feat: 添加日志记录功能，输出至 app.log 文件

.gitignore

@@ -12,3 +12,5 @@ config_test.json
 /novel_generator/__pycache__
 /ui/__pycache__
 .idea/
+/novel
+app.log

novel_generator/architecture.py

@@ -16,6 +16,13 @@ from prompt_definitions import (
     plot_architecture_prompt,
     create_character_state_prompt
 )
+logging.basicConfig(
+    filename='app.log',      # 日志文件名
+    filemode='a',            # 追加模式（'w' 会覆盖）
+    level=logging.INFO,      # 记录 INFO 及以上级别的日志
+    format='%(asctime)s - %(name)s - %(levelname)s - %(message)s',
+    datefmt='%Y-%m-%d %H:%M:%S'
+)
 from utils import clear_file_content, save_string_to_txt
 
 def load_partial_architecture_data(filepath: str) -> dict:

novel_generator/blueprint.py

@@ -10,7 +10,13 @@ from novel_generator.common import invoke_with_cleaning
 from llm_adapters import create_llm_adapter
 from prompt_definitions import chapter_blueprint_prompt, chunked_chapter_blueprint_prompt
 from utils import read_file, clear_file_content, save_string_to_txt
-
+logging.basicConfig(
+    filename='app.log',      # 日志文件名
+    filemode='a',            # 追加模式（'w' 会覆盖）
+    level=logging.INFO,      # 记录 INFO 及以上级别的日志
+    format='%(asctime)s - %(name)s - %(levelname)s - %(message)s',
+    datefmt='%Y-%m-%d %H:%M:%S'
+)
 def compute_chunk_size(number_of_chapters: int, max_tokens: int) -> int:
     """
     基于“每章约100 tokens”的粗略估算，
@@ -18,7 +24,7 @@ def compute_chunk_size(number_of_chapters: int, max_tokens: int) -> int:
       chunk_size = (floor(max_tokens/100/10)*10) - 10
     并确保 chunk_size 不会小于1或大于实际章节数。
     """
-    tokens_per_chapter = 100.0
+    tokens_per_chapter = 200.0
     ratio = max_tokens / tokens_per_chapter
     ratio_rounded_to_10 = int(ratio // 10) * 10
     chunk_size = ratio_rounded_to_10 - 10

novel_generator/chapter.py

@@ -22,6 +22,13 @@ from novel_generator.vectorstore_utils import (
     get_relevant_context_from_vector_store,
     load_vector_store  # 添加导入
 )
+logging.basicConfig(
+    filename='app.log',      # 日志文件名
+    filemode='a',            # 追加模式（'w' 会覆盖）
+    level=logging.INFO,      # 记录 INFO 及以上级别的日志
+    format='%(asctime)s - %(name)s - %(levelname)s - %(message)s',
+    datefmt='%Y-%m-%d %H:%M:%S'
+)
 
 def get_last_n_chapters_text(chapters_dir: str, current_chapter_num: int, n: int = 3) -> list:
     """

novel_generator/common.py

@@ -7,7 +7,13 @@ import logging
 import re
 import time
 import traceback
-
+logging.basicConfig(
+    filename='app.log',      # 日志文件名
+    filemode='a',            # 追加模式（'w' 会覆盖）
+    level=logging.INFO,      # 记录 INFO 及以上级别的日志
+    format='%(asctime)s - %(name)s - %(levelname)s - %(message)s',
+    datefmt='%Y-%m-%d %H:%M:%S'
+)
 def call_with_retry(func, max_retries=3, sleep_time=2, fallback_return=None, **kwargs):
     """
     通用的重试机制封装。

novel_generator/finalization.py

@@ -11,7 +11,13 @@ from prompt_definitions import summary_prompt, update_character_state_prompt
 from novel_generator.common import invoke_with_cleaning
 from utils import read_file, clear_file_content, save_string_to_txt
 from novel_generator.vectorstore_utils import update_vector_store
-
+logging.basicConfig(
+    filename='app.log',      # 日志文件名
+    filemode='a',            # 追加模式（'w' 会覆盖）
+    level=logging.INFO,      # 记录 INFO 及以上级别的日志
+    format='%(asctime)s - %(name)s - %(levelname)s - %(message)s',
+    datefmt='%Y-%m-%d %H:%M:%S'
+)
 def finalize_chapter(
     novel_number: int,
     word_number: int,

novel_generator/knowledge.py

@@ -16,11 +16,17 @@ from langchain.docstore.document import Document
 # 禁用特定的Torch警告
 warnings.filterwarnings('ignore', message='.*Torch was not compiled with flash attention.*')
 os.environ["TOKENIZERS_PARALLELISM"] = "false"
-
+logging.basicConfig(
+    filename='app.log',      # 日志文件名
+    filemode='a',            # 追加模式（'w' 会覆盖）
+    level=logging.INFO,      # 记录 INFO 及以上级别的日志
+    format='%(asctime)s - %(name)s - %(levelname)s - %(message)s',
+    datefmt='%Y-%m-%d %H:%M:%S'
+)
 def advanced_split_content(content: str, similarity_threshold: float = 0.7, max_length: int = 500) -> list:
     """使用基本分段策略"""
-    nltk.download('punkt', quiet=True)
-    nltk.download('punkt_tab', quiet=True)
+    # nltk.download('punkt', quiet=True)
+    # nltk.download('punkt_tab', quiet=True)
     sentences = nltk.sent_tokenize(content)
     if not sentences:
         return []

novel_generator/vectorstore_utils.py

@@ -13,7 +13,13 @@ import ssl
 import requests
 import warnings
 from langchain_chroma import Chroma
-
+logging.basicConfig(
+    filename='app.log',      # 日志文件名
+    filemode='a',            # 追加模式（'w' 会覆盖）
+    level=logging.INFO,      # 记录 INFO 及以上级别的日志
+    format='%(asctime)s - %(name)s - %(levelname)s - %(message)s',
+    datefmt='%Y-%m-%d %H:%M:%S'
+)
 # 禁用特定的Torch警告
 warnings.filterwarnings('ignore', message='.*Torch was not compiled with flash attention.*')
 os.environ["TOKENIZERS_PARALLELISM"] = "false"  # 禁用tokenizer并行警告
@@ -146,8 +152,8 @@ def split_text_for_vectorstore(chapter_text: str, max_length: int = 500, similar
     if not chapter_text.strip():
         return []
     
-    nltk.download('punkt', quiet=True)
-    nltk.download('punkt_tab', quiet=True)
+    # nltk.download('punkt', quiet=True)
+    # nltk.download('punkt_tab', quiet=True)
     sentences = nltk.sent_tokenize(chapter_text)
     if not sentences:
         return []