment>Docstring incorrectly states that strict mode raises ValueError for missing paths even though traverse_path actually propagates KeyError/IndexError/TypeError, which misleads callers about which exceptions to handle.</comment>
<file context>
@@ -1,48 +1,57 @@
Value at path or default if not found
+
+ Raises:
+ ValueError: In strict mode (via mapping_context), if path not found
+
+ Note:
</file context>
Suggested change
| ValueError: In strict mode (via mapping_context), if path not found | |
| KeyError | IndexError | TypeError: In strict mode, traverse_path propagates the underlying missing-path error |
✅ Addressed in 694acea
chidian/drop.py
Outdated
| raise _DropSignal(0) | ||
| else: | ||
| # GRANDPARENT or higher - propagate up | ||
| raise _DropSignal(item.value - 2) |
There was a problem hiding this comment.
DROP.GRANDPARENT and higher applied directly inside a list only remove ancestors one level too shallow because the propagation subtracts two levels.
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. At chidian/drop.py, line 121:
<comment>DROP.GRANDPARENT and higher applied directly inside a list only remove ancestors one level too shallow because the propagation subtracts two levels.</comment>
<file context>
@@ -0,0 +1,137 @@
+ raise _DropSignal(0)
+ else:
+ # GRANDPARENT or higher - propagate up
+ raise _DropSignal(item.value - 2)
+
+ try:
</file context>
✅ Addressed in 694acea
| pass | ||
| elif signal.levels == 1: | ||
| # Remove this dict from its parent | ||
| raise _DropSignal(0) |
There was a problem hiding this comment.
DROP.PARENT used directly in a list removes only the list instead of its parent container, contradicting the documented semantics.
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. At chidian/drop.py, line 98:
<comment>DROP.PARENT used directly in a list removes only the list instead of its parent container, contradicting the documented semantics.</comment>
<file context>
@@ -0,0 +1,137 @@
+ pass
+ elif signal.levels == 1:
+ # Remove this dict from its parent
+ raise _DropSignal(0)
+ else:
+ # Propagate further up
</file context>
✅ Addressed in 694acea
- Update version in pyproject.toml - Update README.md documentation - Remove ambiguous examples - Modify core.py and drop.py modules - KEEP now processes DROP sentiinels - Fixed DROP-level propagation
There was a problem hiding this comment.
1 issue found across 6 files (reviewed changes from recent commits).
Prompt for AI agents (all 1 issues)
Check if these issues are valid — if so, understand the root cause of each and fix them.
<file name="chidian/core.py">
<violation number="1" location="chidian/core.py:33">
`grab` still raises `ValueError` in strict mode (invalid path syntax or traversing None), but the updated Raises section now claims only KeyError/IndexError/TypeError are possible, which misleads callers about the function’s behavior.</violation>
</file>
Reply to cubic to teach it or ask questions. Re-run a review with @cubic-dev-ai review this PR
| Raises: | ||
| KeyError: In strict mode, if a dict key is not found | ||
| IndexError: In strict mode, if a list index is out of range | ||
| TypeError: In strict mode, if a type mismatch occurs during traversal |
There was a problem hiding this comment.
grab still raises ValueError in strict mode (invalid path syntax or traversing None), but the updated Raises section now claims only KeyError/IndexError/TypeError are possible, which misleads callers about the function’s behavior.
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. At chidian/core.py, line 33:
<comment>`grab` still raises `ValueError` in strict mode (invalid path syntax or traversing None), but the updated Raises section now claims only KeyError/IndexError/TypeError are possible, which misleads callers about the function’s behavior.</comment>
<file context>
@@ -28,12 +28,14 @@ def grab(
- ValueError: In strict mode (via mapping_context), if path not found
+ KeyError: In strict mode, if a dict key is not found
+ IndexError: In strict mode, if a list index is out of range
+ TypeError: In strict mode, if a type mismatch occurs during traversal
Note:
</file context>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Refactors
chidianto be more focused on what it was created to do:dict-like transformations that make code more readable and usable.Incorporates some old patterns and learnings from pydian. Removes tabular abstractions, adds-back validation structure.
Summary by cubic
Reset chidian to focus on simple dict-to-dict transformations. Adds a lightweight API (grab, mapper, DROP/KEEP, validation) and removes table/DSL/legacy mapper to ship 0.2.0.
New Features
Migration
@mapperand the new validation module.Written for commit 694acea. Summary will update automatically on new commits.