barrett/cp.nvim
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

fix(ci): regex

Browse source
This commit is contained in:
Barrett Ruth 2025-10-23 22:24:39 -04:00
parent ce975d0f1e
commit a9ac06de83
1 changed files with 6 additions and 6 deletions
Download patch file Download diff file Expand all files Collapse all files

12
scrapers/cses.py
View file

@ -72,7 +72,7 @@ async def fetch_text(client: httpx.AsyncClient, path: str) -> str:
CATEGORY_BLOCK_RE = re.compile( CATEGORY_BLOCK_RE = re.compile(
r'<h2>(?P<cat>[^<]+)</h2>\s*<ul class="task-list">(?P<body>.*?)</ul>', r'<h2>(?P<cat>[^<]+)</h2>\s*<ul\s+class="task-list">(?P<body>.*?)</ul>',
re.DOTALL, re.DOTALL,
) )
TASK_LINK_RE = re.compile( TASK_LINK_RE = re.compile(
@ -81,15 +81,15 @@ TASK_LINK_RE = re.compile(
) )
TITLE_RE = re.compile( TITLE_RE = re.compile(
r'<div class="title-block">.*?<h1>(?P<title>[^<]+)</h1>', re.DOTALL r'<div\s+class="title-block">.*?<h1>(?P<title>[^<]+)</h1>', re.DOTALL
) )
TIME_RE = re.compile(r"<li><b>Time limit:</b>\s*([0-9.]+)\s*s</li>") TIME_RE = re.compile(r"<li>\s*<b>Time limit:</b>\s*([0-9.]+)\s*s\s*</li>")
MEM_RE = re.compile(r"<li><b>Memory limit:</b>\s*(\d+)\s*MB</li>") MEM_RE = re.compile(r"<li>\s*<b>Memory limit:</b>\s*(\d+)\s*MB\s*</li>")
SIDEBAR_CAT_RE = re.compile( SIDEBAR_CAT_RE = re.compile(
r'<div class="nav sidebar">.*?<h4>(?P<cat>[^<]+)</h4>', re.DOTALL r'<div\s+class="nav sidebar">.*?<h4>(?P<cat>[^<]+)</h4>', re.DOTALL
) )
MD_BLOCK_RE = re.compile(r'<div class="md">(.*?)</div>', re.DOTALL | re.IGNORECASE) MD_BLOCK_RE = re.compile(r'<div\s+class="md">(.*?)</div>', re.DOTALL | re.IGNORECASE)
EXAMPLE_SECTION_RE = re.compile( EXAMPLE_SECTION_RE = re.compile(
r"<h[1-6][^>]*>\s*example[s]?:?\s*</h[1-6]>\s*(?P<section>.*?)(?=<h[1-6][^>]*>|$)", r"<h[1-6][^>]*>\s*example[s]?:?\s*</h[1-6]>\s*(?P<section>.*?)(?=<h[1-6][^>]*>|$)",
re.DOTALL | re.IGNORECASE, re.DOTALL | re.IGNORECASE,