GuXiaopeng commited on
Commit
717a03f
·
verified ·
1 Parent(s): 1edc3fd

Rename app_config.py to config.py

Browse files
Files changed (1) hide show
  1. app_config.py → config.py +31 -10
app_config.py → config.py RENAMED
@@ -4,15 +4,20 @@ import streamlit as st
4
 
5
  def load_config():
6
  base_config = {
 
7
  "lit_target_papers": int(os.environ.get("LIT_TARGET_PAPERS", "250")),
8
  "lit_per_page": int(os.environ.get("LIT_PER_PAGE", "100")),
9
  "lit_max_workers": int(os.environ.get("LIT_MAX_WORKERS", "5")),
10
  "lit_max_page_batches": int(os.environ.get("LIT_MAX_PAGE_BATCHES", "12")),
11
  "chunk_words_default": int(os.environ.get("CHUNK_WORDS_DEFAULT", "1500")),
12
  "chunk_words_cyrillic": int(os.environ.get("CHUNK_WORDS_CYRILLIC", "1000")),
 
 
13
  "deepseek_api_url": os.environ.get("DEEPSEEK_API_URL", "https://api.deepseek.com/chat/completions"),
14
  "openalex_api_url": os.environ.get("OPENALEX_API_URL", "https://api.openalex.org/works"),
15
  "s2_api_url": os.environ.get("S2_API_URL", "https://api.semanticscholar.org/graph/v1"),
 
 
16
  "ui_labels": {
17
  "中文": {"title": "📊 科研论文智能体", "btn_generate": "🚀 生成初稿", "btn_download": "📥 下载文件", "topic_label": "📝 论文名称", "keywords_label": "🔍 关键字", "words_label": "📏 目标总字数 (12000-25000字)", "success": "🎉 流水线推演完毕!双语文档已就绪。"},
18
  "English": {"title": "📊 Article AI Agent", "btn_generate": "🚀 Generate Draft", "btn_download": "📥 Download", "topic_label": "📝 Article Title", "keywords_label": "🔍 Keywords", "words_label": "📏 Target Word Count", "success": "🎉 Task complete!"},
@@ -25,6 +30,28 @@ def load_config():
25
  "Русский": "Russian (Русский)",
26
  "Қазақша": "Kazakh (Қазақша)"
27
  },
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
28
  }
29
 
30
  generic_prompts = {
@@ -59,19 +86,13 @@ def load_config():
59
  "empirical_data_logic": "Maintain strict adherence to APA 7th edition reporting standards. Include plausible degrees of freedom, exact p-values, confidence intervals, and effect sizes.",
60
  }
61
 
62
- # 新增:提示词模板(用于减少硬编码)
63
- targeted_directives = {
64
- "introduction": "你必须审视 [TARGETED INJECTION] 中列出的核心文献的局限性与未来展望 (Future Gaps)。禁止泛泛而谈研究背景。你必须将本研究定位为对这些明确 Gaps 的直接回应,建立极强的学术对话感。",
65
- "literature": "严禁简单罗列前人观点或凭空捏造机制。你必须明确使用 [TARGETED INJECTION] 中提供的理论视角 (Theoretical Lens) 作为推导的底层逻辑。你的论证过程必须引用这些核心文献已验证的因果路径,说明新假设是如何在它们的肩膀上向前推演的。",
66
- "methodology": "在撰写测量工具 (Measures) 时,你必须明确声明我们的问卷题项改编自 [TARGETED INJECTION] 中记录的核心文献,以此证明量表具有极高的历史信度与效度。将注入的 Items 自然融入行文中。",
67
- "results": "【绝对数据忠诚指令】你现在是一个没有感情的统计报告员。1. 必须逐字逐句将 <USER_ACTUAL_DATA> 翻译为规范的学术段落。2. 绝对禁止生成任何表格中不存在的 P值、T值或 R²。3. 必须逐一核对 【ESTABLISHED HYPOTHESES】,并在段末明确写出判定 (如 'H1 is supported')。禁止在此进行理论探讨。",
68
- "discussion": "【理论回声与升华指令】本章是全篇灵魂。1. 不要简单重复结果,必须将本研究成立的假设与 [TARGETED INJECTION] 中核心文献的发现进行结构化对比,指出是印证了前人还是得出了矛盾/新边界。2. 基于对比提供高度可操作的管理启示。3. 诚实反思局限并提供具体研究方向。"
69
- }
70
-
71
  default_config = base_config.copy()
72
  default_config.update(generic_prompts)
73
- default_config["targeted_directives"] = targeted_directives # 注入模板
74
  return default_config, amj_prompts, generic_prompts
75
 
 
 
 
 
76
  DEFAULT_CONFIG, AMJ_PROMPTS, GENERIC_PROMPTS = load_config()
77
  APP_CONFIG = DEFAULT_CONFIG.copy()
 
4
 
5
  def load_config():
6
  base_config = {
7
+ # --- 基础运行参数 ---
8
  "lit_target_papers": int(os.environ.get("LIT_TARGET_PAPERS", "250")),
9
  "lit_per_page": int(os.environ.get("LIT_PER_PAGE", "100")),
10
  "lit_max_workers": int(os.environ.get("LIT_MAX_WORKERS", "5")),
11
  "lit_max_page_batches": int(os.environ.get("LIT_MAX_PAGE_BATCHES", "12")),
12
  "chunk_words_default": int(os.environ.get("CHUNK_WORDS_DEFAULT", "1500")),
13
  "chunk_words_cyrillic": int(os.environ.get("CHUNK_WORDS_CYRILLIC", "1000")),
14
+
15
+ # --- API 节点配置 ---
16
  "deepseek_api_url": os.environ.get("DEEPSEEK_API_URL", "https://api.deepseek.com/chat/completions"),
17
  "openalex_api_url": os.environ.get("OPENALEX_API_URL", "https://api.openalex.org/works"),
18
  "s2_api_url": os.environ.get("S2_API_URL", "https://api.semanticscholar.org/graph/v1"),
19
+
20
+ # --- UI 多语言标签 ---
21
  "ui_labels": {
22
  "中文": {"title": "📊 科研论文智能体", "btn_generate": "🚀 生成初稿", "btn_download": "📥 下载文件", "topic_label": "📝 论文名称", "keywords_label": "🔍 关键字", "words_label": "📏 目标总字数 (12000-25000字)", "success": "🎉 流水线推演完毕!双语文档已就绪。"},
23
  "English": {"title": "📊 Article AI Agent", "btn_generate": "🚀 Generate Draft", "btn_download": "📥 Download", "topic_label": "📝 Article Title", "keywords_label": "🔍 Keywords", "words_label": "📏 Target Word Count", "success": "🎉 Task complete!"},
 
30
  "Русский": "Russian (Русский)",
31
  "Қазақша": "Kazakh (Қазақша)"
32
  },
33
+
34
+ # --- 🚨 新增:系统全局核心配置 (去硬编码) ---
35
+ "workspace_dir": ".workspace",
36
+ "max_extract_chars": 150000,
37
+
38
+ # --- 🚨 新增:Word 渲染器样式配置 ---
39
+ "docx_styles": {
40
+ "font_name": "Times New Roman",
41
+ "font_size_pt": 12,
42
+ "code_font": "Courier New",
43
+ "code_size_pt": 11
44
+ },
45
+
46
+ # --- 🚨 新增:全局拦截与保护黑名单 ---
47
+ "blacklists": {
48
+ "meta_sections": [
49
+ "abstract", "keyword", "keywords", "摘要", "关键词", "关键字",
50
+ "reference", "references", "reference list", "参考文献",
51
+ "acknowledgment", "acknowledgements", "致谢",
52
+ "аннотация", "ключевые слова", "түйіндеме", "кілт сөздер"
53
+ ]
54
+ }
55
  }
56
 
57
  generic_prompts = {
 
86
  "empirical_data_logic": "Maintain strict adherence to APA 7th edition reporting standards. Include plausible degrees of freedom, exact p-values, confidence intervals, and effect sizes.",
87
  }
88
 
 
 
 
 
 
 
 
 
 
89
  default_config = base_config.copy()
90
  default_config.update(generic_prompts)
 
91
  return default_config, amj_prompts, generic_prompts
92
 
93
+ # =====================================================================
94
+ # 🚨 核心重点:下面这两行代码绝对不能有任何缩进,必须紧贴屏幕最左侧!
95
+ # 这是让 APP_CONFIG 变成全局变量,能被外部模块识别到的关键。
96
+ # =====================================================================
97
  DEFAULT_CONFIG, AMJ_PROMPTS, GENERIC_PROMPTS = load_config()
98
  APP_CONFIG = DEFAULT_CONFIG.copy()