fast-soup 1.1.0

BeautifulSoup interface for lxml

FastSoup

---

BeautifulSoup interface for lxml

Key features

• FAST search in tree

• FAST serialize to str

• BeautifulSoup4 interface to interact with object:

  • Search: find, find_all, find_next, find_next_sibling

  • Text: get_text, string

  • Tag: name, get, clear, __getitem__, __str__, __repr__, append, new_tag, extract, replace_with

Install

pip install fast-soup==1.1.0

How to use

from fast_soup import FastSoup

content = ...  # read some html content
soup = FastSoup(content)

# interact like BS4 object
result = soup.find('a', id='my_link')

# interact like lxml object
el = result.unwrap()

FAQ

Q: BS4 already implement lxml parser. Why i should use FastSoup?

A: Yes, BS4 implement parser, and it’s just building the tree. All next interactions proceed with “Python speed”: searching, serialization. FastSoup internally use lxml and guarantee “C speed”.

Q: How FastSoup speedup works?

A: FastSoup just build xpath and execute them. For prevent rebuilding LRU cache used.

Q: Why you don’t support whole interface? This will be soon?

A: I wrote functions which speed up parsing in my projects. Just create a issue or pull request and i think we find the solution ;)

Miscellaneous

You can got power of BeautifulSoup when wrap your lxml objects, e.g:

from fast_soup import Tag

content = ...  # some bytes ready to parse
context = lxml.etree.iterparse(
    io.BytesIO(content),  ...
)
for event, elem in context:
    tag = Tag(elem)

    tag_text = tag.get_text()
    tag_attr = tag['attribute']