A python based HTML to text converter with minimal support for CSS.

• Python 3.4+ (preferred) or Python 2.7+
• lxml

The command line client converts text files or text retrieved from Web pages to the corresponding text representation.

Installation

sudo python3 setup.py install

Command line parameters

usage: inscript.py [-h] [-o OUTPUT] [-e ENCODING] input

Converts HTML from file or url to a clean text version

positional arguments:
  input                 Html input either from a file or an url

optional arguments:
  -h, --help            show this help message and exit
  -o OUTPUT, --output OUTPUT
                        Output file (default:stdout).
  -e ENCODING, --encoding ENCODING
                        Content encoding for files (default:utf-8)
  -i, --image-captions  Display image captions (default:false).
  -d, --deduplicate-image-captions
                        Deduplicate image captions (default:false).

Examples

# convert the given page to text and output the result to the screen
python3 inscriptis.py http://www.htwchur.ch

# convert the file to text and save the output to output.txt
python3 inscriptis.py htwchur.html -o htwchur.txt

import urllib.request
from inscriptis import get_text

url = "http://www.informationscience.ch"
html = urllib.request.urlopen(url).read()

text = get_text(html)

print(text)

Test cases concerning the html to text conversion are located in the tests/html directory and consist of two files:

1. test-name.html and
2. test-name.txt

the latter one containing the reference text output for the given html file.

inscriptis offers a small benchmarking script that can compare different HTML to txt convertion approaches. The script will run the different approaches on a list of URLs, url_list.txt, and save the text output into a time stampped folder in benchmarking/benchmarking_results for manual comparison. Additionally the processing speed of every approach per URL is measured and saved in a text file called speed_comparisons.txt in the respective time stampped folder.

To run the benchmarking script execute run_benchmarking.py from within the folder benchmarking. In def pipeline() set the which HTML -> Text algorithms to be executed by modifying

run_lynx = True
run_justext = True
run_html2text = True
run_beautifulsoup = True
run_inscriptis = True

In url_list.txt the URLs to be parsed can be specified by adding them to the file, one per line with no additional formatting. URLs need to be complete (including http:// or https://) e.g.

http://www.informationscience.ch
https://en.wikipedia.org/wiki/Information_science
...