Merge fb432792b9 into 0752ff0472

Fix ZWJ and edge cases

CHANGELOG.md

@@ -5,6 +5,13 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [14.3.2] - 2026-02-01
+
+### Fixed
+
+- Fixed solo ZWJ crash https://github.com/Textualize/rich/pull/3953
+- Fixed control codes reporting width of 1 https://github.com/Textualize/rich/pull/3953
+
 ## [14.3.1] - 2026-01-24
 
 ### Fixed

pyproject.toml

@@ -2,7 +2,7 @@
 name = "rich"
 homepage = "https://github.com/Textualize/rich"
 documentation = "https://rich.readthedocs.io/en/latest/"
-version = "14.3.1"
+version = "14.3.2"
 description = "Render rich text, tables, progress bars, syntax highlighting, markdown and more to the terminal"
 authors = ["Will McGugan <willmcgugan@gmail.com>"]
 license = "MIT"

rich/cells.py

@@ -55,23 +55,26 @@ def get_character_cell_size(character: str, unicode_version: str = "auto") -> in
         int: Number of cells (0, 1 or 2) occupied by that character.
     """
     codepoint = ord(character)
+    if codepoint and codepoint < 32 or 0x07F <= codepoint < 0x0A0:
+        return 0
     table = load_cell_table(unicode_version).widths
-    if codepoint > table[-1][1]:
+
+    last_entry = table[-1]
+    if codepoint > last_entry[1]:
         return 1
+
     lower_bound = 0
     upper_bound = len(table) - 1
-    index = (lower_bound + upper_bound) // 2
-    while True:
+
+    while lower_bound <= upper_bound:
+        index = (lower_bound + upper_bound) >> 1
         start, end, width = table[index]
         if codepoint < start:
             upper_bound = index - 1
         elif codepoint > end:
             lower_bound = index + 1
         else:
-            return 0 if width == -1 else width
-        if upper_bound < lower_bound:
-            break
-        index = (lower_bound + upper_bound) // 2
+            return width
     return 1
 
 
@@ -135,12 +138,14 @@ def _cell_len(text: str, unicode_version: str) -> int:
 
     SPECIAL = {"\u200d", "\ufe0f"}
 
-    iter_characters = iter(text)
+    index = 0
+    character_count = len(text)
 
-    for character in iter_characters:
+    while index < character_count:
+        character = text[index]
         if character in SPECIAL:
             if character == "\u200d":
-                next(iter_characters)
+                index += 1
             elif last_measured_character:
                 total_width += last_measured_character in cell_table.narrow_to_wide
                 last_measured_character = None
@@ -148,6 +153,7 @@ def _cell_len(text: str, unicode_version: str) -> int:
             if character_width := get_character_cell_size(character, unicode_version):
                 last_measured_character = character
                 total_width += character_width
+        index += 1
 
     return total_width
 

rich/style.py

@@ -1,8 +1,9 @@
 import sys
 from functools import lru_cache
+from itertools import count
 from operator import attrgetter
 from pickle import dumps, loads
-from random import randint
+from random import getrandbits
 from typing import Any, Dict, Iterable, List, Optional, Type, Union, cast
 
 from . import errors
@@ -18,6 +19,9 @@ _hash_getter = attrgetter(
 StyleType = Union[str, "Style"]
 
 
+_id_generator = count(getrandbits(24))
+
+
 class _Bit:
     """A descriptor to get/set a style attribute bit."""
 
@@ -195,7 +199,7 @@ class Style:
         self._link = link
         self._meta = None if meta is None else dumps(meta)
         self._link_id = (
-            f"{randint(0, 999999)}{hash(self._meta)}" if (link or meta) else ""
+            f"{next(_id_generator)}{hash(self._meta)}" if (link or meta) else ""
         )
         self._hash: Optional[int] = None
         self._null = not (self._set_attributes or color or bgcolor or link or meta)
@@ -245,7 +249,7 @@ class Style:
         style._attributes = 0
         style._link = None
         style._meta = dumps(meta)
-        style._link_id = f"{randint(0, 999999)}{hash(style._meta)}"
+        style._link_id = f"{next(_id_generator)}{hash(style._meta)}"
         style._hash = None
         style._null = not (meta)
         return style
@@ -483,7 +487,7 @@ class Style:
         style._attributes = self._attributes
         style._set_attributes = self._set_attributes
         style._link = self._link
-        style._link_id = f"{randint(0, 999999)}" if self._link else ""
+        style._link_id = f"{next(_id_generator)}" if self._link else ""
         style._null = False
         style._meta = None
         style._hash = None
@@ -635,7 +639,7 @@ class Style:
         style._attributes = self._attributes
         style._set_attributes = self._set_attributes
         style._link = self._link
-        style._link_id = f"{randint(0, 999999)}" if self._link else ""
+        style._link_id = f"{next(_id_generator)}" if self._link else ""
         style._hash = self._hash
         style._null = False
         style._meta = self._meta
@@ -681,7 +685,7 @@ class Style:
         style._attributes = self._attributes
         style._set_attributes = self._set_attributes
         style._link = link
-        style._link_id = f"{randint(0, 999999)}" if link else ""
+        style._link_id = f"{next(_id_generator)}" if link else ""
         style._hash = None
         style._null = False
         style._meta = self._meta

tests/test_cells.py

@@ -187,3 +187,20 @@ def test_nerd_font():
     """Regression test for https://github.com/Textualize/rich/issues/3943"""
     # Not allocated by unicode, but used by nerd fonts
     assert cell_len("\U000f024d") == 1
+
+
+def test_zwj():
+    """Test special case of zero width joiners"""
+    assert cell_len("") == 0
+    assert cell_len("\u200d") == 0
+    assert cell_len("1\u200d") == 1
+    # This sequence should really produce 2, but it aligns with with wcwidth
+    # What gets written to the terminal is anybody's guess, I've seen multiple variations
+    assert cell_len("1\u200d2") == 1
+
+
+def test_non_printable():
+    """Non printable characters should report a width of 0."""
+    for ordinal in range(31):
+        character = chr(ordinal)
+        assert cell_len(character) == 0