Download helpers

from IPython.display import Markdown,HTML
from fastcore.test import *

---

source

clean_md

 clean_md (text, rm_comments=True, rm_details=True)

Remove comments and <details> sections from text

---

source

read_md

 read_md (url, rm_comments=True, rm_details=True,
          params:QueryParamTypes|None=None, headers:HeaderTypes|None=None,
          cookies:CookieTypes|None=None, auth:AuthTypes|None=None,
          proxy:ProxyTypes|None=None, follow_redirects:bool=False,
          verify:ssl.SSLContext|str|bool=True,
          timeout:TimeoutTypes=Timeout(timeout=5.0), trust_env:bool=True)

Read text from url and clean with clean_docs

mdurl = 'https://claudette.answer.ai/index.html.md'
md = read_md(mdurl)
# Markdown(md)

---

source

html2md

 html2md (s:str, ignore_links=True)

Convert s from HTML to markdown

---

source

read_html

 read_html (url, sel=None, rm_comments=True, rm_details=True, multi=False,
            wrap_tag=None, ignore_links=True)

Get url, optionally selecting CSS selector sel, and convert to clean markdown

	Type	Default	Details
url			URL to read
sel	NoneType	None	Read only outerHTML of CSS selector sel
rm_comments	bool	True	Removes HTML comments
rm_details	bool	True	Removes <details> tags
multi	bool	False	Get all matches to sel or first one
wrap_tag	NoneType	None	If multi, each selection wrapped with content
ignore_links	bool	True

# test single class selector
listings = read_html('https://www.answer.ai/', sel='.listing-description')
assert len(listings) < 500

# Test multi class selector
listings = read_html('https://www.answer.ai/', sel='.listing-description', multi=True)
assert len(listings) > 1000 # returns more than single so selecting multi

# Test multi_wrap_tag
listings = read_html('https://www.answer.ai/', sel='.listing-description', multi=True, wrap_tag='document')
assert '<document>' in listings and '</document>' in listings

read_html('https://www.answer.ai/', sel='.listing-description', ignore_links=False)

'[My experience learning GPU programming, and implementing a new GPU education app in the process](./posts/2025-03-17-gpu-programming-scratch.html)\n\n'

# test tag css selectors
assert len(read_html('https://www.answer.ai/', sel='div.listing-description', multi=True)) > 1000
assert len(read_html('https://www.answer.ai/', sel='div', multi=True)) > 1000

htmlurl = 'https://hypermedia.systems/hypermedia-a-reintroduction/'
hmd = read_html(htmlurl)
assert len(hmd) > 100
# Markdown(hmd)

---

source

get_llmstxt

 get_llmstxt (url, optional=False, n_workers=None)

Get llms.txt file from and expand it with llms_txt.create_ctx()

# print(get_llmstxt('https://llmstxt.org/llms.txt'))

---

source

split_url

 split_url (url)

Split url into base, path, and file name, normalising name to ‘/’ if empty

urls = ('https://claudette.answer.ai/path/', 'https://claudette.answer.ai/', 'https://llmstxt.org', 'https://llmstxt.org/')

[split_url(o) for o in urls]

[('https://claudette.answer.ai', '', '/path'),
 ('https://claudette.answer.ai', '/', ''),
 ('https://llmstxt.org', '/', ''),
 ('https://llmstxt.org', '/', '')]

---

source

find_docs

 find_docs (url)

If available, return LLM-friendly llms.txt context or markdown file location from url

fl_url = 'https://answerdotai.github.io/fastlite'

find_docs(fl_url)

'https://answerdotai.github.io/fastlite/llms.txt'

for o in urls: print(find_docs(o))

https://claudette.answer.ai/llms.txt
https://claudette.answer.ai/llms.txt
https://llmstxt.org/llms.txt
https://llmstxt.org/llms.txt

suffixes = ["/", "/tmp", "/tmp/tmp/"]
for suff in suffixes:
    for o in urls:  test_eq(find_docs(o), find_docs(o+suff))

test_eq(find_docs("https://github.com"), "https://github.com/llms.txt")
test_eq(find_docs("https://github.com/AnswerDotAI"), "https://github.com/llms.txt")
test_eq(find_docs("https://github.com/AnswerDotAI/"), "https://github.com/llms.txt")

---

source

read_docs

 read_docs (url, optional=False, n_workers=None, rm_comments=True,
            rm_details=True)

If available, return LLM-friendly llms.txt context or markdown file response for url