rdfexamplechecker.py

# -*- coding: UTF-8 -*-
import urllib2
from lxml import etree
from lxml.cssselect import CSSSelector as css
from rdflib import Graph


def check_examples(url, verbose=False):
    for title, prologue, text, fmt in find_examples(url):
        header = 'Example "%s" as %s: ' % (title, fmt)
        fill = 72 if verbose else 40
        try:
            g = Graph().parse(data=prologue+text, format=fmt)
        except Exception, e:
            print header,
            print "Error:"
            print "-"*fill
            print unicode(e).encode('utf-8')
            if verbose:
                print "-"*fill
                print text
            print "="*fill
            print
        else:
            if verbose:
                print header,
                print "Ok!"
                print "="*fill
                print


def find_examples(url):
    parser = etree.HTMLParser()
    doc = etree.parse(urllib2.urlopen(url), parser)
    prefixes = []
    for i, elem in enumerate(css("pre")(doc)):
        title = i+1
        rows = [txt for txt in elem.xpath('text()')]
        prefixes += [row for row in rows if "@prefix" in row and row not in prefixes]
        text = "\n".join(rows).encode('utf-8')
        text = text.replace(' ', ' ') # nbsp
        guessed_fmt = 'xml' if ("xmlns:" in text or "rdf:about" in text) else 'n3'
        if guessed_fmt == 'n3':
            prologue = '\n'.join(pfx for pfx in prefixes if pfx not in text) + '\n'
        else:
            prologue = ''
        yield title, prologue, text, guessed_fmt


if __name__ == '__main__':
    from sys import argv
    args = argv[1:]
    verbose = '-v' in args
    if verbose: args.remove('-v')
    if args:
        url = args.pop()
        check_examples(url, verbose)