Standalone emoji support

.github/workflows/ci.yml

@@ -108,14 +108,18 @@ jobs:
 
           python -Im pip install tox
 
-      - name: Build wheel
+      - name: Prepare sdist and source-dir
         shell: bash
         run: |
           python -Im pip install build
-          python -Im build --wheel
+          python -Im build
+
+          mkdir source-dir
+          tar -xzvf dist/wcwidth-*.tar.gz -C source-dir --strip-components=1
 
       - name: Fetch test data files
         if: matrix.python-version == '3.14' && matrix.os == 'ubuntu-latest'
+        continue-on-error: true
         shell: bash
         run: |
           python -Im tox -e fetch

docs/intro.rst

@@ -455,10 +455,10 @@ languages.
 History
 =======
 
-0.5.2 *unreleased*
-  * **Bugfix** Specification and result of category ``Mc`` (`Spacing Combining Mark`_), approx. 443
-    codepoints, has a more nuanced specification_, and may be categorized as both zero or wide.
-    `PR #200`.
+0.5.2 *2026-01-29*
+  * **Bugfix** Measurement of category ``Mc`` (`Spacing Combining Mark`_), approx.  443, has a more
+    nuanced specification_, and may be categorized as either zero or wide. `PR #200`_.
+  * **Bugfix** Measurement of "standalone" modifiers and regional indicators, `PR #202`_.
   * **Updated** Data files used in some automatic tests are no longer distributed. `PR #199`_
 
 0.5.1 *2026-01-27*
@@ -666,6 +666,7 @@ https://www.cl.cam.ac.uk/~mgk25/ucs/wcwidth.c::
 .. _`PR #196`: https://github.com/jquast/wcwidth/pull/196
 .. _`PR #199`: https://github.com/jquast/wcwidth/pull/199
 .. _`PR #200`: https://github.com/jquast/wcwidth/pull/200
+.. _`PR #202`: https://github.com/jquast/wcwidth/pull/202
 .. _`Issue #101`: https://github.com/jquast/wcwidth/issues/101
 .. _`Issue #190`: https://github.com/jquast/wcwidth/issues/190
 .. _`jquast/blessed`: https://github.com/jquast/blessed

docs/specs.rst

@@ -33,13 +33,22 @@ Any characters defined by `General Category`_ codes in `DerivedGeneralCategory.t
   `Prepended_Concatenation_Mark`_ characters, aprox. 147 characters.
 - 'Zl': `U+2028`_ LINE SEPARATOR only
 - 'Zp': `U+2029`_ PARAGRAPH SEPARATOR only
-- 'Sk': `Modifier Symbol`_, aprox. 4 characters of only those where phrase
-  ``'EMOJI MODIFIER'`` is present in comment of `UnicodeData.txt`_.
+- 'Sk': `Modifier Symbol`_, aprox. 1 character with ``'FULLWIDTH'`` in comment
+  of `UnicodeData.txt`_ (see `Width of 2`_). `Emoji Modifier`_ Fitzpatrick
+  symbols (`U+1F3FB`_ through `U+1F3FF`_) are zero-width only when following
+  an emoji base character in sequence; see `Width of 2`_ for standalone.
 
 The NULL character (`U+0000`_).
 
-Any character following ZWJ (`U+200D`_) when in sequence by
-function :func:`wcwidth.wcswidth`.
+Any character following ZWJ (`U+200D`_) when preceded by an emoji
+(`Extended_Pictographic`_ property) or `Regional Indicator`_ in sequence by
+function :func:`wcwidth.wcswidth`. When ZWJ follows a non-emoji character
+(including CJK), only the ZWJ itself is zero-width; the following character
+is measured normally.
+
+The second `Regional Indicator`_ symbol (`U+1F1E6`_ through `U+1F1FF`_) in a
+consecutive pair, when measured in sequence by :func:`wcwidth.wcswidth` or
+:func:`wcwidth.width`. The first indicator of the pair is `Width of 2`_.
 
 `Hangul Jamo`_ Jungseong and "Extended-B" code blocks, `U+1160`_ through
 `U+11FF`_ and `U+D7B0`_ through `U+D7FF`_.
@@ -62,6 +71,15 @@ Any character defined by `East Asian`_ Fullwidth (``F``) or Wide (``W``)
 properties in `EastAsianWidth.txt`_ files, except those that are defined by the
 Category code of `Nonspacing Mark`_ (``Mn``).
 
+`Regional Indicator`_ symbols (`U+1F1E6`_ through `U+1F1FF`_). Though
+classified as Neutral in `EastAsianWidth.txt`_, terminals universally render
+these as double-width. A consecutive pair of Regional Indicators forms a flag
+emoji and is measured as width 2 total (first indicator is 2, second is 0).
+
+`Emoji Modifier`_ Fitzpatrick symbols (`U+1F3FB`_ through `U+1F3FF`_) when
+measured standalone (not following an emoji base character). When following
+an emoji base, they combine with the base and add 0 to total width.
+
 Any characters of `Modifier Symbol`_ category, ``'Sk'`` where ``'FULLWIDTH'`` is
 present in comment of `UnicodeData.txt`_, aprox. 3 characters.
 
@@ -105,4 +123,11 @@ by a Nukta (``Mn``) and then a vowel sign (``Mc``) is measured as base + 1.
 .. _`U+D7FF`: https://codepoints.net/U+D7FF
 .. _`UnicodeData.txt`: https://www.unicode.org/Public/UCD/latest/ucd/UnicodeData.txt
 .. _`East Asian`: https://www.unicode.org/reports/tr11/
+.. _`U+1F1E6`: https://codepoints.net/U+1F1E6
+.. _`U+1F1FF`: https://codepoints.net/U+1F1FF
+.. _`U+1F3FB`: https://codepoints.net/U+1F3FB
+.. _`U+1F3FF`: https://codepoints.net/U+1F3FF
+.. _`Regional Indicator`: https://www.unicode.org/charts/PDF/U1F100.pdf
+.. _`Emoji Modifier`: https://unicode.org/reports/tr51/#Emoji_Modifiers
+.. _`Extended_Pictographic`: https://www.unicode.org/reports/tr51/#def_extended_pictographic
 .. _`Nonspacing Mark`: https://www.unicode.org/versions/latest/core-spec/chapter-4/#G134153

tests/test_benchmarks.py

@@ -61,6 +61,31 @@ def test_wcswidth_emoji_sequence(benchmark):
     benchmark(wcwidth.wcswidth, text)
 
 
+# Regional Indicator benchmarks - paired flags and unpaired RI
+RI_FLAGS_PAIRED = '🇺🇸🇬🇧🇫🇷🇩🇪🇯🇵' * 100
+RI_FLAGS_UNPAIRED = '🇺🇸🇬🇧🇫' * 100
+
+
+def test_wcswidth_ri_flags_paired(benchmark):
+    """Benchmark wcswidth() with paired regional indicator flags."""
+    benchmark(wcwidth.wcswidth, RI_FLAGS_PAIRED)
+
+
+def test_wcswidth_ri_flags_unpaired(benchmark):
+    """Benchmark wcswidth() with mixed paired and unpaired regional indicators."""
+    benchmark(wcwidth.wcswidth, RI_FLAGS_UNPAIRED)
+
+
+def test_width_ri_flags_paired(benchmark):
+    """Benchmark width() with paired regional indicator flags."""
+    benchmark(wcwidth.width, RI_FLAGS_PAIRED)
+
+
+def test_width_ri_flags_unpaired(benchmark):
+    """Benchmark width() with mixed paired and unpaired regional indicators."""
+    benchmark(wcwidth.width, RI_FLAGS_UNPAIRED)
+
+
 # NFC vs NFD comparison - text with combining marks
 DIACRITICS_COMPOSED = 'café résumé naïve ' * 100
 DIACRITICS_DECOMPOSED = unicodedata.normalize('NFD', DIACRITICS_COMPOSED)

tests/test_emojis.py

@@ -31,7 +31,7 @@ def emoji_zwj_sequence():
               "\u200d"       # Joiner, Category Cf, East Asian Width property 'N'  -- ZERO WIDTH JOINER
               "\U0001f4bb")  # Fused, Category So, East Asian Width property 'W' -- PERSONAL COMPUTER
     # This test adapted from https://www.unicode.org/L2/L2023/23107-terminal-suppt.pdf
-    expect_length_each = (2, 0, 0, 2)
+    expect_length_each = (2, 2, 0, 2)
     expect_length_phrase = 2
 
     # exercise,
@@ -49,7 +49,7 @@ def test_unfinished_zwj_sequence():
     phrase = ("\U0001f469"   # Base, Category So, East Asian Width property 'W' -- WOMAN
               "\U0001f3fb"   # Modifier, Category Sk, East Asian Width property 'W' -- EMOJI MODIFIER FITZPATRICK TYPE-1-2
               "\u200d")      # Joiner, Category Cf, East Asian Width property 'N'  -- ZERO WIDTH JOINER
-    expect_length_each = (2, 0, 0)
+    expect_length_each = (2, 2, 0)
     expect_length_phrase = 2
 
     # exercise,
@@ -67,7 +67,7 @@ def test_non_recommended_zwj_sequence():
     phrase = ("\U0001f469"   # Base, Category So, East Asian Width property 'W' -- WOMAN
               "\U0001f3fb"   # Modifier, Category Sk, East Asian Width property 'W' -- EMOJI MODIFIER FITZPATRICK TYPE-1-2
               "\u200d")      # Joiner, Category Cf, East Asian Width property 'N'  -- ZERO WIDTH JOINER
-    expect_length_each = (2, 0, 0)
+    expect_length_each = (2, 2, 0)
     expect_length_phrase = 2
 
     # exercise,
@@ -87,7 +87,7 @@ def test_another_emoji_zwj_sequence():
         "\u200D"        # ZERO WIDTH JOINER
         "\u2640"        # FEMALE SIGN
         "\uFE0F")       # VARIATION SELECTOR-16
-    expect_length_each = (1, 0, 0, 1, 0)
+    expect_length_each = (1, 2, 0, 1, 0)
     expect_length_phrase = 2
 
     # exercise,
@@ -120,7 +120,7 @@ def test_longer_emoji_zwj_sequence():
               "\U0001F3FD"   # 'Sk', 'W' -- EMOJI MODIFIER FITZPATRICK TYPE-4
               ) * 2
     # This test adapted from https://www.unicode.org/L2/L2023/23107-terminal-suppt.pdf
-    expect_length_each = (2, 0, 0, 1, 0, 0, 2, 0, 2, 0) * 2
+    expect_length_each = (2, 2, 0, 1, 0, 0, 2, 0, 2, 2) * 2
     expect_length_phrase = 4
 
     # exercise,
@@ -191,6 +191,56 @@ def measure_all():
     assert len(sequences) >= 742
 
 
+@pytest.mark.skipif(NARROW_ONLY, reason="Test cannot verify on python 'narrow' builds")
+def test_regional_indicator_single():
+    """Single Regional Indicator symbol is width 2."""
+    assert wcwidth.wcwidth('\U0001F1FA') == 2
+    assert wcwidth.wcswidth('\U0001F1FA') == 2
+
+
+@pytest.mark.skipif(NARROW_ONLY, reason="Test cannot verify on python 'narrow' builds")
+def test_regional_indicator_pair():
+    """Flag pair (two Regional Indicators) is width 2, not 4."""
+    assert wcwidth.wcswidth('\U0001F1FA\U0001F1F8') == 2
+
+
+@pytest.mark.skipif(NARROW_ONLY, reason="Test cannot verify on python 'narrow' builds")
+def test_regional_indicator_three():
+    """Three Regional Indicators: one pair (2) + one single (2) = 4."""
+    assert wcwidth.wcswidth('\U0001F1FA\U0001F1F8\U0001F1E6') == 4
+
+
+@pytest.mark.skipif(NARROW_ONLY, reason="Test cannot verify on python 'narrow' builds")
+def test_regional_indicator_four():
+    """Four Regional Indicators: two pairs = 2 + 2 = 4."""
+    assert wcwidth.wcswidth(
+        '\U0001F1FA\U0001F1F8\U0001F1E6\U0001F1FA') == 4
+
+
+@pytest.mark.skipif(NARROW_ONLY, reason="Test cannot verify on python 'narrow' builds")
+def test_zwj_after_non_emoji():
+    """ZWJ after non-emoji unconditionally consumes next character."""
+    # This does *not* match most terminal behavior -- it is a negative test,
+    # they fail because our library doesn't handle 'glitch' emoji as an
+    # optimization. Non-emoji + ZWJ is undefined per Unicode UAX #29 GB11.
+    assert wcwidth.wcswidth('xx\u200d\U0001F384') == 2
+    assert wcwidth.wcswidth('a\u200d\U0001F600') == 1
+    assert wcwidth.wcswidth('\u4e16\u200d\U0001F600') == 2
+
+
+@pytest.mark.skipif(NARROW_ONLY, reason="Test cannot verify on python 'narrow' builds")
+def test_fitzpatrick_standalone():
+    """Standalone Fitzpatrick modifier is width 2."""
+    assert wcwidth.wcwidth('\U0001F3FB') == 2
+    assert wcwidth.wcswidth('\U0001F3FB') == 2
+
+
+@pytest.mark.skipif(NARROW_ONLY, reason="Test cannot verify on python 'narrow' builds")
+def test_fitzpatrick_after_emoji():
+    """Fitzpatrick modifier after emoji base combines, total width 2."""
+    assert wcwidth.wcswidth('\U0001F469\U0001F3FB') == 2
+
+
 def test_vs16_effect():
     """Verify effect of VS-16 (always active with latest Unicode version)."""
     phrase = ("\u2640"        # FEMALE SIGN

tests/test_width.py

@@ -437,3 +437,15 @@ def test_soft_hyphen_exception():
     """U+00AD SOFT HYPHEN remains width 1 for ISO-8859-1 compatibility."""
     result = wcwidth.wcwidth('\u00AD')
     assert result == 1
+
+
+def test_fitzpatrick_modifier_after_emoji():
+    """Fitzpatrick modifier following emoji base adds zero-width in width()."""
+    result = wcwidth.width('\U0001F469\U0001F3FB')
+    assert result == 2
+
+
+def test_fitzpatrick_modifier_standalone_width():
+    """Standalone Fitzpatrick modifier, however, is wide character in width()."""
+    result = wcwidth.width('\U0001F3FB')
+    assert result == 2

tox.ini

@@ -107,7 +107,7 @@ commands = python {toxinidir}/bin/update-tables.py {posargs:--fetch-all-versions
 basepython = python3.14
 usedevelop = true
 deps = -r requirements-update.txt
-commands = python {toxinidir}/bin/update-tables.py {posargs:--only-fetch}
+commands = - python {toxinidir}/bin/update-tables.py {posargs:--only-fetch}
 
 [testenv:autopep8]
 basepython = python3.14

wcwidth/table_wide.py

@@ -90,6 +90,7 @@
         (0x1f0cf, 0x1f0cf,),  # Playing Card Black Joker
         (0x1f18e, 0x1f18e,),  # Negative Squared Ab
         (0x1f191, 0x1f19a,),  # Squared Cl              ..Squared Vs
+        (0x1f1e6, 0x1f1ff,),  # Regional Indicator Symbo..Regional Indicator Symbo
         (0x1f200, 0x1f202,),  # Square Hiragana Hoka    ..Squared Katakana Sa
         (0x1f210, 0x1f23b,),  # Squared Cjk Unified Ideo..Squared Cjk Unified Ideo
         (0x1f240, 0x1f248,),  # Tortoise Shell Bracketed..Tortoise Shell Bracketed
@@ -104,6 +105,7 @@
         (0x1f3e0, 0x1f3f0,),  # House Building          ..European Castle
         (0x1f3f4, 0x1f3f4,),  # Waving Black Flag
         (0x1f3f8, 0x1f3fa,),  # Badminton Racquet And Sh..Amphora
+        (0x1f3fb, 0x1f3ff,),  # Emoji Modifier Fitzpatri..Emoji Modifier Fitzpatri
         (0x1f400, 0x1f43e,),  # Rat                     ..Paw Prints
         (0x1f440, 0x1f440,),  # Eyes
         (0x1f442, 0x1f4fc,),  # Ear                     ..Videocassette

wcwidth/table_zero.py

@@ -345,7 +345,9 @@
         (0x1e6f5, 0x1e6f5,),  # Tai Yo Sign Om
         (0x1e8d0, 0x1e8d6,),  # Mende Kikakui Combining ..Mende Kikakui Combining
         (0x1e944, 0x1e94a,),  # Adlam Alif Lengthener   ..Adlam Nukta
-        (0x1f3fb, 0x1f3ff,),  # Emoji Modifier Fitzpatri..Emoji Modifier Fitzpatri
+        # Emoji Modifier Fitzpatrick types (U+1F3FB..U+1F3FF) excluded:
+        # standalone they display as wide (2 cells), only zero-width
+        # when following an emoji base character in sequence.
         (0xe0000, 0xe0fff,),  # (nil)
     ),
 }