Expand terminal escape sequence for three more ECMA-48 "families"

tests/test_clip.py

@@ -23,7 +23,7 @@
     ('\x1b[1m\U0001F468\u200D\U0001F469\u200D\U0001F467\x1b[0m',
      '\U0001F468\u200D\U0001F469\u200D\U0001F467'),
     ('\x1b', '\x1b'),
-    ('a\x1bb', 'a\x1bb'),
+    ('a\x1bb', 'a'),
     ('\x1b[', ''),
     ('text\x1b[mmore', 'textmore'),
 ]
@@ -148,7 +148,7 @@ def test_clip_sequences_between_chars():
     assert clip('a\x1b[31mb\x1b[0mc', 1, 2) == '\x1b[31mb\x1b[0m'
 
 
-def test_clip_sequences_lone_esc():
+def test_clip_sequences_fs_escape():
     assert clip('a\x1bb', 0, 2) == 'a\x1bb'
 
 
@@ -236,6 +236,7 @@ def test_clip_tab_with_sequences():
     ('abc\rde', 0, 5, 'abc\rde'),
     ('\a\b\rHello', 0, 5, '\a\b\rHello'),
     ('ab\x01\x02cd', 0, 4, 'ab\x01\x02cd'),
+    ('ab\x1b\x00cd', 0, 4, 'ab\x1b\x00cd'),
 ]
 
 

tests/test_textwrap.py

@@ -221,8 +221,8 @@ def test_wrap_unicode(benchmark, text, w, expected):
     # Sequences in long word breaking - red starts after 'x', continues across lines
     ('x\x1b[31mabcdefghij\x1b[0m', 3,
      ['x\x1b[31mab\x1b[0m', '\x1b[31mcde\x1b[0m', '\x1b[31mfgh\x1b[0m', '\x1b[31mij\x1b[0m']),
-    # Lone ESC - not a valid SGR sequence, stays with preceding text
-    ('abc\x1bdefghij', 3, ['abc\x1b', 'def', 'ghi', 'j']),
+    # Fs sequence (ESC d) - zero-width, stays with preceding text
+    ('abc\x1bdefghij', 3, ['abc\x1bd', 'efg', 'hij']),
 ]
 
 SEQUENCE_CASES_NO_PROPAGATE = [
@@ -521,3 +521,9 @@ def test_wrap_replace_whitespace_false_newlines_zero_width():
     """Newlines have zero display width, so more text fits per line than stdlib."""
     assert wrap('hello\nworld foo\nbar', 10, replace_whitespace=False) == [
         'hello\nworld', 'foo\nbar']
+
+
+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']

tests/test_width.py

@@ -4,7 +4,7 @@
 
 # local
 import wcwidth
-from wcwidth.escape_sequences import ZERO_WIDTH_PATTERN
+from wcwidth.escape_sequences import ZERO_WIDTH_PATTERN, INDETERMINATE_EFFECT_SEQUENCE
 
 BASIC_WIDTH_CASES = [
     ('', 0, 'empty'),
@@ -29,7 +29,7 @@ def test_width_basic(text, expected, name):
     ('\x1b[31mred\x1b[0m', 3, 'SGR_sequence'),
     ('hello\x80world', 10, 'C1_control'),
     ('\x1b', 0, 'lone_ESC'),
-    ('a\x1bb', 2, 'lone_ESC_between'),
+    ('a\x1bb', 1, 'fs_sequence_between'),
 ]
 
 
@@ -65,7 +65,7 @@ def test_width_control_codes_strict_raises(text, name):
     ('\x1b[31mred\x1b[0m', 3, 'SGR_sequence'),
     ('a\x1b[2Cb', 4, 'cursor_right'),
     ('\x1b', 0, 'lone_ESC'),
-    ('a\x1bb', 2, 'lone_ESC_between'),
+    ('a\x1bb', 1, 'fs_sequence_between'),
     ('\x1b!', 1, 'ESC_unrecognized'),
 ]
 
@@ -88,6 +88,7 @@ def test_width_control_codes_strict_allowed(text, expected, name):
     ('\x1b[1X', 'erase_chars'),
     ('\x1b[1S', 'parm_index'),
     ('\x1b[1T', 'parm_rindex'),
+    ('\x1bc', 'full_reset'),
 ]
 
 
@@ -449,3 +450,88 @@ def test_fitzpatrick_modifier_standalone_width():
     """Standalone Fitzpatrick modifier, however, is wide character in width()."""
     result = wcwidth.width('\U0001F3FB')
     assert result == 2
+
+
+FS_SEQUENCE_CASES = [
+    ('\x1bc', 'ris'),
+    ('\x1bl', 'memory_lock'),
+    ('\x1bm', 'memory_unlock'),
+    ('\x1bn', 'ls2'),
+    ('\x1bo', 'ls3'),
+    ('\x1b|', 'ls3r'),
+    ('\x1b}', 'ls2r'),
+    ('\x1b~', 'ls1r'),
+]
+
+
+@pytest.mark.parametrize('seq,name', FS_SEQUENCE_CASES)
+def test_fs_sequences_matched(seq, name):
+    """Fs (independent function) sequences are matched as zero-width."""
+    segments = list(wcwidth.iter_sequences(seq))
+    assert segments == [(seq, True)]
+    assert wcwidth.width(seq) == 0
+
+
+FP_SEQUENCE_CASES = [
+    ('\x1b7', 'decsc'),
+    ('\x1b8', 'decrc'),
+    ('\x1b=', 'deckpam'),
+    ('\x1b>', 'deckpnm'),
+    ('\x1b0', 'fp_0'),
+    ('\x1b1', 'fp_1'),
+    ('\x1b9', 'fp_9'),
+]
+
+
+@pytest.mark.parametrize('seq,name', FP_SEQUENCE_CASES)
+def test_fp_sequences_matched(seq, name):
+    """Fp (private use) sequences are matched as zero-width."""
+    segments = list(wcwidth.iter_sequences(seq))
+    assert segments == [(seq, True)]
+    assert wcwidth.width(seq) == 0
+
+
+NF_SEQUENCE_CASES = [
+    ('\x1b F', 's7c1t'),
+    ('\x1b G', 's8c1t'),
+    ('\x1b#3', 'decdhl_top'),
+    ('\x1b#4', 'decdhl_bottom'),
+    ('\x1b#5', 'decswl'),
+    ('\x1b#6', 'decdwl'),
+    ('\x1b#8', 'decaln'),
+    ('\x1b%G', 'utf8_designate'),
+    ('\x1b%@', 'iso2022_return'),
+]
+
+
+@pytest.mark.parametrize('seq,name', NF_SEQUENCE_CASES)
+def test_nf_sequences_matched(seq, name):
+    """NF (multi-byte) escape sequences are matched as zero-width."""
+    segments = list(wcwidth.iter_sequences(seq))
+    assert segments == [(seq, True)]
+    assert wcwidth.width(seq) == 0
+
+
+def test_fs_sequence_embedded_in_text():
+    """Fs sequence surrounded by text is correctly segmented."""
+    segments = list(wcwidth.iter_sequences('abc\x1bcdef'))
+    assert segments == [('abc', False), ('\x1bc', True), ('def', False)]
+    assert wcwidth.width('abc\x1bcdef') == 6
+
+
+def test_nf_sequence_embedded_in_text():
+    """NF sequence surrounded by text is correctly segmented."""
+    segments = list(wcwidth.iter_sequences('abc\x1b#8def'))
+    assert segments == [('abc', False), ('\x1b#8', True), ('def', False)]
+    assert wcwidth.width('abc\x1b#8def') == 6
+
+
+def test_screen_title_sequences():
+    """Screen/tmux title sequence ESC k hello ST."""
+    segments = list(wcwidth.iter_sequences('\x1bkhello\x1b\\'))
+    assert segments[0] == ('\x1bk', True)
+
+
+def test_ris_indeterminate():
+    """RIS (ESC c) is flagged as indeterminate effect."""
+    assert INDETERMINATE_EFFECT_SEQUENCE.match('\x1bc')

wcwidth/escape_sequences.py

@@ -21,12 +21,16 @@
     r'\x1bP[^\x1b\x07]*(?:\x07|\x1b\\)|'
     # PM sequences
     r'\x1b\^[^\x1b\x07]*(?:\x07|\x1b\\)|'
-    # Character set designation
+    # Character set designation (subset of nF, handled separately for clarity)
     r'\x1b[()].|'
-    # Fe sequences
+    # nF sequences: ESC + one or more intermediate bytes (0x20-0x2F) + final byte (0x30-0x7E)
+    r'\x1b[\x20-\x2f]+[\x30-\x7e]|'
+    # Fe sequences (C1 controls)
     r'\x1b[\x40-\x5f]|'
-    # Fp sequences
-    r'\x1b[78=>g]'
+    # Fp sequences (private use)
+    r'\x1b[\x30-\x3f]|'
+    # Fs sequences (independent functions)
+    r'\x1b[\x60-\x7e]'
 )
 
 # Cursor right movement: CSI [n] C, parameter may be parsed by width()
@@ -65,5 +69,6 @@
         r'\x1b8',                    # restore_cursor
         r'\x1bD',                    # scroll_forward (index)
         r'\x1bM',                    # scroll_reverse (reverse index)
+        r'\x1bc',                    # full_reset (RIS)
     ))
 )