# $Id$
# generate comments page for an infogami site

import sys, re, time
import infogamibot
import urlparse, urllib2

# --------------------------------------------------------------------
# configuration

INDEX = "_special/index"

# cache entries older than this are checked
CACHE_AGE = 10 * 60 # 10 minutes

# comments older than this are left out
MAX_AGE = 10 * 24 * 3600 # 10 days

# comments older than this are shown with titles only
FRESH_AGE = 3 * 24 * 3600 # 1 day
FRESH_COUNT = 5 # at least this many

# --------------------------------------------------------------------

now = time.time()

MAX_AGE = time.gmtime(now - MAX_AGE)
FRESH_AGE = time.gmtime(now - FRESH_AGE)

ET = infogamibot.ET

def gettext(elem):
    text = elem.text or ""
    for e in elem:
        text += gettext(e)
        if e.tail:
            text += e.tail
    return text

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

bot = infogamibot.Bot(site, cache=CACHE_AGE)

# get list of comments
comment_list = []
for page, title in bot.get_index(INDEX):
    if page.startswith("/%2F_"):
        continue # skip toplevel comments
    m = re.match(".*/_comments/(\w+)$", page)
    if not m:
        continue
    comment_id = m.group(1).zfill(10)
    comment_list.append((comment_id, page))

comment_list.sort()
comment_list.reverse()

# get comment information, starting with newest and going backwards
comments = []
titles = {}

for _, page in comment_list:
    # print "---", "fetch", page
    try:
        result = bot.get_html(page) # get comment timestamp
    except urllib2.HTTPError, v:
        if v.code == 404:
            continue
        print "===", "error reading", bot.site + page, "-", v
        continue
    except SyntaxError, v:
        print "===", "error parsing", bot.site + page, "-", v
        continue
    # check age
    mtime = result.headers.getdate("last-modified")
    if len(comments) > 10 and mtime < MAX_AGE:
        break
    # extract comment text
    text = author = ""
    for elem in result.getiterator():
        if elem.get("class") != "page":
            continue
        for e in elem.getiterator():
            if e.get("class") == "dateline" and not author:
                author = gettext(e).strip()
                e.clear()
        text = gettext(elem).strip()
        break
    # locate actual page
    page = page.split("/_comments")[0]
    comments.append((mtime, page, text, author))
    if page not in titles:
        # print "---", "fetch", page, "title"
        try:
            result = bot.get_body(page)
        except urllib2.HTTPError, v:
            titles[page] = "(unknown)"
        else:
            if result.title is not None:
                titles[page] = infogamibot.escape(result.title.text)
            else:
                print "===", "unknown title for", page
                titles[page] = "(unknown)"

comments.sort()
comments.reverse()

body = [
    "This list is updated by an off-site script, at irregular intervals.\n",
    "\n",
]

last_comment = None

for index, (date, page, text, author) in enumerate(comments):
    if not date:
        continue
    timestamp = "%04d-%02d-%02d %02d:%02d" % date[:5]
    body.append("%s - [%s](%s#end)  \n" % (
        timestamp, titles[page], page
        ))
    if index < FRESH_COUNT or date >= FRESH_AGE:
        text = " ".join(text.split()[:20]) + "..."
        m = re.search("by\s(\w+)\s+#", author)
        if m:
            text = "**" + m.group(1) + ":** " + infogamibot.escape(text)
        if not last_comment:
            last_comment = text # most recent comment
        body.append("> " + text + "\n\n")

if bot.update("comments", None, "".join(body)) is not None:
    print "---", bot.site + "/comments", "UPDATED!"
    print
    print last_comment