Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-08-07 09:01:38 -04:00
Code Issues Packages Projects Releases Wiki Activity

Edit book: Fix searching marking incorrect text in the presence of non-BMP unicode characters. Fixes #2075970 [regex does not properly select match](https://bugs.launchpad.net/calibre/+bug/2075970)

Browse Source
This commit is contained in:
Kovid Goyal 2024-08-15 08:49:53 +05:30
parent 9f76ddb56f
commit da187adea6
No known key found for this signature in database
GPG Key ID: 06BC317B515ACE7C
1 changed files with 10 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
src/calibre/gui2/tweak_book/editor/text.py
View File

@ -50,6 +50,13 @@ from calibre.utils.titlecase import titlecase
from polyglot.builtins import as_unicode from polyglot.builtins import as_unicode
def adjust_for_non_bmp_chars(raw: str, start: int, end: int) -> tuple[int, int]:
adjusted_start = utf16_length(raw[:start])
end = adjusted_start + utf16_length(raw[start:end])
start = adjusted_start
return start, end
def get_highlighter(syntax): def get_highlighter(syntax):
if syntax: if syntax:
try: try:
@ -395,6 +402,7 @@ class TextEdit(PlainTextEdit):
start, end = m.span() start, end = m.span()
if start == end: if start == end:
return False return False
start, end = adjust_for_non_bmp_chars(raw, start, end)
if wrap: if wrap:
if reverse: if reverse:
textpos = c.anchor() textpos = c.anchor()
@ -486,7 +494,7 @@ class TextEdit(PlainTextEdit):
start, end = m.span() start, end = m.span()
if start == end: if start == end:
return False return False
end = start + utf16_length(raw[start:end]) start, end = adjust_for_non_bmp_chars(raw, start, end)
if wrap and not complete: if wrap and not complete:
if reverse: if reverse:
textpos = c.anchor() textpos = c.anchor()
@ -532,7 +540,7 @@ class TextEdit(PlainTextEdit):
start, end = m.span() start, end = m.span()
if start == end: if start == end:
return False return False
end = start + utf16_length(raw[start:end]) start, end = adjust_for_non_bmp_chars(raw, start, end)
if reverse: if reverse:
start, end = end, start start, end = end, start
c.clearSelection() c.clearSelection()