From 528f92f7a2a26cade1c57ccf26b0ba6524e7cae5 Mon Sep 17 00:00:00 2001
From: TnR2 <115166373+TnR2@users.noreply.github.com>
Date: 星期三, 01 十月 2025 14:45:17 +0800
Subject: [PATCH] fix: handle empty strings after event removal in transcription processing (def rich_transcription_postprocess(s)) (#2681)
---
fun_text_processing/inverse_text_normalization/fr/verbalizers/whitelist.py | 8 +++-----
1 files changed, 3 insertions(+), 5 deletions(-)
diff --git a/fun_text_processing/inverse_text_normalization/fr/verbalizers/whitelist.py b/fun_text_processing/inverse_text_normalization/fr/verbalizers/whitelist.py
index c94eebf..889dc88 100644
--- a/fun_text_processing/inverse_text_normalization/fr/verbalizers/whitelist.py
+++ b/fun_text_processing/inverse_text_normalization/fr/verbalizers/whitelist.py
@@ -1,5 +1,3 @@
-
-
import pynini
from fun_text_processing.inverse_text_normalization.fr.graph_utils import (
DAMO_CHAR,
@@ -21,9 +19,9 @@
graph = (
pynutil.delete("name:")
+ delete_space
- + pynutil.delete("\"")
+ + pynutil.delete('"')
+ pynini.closure(DAMO_CHAR - " ", 1)
- + pynutil.delete("\"")
+ + pynutil.delete('"')
)
- graph = graph @ pynini.cdrewrite(pynini.cross(u"\u00A0", " "), "", "", DAMO_SIGMA)
+ graph = graph @ pynini.cdrewrite(pynini.cross("\u00A0", " "), "", "", DAMO_SIGMA)
self.fst = graph.optimize()
--
Gitblit v1.9.1