#!/usr/bin/env python2.5
# -*- encoding: utf-8 -*-
#
# MuninStatus.py: récupère la page munin et la filtre pour ne garder que les liens intéressants
# 
# Copyright (c) 2009, Nicolas Dandrimont <Nicolas.Dandrimont@crans.org>
#
# This file is released under the GNU General Public License version 2
#

import sys
import urllib

try:
    import BeautifulSoup
except ImportError:
    print "munin.py nécessite BeautifulSoup (paquet python-beautifulsoup)"
    sys.exit(2)

URL = "http://munin.crans.org/"
ENCODING = "utf-8"

ATTRS = {
            'class': ('crit', 'warn'),
        }

def to_unicode(input, encodings=None):
    if not encodings:
        encodings = ['utf-8', 'iso-8859-15']

    for encoding in encodings:
        try:
            return input.decode(encoding)
        except UnicodeDecodeError:
            pass

    return input.decode(encodings[-1], 'ignore')

def family(nodes):
    parents = sum((node.findParents() for node in nodes), [])
    children = sum((node.findChildren() for node in nodes), [])
    return nodes + parents + children

def keep_item_headings(headings, items, level = 1):
    kept = []

    def level_up(heading):
        parent = heading
        for _ in range(level):
            parent = parent.parent
        return parent

    for heading in headings:
        parent = level_up(heading)

        found_parent = False
        for parents in [item.findParents() for item in items]:
            if parent in parents:
                found_parent = True
                break
        
        if found_parent:
            kept.append(heading)

    return kept

def execute(macro, _):
    munin = BeautifulSoup.BeautifulSoup(urllib.urlopen(URL).read(), fromEncoding = ENCODING)

    warning_items = munin.findAll(attrs=ATTRS)
    warning_hosts = keep_item_headings(munin.findAll(attrs = "host"), warning_items)
    warning_domains = keep_item_headings(munin.findAll(attrs = "domain"), warning_items)

    keep = family(warning_items + warning_hosts + warning_domains)

    for item in munin.findChildren():
        if item not in keep:
            item.extract()

    for tag in munin.findAll(href=True):
        tag["href"] = u'http://munin.crans.org/' + tag["href"]
    
    if munin.html and munin.html.body:
        return to_unicode(munin.html.body.prettify())
    else:
        return u"Munin n'a rien à dire."