"""
《乱世佳人》文学文本分析工具 —— 专业增强版
功能亮点：
- ✨ 深色/浅色主题切换（护眼模式）
- 📤 一键导出 Markdown 分析报告
- 😊 基于 VADER 的文学情感分析（比 TextBlob 更准）
- 🔗 高效人物共现关系构建（使用 itertools 优化）
- ⚡ 多线程 + 缓存，避免重复计算
- 🎨 精致 UI：卡片式摘要、分隔线、加载指示器
作者：AI 助手 | 日期：2025-12-09
"""
import tkinter as tk
from tkinter import ttk, scrolledtext, filedialog, messagebox
import re
from collections import Counter
import threading
import hashlib
import os
from itertools import combinations
# === 第三方库 ===
try:
from ttkthemes import ThemedTk
except ImportError:
raise ImportError
("请安装 ttkthemes: pip install ttkthemes"
)
import nltk
from nltk.corpus import stop
words
from nltk.tokenize import word_tokenize, sent_tokenize
nltk.download
('punkt', quiet=True
)
nltk.download
('stop
words', quiet=True
)
# 使用 VADER 替代 TextBlob，更适合文学/口语化文本
try:
from vaderSentiment.vaderSentiment import SentimentIntensityAnalyzer
VADER_AVAILABLE = True
except ImportError:
VADER_AVAILABLE = False
print
("警告：未安装 vaderSentiment，将回退到基础情感提示。"
)
from wordcloud import WordCloud
import matplotlib.pyplot as plt
import networkx as nx
# === 全局配置 ===
CHARACTERS = [
"scarlett", "rhett", "melanie", "ashley", "ellen", "gerald",
"pitty", "archie", "frank", "bonnie", "suellen", "mammy"
]
ANALYZERS = [
{"name": "basic", "title": "📊 基本统计"},
{"name": "freq", "title": "📈 词频分析"},
{"name": "sentiment", "title": "😊 情感分析"},
{"name": "chars", "title": "👥 人物识别"},
{"name": "relations", "title": "🔗 人物关系"},
]
UI_CONFIG = {
"font_normal":
("Segoe UI", 10
),
"font_bold":
("Segoe UI", 10, "bold"
),
"card_padding":
(12, 8
),
"button_font":
("Segoe UI", 9
),
}
# === 工具函数 ===
def clean_text
(text
):
"""
清理原始文本：转小写、移除非字母字符、分词、过滤停用词。
Args:
text
(str
): 原始输入文本
Returns:
list[str]: 清洗后的单词列表
"""
text = re.sub
(r'[^a-z\s]', ' ', text.lower
())
tokens = [
w for w in word_tokenize
(text
)
if w not in stop
words.
words('english'
) and len
(w
) > 2
]
return tokens
def split_sentences
(text
):
"""
使用 NLTK 将英文文本按句子分割，并转为小写。
Args:
text
(str
): 原始文本
Returns:
list[str]: 小写句子列表
"""
return sent_tokenize
(text.lower
())
def validate_input
(text
):
"""
验证输入是否为有效英文文学文本。
条件：
- 非空且长度 ≥10
- 英文字母占比 ≥30%
- 英文字母总数 ≥10
Returns:
bool: 是否有效
"""
if not
(text := text.strip
()):
return False
if len
(text
) < 10:
return False
en_chars = sum
(1 for c in text if 'a' <= c.lower
() <= 'z'
)
alpha_total = max
(len
([c for c in text if c.isalpha
()]
), 1
)
return en_chars >= 10 and en_chars / alpha_total >= 0.3
def read_file_smart
(path
):
"""尝试多种编码读取文件，优先 UTF-8"""
for enc in ['utf-8-sig', 'utf-8', 'gbk', 'latin1']:
try:
with open
(path, encoding=enc
) as f:
return f.read
()
except UnicodeDecodeError:
continue
raise ValueError
("无法读取文件编码，请使用纯文本
(.txt
)文件。"
)
def clean_markup
(text
):
"""
移除常见标记语言残留，提升分析准确性。
- HTML 标签：<div>...</div>
- Markdown 链接：[text]
(url
) 或图片 ![alt]
(img
)
"""
# 移除 HTML 标签
text = re.sub
(r'<[^>]+>', ' ', text
)
# 移除 Markdown 链接和图片语法
text = re.sub
(r'\[.*?\]\
(.*?\
)|!\[.*?\]\
(.*?\
)', ' ', text
)
# 合并多个空白符为单个空格
return re.sub
(r'\s+', ' ', text
).strip
()
# === 缓存系统 ===
class Cache:
"""轻量级内存缓存，避免重复分析相同文本"""
def __init__
(self
):
self._cache = {}
def key
(self, text
):
"""生成文本的短哈希键（16位）"""
return hashlib.md5
(text.encode
('utf-8', errors='ignore'
)).hexdigest
()[:16]
def get
(self, k, name
):
return self._cache.get
((k, name
))
def
set(self, k, name, data
):
self._cache[
(k, name
)] = data
CACHE = Cache
()
# === 分析引擎（独立模块）===
def basic_stats
(text
):
"""返回基本统计信息：总词数、唯一词数"""
words = clean_text
(text
)
return {"total": len
(words), "unique": len
(set(words))}
def word_freq
(text
):
"""返回高频词列表及完整计数器（用于词云）"""
words = clean_text
(text
)
counter = Counter
(words)
return {"top": counter.most_common
(30
), "counter": counter}
def sentiment
(text
):
"""使用 VADER 进行情感分析（若不可用则返回提示）"""
if not VADER_AVAILABLE:
return {
"mood": "⚠️ 模型缺失",
"desc": "请安装 vaderSentiment 以启用精准情感分析：pip install vaderSentiment",
"
compound": None,
"detail": {}
}
analyzer = SentimentIntensityAnalyzer
()
scores = analyzer.polarity_scores
(text
)
compound = scores['
compound']
if
compound >= 0.05:
mood = "积极乐观 😊"
elif
compound <= -0.05:
mood = "消极悲观 😞"
else:
mood = "中性平和 😐"
return {
"mood": mood,
"desc": "基于 VADER 模型（专为文学、口语化文本设计）",
"
compound": round
(compound, 3
),
"detail": scores
}
def characters
(text
):
"""统计预设人物在文本中的出现次数"""
low = text.lower
()
return {c: low.count
(c
) for c in CHARACTERS if c in low}
def relations
(text
):
"""
构建人物共现关系网络。
优化点：
- 仅当句子提及 ≥2 人时才计算组合
- 使用 itertools.combinations 避免手写双循环
"""
pairs = Counter
()
for s in split_sentences
(text
):
mentioned = [c for c in CHARACTERS if c in s]
if len
(mentioned
) < 2:
continue
# 对提及人物排序后生成无向边
for a, b in combinations
(sorted
(mentioned
), 2
):
pairs[
(a, b
)] += 1
return {"pairs": pairs}
# === 摘要生成器 ===
def generate_summary
(name, data
):
"""根据分析结果生成用户友好的自然语言摘要"""
if name == "basic":
total, unique = data["total"], data["unique"]
div = unique / total if total else 0
level = '较高' if div > 0.3 else '中等' if div > 0.2 else '较低'
return
(
f"📌 这段文本共包含 **{total:,} 个单词**，其中 **{unique:,} 个不重复单词**。\n"
f"词汇丰富度为 **{div:.1%}** —— **{level}**。\n"
f"📚 小知识：经典文学作品通常在 20%~40% 之间。"
)
elif name == "freq":
top10 = [w for w, _ in data["top"][:10]]
return
(
f"🔤 **高频词前10名**：{' · '.join
(top10
)}\n"
f"这些词反映核心主题（已过滤停用词）。\n"
f"若出现人名（如 scarlett），说明聚焦人物互动。"
)
elif name == "sentiment":
if data["
compound"] is None:
return data["desc"]
return
(
f"🎭 **情绪基调**：{data['mood']}\n"
f"{data['desc']}\n"
f"📊 复合极性值：{data['
compound']}（-1~1，越正越积极）\n"
f"🔍 细项：正面={data['detail']['pos']:.2f}, "
f"负面={data['detail']['neg']:.2f}, "
f"中性={data['detail']['neu']:.2f}"
)
elif name == "chars":
if not data:
return "🔍 **未检测到预设人物**。\n请确保包含如 Scarlett、Rhett 等名字（不区分大小写）。"
main = max
(data, key=data.get
)
total_mentions = sum
(data.values
())
return
(
f"👑 **核心人物**：**{main.capitalize
()}**（被提及 {data[main]} 次）\n"
f"共识别 {len
(data
)} 位角色，总计 {total_mentions} 次提及。\n"
f"当前段落可能围绕 {main.capitalize
()} 展开情节。"
)
elif name == "relations":
pairs = [
(a, b, w
) for
(a, b
), w in data["pairs"].items
() if w >= 2]
if not pairs:
return "🕸️ **未发现强人物关联**。\n建议分析更长段落（如一整章）以捕捉互动。"
a, b, w = max
(pairs, key=lambda x: x[2]
)
return
(
f"💞 **最紧密关系**：**{a.capitalize
()} ↔ {b.capitalize
()}**（共现 {w} 次）\n"
f"两人可能频繁互动（情侣/对手/亲属）。\n"
f"点击下方按钮查看完整关系网络图。"
)
return str
(data
)
# === 结果渲染器（策略模式）===
class ResultRenderer:
"""将不同分析结果的 UI 渲染逻辑解耦"""
@staticmethod
def _create_summary_card
(parent, summary, is_dark
):
"""创建统一风格的摘要卡片"""
bg = "#2d2d2d" if is_dark else "#f8f9fa"
fg = "white" if is_dark else "black"
card = tk.Frame
(parent, bg=bg, relief="groove", bd=1
)
card.pack
(fill=tk.X, padx=10, pady=
(10, 5
))
txt = tk.Text
(
card, wrap=tk.WORD, bg=bg, fg=fg, relief="flat",
height=4, font=UI_CONFIG["font_normal"]
)
txt.insert
(tk.END, summary
)
# 支持 **加粗** 语法
start = 0
while True:
start = txt.search
("**", f"1.{start}", tk.END
)
if not start: break
end = txt.search
("**", f"{start}+2c", tk.END
)
if not end: break
txt.delete
(f"{start}", f"{start}+2c"
)
txt.delete
(f"{end}-2c", f"{end}"
)
txt.tag_add
("bold", f"{start}", f"{end}-2c"
)
start = txt.index
(f"{end}-2c"
)
bold_font = tk.font.Font
(txt, family="Segoe UI", size=10, weight="bold"
)
txt.tag_configure
("bold", font=bold_font
)
txt.config
(state='disabled'
)
txt.pack
(padx=12, pady=8, fill=tk.X
)
return card
@staticmethod
def render_basic
(tab, data, app
):
summary = generate_summary
("basic", data
)
ResultRenderer._create_summary_card
(tab, summary, app.is_dark
)
@staticmethod
def render_freq
(tab, data, app
):
summary = generate_summary
("freq", data
)
ResultRenderer._create_summary_card
(tab, summary, app.is_dark
)
list_frame = ttk.Frame
(tab
)
list_frame.pack
(fill=tk.BOTH, expand=True, padx=10
)
txt = scrolledtext.ScrolledText
(list_frame, width=30, font=
("Consolas", 9
))
for w, c in data["top"]:
txt.insert
(tk.END, f"{w:<15} : {c}\n"
)
txt.config
(state='disabled'
)
txt.pack
(side=tk.LEFT, fill=tk.BOTH, expand=True
)
btn = ttk.Button
(list_frame, text="🖼️ 生成词云", command=lambda: app.show_wordcloud
(data["counter"]
))
btn.pack
(side=tk.RIGHT, padx=10, pady=10
)
@staticmethod
def render_sentiment
(tab, data, app
):
summary = generate_summary
("sentiment", data
)
ResultRenderer._create_summary_card
(tab, summary, app.is_dark
)
@staticmethod
def render_chars
(tab, data, app
):
summary = generate_summary
("chars", data
)
ResultRenderer._create_summary_card
(tab, summary, app.is_dark
)
if data:
txt = scrolledtext.ScrolledText
(tab, height=6, font=
("Consolas", 9
))
for c, n in sorted
(data.items
(), key=lambda x: -x[1]
):
txt.insert
(tk.END, f"{c.capitalize
():<12} : {n} 次\n"
)
txt.config
(state='disabled'
)
txt.pack
(fill=tk.BOTH, expand=True, padx=10
)
@staticmethod
def render_relations
(tab, data, app
):
summary = generate_summary
("relations", data
)
ResultRenderer._create_summary_card
(tab, summary, app.is_dark
)
pairs = [
(a, b, w
) for
(a, b
), w in data["pairs"].items
() if w >= 2]
if pairs:
btn = ttk.Button
(tab, text="🌐 查看人物关系图", command=lambda: app.show_relations
(data["pairs"]
))
btn.pack
(pady=10
)
# === 主应用类 ===
class App:
def __init__
(self, root
):
self.root = root
self.is_dark = False
self.spinner = None
root.title
("《乱世佳人》文本分析工具 — 专业增强版"
)
root.geometry
("920x800"
)
self.
setup_ui
()
def
setup_ui
(self
):
# 菜单栏
menubar = tk.Menu
(self.root
)
view_menu = tk.Menu
(menubar, tearoff=0
)
view_menu.add_command
(label="切换深色/浅色主题", command=self.toggle_theme
)
menubar.add_cascade
(label="视图", menu=view_menu
)
self.root.config
(menu=menubar
)
style = ttk.Style
()
style.configure
("TButton", font=UI_CONFIG["button_font"]
)
style.configure
("TNotebook.Tab", padding=[12, 6]
)
# 控制区
ctrl = ttk.Frame
(self.root
)
ctrl.pack
(pady=8, padx=10, fill=tk.X
)
ttk.Button
(ctrl, text="📂 加载文件", command=self.load_file
).pack
(side=tk.LEFT
)
ttk.Button
(ctrl, text="🧹 清空", command=self.clear
).pack
(side=tk.LEFT, padx=5
)
self.analyze_btn = ttk.Button
(ctrl, text="▶️ 分析全部", state='disabled', command=self.start_analysis
)
self.analyze_btn.pack
(side=tk.LEFT, padx=5
)
self.export_btn = ttk.Button
(ctrl, text="📤 导出报告", state='disabled', command=self.export_report
)
self.export_btn.pack
(side=tk.LEFT, padx=5
)
self.status = ttk.Label
(ctrl, text="📝 请粘贴英文小说原文或加载 .txt 文件", foreground="gray"
)
self.status.pack
(side=tk.RIGHT
)
# 输入区
ttk.Label
(self.root, text="📖 英文小说原文（建议 ≥100 字）:", font=UI_CONFIG["font_bold"]
).pack
(anchor='w', padx=10, pady=
(10, 0
))
self.text_area = scrolledtext.ScrolledText
(self.root, height=5, font=
("Consolas", 10
))
self.text_area.pack
(padx=10, fill=tk.BOTH
)
self.text_area.bind
('<KeyRelease>', self.on_text_change
)
# 分隔线
ttk.Separator
(self.root, orient='horizontal'
).pack
(fill='x', padx=10, pady=5
)
# 结果区
self.notebook = ttk.Notebook
(self.root
)
self.notebook.pack
(padx=10, pady=
(0, 10
), fill=tk.BOTH, expand=True
)
self.tabs = {}
for a in ANALYZERS:
frame = ttk.Frame
(self.notebook
)
self.notebook.add
(frame, text=a["title"]
)
self.tabs[a["name"]] = frame
def toggle_theme
(self
):
"""切换深色/浅色主题并同步更新摘要卡片"""
self.is_dark = not self.is_dark
theme = "azure-dark" if self.is_dark else "azure"
self.root.
set_theme
(theme
)
# 重新渲染所有标签页以应用新主题
raw = self.text_area.get
(1.0, tk.END
)
text = clean_markup
(raw
)
if hasattr
(self, 'current_hash'
):
self._refresh_ui_with_theme
(text
)
def _refresh_ui_with_theme
(self, text
):
"""仅用于主题切换时刷新 UI（保留分析结果）"""
for name in ANALYZERS:
data = CACHE.get
(self.current_hash, name["name"]
)
if data is not None:
for w in self.tabs[name["name"]].winfo_children
():
w.destroy
()
getattr
(ResultRenderer, f"render_{name['name']}"
)(self.tabs[name["name"]], data, self
)
def on_text_change
(self, event=None
):
raw = self.text_area.get
(1.0, tk.END
)
text = clean_markup
(raw
)
valid = validate_input
(text
)
state = 'normal' if valid else 'disabled'
self.analyze_btn.config
(state=state
)
self.export_btn.config
(state=state
)
self.status.config
(
text="✅ 内容有效 · 可点击「分析全部」" if valid else
"📝 请粘贴英文小说原文或加载文件" if not raw.strip
() else
"⚠️ 当前内容暂不符合分析要求",
foreground="green" if valid else "gray" if not raw.strip
() else "orange"
)
def load_file
(self
):
path = filedialog.askopenfilename
(filetypes=[
("Text files", "*.txt"
)]
)
if not path: return
threading.Thread
(target=self._load_bg, args=
(path,
), daemon=True
).start
()
def _load_bg
(self, path
):
try:
content = read_file_smart
(path
)
self.root.after
(0, lambda: self.text_area.delete
(1.0, tk.END
) or self.text_area.insert
(tk.END, content
))
except Exception as e:
self.root.after
(0, lambda: messagebox.showerror
("错误", str
(e
)))
def clear
(self
):
self.text_area.delete
(1.0, tk.END
)
self.on_text_change
()
def start_analysis
(self
):
raw = self.text_area.get
(1.0, tk.END
)
text = clean_markup
(raw
)
if not validate_input
(text
):
messagebox.showwarning
("输入无效", "请确保输入足够长的英文文本。"
)
return
new_hash = CACHE.key
(text
)
if getattr
(self, 'current_hash', None
) == new_hash:
self.status.config
(text="ℹ️ 内容未变，跳过重复分析", foreground="blue"
)
return
self.current_hash = new_hash
# 显示加载指示器
if self.spinner:
self.spinner.destroy
()
self.spinner = ttk.Progressbar
(self.root, mode='indeterminate'
)
self.spinner.pack
(pady=2
)
self.spinner.start
()
self.analyze_btn.config
(state='disabled'
)
threading.Thread
(target=self._analyze_bg, args=
(text,
), daemon=True
).start
()
def _analyze_bg
(self, text
):
"""后台执行全部分析任务"""
results = {}
analysis_funcs = {
"basic": basic_stats,
"freq": word_freq,
"sentiment": sentiment,
"chars": characters,
"relations": relations
}
for name in analysis_funcs:
cached = CACHE.get
(self.current_hash, name
)
if cached is not None:
results[name] = cached
else:
res = analysis_funcs[name]
(text
)
CACHE.
set(self.current_hash, name, res
)
results[name] = res
self.root.after
(0, lambda: self._update_ui
(results
))
def _update_ui
(self, results
):
"""主线程更新 UI"""
if self.spinner:
self.spinner.stop
()
self.spinner.destroy
()
self.spinner = None
for name, data in results.items
():
tab = self.tabs[name]
for w in tab.winfo_children
():
w.destroy
()
getattr
(ResultRenderer, f"render_{name}"
)(tab, data, self
)
self.status.config
(text="🎉 分析完成！切换标签页查看结果", foreground="green"
)
self.analyze_btn.config
(state='normal'
)
self.export_btn.config
(state='normal'
)
def export_report
(self
):
raw = self.text_area.get
(1.0, tk.END
)
text = clean_markup
(raw
)
if not validate_input
(text
) or not hasattr
(self, 'current_hash'
):
messagebox.showwarning
("导出失败", "请先完成有效分析。"
)
return
path = filedialog.asksaveasfilename
(
defaultextension=".md",
filetypes=[
("Markdown 文件", "*.md"
),
("所有文件", "*.*"
)]
)
if not path: return
try:
md = f"# 《乱世佳人》文本分析报告\n\n> 基于 {len
(text
)} 字符的英文文本\n\n---\n"
for a in ANALYZERS:
data = CACHE.get
(self.current_hash, a["name"]
)
if data is not None:
summary = generate_summary
(a["name"], data
)
summary_clean = summary.replace
("**", ""
)
md += f"\n## {a['title']}\n\n{summary_clean}\n"
with open
(path, 'w', encoding='utf-8'
) as f:
f.write
(md
)
messagebox.showinfo
("导出成功", f"报告已保存至：\n{os.path.abspath
(path
)}"
)
except Exception as e:
messagebox.showerror
("导出失败", f"保存时出错：\n{str
(e
)}"
)
# === 可视化方法 ===
def show_wordcloud
(self, counter
):
try:
wc = WordCloud
(width=800, height=400, background_color='white'
).generate_from_frequencies
(counter
)
plt.figure
(figsize=
(10, 5
))
plt.imshow
(wc, interpolation='bilinear'
)
plt.axis
("off"
)
plt.title
("高频词云图", fontsize=14
)
plt.tight_layout
()
plt.show
()
except Exception as e:
messagebox.showerror
("错误", f"词云生成失败：{e}"
)
def show_relations
(self, pairs
):
G = nx.Graph
()
for
(a, b
), w in pairs.items
():
if w >= 2:
G.add_edge
(a.capitalize
(), b.capitalize
(), weight=w
)
if not G.nodes
():
messagebox.showinfo
("提示", "无足够共现数据（需 ≥2 次）"
)
return
plt.figure
(figsize=
(10, 8
))
pos = nx.spring_layout
(G, k=3, iterations=100
)
nx.draw_networkx_nodes
(G, pos, node_size=1200, node_color='lightcoral'
)
nx.draw_networkx_labels
(G, pos, font_size=11, font_weight='bold'
)
weights = [d['weight'] for u, v, d in G.edges
(data=True
)]
nx.draw_networkx_edges
(G, pos, width=[w*1.2 for w in weights], alpha=0.6
)
plt.title
("人物共现关系网络（阈值：≥2 次）", fontsize=14
)
plt.axis
('off'
)
plt.tight_layout
()
plt.show
()
# === 启动入口 ===
if __name__ == "__main__":
root = ThemedTk
(theme="azure"
)
App
(root
)
root.mainloop
()
这个打分
扫一扫
770
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
