# $Id$
# get text and xml versions of selected pages

import os, sys
import infogamibot
import urllib2

ET = infogamibot.ET

try:
    site = sys.argv[1]
    if "@" not in site:
        raise IndexError
except IndexError:
    print "usage: infogami-backup user:pass@site [index]"
    sys.exit(1)

try:
    index = sys.argv[2]
except:
    index = "/_special/index"

# use 5 minute caching to force if-modified-since behaviour and
# allow restarts
bot = infogamibot.Bot(site, cache=300)

backup_dir = bot.site
if backup_dir.startswith("http://"):
    backup_dir = backup_dir[7:]
backup_dir = backup_dir.split(".")[0]

if not os.path.isdir(backup_dir):
    os.makedirs(backup_dir)

indexes = bot.get_index(index)
indexes.append((index, "Index"))

for page, title in indexes:

    if page.startswith("http://"):
        continue

    name, e = os.path.splitext(page.lstrip("/"))

    parts = page.split("/")
    if "_comments" in parts or "_special" in parts:
        continue

    # get rendered page
    try:
        result = bot.get_body(page)
    except urllib2.HTTPError, v:
        if v.code == 404:
            continue # deleted pages are included in the main index
        else:
            raise
    html = ET.Element("html")
    head = ET.SubElement(html, "head")
    if result.title is not None:
        head.append(result.title)
    if result.body is not None:
        html.append(result.body)
    try:
        f = open(os.path.join(backup_dir, name + ".xml"), "w")
    except IOError, v:
        print page, "skipped:", v
        continue
    ET.ElementTree(html).write(f)
    f.close()
    print page, "->", f.name, "ok"

    # get source
    result = bot.get_source(page)
    f = open(os.path.join(backup_dir, name + ".txt"), "w")
    f.write((result.title.encode("utf-8") or "") + "\n\n")
    if result.body:
        f.write(result.body.encode("utf-8"))
    f.close()
    print page, "->", f.name, "ok"