Files

130 lines
5.2 KiB
Python
Raw Permalink Normal View History

# -*- coding: utf-8 -*-
"""线性代数各讲:难度分布 + 多问结构 + 文件体积 自检。
用法:
python stat_difficulty.py # 扫描全部线代在学练习讲
python stat_difficulty.py <讲目录> ... # 只看指定讲
python stat_difficulty.py <单个.md> # 单文件明细
约定(2026-10-25 定):
题目标题 = `### 题 N · X.Y:短标题(★☆☆|★★☆|★★★)`
道题标签取各小问中的最高难度;(Ⅰ)应当永远是 ★☆☆ 的入口问。
单个 .md 不超过 13000 字符,超了就要拆。
"""
import os
import re
import sys
ROOT = os.path.abspath(os.path.join(os.path.dirname(__file__), '..', '..'))
BASE = os.path.join(ROOT, '基于考研的学习', '数学一', '在学练习')
LIMIT = 13000
STAR = {'★☆☆': 1, '★★☆': 2, '★★★': 3}
INV = {1: '★☆☆', 2: '★★☆', 3: '★★★'}
def iter_point_files(target):
"""返回 [(目录名, [点文件路径])]"""
if os.path.isfile(target):
yield (os.path.basename(os.path.dirname(target)), [target])
return
for name in sorted(os.listdir(target)):
p = os.path.join(target, name)
if os.path.isfile(p) and re.match(r'^\d\d[^\d].*\.md$', name) and '速查' not in name and '盘点' not in name:
yield (os.path.basename(target.rstrip('/\\')), [p])
break_ = False
# 目录情形:收集所有 02..09 点文件
if os.path.isdir(target):
files = [os.path.join(target, f) for f in sorted(os.listdir(target))
if re.match(r'^\d\d[^\d].*\.md$', f) and '速查' not in f and '盘点' not in f]
yield (os.path.basename(target.rstrip('/\\')), files)
def analyze(path):
"""返回 (题数, {星级:数量}, [每题问数], 字符数, [非★☆☆起手的标题])"""
t = open(path, encoding='utf-8').read()
cnt = {1: 0, 2: 0, 3: 0}
asks = []
titles = []
for blk in re.split(r'^### 题 ', t, flags=re.M)[1:]:
head = blk.split('\n')[0]
titles.append(head)
# 只统计「题干」中的小问:题目标题之后、"怎么想/解"之前的部分。
# 解析里往往会重复出现(Ⅰ)(Ⅱ)等标号,计入会严重高估。
stem = re.split(r'\n\*\*怎么想\*\*|\n\*\*解\*\*|\n> 考法来源', blk)[0]
body = blk.split('\n---\n')[0]
# 题干在未出现"怎么想"时就是整段 body 的前半:退回到 body
cut = min(len(stem), len(body))
seg = stem[:cut] if stem.strip() else body
n = len(re.findall(r'([ⅠⅡⅢ])|\([123]\)', seg))
asks.append(n)
for k, s in (('★★★', 3), ('★★☆', 2), ('★☆☆', 1)):
if k in head:
cnt[s] += 1
break
return (len(titles), cnt, asks, len(t), titles)
def report(lecture, files):
tot = {1: 0, 2: 0, 3: 0}
allasks = []
over = []
print('=' * 62)
print(lecture)
print('=' * 62)
for f in files:
n, cnt, asks, chars, titles = analyze(f)
if not n:
continue
allasks += asks
for k in tot:
tot[k] += cnt[k]
flag = ' ⚠ 超 %d' % LIMIT if chars > LIMIT else ''
multi = sum(1 for a in asks if a >= 1)
print(' %-42s %2d题 %2d字符 多问%2d/%2d%s'
% (os.path.basename(f)[:42], n, chars, multi, n, flag))
if chars > LIMIT:
over.append(os.path.basename(f))
for h, a in zip(titles, asks):
if a == 0:
print(' · 单问题: %s' % h[:52])
total = sum(tot.values())
if not total:
print(' (无题目)')
return
pct = lambda k: round(100 * tot[k] / total)
print(' ' + '-' * 58)
print(' 合计 %d 题 | ★☆☆ %d (%d%%) ★★☆ %d (%d%%) ★★★ %d (%d%%)'
% (total, tot[1], pct(1), tot[2], pct(2), tot[3], pct(3)))
if allasks:
print(' 每道题小问数: 平均 %.1f | 单问 %d 题, 2问 %d 题, 3问+ %d 题'
% (sum(allasks) / len(allasks),
sum(1 for a in allasks if a == 0),
sum(1 for a in allasks if a == 1),
sum(1 for a in allasks if a >= 2)))
if over:
print(' ⚠ 超 13000 字符需拆分: ' + ', '.join(over))
print()
def main():
args = sys.argv[1:]
if not args:
args = [os.path.join(BASE, d) for d in sorted(os.listdir(BASE))
if d.startswith('线性代数-第') and os.path.isdir(os.path.join(BASE, d))]
for a in args:
target = a if os.path.isfile(a) or os.path.isdir(a) else os.path.join(BASE, a)
if os.path.isfile(target):
n, cnt, asks, chars, titles = analyze(target)
print('%s' % os.path.basename(target))
print(' %d 题, %d 字符%s' % (n, chars, ' ⚠ 超上限' if chars > LIMIT else ''))
for h, q in zip(titles, asks):
print(' %2d问 %s' % (q, h))
continue
files = sorted(os.path.join(target, f) for f in os.listdir(target)
if re.match(r'^\d\d[^\d].*\.md$', f) and '速查' not in f and '盘点' not in f)
report(os.path.basename(target.rstrip('/\\')), files)
if __name__ == '__main__':
main()