wcwidth 0.8.2__tar.gz → 0.8.3__tar.gz

{wcwidth-0.8.2 → wcwidth-0.8.3}/PKG-INFO

@@ -1,6 +1,6 @@
-Metadata-Version: 2.4
+Metadata-Version: 2.5
 Name: wcwidth
-Version: 0.8.2
+Version: 0.8.3
 Summary: Measures the displayed width of unicode strings in a terminal
 Project-URL: Homepage, https://github.com/jquast/wcwidth
 Author-email: Jeff Quast <contact@jeffquast.com>
@@ -176,7 +176,7 @@ width()
 -------
 
 Use function `width()`_ to measure a string with improved handling of ``control_codes`` and
-measurement Corrections_ through ``term_program``:.
+measurement Corrections_ through ``term_program``:
 
 .. code-block:: python
 
@@ -511,8 +511,7 @@ recognized names:
 .. END_LIST_TERM_PROGRAMS
 
 ``term_program=False`` (the default for `width()`_, `ljust()`_, `rjust()`_, `center()`_,
-`wrap()`_, and `clip()`_) disables terminal corrections.  `wcstwidth()`_ defaults to
-``term_program=True`` for auto-detection.
+`wrap()`_, and `clip()`_) disables terminal corrections.
 
 For automatic tests and other purposes that require cross-environment consistency, set static values
 or unset ``TERM`` and ``TERM_PROGRAM`` environment values, such as in ``conftest.py`` with pytest:
@@ -588,16 +587,6 @@ To upgrade requirements for building documentation, run::
 
    tox -e update_requirements_docs
 
-Utilities
----------
-
-Supplementary tools for browsing and testing terminals for wide unicode
-characters are found in the `bin/`_ of this project's source code.  Just ensure
-to first ``pip install -r requirements-develop.txt`` from this projects main
-folder. For example, an interactive browser for testing::
-
-  python ./bin/wcwidth-browser.py
-
 ====
 Uses
 ====
@@ -685,9 +674,15 @@ languages:
 History
 =======
 
+0.8.3 *2026-08-28*
+  * **Bugfix** Do not hang on `wrap()`_ calls of width 1 with text containing OSC8 hyperlinks and
+    wide characters, `PR #231`_.
+  * **Bugfix** `clip()`_ with ``propagate_sgr=True``, should match behavior of `propagate_sgr()`_,
+    `PR #235`_.
+
 0.8.2 *2026-06-29*
-  * **Bugfix** Do not raise IndexError when given legacy POSIX ``n`` argument to `wcswidth()`_ or
-    `wcstwidth()`_ exceed string length without raising IndexError
+  * **Bugfix** Do not raise IndexError when legacy POSIX ``n`` argument to `wcswidth()`_ or
+    `wcstwidth()`_ exceeds the given string length. `PR #230`_
 
 0.8.1 *2026-06-08*
   * **Improved** `wcstwidth()`_ with new ``zeroer``, ``narrow_wider``, and ``narrow_zeroer``
@@ -942,6 +937,9 @@ https://www.cl.cam.ac.uk/~mgk25/ucs/wcwidth.c::
 .. _`PR #223`: https://github.com/jquast/wcwidth/pull/223
 .. _`PR #224`: https://github.com/jquast/wcwidth/pull/224
 .. _`PR #226`: https://github.com/jquast/wcwidth/pull/226
+.. _`PR #230`: https://github.com/jquast/wcwidth/pull/230
+.. _`PR #231`: https://github.com/jquast/wcwidth/pull/231
+.. _`PR #235`: https://github.com/jquast/wcwidth/pull/235
 .. _`Issue #101`: https://github.com/jquast/wcwidth/issues/101
 .. _`Issue #155`: https://github.com/jquast/wcwidth/issues/155
 .. _`Issue #190`: https://github.com/jquast/wcwidth/issues/190

{wcwidth-0.8.2 → wcwidth-0.8.3}/docs/intro.rst

@@ -145,7 +145,7 @@ width()
 -------
 
 Use function `width()`_ to measure a string with improved handling of ``control_codes`` and
-measurement Corrections_ through ``term_program``:.
+measurement Corrections_ through ``term_program``:
 
 .. code-block:: python
 
@@ -480,8 +480,7 @@ recognized names:
 .. END_LIST_TERM_PROGRAMS
 
 ``term_program=False`` (the default for `width()`_, `ljust()`_, `rjust()`_, `center()`_,
-`wrap()`_, and `clip()`_) disables terminal corrections.  `wcstwidth()`_ defaults to
-``term_program=True`` for auto-detection.
+`wrap()`_, and `clip()`_) disables terminal corrections.
 
 For automatic tests and other purposes that require cross-environment consistency, set static values
 or unset ``TERM`` and ``TERM_PROGRAM`` environment values, such as in ``conftest.py`` with pytest:
@@ -557,16 +556,6 @@ To upgrade requirements for building documentation, run::
 
    tox -e update_requirements_docs
 
-Utilities
----------
-
-Supplementary tools for browsing and testing terminals for wide unicode
-characters are found in the `bin/`_ of this project's source code.  Just ensure
-to first ``pip install -r requirements-develop.txt`` from this projects main
-folder. For example, an interactive browser for testing::
-
-  python ./bin/wcwidth-browser.py
-
 ====
 Uses
 ====
@@ -654,9 +643,15 @@ languages:
 History
 =======
 
+0.8.3 *2026-08-28*
+  * **Bugfix** Do not hang on `wrap()`_ calls of width 1 with text containing OSC8 hyperlinks and
+    wide characters, `PR #231`_.
+  * **Bugfix** `clip()`_ with ``propagate_sgr=True``, should match behavior of `propagate_sgr()`_,
+    `PR #235`_.
+
 0.8.2 *2026-06-29*
-  * **Bugfix** Do not raise IndexError when given legacy POSIX ``n`` argument to `wcswidth()`_ or
-    `wcstwidth()`_ exceed string length without raising IndexError
+  * **Bugfix** Do not raise IndexError when legacy POSIX ``n`` argument to `wcswidth()`_ or
+    `wcstwidth()`_ exceeds the given string length. `PR #230`_
 
 0.8.1 *2026-06-08*
   * **Improved** `wcstwidth()`_ with new ``zeroer``, ``narrow_wider``, and ``narrow_zeroer``
@@ -911,6 +906,9 @@ https://www.cl.cam.ac.uk/~mgk25/ucs/wcwidth.c::
 .. _`PR #223`: https://github.com/jquast/wcwidth/pull/223
 .. _`PR #224`: https://github.com/jquast/wcwidth/pull/224
 .. _`PR #226`: https://github.com/jquast/wcwidth/pull/226
+.. _`PR #230`: https://github.com/jquast/wcwidth/pull/230
+.. _`PR #231`: https://github.com/jquast/wcwidth/pull/231
+.. _`PR #235`: https://github.com/jquast/wcwidth/pull/235
 .. _`Issue #101`: https://github.com/jquast/wcwidth/issues/101
 .. _`Issue #155`: https://github.com/jquast/wcwidth/issues/155
 .. _`Issue #190`: https://github.com/jquast/wcwidth/issues/190

{wcwidth-0.8.2 → wcwidth-0.8.3}/docs/specs.rst

@@ -81,6 +81,14 @@ Width of 1
 String characters are measured width of 1 when they are not
 measured as `Width of 0`_ or `Width of 2`_.
 
+Any character with `U+FE0E`_ (Variation Selector 15) defined as ``text style``
+in `emoji-variation-sequences.txt`_, per `UTS #51`_ and `Unicode Standard
+Section 23.4`_: VS15 subtracts 1 cell from the wide character it directly
+follows, making the pair width 1. Narrow characters are unchanged.
+
+.. note:: Variation Selector 15 width specification is contested, see
+          https://github.com/jquast/wcwidth/issues/211
+
 Width of 2
 ----------
 
@@ -105,11 +113,6 @@ in `emoji-variation-sequences.txt`_, per `UTS #51`_ and `Unicode Standard
 Section 23.4`_: VS16 adds 1 cell to the narrow character it directly follows,
 making the pair width 2. Wide characters are unchanged.
 
-Any character with `U+FE0E`_ (Variation Selector 15) defined as ``text style``
-in `emoji-variation-sequences.txt`_, per `UTS #51`_ and `Unicode Standard
-Section 23.4`_: VS15 subtracts 1 cell from the wide character it directly
-follows, making the pair width 1. Narrow characters are unchanged.
-
 Any character of non-zero width followed by an ``Mc`` (`Spacing Combining Mark`_)
 character when measured in sequence by :func:`wcwidth.wcswidth` or
 :func:`wcwidth.width`. The ``Mc`` character caps the cluster width at 2,

{wcwidth-0.8.2 → wcwidth-0.8.3}/pyproject.toml

@@ -4,7 +4,7 @@ requires = [ "hatchling" ]
 
 [project]
 name = "wcwidth"
-version = "0.8.2"  # don't forget to also update wcwidth/__init__.py:__version__
+version = "0.8.3"  # don't forget to also update wcwidth/__init__.py:__version__
 description = "Measures the displayed width of unicode strings in a terminal"
 readme = "README.rst"
 keywords = [

{wcwidth-0.8.2 → wcwidth-0.8.3}/tests/test_clip.py

@@ -4,7 +4,7 @@
 import pytest
 
 # local
-from wcwidth import clip, width, strip_sequences
+from wcwidth import clip, width, propagate_sgr, strip_sequences
 
 STRIP_SEQUENCES_CASES = [
     ('', ''),
@@ -132,6 +132,86 @@ def test_clip_sequences_only():
     assert repr(clip('\x1b[31m\x1b[0m', 0, 10, propagate_sgr=False)) == repr('\x1b[31m\x1b[0m')
 
 
+SGR_PATH_KWARGS = [{}, {'overtyping': True}, {'control_codes': 'ignore'}]
+
+
+@pytest.mark.parametrize('kwargs', SGR_PATH_KWARGS)
+def test_clip_sgr_inside_window_preserved(kwargs):
+    """Style changes within (start, end) are emitted at their original position."""
+    # nothing is clipped away, so nothing changes,
+    assert (clip('\x1b[1mbold\x1b[m normal', 0, 20, **kwargs) ==
+            '\x1b[1mbold\x1b[m normal')
+    # and the same holds when the tail is clipped, the trailing reset is
+    # synthesized because the final style is still active.
+    assert (clip('\x1b[31mred\x1b[32mgreen\x1b[0m', 0, 4, **kwargs) ==
+            '\x1b[31mred\x1b[32mg\x1b[0m')
+
+
+@pytest.mark.parametrize('kwargs', SGR_PATH_KWARGS)
+def test_clip_sgr_before_start_folded_into_prefix(kwargs):
+    """Style set before *start* is synthesized as a prefix, not emitted twice."""
+    assert clip('\x1b[1;34mHello world\x1b[0m', 6, 11, **kwargs) == '\x1b[1;34mworld\x1b[0m'
+    # the style in effect at *start* is the most recent one,
+    assert clip('\x1b[31mred\x1b[32mgreen\x1b[0m', 4, 8, **kwargs) == '\x1b[32mreen\x1b[0m'
+
+
+@pytest.mark.parametrize('kwargs', SGR_PATH_KWARGS)
+def test_clip_sgr_after_end_discarded(kwargs):
+    """Style set at or beyond *end* does not appear in the result."""
+    assert clip('hello\x1b[31m world\x1b[0m', 0, 5, **kwargs) == 'hello'
+
+
+@pytest.mark.parametrize('kwargs', SGR_PATH_KWARGS)
+@pytest.mark.parametrize('text', [
+    'plain text',
+    '\x1b[31mred',
+    '\x1b[31mred\x1b[0m',
+    '\x1b[1mbold\x1b[m normal',
+    '\x1b[31mred\x1b[32mgreen\x1b[0m',
+    'plain \x1b[4munderline',
+])
+def test_clip_matches_propagate_sgr(text, kwargs):
+    """Clipping nothing away agrees with :func:`propagate_sgr`."""
+    assert clip(text, 0, width(text), **kwargs) == propagate_sgr([text])[0]
+
+
+@pytest.mark.parametrize('kwargs', SGR_PATH_KWARGS)
+@pytest.mark.parametrize('text', [
+    '\x1b[1m\x1b[31mbold red\x1b[0m',
+    '\x1b[31mred\x1b[32mgreen\x1b[0m',
+    '\x1b[?25l\x1b[31mred',
+])
+def test_clip_sgr_result_is_self_contained(text, kwargs):
+    """The result carries its own style: propagating it again is a no-op."""
+    result = clip(text, 0, width(text), **kwargs)
+    assert propagate_sgr([result])[0] == result
+    assert strip_sequences(result) == strip_sequences(text)
+
+
+@pytest.mark.parametrize('kwargs', SGR_PATH_KWARGS)
+def test_clip_sgr_after_non_sgr_sequence(kwargs):
+    """A non-SGR sequence before the first visible cell does not swallow the style."""
+    assert clip('\x1b[?25l\x1b[31mred', 0, 10, **kwargs) == '\x1b[31m\x1b[?25lred\x1b[0m'
+
+
+@pytest.mark.parametrize('kwargs', [{}, {'overtyping': True}])
+def test_clip_sgr_inside_hyperlink_tracked(kwargs):
+    """Style changed by hyperlink inner text still terminates with a reset."""
+    text = '\x1b]8;;http://x\x07a\x1b[31mb\x1b]8;;\x07c'
+    assert clip(text, 0, 10, **kwargs) == text + '\x1b[0m'
+
+
+@pytest.mark.parametrize('kwargs', [{}, {'overtyping': True}])
+def test_clip_sgr_inside_hyperlink_untracked(kwargs):
+    """With ``propagate_sgr=False``, hyperlink inner style is neither tracked nor reset."""
+    text = '\x1b]8;;http://x\x07a\x1b[31mb\x1b]8;;\x07c'
+    # sequences pass through verbatim, no synthesized prefix or trailing reset,
+    assert clip(text, 0, 10, propagate_sgr=False, **kwargs) == text
+    # and that holds when 'c', the only cell outside the hyperlink, is clipped away.
+    assert (clip(text, 0, 2, propagate_sgr=False, **kwargs) ==
+            '\x1b]8;;http://x\x07a\x1b[31mb\x1b]8;;\x07')
+
+
 def test_clip_sequences_osc_hyperlink():
     assert repr(clip('\x1b]8;;https://example.com\x07link\x1b]8;;\x07', 0, 4)) == repr(
         '\x1b]8;;https://example.com\x07link\x1b]8;;\x07'

{wcwidth-0.8.2 → wcwidth-0.8.3}/tests/test_textwrap.py

@@ -387,6 +387,16 @@ OSC_END_BEL = '\x1b]8;;\x07'
             '\x1b]8;foo=bar:id=mylink;http://example.com\x1b\\Click\x1b]8;;\x1b\\',
             '\x1b]8;foo=bar:id=mylink;http://example.com\x1b\\here\x1b]8;;\x1b\\',
         ],
+    ),
+    (   # wide grapheme after OSC 8 open at width 1 (must not hang)
+        '\x1b]8;;u\x07😀',
+        1,
+        ['\x1b]8;id=00000001;u\x07😀\x1b]8;;\x07'],
+    ),
+    (   # CJK wide char after OSC 8 open at width 1 (must not hang)
+        '\x1b]8;;u\x07あ',
+        1,
+        ['\x1b]8;id=00000001;u\x07あ\x1b]8;;\x07'],
     ),])
 def test_wrap_hyperlink_word_boundary(text, w, expected):
     """OSC hyperlink sequences should act as word boundaries."""
@@ -491,3 +501,13 @@ def test_wrap_bare_esc():
     """Bare ESC not part of a recognized sequence is treated as zero-width."""
     assert wrap('ab\x1bcd ef', 5) == ['ab\x1bcd', 'ef']
     assert wrap('ab\x1b\x00cdef', 3) == ['ab\x1b\x00c', 'def']
+
+
+def test_wrap_bare_esc_at_line_start():
+    """
+    Exercises the bare-ESC safety break in _find_first_visible_break.
+
+    A bare \\x1b that does not match ZERO_WIDTH_PATTERN triggers a safety 'break' statement but is
+    not found in any practical terminal sequence or string (ESC followed by NUL).
+    """
+    assert wrap('\x1b\x00あ', 1) == ['\x1b', '\x00', 'あ']

{wcwidth-0.8.2 → wcwidth-0.8.3}/tox.ini

@@ -1,5 +1,5 @@
 [tox]
-envlist = update, fetch, compile, autopep8, docformatter, docformatter_check, isort, isort_check, pylint, pylint_tests, flake8, flake8_tests, pydocstyle, mypy, codespell, format, lint, docs, verify_tables, py{38, 39, 310, 311, 312, 313, 314}
+envlist = update, fetch, compile, autopep8, docformatter, docformatter_check, isort, isort_check, pylint, pylint_tests, flake8, flake8_tests, pydocstyle, mypy, codespell, twine_check, format, lint, docs, verify_tables, py{38, 39, 310, 311, 312, 313, 314}
 skip_missing_interpreters = true
 
 [base]
@@ -122,13 +122,13 @@ commands =
         wcwidth/ bin/ tests/
 
 [testenv:isort]
-deps = isort
+deps = isort==8.0.1
 basepython = python3.13
-commands = isort --quiet --apply --recursive wcwidth tests bin
+commands = isort --quiet wcwidth tests bin
 
 [testenv:pylint]
 basepython = python3.13
-deps = pylint
+deps = pylint==4.0.7
 commands = pylint --rcfile={toxinidir}/.pylintrc \
            --ignore=tests,docs,conf.py,build,distutils,.pyenv,.git,.tox \
            {posargs:{toxinidir}}/wcwidth
@@ -183,7 +183,7 @@ commands = docformatter --check --diff --recursive --pre-summary-newline \
 
 [testenv:isort_check]
 basepython = python3.14
-deps = isort
+deps = {[testenv:isort]deps}
 commands = isort --diff --check-only wcwidth tests bin
 
 [testenv:flake8_tests]
@@ -194,7 +194,7 @@ commands = flake8 --ignore=E501,W504,F401 tests/
 [testenv:pylint_tests]
 basepython = python3.13
 deps = pytest
-       pylint
+       {[testenv:pylint]deps}
 commands = pylint --rcfile={toxinidir}/.pylintrc \
            --disable=protected-access,unused-argument,missing-function-docstring,line-too-long \
            tests
@@ -217,6 +217,18 @@ commands = codespell --skip="*.pyc,htmlcov,_build,build,*.egg-info,.tox,data,./t
                      --uri-ignore-words-list '*' \
                      --summary --count
 
+[testenv:twine_check]
+# Verify the built distributions the way upload.pypi.org does: 'twine check'
+# runs the README through readme_renderer, which halts on any reStructuredText
+# error (such as a reference with no matching target).  Neither doc8 (D000 is
+# disabled) nor the python linters inspect README.rst.
+basepython = python3.14
+skip_install = true
+deps = build
+       twine
+commands = python -m build --outdir {envtmpdir}/dist
+           python -c "import glob, subprocess, sys; subprocess.check_call(['twine', 'check', '--strict'] + sorted(glob.glob(sys.argv[1] + '/dist/*')))" {envtmpdir}
+
 [testenv:format]
 basepython = python3.13
 deps = {[testenv:isort]deps}
@@ -234,6 +246,7 @@ deps = {[testenv:flake8]deps}
        {[testenv:pydocstyle]deps}
        {[testenv:pylint_tests]deps}
        {[testenv:codespell]deps}
+       {[testenv:twine_check]deps}
 commands = {[testenv:compile]commands}
 	   {[testenv:flake8]commands}
            {[testenv:mypy]commands}