cssstylelist.py (HTML safe)

#! /usr/bin/env python3
# file: cssstylelist.py
# Make a list of every style used in html and returns that
# From the CLI, type `./cssstylelist.py > file_with_a_list.txt
# Feel free to address any complain to @gabalese

import os, glob, sys
try:
    from lxml import etree as ET
except ImportError:
    import xml.etree.ElementTree as ET
    print("Failed to import lxml: running with standard xml.etree instead.")

path = "OEBPS/Text" # your mileage may vary
list = []
new_list = []

def cssList():
    global list
    global new_list
    parser = ET.XMLParser(recover=True) # this makes the XML parsing "tolerant" of ill-formedness
    for infile in glob.glob(os.path.join(path, '*html')):
        try:
            html = ET.parse(infile, parser).getroot()
        except:
            print("ERROR: Unable to parse " + infile)
            print("This is likely to happen with ill-formed xhtml files.")
            sys.exit(1)
        for i in html.iter():
            list.append(i.get("class"))

    for i in list:
        if i not in new_list:
            if i is not None:
                new_list.append(i)

    return new_list



if __name__ == "__main__":
    for item in (cssList()):
        print(item)