KaraKaraWitch commited on
Commit
f1322a4
·
verified ·
1 Parent(s): cb2f5bb

Upload scripts/MDNClean.py with huggingface_hub

Browse files
Files changed (1) hide show
  1. scripts/MDNClean.py +111 -0
scripts/MDNClean.py ADDED
@@ -0,0 +1,111 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import asyncio
2
+ import pathlib
3
+ import re
4
+ import markdownify
5
+ from markdownify import chomp
6
+ import urllib.parse
7
+ import orjson
8
+ import tqdm
9
+ import typer
10
+ import httpx
11
+ from bs4 import BeautifulSoup, Tag
12
+
13
+ app = typer.Typer()
14
+
15
+ class MDNConverter(markdownify.MarkdownConverter):
16
+ def convert_a(self, el, text, convert_as_inline):
17
+ prefix, suffix, text = chomp(text)
18
+ if not text:
19
+ return ""
20
+ return "%s%s%s" % (prefix, text, suffix)
21
+
22
+ def convert_pre(self, el:Tag, text, convert_as_inline):
23
+ if not text:
24
+ return ''
25
+
26
+ code_language = self.options['code_language']
27
+ #
28
+ if el.previous_sibling and el.previous_sibling.name and el.previous_sibling.name == "div":
29
+ if "example-header" in el.previous_sibling.get("class",[]):
30
+ code_language = el.previous_sibling.find('span').get_text()
31
+ el.previous_sibling.decompose()
32
+
33
+ # if self.options['code_language_callback']:
34
+ # code_language = self.options['code_language_callback'](el) or code_language
35
+
36
+ return '\n```%s\n%s\n```\n' % (code_language, text)
37
+
38
+ inst = MDNConverter()
39
+
40
+ rgx = re.compile(r"(.+)\n```\1")
41
+
42
+ def quote_fx(matchobj):
43
+ return f"\n```{matchobj.group(1)}"
44
+
45
+ def soup_processor(soup:BeautifulSoup):
46
+ base_article = soup.select_one("main#content > article")
47
+ [script.decompose() for script in base_article.find_all("script")]
48
+ if not base_article:
49
+ return None
50
+ mod_time = base_article.find("aside",attrs={"class":"metadata"})
51
+ if mod_time:
52
+ mod_time.decompose()
53
+ hydro = base_article.select_one("#hydration")
54
+ if hydro:
55
+ hydro.decompose()
56
+
57
+ r = inst.convert_soup(base_article).replace("\n\n\n\n","\n\n").replace("\n\n\n","\n\n")
58
+ r = rgx.sub(quote_fx, r).replace("\n \n","\n\n").replace("\n\n```\n","\n```\n\n").rstrip().replace("\t"," ")
59
+
60
+ data = {"title":soup.find("title").get_text().split("|")[0].strip(),"text":r}
61
+ return data
62
+
63
+
64
+
65
+
66
+ @app.command()
67
+ def clean(root:pathlib.Path, output_jsonl:pathlib.Path):
68
+ with open(output_jsonl.resolve(),"wb") as f:
69
+ for file in root.rglob("*.html"):
70
+ if file.is_file() and "docs/MDN/" not in str(file):
71
+ soup = BeautifulSoup(file.read_text(encoding="utf-8"),"lxml")
72
+ print(file)
73
+ r = soup_processor(soup)
74
+ f.write(orjson.dumps(r)+b"\n")
75
+
76
+
77
+
78
+ async def async_executor(urls:list[str], root:pathlib.Path):
79
+ sem = asyncio.Semaphore(64)
80
+ client = httpx.AsyncClient(timeout=None)
81
+ client.headers["user-agent"] = "curl/7.54.1 (compatible; kinda like curl) " + client.headers["user-agent"] + " mdncrawl/1.0.0 (SiteMap Crawling; +shinon@recursal.ai; Contact if bot errors.)"
82
+ pbar = tqdm.tqdm()
83
+ async def url_executor(url:str):
84
+ pth = urllib.parse.urlparse(url).path[1:]
85
+ url_write = (root / pth).resolve()
86
+ # print(url_write)
87
+ url_write.parent.mkdir(exist_ok=True,parents=True)
88
+ async with sem:
89
+ response = await client.get(url)
90
+ if response.status_code == 200:
91
+ if url_write.suffix != ".html":
92
+ url_write = url_write.with_suffix(f"{url_write.suffix}.html")
93
+ url_write.write_bytes(response.content)
94
+ pbar.update(1)
95
+ else:
96
+ print(response.status_code)
97
+ loop = asyncio.get_running_loop()
98
+ workers = [loop.create_task(url_executor(url)) for url in urls]
99
+ await asyncio.gather(*workers)
100
+ pass
101
+
102
+ @app.command()
103
+ def sitemap(xml_file:pathlib.Path, root:pathlib.Path):
104
+ root = root.resolve()
105
+ soup = BeautifulSoup(xml_file.read_text(encoding="utf-8"), "xml")
106
+ urls = [i.get_text() for i in soup.find_all("loc")]
107
+ # print(urls)
108
+ asyncio.run(async_executor(urls, root))
109
+
110
+ if __name__ == "__main__":
111
+ app()