Dymas/abogen
1
0
Fork 0
mirror of https://github.com/denizsafak/abogen.git synced 2026-07-18 13:40:27 +02:00
Code Issues Packages Projects Releases Wiki Activity

fix: Refactor regex patterns for number ranges and fractions to improve accuracy

Browse Source
This commit is contained in:
JB
2025-10-30 06:54:33 -07:00
parent 013c5c2dbb
commit 9acec3c309
1 changed files with 4 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files
abogen/kokoro_text_normalization.py
+4 -2
View File
@@ -94,13 +94,15 @@ CONTRACTIONS_EXACT = {
# For ambiguous 'd and 's we handle separately
_NUMBER_WITH_GROUP_RE = re.compile(r"(?<![\w\d])(-?\d{1,3}(?:,\d{3})+)(?![\w\d])")
_NUMBER_RANGE_SEPARATORS = "-‐‑–—−"
_NUMBER_RANGE_CLASS = re.escape(_NUMBER_RANGE_SEPARATORS)
_WIDE_RANGE_SEPARATORS = {"–", "—"}
_NUMBER_RANGE_RE = re.compile(
rf"(?<!\w)(?P<left>-?\d+)(?P<sep>\s*[{_NUMBER_RANGE_SEPARATORS}]\s*)(?P<right>-?\d+)(?![\w{_NUMBER_RANGE_SEPARATORS}/])"
rf"(?<!\w)(?P<left>-?\d+)(?P<sep>\s*[{_NUMBER_RANGE_CLASS}]\s*)(?P<right>-?\d+)(?![\w{_NUMBER_RANGE_CLASS}/])"
)
_FRACTION_SLASHES = "/⁄"
_FRACTION_SLASH_CLASS = re.escape(_FRACTION_SLASHES)
_FRACTION_RE = re.compile(
rf"(?<!\w)(?P<numerator>-?\d+)\s*[{_FRACTION_SLASHES}]\s*(?P<denominator>-?\d+)(?![\w{_FRACTION_SLASHES}])"
rf"(?<!\w)(?P<numerator>-?\d+)\s*[{_FRACTION_SLASH_CLASS}]\s*(?P<denominator>-?\d+)(?![\w{_FRACTION_SLASH_CLASS}])"
)