nikola_check

#!/usr/bin/env python
import os
import sys
import urllib
from urlparse import urlparse

import lxml.html

existing_targets = set([])

def analize(task):
    try:
        filename = task.split(":")[-1]
        d = lxml.html.fromstring(open(filename).read())
        for l in d.iterlinks():
            target = l[0].attrib[l[1]]
            if target == "#":
                continue
            parsed = urlparse(target)
            if parsed.scheme:
                continue
            if parsed.fragment:
                target = target.split('#')[0]
            target_filename = os.path.abspath(os.path.join(os.path.dirname(filename), urllib.unquote(target)))
            if target_filename not in existing_targets:
                if os.path.exists(target_filename):
                    existing_targets.add(target_filename)
                else:
                    print "In %s broken link: " % filename, target
                    if '--find-sources' in sys.argv:
                        print "Possible sources:"
                        print os.popen('doit list --deps %s' % task, 'r').read()
                        print "===============================\n"

    except Exception as exc:
        print "Error with:", filename, exc

def scan_links():
    for task in os.popen('doit list --all', 'r').readlines():
        task = task.strip()
        if task.split(':')[0] in (
            'render_tags',
            'render_archive',
            'render_galleries',
            'render_indexes',
            'render_pages',
            'render_site') and '.html' in task:
            analize(task)

def scan_files():
    task_fnames = set([])
    real_fnames = set([])
    # First check that all targets are generated in the right places
    for task in os.popen('doit list --all', 'r').readlines():
        task = task.strip()
        if 'output' in task and ':' in task:
            fname = task.split(':')[-1]
            task_fnames.add(fname)
     # And now check that there are no non-target files
    for root, dirs, files in os.walk('output'):
        for src_name in files:
            fname = os.path.join(root, src_name)
            real_fnames.add(fname)

    only_on_output = list(real_fnames - task_fnames)
    if only_on_output:
        only_on_output.sort()
        print "\nFiles from unknown origins:\n"
        for f in only_on_output:
            print f

    only_on_input = list(task_fnames - real_fnames)
    if only_on_input:
        only_on_input.sort()
        print "\nFiles not generated:\n"
        for f in only_on_input:
            print f


if __name__ == '__main__':
    if '--help' in sys.argv or len(sys.argv) == 1:
        print "Usage: nikola_check [--check-links [--find-sources]] [--check-files]"
        sys.exit()
    elif '--check-links' in sys.argv:
        scan_links()
    elif '--check-files' in sys.argv:
        scan_files()