#-*- coding: utf8 -*-

# microblog.py - levanta los feeds de twitter e identi.ca y deja algo lindo
#
# Copyright (C) 2009 - Facundo Batista <facundo@taniquetil.com.ar>
#
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program.  If not, see <http://www.gnu.org/licenses/>.

from __future__ import division
from __future__ import with_statement

import codecs
import herrams
import htmlentitydefs
import operator
import re
import sys
import time

import feedparser


class RetrievingError(Exception):
    '''A problem retrieving the feed data.'''

RSS_IDENT = "https://identi.ca/api/statuses/user_timeline/facundobatista.rss"
RSS_TWITT = "http://twitter.com/statuses/user_timeline/40733398.rss"
RE_UNQUOTE = re.compile("&(.*?);")
CUANTAS = 10

NO_TAGS = (
    re.compile("</?span.*?>"),
    re.compile("</?a.*?>"),
)

def filtra_tags(txt):
    for filtro in NO_TAGS:
        txt = filtro.sub("", txt)
    return txt

def real_unquote(txt):
    def r(m):
        torep = m.groups()[0]
        if torep[0:2] == "#x":
            caracter = unichr(int(torep[2:], 16))
        else:
            caracter = unichr(htmlentitydefs.name2codepoint.get(torep, 32))
        return caracter
    return RE_UNQUOTE.sub(r, txt)

def get_identica():
    todo = feedparser.parse(RSS_IDENT)
    title = getattr(todo.feed, "title", None)
    if not title:
        raise RetrievingError("No title in identica")
    assert title == "facundobatista timeline"

    for entry in todo.entries[:CUANTAS]:
        texto = entry.summary
        # FIXME: ver de usar HTMLParser.HTMLParser().unescape('&uacute;')
        texto = real_unquote(texto)
        cuando = entry.updated_parsed
        link = entry.links[0]["href"]
        yield Post(cuando, texto, link)

def get_twitter():
    todo = feedparser.parse(RSS_TWITT)
    title = getattr(todo.feed, "title", None)
    if not title:
        raise RetrievingError("No title in twitter")
    assert title == "Twitter / facundobatista"

    for entry in todo.entries[:CUANTAS]:
        texto = entry.summary
        if texto.startswith("facundobatista: "):
            texto = texto[16:]
        cuando = entry.updated_parsed
        link = entry.links[0]["href"]
        yield Post(cuando, texto, link)


class Post(object):
    def __init__(self, cuando, texto, link):
        self.cuando = cuando
        self.texto = filtra_tags(texto)
        self.link = link

    @property
    def linea(self):
        desde = herrams.fecha2desde(self.cuando)
        t = u"%s (hace %s)" % (self.texto, desde)
        return t


def get_lineas_mixer():
    try:
        twitt = dict((p.texto, p) for p in get_twitter())
#        ident = dict((p.texto, p) for p in get_identica())
    except RetrievingError:
        sys.exit()

    # mezclamos! el algoritmo es:
    #   ponemos primero lo de twitter, segundo lo de identica, así
    #   si hay repetidos nos queda el link a este segundo
    todas = twitt
#    todas.update(ident)

    # ordenamos y nos quedamos con las más nuevas
    orden = sorted(todas.itervalues(), key=operator.attrgetter("cuando"),
                                                                reverse=True)
    return orden[:CUANTAS]

def main(logtype, nomarch_all, facundario):
    """Log everything according to the type, and also maybe to facundario."""
    posts = get_lineas_mixer()

    if logtype == 'html':
        lines = []
        for post in posts:
            txt = post.linea.encode("ascii", "xmlcharrefreplace")
            t = u"<li><a href=%s>%s</a></li>" % (post.link, txt)
            lines.append(t)
        text = "\n".join(lines) + "\n"
    elif logtype == 'plain':
        text = "\n".join(p.linea for p in posts) + "\n"
    else:
        raise ValueError("Bad logtype: " + repr(logtype))

    with codecs.open(nomarch_all, "w", "utf8") as fh:
        fh.write(text)

    # facundario
    if facundario is not None:
        lines = []
        for post in posts:
            if '#facundario' not in post.texto:
                continue
            clean = post.texto.replace("#facundario", "").strip()
            tstamp = time.strftime("%Y-%m-%d %H:%M", post.cuando)
            lines.append((tstamp, "%s %s\n" % (tstamp, clean)))

        # get all the timestamps in the destination file
        with codecs.open(facundario, "r", "utf8") as fh:
            old_lines = set(" ".join(l.split()[0:2]) for l in fh.readlines())

        # add it if new, comparing the timestamp (so we can fix text)
        new_lines = [line for tstamp, line in lines if tstamp not in old_lines]
        if new_lines:
            text = "".join(sorted(new_lines))
            with codecs.open(facundario, "a", "utf8") as fh:
                fh.write(text)



HELP = """
Usar microblog.py {--plain|--html} arch_output [--facundario other_arch]
"""

if __name__ == "__main__":
    if len(sys.argv) < 3:
        print HELP
        sys.exit()

    opc = sys.argv[1]
    if opc not in ("--plain", "--html"):
        print HELP
        sys.exit()
    logtype = opc[2:]
    nomarch_all = sys.argv[2]

    facundario = None
    if len(sys.argv) == 5:
        if sys.argv[3] != '--facundario':
            print HELP
            sys.exit()
        facundario = sys.argv[4]

    main(logtype, nomarch_all, facundario)