commit eff25c5d
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Wed Feb 12 23:09:01 2025 +0100
commit 2508facf
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Wed Feb 12 23:06:38 2025 +0100
commit 242ec2ea
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Wed Feb 12 22:44:31 2025 +0100
commit a402e7fb
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Wed Feb 12 22:23:12 2025 +0100

    Doc updates

    README.md
        - Updated docs on:
          - The `np_format_items_with_indenters` parameter
          - The `Numpy{ComplexNumber,Void{}Indenter` types
          - Note for 'technical details' -> 'indenter types'
        - Fixed typos
        - Added named anchors for the notes for more persistent links
          (instead of '#notes', '#note-1', etc. which are prone to
          breaking when notes and added and moved)

    ~[.api].linewrap_collection()
        Added parameter `np_format_items_with_indenters` and docs
        therefor

commit 850898fe
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Wed Feb 12 21:49:04 2025 +0100
commit ddb2fb2a
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Wed Feb 12 18:13:16 2025 +0100
commit 132ef034
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Wed Feb 12 18:07:28 2025 +0100
commit 3b96ff7d
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Wed Feb 12 17:24:08 2025 +0100

    Standalone `np.void` representation

    pyproject.toml
        Registered `NumpyVoidIndenter` under
        `project.entry-points.polyethylene.repr_indenter_getters`
        (entry point used by `~[.plugins].get_repr_indenter()`,
        `~[.api].linewrap_repr()`, etc.)

    ~._numbers.get_indenter()
        Added more metadata

    ~._numpy.NumpyVoidIndenter
        New indenter for standalone `np.void` objects

    ~._numpy.DataTypeIndenter.create_child_indenters()
        Fixed typo in error message

    ~.tests.test_numpy.test_{numpy_void_formatting,recursive_void}()
        New tests

commit 93d26eb2
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Wed Feb 12 05:58:42 2025 +0100
commit 7f382404
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Wed Feb 12 01:08:29 2025 +0100

    `numpy`-numeric-type formatting

    pyproject.toml
        Registered `~._numpy.NumpyComplexNumberIndenter` to
        `~._numbers.get_indenter()` via entry point

    ~._numbers.ComplexNumberIndenter
        Refactored internal method for generating representations

    ~._numpy.NumpyComplexNumberIndenter
        New subtype of the above specifically for generating
        representations for standalone `np.number` and `np.bool_`
        instances

    ~.constants.NORMALIZABLE_NAMES
        Registered normalization of `np.bool_` ('numpy.bool' ->
        'numpy.bool_')

    ~.tests.test_numpy
        test_complex_structured_dtype_evaluability()
            Refactored to use
            `~.tests.utils.build_param_resolution_test()`
        test_type_preservation_in_structured_types()
            Added `.can_eval` checks since representations of standalone
            `numpy` numerals are now properly handled
        test_numpy_{bool,int,[long_]{float,complex}}_formatting()
            New tests for the representation or standalone `numpy`
            numerals

    ~.tests.utils.build_param_resolution_test()
        Added to `~~.__all__`

commit ff08e6b3
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Tue Feb 11 20:35:18 2025 +0100

    Handling of unstructured `np.void` types

    ~._numpy.NDimensionalArrayIndenter
        Added handling of unstructured `np.void` types to the internal
        indenter type handling void objects

    ~.tests.test_numpy.test_unstructured_dtype_evaluability()
        New test for the above

commit 56a67140
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Tue Feb 11 19:14:45 2025 +0100
commit 243c1dba
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Tue Feb 11 17:26:13 2025 +0100

    Various fixes in `~._numpy`

    ~~.NDimensionalArrayIndenter
        can_eval
            Fixed bug where shaped field in structured types impairs
            evaluation of evaluability
            (fixes broken subtest in
            `~.tests.test_numpy
            .test_complex_structured_dtype_evaluability()`)
        <Others>
            Fixed handling of `np.datetime64` fields in structured data
            types (so that they are formatted as in the array `repr()`)

    ~.tests.test_numpy.test_datetime_in_structured_data_types()
        New test for the `datetime64` fix

commit 29d741aa
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Tue Feb 11 17:13:28 2025 +0100
commit def67ce8
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Tue Feb 11 16:33:59 2025 +0100

    Complex data type handling in `~._numpy`

    ~~.DataTypeIndenter
        Fixed bugs when formatting structured void types where:
        - Shaped fields are not accounted for
        - Nested structures result in error

    ~~.NDimensionalArrayIndenter
        Fixed internal type used thereby for formatting structured data
        types:
        - Added doctest thereto
        - Now directly formatting `np.void` items (with `numpy`-native
          fields) instead of casting them into Python objects with
          `.item()` first
        - Fixed handling of shaped fields

    ~.tests.test_numpy.test_complex_structured_dtype_formatting()
    ~.tests.test_numpy.test_complex_structured_dtype_evaluability()
    ~.tests.test_numpy.test_type_preservation_in_structured_types()
        New tests for the above

commit d33345e7
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Tue Feb 11 15:48:51 2025 +0100
commit 3bcd8aac
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Tue Feb 11 00:01:56 2025 +0100

    Various fixes

    ~._numpy
        NDimensionalArrayIndenter.np_format_items_with_indenters
            Fixed bug where the aliases 'int' and 'int_kind' include
            `np.timedelta64`

        <Others>
            Fixed type hints of `.get_indenter_class()` of some internal
            indenter classes

    ~._type_hierarchy.TypeHierarchy.types
        Now pruning dead references to GC-ed types, which resulted in
        the type hierarchy being polluted by `None`

    ~._resolved_value.resolve_from_mapping(key=None)
        Changed order of attribute access on the internal
        `TypeHierarchy` so that the above doesn't result in
        inconsistencies

commit 19f379c4
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Mon Feb 10 20:04:17 2025 +0100

    More consistent alias resolution

    ~._resolved_value.ResolvedDescriptor.default_fset()
        Improved the resolution of aliases for mappings:
        - Between an alias and the alias target, prioritize the alias
          target
        - Between a broader alias and a narrower alias, prioritize the
          narrower alias

    ~.tests.test_misc.test_type_checking_type_classification()
        Test for the above

commit 2dd4746d
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Sun Feb 9 04:22:52 2025 +0100

    Added tools for debugging

    ~._debug_tools
        New internal submodule for debugging tools

    ~~.DebugAttributeAccessMeta
        New metaclass for tracking attribute access

commit e3b41bf9
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Sun Feb 9 07:00:11 2025 +0100
commit 0c5704ad
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Sun Feb 9 04:48:36 2025 +0100
commit 2c60e31f
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Fri Feb 7 12:47:30 2025 +0100
commit eb609f15
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Wed Feb 5 16:24:16 2025 +0100
commit 35b52a58
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Wed Feb 5 14:28:21 2025 +0100
commit 4c43b1f7
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Tue Feb 4 19:24:00 2025 +0100

    Removed type qualifiers for `numpy` array items

    ~~.NDimensionalArrayIndenter.create_child_indenters()
        - Fixed bug where the items of item-container arrays are not
          properly formatted with indenters
        - If `.np_format_items_with_indenters` resolves to true:
          - Indenters for `numpy`-native (byte) strings, booleans, and
            numeric types no longer include the '<type_name>(' and ')'
            affixes;
            i.e. the `numpy`-native base data types now behave like
            their `python`-native equivalents (`str`, `bytes`, `bool`,
            etc.)
          - If the array is of an integral type (incl. booleans and time
            deltas) and a `number_format` is specified, the `dtype` is
            always shown since float-like formatting options can obscure
            the data type

    ~._numbers.ComplexNumberIndenter
        _BaseTypes, _get_type_cast_toolbox()
            New (internal) hooks so that the above can be implemented

    ~.tests.test_numpy
        test_indenter_based_{numeric,string}_item_formatting()
            New tests for handling of item formatting when
            `.np_format_items_with_indenters` is true
        test_item_container_array()
            Fixed erroneous test: `object()` is not evaluable
        test_np_format_items_with_indenters_resolution()
            New test for resolution of `np_format_items_with_indenters`

commit 08474682
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Tue Feb 4 17:26:01 2025 +0100
commit 6e02c873
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Tue Feb 4 16:55:37 2025 +0100

    Various fixes in `~._numpy`

    - Fixed bug where attempting to retrieve the signature of
      `np.set_printoptions()` when `numpy` is not available results in
      an import-time error
    - Suppressed `DeprecationWarning` emitted (conversion of abstract
      `numpy` types like `np.integer` to `np.dtype`s deprecated) when
      setting up aliases for
      `~~.NDimensionalArrayIndenter.np_format_items_with_indenters`

commit 2965e354
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Tue Feb 4 16:26:53 2025 +0100

    WIP: `~._numpy`: customizable item formatting

    ~~.NDimensionalArrayIndenter
        __doc__
            Updated
        np_format_items_with_indenters
            New parameter/resolved descriptor

    ~.constants.DEFAULTS['np_format_items_with_indenters']
        New constant

commit b202e0ae
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Tue Feb 4 14:04:29 2025 +0100
commit 2f7a2ae3
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Tue Feb 4 13:19:47 2025 +0100

    New tests in `~.tests.test_numpy` for edge cases

    ~~.NDimensionalArrayIndenter
        .{get_indenter_class,create_child_indenters}()
            Fixed bug where item-container arrays of structured
            `np.void` is not referred to the appropriate `Indenter`
            subclass and does not have proper recursion handling

    ~~.DataTypeIndenter.create_child_indenters()
        Fixed formatting of data types corresponding to structured
        `np.void` types

    ~~.test_structured_dtype_formatting()
    ~~.test_recursive_array_with_structured_type()
        New tests for the above fixes

    ~~.test_recursive_array()
    ~~.test_indirect_recursive_array()
        Now checking against the array `repr()`s

commit a9ff230f
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Tue Feb 4 12:55:20 2025 +0100
commit 529be1ae
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Wed Jan 22 11:12:13 2025 +0100

    Insulation against `np` print options

    ~._{collections,numpy}.NDimensionalArrayIndenter
        Now insulated against `np.[set_]printoptions()` so that array
        items are always formatted in a `repr()`-friendly form

    ~.tests.test_numpy.test_formatter_insulation()
        Test for the above

    ~.utils.strip_affixes()
        Migrated from private helper function in `~.api`

commit af6371d2
Author: Terence S.-C. Tsang <t.tsang.sci@gmail.com>
Date:   Wed Jan 22 08:12:37 2025 +0100

    Minor fixes in `~._numpy`

    ~.__version__
        Bumped to 0.10.0

    ~~.DataTypeIndenter
        .is_structured
            New descriptor
        <Others>
            Fixed internal method used for determining the short
            representation of an `np.dtype` (currently not used due to
            the availability of a curresponding `np` internal method)

    ~.tests.test_numpy.test_dtype_{abbr,implcation}_emulation()
        New tests to ensure that the vanilla implmentations of
        `DataTypeIndenter` methods are consistent with their `np`
        equivalents, as long as they are available