Skip to main content

Select and reverse-Markdown (html2text) web page fragments.

Project description


Markdown gives you HTML from plain text and html2text reverses the process. If you want the plain text version of a specific section of a web page (an HTML fragment), you would normally do the selecting (parsing) first, then generate the Markdown text to preserve some formatting.

I made frag2text to:

  • easily select a web page fragment by CSS selector or XPath expression
  • get the plain text of the fragment with some formatting intact for later use with Markdown
  • not shell out to a another program (like lynx -dump)
  • not parse HTML or text directly
  • use html5lib for robust parsing
  • have a simple python module that is easy to maintain (encapsulating the tricky business of subclassing parsers, treebuilders and serializers, by keeping it simple – this shouldn’t require BeautifulSoup)

This is a problem I’ve tinkered with for some time and solved in many different ways. It seems trivial but gets ridiculous quickly. If you have any suggestions or want to share your experiences with other tools, please let me know.


$ pip install frag2text



>>> from frag2text import frag2text
>>> help(frag2text)
Help on function frag2text in module frag2text:

frag2text(endpoint, stype, selector, clean=False, raw=False, verbose=False)
    returns Markdown text of selected fragment.

        endpoint: URL, file, or HTML string
        stype: { 'css' | 'xpath' }
        selector: CSS selector or XPath expression
        Markdown text
        clean: cleans fragment (lxml.html.clean defaults)
        raw: returns raw HTML fragment
        verbose: show http status, encoding, headers


$ -h
usage: [-h] [-c] [-r] [-v] endpoint {css,xpath} selector

reverse Markdown (html2text) HTML fragments.

positional arguments:
  endpoint       URL, file, or HTML string
  {css,xpath}    fragment selector type
  selector       CSS select statement or XPath expression

optional arguments:
  -h, --help     show this help message and exit
  -c, --clean    clean fragment (lxml.html.clean defaults)
  -r, --raw      output raw fragment
  -v, --verbose  print status, encoding, headers



from frag2text import frag2text

info = frag2text('', 'css', '.infobox')


$ "<ht?+><borkt><h1>hello" xpath //h1
# hello

CSS select

$ css .infobox
_[Amanita muscaria](/wiki/Amanita_muscaria)_
Albin Schmalfuß, 1897
[Scientific classification](/wiki/Biological_classification)
Kingdom: | [Fungi](/wiki/Fungi)
Division: | [Basidiomycota](/wiki/Basidiomycota)
Class: | [Agaricomycetes](/wiki/Agaricomycetes)
Order: | [Agaricales](/wiki/Agaricales)
Family: | [Amanitaceae](/wiki/Amanitaceae)
Genus: | _**Amanita**_
[Pers.](/wiki/Christian_Hendrik_Persoon) (1794)
[Type species](/wiki/Type_species)
_[Amanita muscaria](/wiki/Amanita_muscaria)_
([L.](/wiki/Linnaeus)) [Lam.](/wiki/Lam.) (1783)
[c.600 species](/wiki/List_of_Amanita_species)

XPath expression

$ xpath '//p[1]'

The [genus](/wiki/Genus) _**Amanita**_ contains about 600 [species](/wik
i/Species) of [agarics](/wiki/Agarics) including some of the most [toxic
](/wiki/Toxic) known [mushrooms](/wiki/Mushrooms) found worldwide, as we
ll as some well-regarded edible species. This genus is responsible for a
pproximately 95% of the fatalities resulting from [mushroom poisoning](/
wiki/Mushroom_poisoning), with the [death cap](/wiki/Death_cap) accounti
ng for about 50% on its own. The most potent toxin present in these mush
rooms is α[-amanitin](/wiki/%CE%91-amanitin).

Release History

0.0.5 (2015-02-18)

  • handle XPathEvalError, SelectorSyntaxError and Nothing found.
  • join list of selected fragments, was only returning first.
  • do not exit early on error.

0.0.1 (2015-01-14)

  • It seems to work!

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

frag2text-0.0.6.tar.gz (5.9 kB view hashes)

Uploaded source

Built Distribution

frag2text-0.0.6-py2.py3-none-any.whl (8.0 kB view hashes)

Uploaded 2 7

Supported by

AWS AWS Cloud computing Datadog Datadog Monitoring Facebook / Instagram Facebook / Instagram PSF Sponsor Fastly Fastly CDN Google Google Object Storage and Download Analytics Huawei Huawei PSF Sponsor Microsoft Microsoft PSF Sponsor NVIDIA NVIDIA PSF Sponsor Pingdom Pingdom Monitoring Salesforce Salesforce PSF Sponsor Sentry Sentry Error logging StatusPage StatusPage Status page