From 53923e0d25905f0999464b39202589a3970fc5f5 Mon Sep 17 00:00:00 2001 From: harry Date: Sat, 13 Apr 2024 21:03:55 +0800 Subject: [PATCH] optimize segmentation --- app/config/config.py | 2 +- app/utils/utils.py | 17 ++++++++++++++++- 2 files changed, 17 insertions(+), 2 deletions(-) diff --git a/app/config/config.py b/app/config/config.py index 19d4a77..a8244b4 100644 --- a/app/config/config.py +++ b/app/config/config.py @@ -53,7 +53,7 @@ listen_port = _cfg.get("listen_port", 8080) project_name = _cfg.get("project_name", "MoneyPrinterTurbo") project_description = _cfg.get("project_description", "https://github.com/harry0703/MoneyPrinterTurbo") -project_version = _cfg.get("project_version", "1.1.0") +project_version = _cfg.get("project_version", "1.1.1") reload_debug = False imagemagick_path = app.get("imagemagick_path", "") diff --git a/app/utils/utils.py b/app/utils/utils.py index 9d3e675..12bddf6 100644 --- a/app/utils/utils.py +++ b/app/utils/utils.py @@ -163,12 +163,27 @@ def str_contains_punctuation(word): def split_string_by_punctuations(s): result = [] txt = "" - for char in s: + + previous_char = "" + next_char = "" + for i in range(len(s)): + char = s[i] + if i > 0: + previous_char = s[i - 1] + if i < len(s) - 1: + next_char = s[i + 1] + + if char == "." and previous_char.isdigit() and next_char.isdigit(): + # 取现1万,按2.5%收取手续费, 2.5 中的 . 不能作为换行标记 + txt += char + continue + if char not in const.PUNCTUATIONS: txt += char else: result.append(txt.strip()) txt = "" + return result