"""Regression checks for generated search anchors and asset collection.""" import re import json import sys import tempfile import unittest from collections import Counter from pathlib import Path sys.path.insert(0, str(Path(__file__).resolve().parent)) from site_assets import add_image_sizes, heading_slug, optimize_images, referenced_paths, search_document from build import github_slug, hot_spans, strip_hot class SearchIndexTests(unittest.TestCase): def test_hot_heading_mark(self): # 标题末尾的“不换行空格 + 🔥”不改变 GitHub 锚点;网站上换成空的 span,Docsify 的 id 也不变 self.assertEqual(github_slug('ThreadPoolExecutor\u00a0🔥🔥'), 'threadpoolexecutor') self.assertEqual(github_slug('5. 协程\u00a0🔥'), '5-协程') self.assertEqual(github_slug('ThreadPoolExecutor 🔥'), 'threadpoolexecutor-') self.assertEqual(strip_hot('### 5. 协程\u00a0🔥\n正文 🔥'), '### 5. 协程\n正文 🔥') site = hot_spans('* [三、检索增强生成](#三检索增强生成) 🔥🔥\n## 三、检索增强生成\u00a0🔥🔥') self.assertEqual(site, '* [三、检索增强生成](#三检索增强生成) \n' '## 三、检索增强生成') self.assertEqual(heading_slug(site.split('\n')[1][3:], Counter()), '三、检索增强生成') doc = search_document(Path('x.md'), site) self.assertEqual([s['title'] for s in doc['sections']], ['x', '三、检索增强生成']) self.assertEqual(doc['sections'][0]['text'], '') def test_docsify_numeric_duplicate_and_chinese_punctuation(self): counts = Counter() self.assertEqual(heading_slug('1. 慢开始与拥塞避免', counts), '_1-慢开始与拥塞避免') self.assertEqual(heading_slug('三、检索增强生成', counts), '三、检索增强生成') self.assertEqual(heading_slug('TCP 拥塞控制', counts), 'tcp-拥塞控制') self.assertEqual(heading_slug('TCP 拥塞控制', counts), 'tcp-拥塞控制-1') # marked 先把引号转义成 ",Docsify 再去掉标点,留下 quot(Edge 渲染的 Java 基础页核对过) self.assertEqual(heading_slug('new String("abc")', counts), 'new-stringquotabcquot') self.assertEqual(heading_slug('1\\. 数据', counts), '_1-数据') def test_headings_in_code_do_not_change_real_anchors(self): doc = search_document(Path('示例.md'), '# 示例\n\n```python\n# TCP 拥塞控制\n```\n\n## TCP 拥塞控制\n慢开始\n') sections = [s for s in doc['sections'] if s['id'] == 'tcp-拥塞控制'] self.assertEqual(len(sections), 1) self.assertEqual(sections[0]['text'], '慢开始') self.assertEqual(sections[0]['level'], 2) def test_comments_and_example_asset_urls_are_not_published(self): content = '\n```html\n\n```\n\n[正文](/notes/示例.md)\n' self.assertEqual(set(referenced_paths(content)), {'notes/pics/actual.png', 'notes/示例.md'}) def test_inline_triple_backticks_do_not_hide_later_images(self): # 正文中的行内代码 ```` ```json ```` 不是代码块,后面的图片仍要发布 content = ('例如包在 ```` ```json ```` 代码块里。\n\n\n\n' '- 列表\n\n ```text\n \n ```\n\n\n') self.assertEqual(set(referenced_paths(content)), {'notes/pics/a.png', 'notes/pics/b.png'}) def test_image_manifest_stable_across_file_creation_order(self): from PIL import Image with tempfile.TemporaryDirectory() as directory: root = Path(directory) outputs = [] for index, names in enumerate([['a.png', 'z.png'], ['z.png', 'a.png']]): dist = root / str(index) dist.mkdir() for name in names: Image.new('RGB', (128, 128), 'red' if name == 'a.png' else 'blue').save( dist / name, compress_level=0) report = optimize_images(dist, root / 'cache') self.assertEqual(report['converted'], 2) outputs.append((dist / 'image-variants.json').read_bytes()) self.assertEqual(outputs[0], outputs[1]) def test_image_sizes_keep_aspect_ratio(self): from PIL import Image with tempfile.TemporaryDirectory() as directory: dist = Path(directory) (dist / 'pics').mkdir() Image.new('RGB', (1520, 600)).save(dist / 'pics' / 'a.png') (dist / 'a.md').write_text('\n\n' '\n\n' '```html\n\n```\n') self.assertEqual(add_image_sizes(dist), 2) self.assertEqual((dist / 'a.md').read_text(), '\n\n' '\n\n' '```html\n\n```\n') def test_static_note_pages(self): # 静态页面的标题 id 与网站(Docsify)一致,站内链接换成静态页面,文首目录折叠,只剩一个链接的旧页面跳到目标 import collections, subprocess with tempfile.TemporaryDirectory() as directory: dist = Path(directory) (dist / 'notes').mkdir() titles = ['示例', '1. 爬楼梯', 'HashMap 的 `put()`', '"引号"', '示例', 'A1 实测'] toc = '\n'.join(f'* [{t}](#x{i})' for i, t in enumerate(titles)) body = f'# {titles[0]}\n\n{toc}\n\n\n正文第一段,见 [另一篇](/notes/%E5%8F%A6.md#_1-%E7%88%AC%E6%A5%BC%E6%A2%AF)。\n\n' + \ ''.join(f'## {t}\n\n内容\n\n' for t in titles[1:]) + 'HTML 链接\n\n```java\n# 不是标题\n```\n' (dist / 'notes' / '示例.md').write_text(body) (dist / 'notes' / '算法.md').write_text('见 [算法 - 目录](/notes/算法 - 目录.md)\n') (dist / '_sidebar.md').write_text('- [首页](/)\n\n- **示例分组**\n - [示例](/notes/%E7%A4%BA%E4%BE%8B.md)\n') (dist / 'stubs.json').write_text(json.dumps({'/notes/算法': '/notes/算法 - 目录'}, ensure_ascii=False)) (dist / 'sections.json').write_text(json.dumps({'示例': {'_1-爬楼梯': [['q1', '怎样爬楼梯?', 2]]}}, ensure_ascii=False)) urls = json.loads(subprocess.run(['node', str(Path(__file__).parent / 'static_notes.cjs'), str(dist), 'https://example.com/', str(dist / 'stubs.json'), str(dist / 'sections.json')], check=True, capture_output=True, text=True).stdout) self.assertEqual(urls, ['https://example.com/notes/', 'https://example.com/notes/%E7%A4%BA%E4%BE%8B.html']) self.assertIn('示例', (dist / 'notes' / 'index.html').read_text()) page = (dist / 'notes' / '示例.html').read_text() counts = collections.Counter() expected = [heading_slug(t, counts) for t in titles] self.assertEqual(re.findall(r'本文目录', page) self.assertIn('', page) self.assertIn('href="/#/notes/%E7%A4%BA%E4%BE%8B"', page) self.assertNotIn('1. 爬楼梯\n