nameparser 2.8.1__tar.gz → 2.8.2__tar.gz

{nameparser-2.8.1/nameparser.egg-info → nameparser-2.8.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: nameparser
-Version: 2.8.1
+Version: 2.8.2
 Summary: A simple Python module for parsing human names into their individual components.
 Author-email: Derek Gulbranson <derek73@gmail.com>
 License-Expression: LGPL-2.1-or-later
@@ -40,7 +40,7 @@ composable, and locale packs are opt-in.
 ⚠️ **2.8.1 adds three more DeprecationWarnings** for things 3.0 changes:
 configuring a ``Lexicon`` through its set fields
 (``Lexicon.default().add(titles={"dean"})``), reading the vocabulary
-constants in ``nameparser.config`` (``TITLES`` and the rest, all but
+constants in ``nameparser.config`` (``TITLES`` and the rest; 2.8.2 adds
 ``surnames``), and ``nameparser.config.regexes``. A test run that turns
 warnings into errors (``-W error``, or pytest's ``filterwarnings =
 error``) fails where your code does any of these; see the `migration

{nameparser-2.8.1 → nameparser-2.8.2}/README.rst

@@ -12,7 +12,7 @@ composable, and locale packs are opt-in.
 ⚠️ **2.8.1 adds three more DeprecationWarnings** for things 3.0 changes:
 configuring a ``Lexicon`` through its set fields
 (``Lexicon.default().add(titles={"dean"})``), reading the vocabulary
-constants in ``nameparser.config`` (``TITLES`` and the rest, all but
+constants in ``nameparser.config`` (``TITLES`` and the rest; 2.8.2 adds
 ``surnames``), and ``nameparser.config.regexes``. A test run that turns
 warnings into errors (``-W error``, or pytest's ``filterwarnings =
 error``) fails where your code does any of these; see the `migration

{nameparser-2.8.1 → nameparser-2.8.2}/nameparser/_config_shim.py

@@ -1170,7 +1170,7 @@ class Constants:
         from nameparser.config._compiled import MAIDEN_MARKERS
         from nameparser.config._compiled import (
             AGNOMENS, GLUED_HONORIFICS, SUFFIX_WORDS_AMBIGUOUS)
-        from nameparser.config.surnames import KOREAN_SURNAMES
+        from nameparser.config._ko_surnames import KOREAN_SURNAMES
         from nameparser.config._titles_2_6_0 import TITLES_2_6_0
         from nameparser.config._compiled import (
             AFTER_GIVEN_TITLES, TITLE_MODIFIERS, TITLE_PHRASES, TITLE_STYLES,

{nameparser-2.8.1 → nameparser-2.8.2}/nameparser/_lexicon.py

@@ -799,7 +799,7 @@ class Lexicon:
     #: written wholly in a script :attr:`Policy.segment_scripts
     #: <nameparser.Policy.segment_scripts>` activates. The default
     #: carries the Korean census list
-    #: (:data:`~nameparser.config.surnames.KOREAN_SURNAMES`); Chinese
+    #: (``nameparser/config/_ko_surnames.py``); Chinese
     #: surnames ship in locales.ZH because Han segmentation is opt-in.
     surnames: frozenset[str] = frozenset()
     #: Honorifics, and credential, office and generational words, that
@@ -1209,7 +1209,7 @@ class Lexicon:
 def _default_lexicon() -> Lexicon:
     # The compiled vocabulary, read from _compiled rather than the public
     # nameparser.config modules, whose constants warn at a caller's read
-    # since 2.8.1 (#1007); surnames.py is not deprecated.
+    # since 2.8.1 (#1007), and surnames.py's since 2.8.2.
     from nameparser.config._compiled import BOUND_GIVEN_NAMES
     from nameparser.config._compiled import GIVEN_NAME_ABBREVIATIONS
     from nameparser.config._compiled import CAPITALIZATION_EXCEPTIONS
@@ -1221,7 +1221,7 @@ def _default_lexicon() -> Lexicon:
         AGNOMENS, GLUED_HONORIFICS, SUFFIX_ACRONYMS, SUFFIX_ACRONYMS_AMBIGUOUS,
         SUFFIX_WORDS, SUFFIX_WORDS_AMBIGUOUS,
     )
-    from nameparser.config.surnames import KOREAN_SURNAMES
+    from nameparser.config._ko_surnames import KOREAN_SURNAMES
     from nameparser.config._compiled import (
         AFTER_GIVEN_TITLES, GIVEN_NAME_TITLES, TITLE_MODIFIERS, TITLE_PHRASES,
         TITLE_STYLES, TITLES, TITLES_AMBIGUOUS)

{nameparser-2.8.1 → nameparser-2.8.2}/nameparser/_version.py

@@ -7,7 +7,7 @@
 #: cannot carry the dev marker, so `VERSION >= (2, 8, 1)` is already
 #: true here while 2.8.1 is unreleased. Compare `__version__` instead
 #: where that distinction matters.
-VERSION = (2, 8, 1)
+VERSION = (2, 8, 2)
 #: PEP 440 pre-release/dev segment appended to the numeric version, or
 #: "" for a final release. Joined WITHOUT a dot ("2.0.0rc1", not
 #: "2.0.0.rc1"); setuptools reads __version__ as the package version.

{nameparser-2.8.1 → nameparser-2.8.2}/nameparser/config/__init__.py

@@ -10,10 +10,11 @@ rest of the facade in 3.0. The vocabulary data modules beside them
 ...) hold the word lists the 2.x :class:`~nameparser.Lexicon` is built
 from, named for its fields since 2.2 (#293); since 2.6 each (all but
 ``surnames.py``) re-exports sets compiled from the per-word TOML files
-in ``nameparser/vocab/`` (#737). 3.0 removes them too, all but
-``surnames``, and since 2.8.1 reading one of their constants warns,
+in ``nameparser/vocab/`` (#737). 3.0 removes them too, and makes
+``surnames`` private; since 2.8.1 reading one of their constants warns,
 naming what 3.0 offers in its place (#1007), as reading
-``nameparser.config.regexes`` does (#1008).
+``nameparser.config.regexes`` does (#1008), and since 2.8.2 reading
+``surnames.KOREAN_SURNAMES`` does.
 
 ``RegexTupleManager`` is served from the shim purely for pickle
 compatibility (through ``__getattr__`` since 2.8, so loading such a blob
@@ -25,10 +26,10 @@ blob raises ``AttributeError`` looking up the class, not a clean
 compatibility failure.
 """
 # Maintainer note, deliberately outside the docstring: 3.0 removes the
-# five shim names below and the vocabulary DATA modules but surnames
-# (#1007, #1008, superseding decisions/3-0-reevaluations's 2026-10-28
-# entry, which kept the modules); the modules warn per constant since
-# 2.8.1. Constants and CONSTANTS are imported eagerly -- Constants
+# five shim names below and the vocabulary DATA modules, surnames made
+# private (#1007, #1008, superseding decisions/3-0-reevaluations's
+# 2026-10-28 entry, which kept the modules); the modules warn per
+# constant since 2.8.1, surnames since 2.8.2. Constants and CONSTANTS are imported eagerly -- Constants
 # warns at construction, the CONSTANTS singleton when mutated -- while
 # the three managers resolve through __getattr__ so that each caller's
 # read warns, at the caller's line, and nameparser's own reads (from

{nameparser-2.8.1 → nameparser-2.8.2}/nameparser/config/_invariants.py

@@ -2,7 +2,7 @@
 
 Since the word lists moved to nameparser/vocab/*.toml, whose compiler
 (tools/vocab/build.py) refuses an unnormalized spelling itself, the one
-constant this module still checks is ``surnames.KOREAN_SURNAMES``, a
+constant this module still checks is ``_ko_surnames.KOREAN_SURNAMES``, a
 segmentation list outside the per-word model.
 
 A constant is looked up in normalized form, so a stray capital or a

nameparser-2.8.2/nameparser/config/surnames.py

@@ -0,0 +1,40 @@
+"""The Korean census surnames the default lexicon segments unspaced
+hangul on (#271), deprecated in 2.8.2: 3.0 makes the module private,
+and ``Lexicon.default().surnames`` holds the same words in 2.x and 3.0
+alike."""
+# The list itself lives in _ko_surnames.py, which the library reads, so
+# only a caller's read of this module's constant warns (#1007, as
+# titles.py over _compiled).
+from typing import TYPE_CHECKING
+
+from nameparser.config import _ko_surnames
+from nameparser.config._deprecated import alias_getattr
+
+KOREAN_SURNAMES = _ko_surnames.KOREAN_SURNAMES
+"""
+Korean surnames (#271), used by the 2.0 API's unspaced-name
+segmentation: a hangul token like "김민준" splits into surname + given
+name by longest match. Ships as DEFAULT vocabulary because a hangul
+entry can only ever match hangul text, and hangul text is unambiguously
+Korean. Chinese surnames live in ``nameparser.locales.zh`` instead (Han
+segmentation is opt-in; a zh list corrupts Japanese kanji names).
+
+Deprecated since 2.8.2, and private in 3.0: read
+``Lexicon.default().surnames``, which holds the same words, and extend
+it with ``Lexicon.default() | Lexicon(surnames=frozenset({...}))``.
+"""
+
+# outside the type checker's view, for the reason alias_getattr's
+# docstring gives: a module __getattr__ mypy can see answers every
+# missing attribute of the module
+if not TYPE_CHECKING:
+    # 2.8.2: the constant is served by __getattr__ and warns at the line
+    # that reads it, naming the Lexicon field 3.0 keeps public
+    __getattr__, __dir__ = alias_getattr(__name__, {}, removed={
+        "KOREAN_SURNAMES": (
+            "Lexicon.default().surnames holds the same words, now and in 3.0"),
+    }, namespace=globals())
+
+# star imports read __all__ and never the module __getattr__ -- see the
+# note in prefixes.py
+__all__ = ["KOREAN_SURNAMES"]

{nameparser-2.8.1 → nameparser-2.8.2/nameparser.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: nameparser
-Version: 2.8.1
+Version: 2.8.2
 Summary: A simple Python module for parsing human names into their individual components.
 Author-email: Derek Gulbranson <derek73@gmail.com>
 License-Expression: LGPL-2.1-or-later
@@ -40,7 +40,7 @@ composable, and locale packs are opt-in.
 ⚠️ **2.8.1 adds three more DeprecationWarnings** for things 3.0 changes:
 configuring a ``Lexicon`` through its set fields
 (``Lexicon.default().add(titles={"dean"})``), reading the vocabulary
-constants in ``nameparser.config`` (``TITLES`` and the rest, all but
+constants in ``nameparser.config`` (``TITLES`` and the rest; 2.8.2 adds
 ``surnames``), and ``nameparser.config.regexes``. A test run that turns
 warnings into errors (``-W error``, or pytest's ``filterwarnings =
 error``) fails where your code does any of these; see the `migration

{nameparser-2.8.1 → nameparser-2.8.2}/nameparser.egg-info/SOURCES.txt

@@ -40,6 +40,7 @@ nameparser/config/__init__.py
 nameparser/config/_compiled.py
 nameparser/config/_deprecated.py
 nameparser/config/_invariants.py
+nameparser/config/_ko_surnames.py
 nameparser/config/_titles_2_6_0.py
 nameparser/config/bound_first_names.py
 nameparser/config/bound_given_names.py

{nameparser-2.8.1 → nameparser-2.8.2}/tests/v2/test_contracts.py

@@ -85,7 +85,8 @@ def test_every_guarded_config_module_is_imported() -> None:
 
     Since the vocabulary moved to nameparser/vocab/*.toml, whose
     compiler checks it (tests/v2/test_vocab.py), the asserts left are
-    on data still written in Python -- ``surnames`` today, which is
+    on data still written in Python -- ``_ko_surnames`` today (the
+    Korean census list, ``surnames.py``'s until 2.8.2), which is
     lazy (imported inside ``_snapshot``/``_default_lexicon``), so its
     guard fires only incidentally, whenever some other test happens to
     build a default Lexicon. Importing them all here makes every guard
@@ -101,11 +102,12 @@ def test_every_guarded_config_module_is_imported() -> None:
     import nameparser.config
 
     config_dir = pathlib.Path(nameparser.config.__file__).parent
-    # private modules excluded: _invariants DEFINES the guard (the
-    # substring hits its own def) and holds no vocabulary of its own
+    # _invariants excluded: it DEFINES the guard (the substring hits its
+    # own def) and holds no vocabulary of its own. Private modules are
+    # read, since the data moved to one in 2.8.2
     guarded = sorted(
         p.stem for p in config_dir.glob("*.py")
-        if not p.stem.startswith("_")
+        if p.stem != "_invariants"
         and "assert_normalized(" in p.read_text(encoding="utf-8"))
     assert guarded, (
         f"no guarded config module found in {config_dir} -- the "

{nameparser-2.8.1 → nameparser-2.8.2}/tests/v2/test_deprecations.py

@@ -434,10 +434,10 @@ def test_a_copy_leaves_the_once_per_line_registry_alone() -> None:
     "from nameparser import HumanName",
     "from nameparser.parser import HumanName",
     "from nameparser.config import CONSTANTS, Constants",
-    # the module imports quietly; its constants warn since 2.8.1 (#1007),
-    # asserted in test_removals_2_8_1.py, and surnames is not removed
+    # the modules import quietly; their constants warn since 2.8.1
+    # (#1007), surnames' since 2.8.2, asserted in test_removals_2_8_1.py
     "import nameparser.config.titles",
-    "from nameparser.config.surnames import KOREAN_SURNAMES",
+    "import nameparser.config.surnames",
     "from nameparser import parse; parse('John Smith')",
 ])
 def test_imports_and_the_core_stay_quiet(code: str) -> None:

{nameparser-2.8.1 → nameparser-2.8.2}/tests/v2/test_removals_2_8_1.py

@@ -7,7 +7,9 @@ line and naming the 3.0 replacement and the migration guide:
   ``Lexicon(...)``, ``add()``, ``remove()`` or ``dataclasses.replace``;
 * the constants of the eight ``nameparser.config`` vocabulary modules
   3.0 deletes (#1007);
-* ``nameparser.config.regexes`` (#1008).
+* ``nameparser.config.regexes`` (#1008);
+* and, since 2.8.2, ``nameparser.config.surnames``, which 3.0 makes
+  private (its words are ``Lexicon.default().surnames``).
 
 The tables are written out literally rather than read from the modules,
 for the reason tests/v2/test_config_aliases.py gives: imported, every
@@ -58,8 +60,8 @@ def _line_after() -> int:
 # -- the vocabulary and regex modules (#1007, #1008) --------------------------
 
 #: (module, constant, a phrase its replacement clause must hold). Every
-#: public name of the nine modules, the 1.x aliases (which
-#: test_config_aliases.py asserts) aside.
+#: public name of the ten modules, the 1.x aliases (which
+#: test_config_aliases.py asserts) aside; surnames since 2.8.2.
 REMOVED_CONSTANTS = [
     ("titles", "GIVEN_NAME_TITLES", 'words(kind="title", addresses="given")'),
     ("titles", "TITLES", 'words(kind="title")'),
@@ -89,6 +91,7 @@ REMOVED_CONSTANTS = [
     ("regexes", "re_bidi", "copy a pattern you reuse"),
     ("regexes", "EMPTY_REGEX", "copy a pattern you reuse"),
     ("regexes", "REGEXES", "copy a pattern you reuse"),
+    ("surnames", "KOREAN_SURNAMES", "Lexicon.default().surnames"),
 ]
 REMOVED_MODULES = sorted({m for m, _, _ in REMOVED_CONSTANTS})
 
@@ -124,7 +127,10 @@ def test_reading_a_removed_constant_warns_at_the_reading_line(
     [w] = _deprecations(record)
     _assert_removal(w, module_name, name, phrase, line)
     # served unchanged: the library's own copy is the same object
-    if module_name != "regexes":
+    if module_name == "surnames":
+        from nameparser.config import _ko_surnames
+        assert value is _ko_surnames.KOREAN_SURNAMES
+    elif module_name != "regexes":
         from nameparser.config import _compiled
         assert value is getattr(_compiled, name)
 
@@ -151,13 +157,28 @@ def test_the_table_covers_every_public_name(module_name: str) -> None:
     assert set(module.__all__) - aliases == listed
 
 
-def test_the_module_imports_quietly_and_surnames_stays() -> None:
+def test_the_modules_import_quietly() -> None:
     with warnings.catch_warnings():
         warnings.simplefilter("error")
         for module_name in REMOVED_MODULES:
             importlib.import_module(f"nameparser.config.{module_name}")
-        from nameparser.config.surnames import KOREAN_SURNAMES
-        assert KOREAN_SURNAMES
+
+
+def test_the_surnames_replacement_runs_quietly_and_holds_the_same_words(
+        ) -> None:
+    # 2.8.2: the message offers Lexicon.default().surnames, the field
+    # 3.0 keeps public, as the code that replaces the read; run the
+    # code it names and hold it to the constant it replaces
+    from nameparser.config import surnames
+    with _caught() as record:
+        held = surnames.KOREAN_SURNAMES
+    [w] = _deprecations(record)
+    code = str(w.message).split("; ", 1)[1].split(" holds", 1)[0]
+    assert code == "Lexicon.default().surnames"
+    with warnings.catch_warnings():
+        warnings.simplefilter("error")
+        offered = eval(code, {"Lexicon": Lexicon})  # noqa: S307
+    assert offered == held
 
 
 def test_an_alias_to_a_removed_constant_warns_once_naming_both() -> None: