rss2maildir.py

   1 #!/usr/bin/python
   2 # coding=utf-8
   3
   4 # rss2maildir.py - RSS feeds to Maildir 1 email per item
   5 # Copyright (C) 2007  Brett Parker <iDunno@sommitrealweird.co.uk>
   6 #
   7 # This program is free software: you can redistribute it and/or modify
   8 # it under the terms of the GNU General Public License as published by
   9 # the Free Software Foundation, either version 3 of the License, or
  10 # (at your option) any later version.
  11 #
  12 # This program is distributed in the hope that it will be useful,
  13 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  14 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  15 # GNU General Public License for more details.
  16 #
  17 # You should have received a copy of the GNU General Public License
  18 # along with this program.  If not, see <http://www.gnu.org/licenses/>.
  19
  20 import sys
  21 import os
  22 import stat
  23 import httplib
  24 import urllib
  25
  26 import feedparser
  27
  28 from email.MIMEMultipart import MIMEMultipart
  29 from email.MIMEText import MIMEText
  30
  31 import datetime
  32 import random
  33 import string
  34 import textwrap
  35
  36 import socket
  37
  38 from optparse import OptionParser
  39 from ConfigParser import SafeConfigParser
  40
  41 from base64 import b64encode
  42 import md5
  43
  44 import cgi
  45 import dbm
  46
  47 import re
  48
  49 from HTMLParser import HTMLParser
  50
  51 class HTML2Text(HTMLParser):
  52     entities = {
  53         u'amp': u'&',
  54         u'lt': u'<',
  55         u'gt': u'>',
  56         u'pound': u'£',
  57         u'copy': u'©',
  58         u'apos': u'\'',
  59         u'quot': u'"',
  60         u'nbsp': u' ',
  61         u'ldquo': u'“',
  62         u'rdquo': u'”',
  63         u'lsquo': u'‘',
  64         u'rsquo': u'’',
  65         u'laquo': u'«',
  66         u'raquo': u'»',
  67         u'lsaquo': u'‹',
  68         u'rsaquo': u'›',
  69         u'bull': u'•',
  70         u'middot': u'·',
  71         u'deg': u'°',
  72         u'helip': u'…',
  73         u'trade': u'™',
  74         u'reg': u'®',
  75         u'agrave': u'à',
  76         u'Agrave': u'À',
  77         u'egrave': u'è',
  78         u'Egrave': u'È',
  79         u'igrave': u'ì',
  80         u'Igrave': u'Ì',
  81         u'ograve': u'ò',
  82         u'Ograve': u'Ò',
  83         u'ugrave': u'ù',
  84         u'Ugrave': u'Ù',
  85         u'aacute': u'á',
  86         u'Aacute': u'Á',
  87         u'eacute': u'é',
  88         u'Eacute': u'É',
  89         u'iacute': u'í',
  90         u'Iacute': u'Í',
  91         u'oacute': u'ó',
  92         u'Oacute': u'Ó',
  93         u'uacute': u'ú',
  94         u'Uacute': u'Ú',
  95         u'yactue': u'ý',
  96         u'Yacute': u'Ý',
  97         u'acirc': u'â',
  98         u'Acirc': u'Â',
  99         u'ecirc': u'ê',
 100         u'Ecirc': u'Ê',
 101         u'icirc': u'î',
 102         u'Icirc': u'Î',
 103         u'ocirc': u'ô',
 104         u'Ocirc': u'Ô',
 105         u'ucirc': u'û',
 106         u'Ucirc': u'Û',
 107         u'atilde': u'ã',
 108         u'Atilde': u'Ã',
 109         u'ntilde': u'ñ',
 110         u'Ntilde': u'Ñ',
 111         u'otilde': u'õ',
 112         u'Otilde': u'Õ',
 113         u'auml': u'ä',
 114         u'Auml': u'Ä',
 115         u'euml': u'ë',
 116         u'Euml': u'Ë',
 117         u'iuml': u'ï',
 118         u'Iuml': u'Ï',
 119         u'ouml': u'ö',
 120         u'Ouml': u'Ö',
 121         u'uuml': u'ü',
 122         u'Uuml': u'Ü',
 123         u'yuml': u'ÿ',
 124         u'Yuml': u'Ÿ',
 125         u'iexcl': u'¡',
 126         u'iquest': u'¿',
 127         u'ccedil': u'ç',
 128         u'Ccedil': u'Ç',
 129         u'oelig': u'œ',
 130         u'OElig': u'Œ',
 131         u'szlig': u'ß',
 132         u'oslash': u'ø',
 133         u'Oslash': u'Ø',
 134         u'aring': u'å',
 135         u'Aring': u'Å',
 136         u'aelig': u'æ',
 137         u'AElig': u'Æ',
 138         u'thorn': u'þ',
 139         u'THORN': u'Þ',
 140         u'eth': u'ð',
 141         u'ETH': u'Ð',
 142         u'mdash': u'—',
 143         u'ndash': u'–',
 144         u'sect': u'§',
 145         u'para': u'¶',
 146         u'uarr': u'↑',
 147         u'darr': u'↓',
 148         u'larr': u'←',
 149         u'rarr': u'→',
 150         u'dagger': u'†',
 151         u'Dagger': u'‡',
 152         u'permil': u'‰',
 153         u'prod': u'∏',
 154         u'infin': u'∞',
 155         u'radic': u'√',
 156         u'there4': u'∴',
 157         u'int': u'∫',
 158         u'asymp': u'≈',
 159         u'ne': u'≠',
 160         u'equiv': '≡',
 161         u'le': u'≤',
 162         u'ge': u'≥',
 163         u'loz': u'⋄',
 164         u'sum': u'∑',
 165         u'part': u'∂',
 166         u'prime': u'′',
 167         u'Prime': u'″',
 168         u'harr': u'↔',
 169         u'micro': u'µ',
 170         u'not': u'¬',
 171         u'plusmn': u'±',
 172         u'divide': u'÷',
 173         u'cent': u'¢',
 174         u'euro': u'€',
 175         }
 176
 177     blockleveltags = [
 178         u'h1',
 179         u'h2',
 180         u'h3',
 181         u'h4',
 182         u'h5',
 183         u'h6',
 184         u'pre',
 185         u'p',
 186         u'ul',
 187         u'ol',
 188         u'dl',
 189         u'li',
 190         u'dt',
 191         u'dd',
 192         u'div',
 193         u'blockquote',
 194         ]
 195
 196     liststarttags = [
 197         u'ul',
 198         u'ol',
 199         u'dl',
 200         ]
 201
 202     cancontainflow = [
 203         u'div',
 204         u'li',
 205         u'dd',
 206         u'blockquote',
 207     ]
 208
 209     def __init__(self,textwidth=70):
 210         self.text = u''
 211         self.curdata = u''
 212         self.textwidth = textwidth
 213         self.opentags = []
 214         self.indentlevel = 0
 215         self.ignorenodata = False
 216         self.listcount = []
 217         self.urls = []
 218         self.images = {}
 219         HTMLParser.__init__(self)
 220
 221     def handle_starttag(self, tag, attrs):
 222         tag_name = tag.lower()
 223         if tag_name in self.blockleveltags:
 224             # handle starting a new block - unless we're in a block element
 225             # that can contain other blocks, we'll assume that we want to close
 226             # the container
 227             if len(self.opentags) > 1 and self.opentags[-1] == u'li':
 228                 self.handle_curdata()
 229
 230             if tag_name == u'ol':
 231                 self.handle_curdata()
 232                 self.listcount.append(1)
 233                 self.listlevel = len(self.listcount) - 1
 234
 235             if tag_name == u'dl':
 236                 self.indentlevel = self.indentlevel + 4
 237
 238             if tag_name in self.liststarttags:
 239                 smallist = self.opentags[-3:-1]
 240                 smallist.reverse()
 241                 for prev_listtag in smallist:
 242                     if prev_listtag in [u'dl', u'ol']:
 243                         self.indentlevel = self.indentlevel + 4
 244                         break
 245                     elif prev_listtag == u'ul':
 246                         self.indentlevel = self.indentlevel + 3
 247                         break
 248
 249             if len(self.opentags) > 0:
 250                 self.handle_curdata()
 251                 if tag_name not in self.cancontainflow:
 252                     self.opentags.pop()
 253             self.opentags.append(tag_name)
 254         else:
 255             if tag_name == "span":
 256                 return
 257             listcount = 0
 258             try:
 259                 listcount = self.listcount[-1]
 260             except:
 261                 pass
 262
 263             if tag_name == u'dd' and len(self.opentags) > 1 \
 264                 and self.opentags[-1] == u'dt':
 265                 self.handle_curdata()
 266                 self.opentags.pop()
 267             elif tag_name == u'dt' and len(self.opentags) > 1 \
 268                 and self.opentags[-1] == u'dd':
 269                 self.handle_curdata()
 270                 self.opentags.pop()
 271             elif tag_name == u'a':
 272                 for attr in attrs:
 273                     if attr[0].lower() == u'href':
 274                         self.urls.append(attr[1].decode('utf-8'))
 275                 self.curdata = self.curdata + u'`'
 276                 self.opentags.append(tag_name)
 277                 return
 278             elif tag_name == u'img':
 279                 self.handle_image(attrs)
 280                 return
 281             elif tag_name == u'br':
 282                 self.handle_br()
 283                 return
 284             else:
 285                 # we don't know the tag, so lets avoid handling it!
 286                 return
 287
 288     def handle_startendtag(self, tag, attrs):
 289         if tag.lower() == u'br':
 290             self.handle_br()
 291         elif tag.lower() == u'img':
 292             self.handle_image(attrs)
 293             return
 294
 295     def handle_br(self):
 296             self.handle_curdata()
 297             self.opentags.append(u'br')
 298             self.handle_curdata()
 299             self.opentags.pop()
 300
 301     def handle_image(self, attrs):
 302         alt = u''
 303         url = u''
 304         for attr in attrs:
 305             if attr[0] == 'alt':
 306                 alt = attr[1].decode('utf-8')
 307             elif attr[0] == 'src':
 308                 url = attr[1].decode('utf-8')
 309         if url:
 310             if alt:
 311                 if self.images.has_key(alt):
 312                     if self.images[alt]["url"] == url:
 313                         self.curdata = self.curdata \
 314                             + u'|%s|' %(alt,)
 315                     else:
 316                         while self.images.has_key(alt):
 317                             alt = alt + "_"
 318                         self.images[alt] = {"url": url}
 319                         self.curdata = self.curdata \
 320                             + u'|%s|' %(alt,)
 321                 else:
 322                     self.images[alt] = {"url": url}
 323                     self.curdata = self.curdata \
 324                         + u'|%s|' %(alt,)
 325             else:
 326                 if self.images.has_key(url):
 327                     self.curdata = self.curdata \
 328                         + u'|%s|' %(url,)
 329                 else:
 330                     self.images[url] = {}
 331                     self.images[url]["url"] =url
 332                     self.curdata = self.curdata \
 333                         + u'|%s|' %(url,)
 334
 335     def handle_curdata(self):
 336
 337         if len(self.opentags) == 0:
 338             return
 339
 340         tag_thats_done = self.opentags[-1]
 341
 342         if len(self.curdata) == 0:
 343             return
 344
 345         if tag_thats_done == u'br':
 346             if len(self.text) == 0 or self.text[-1] != '\n':
 347                 self.text = self.text + '\n'
 348                 self.ignorenodata = True
 349             return
 350
 351         if len(self.curdata.strip()) == 0:
 352             return
 353
 354         if tag_thats_done in self.blockleveltags:
 355             newlinerequired = self.text != u''
 356             if self.ignorenodata:
 357                 newlinerequired = False
 358             self.ignorenodata = False
 359             if newlinerequired:
 360                 if tag_thats_done in [u'dt', u'dd', u'li'] \
 361                     and len(self.text) > 1 \
 362                     and self.text[-1] != u'\n':
 363                         self.text = self.text + u'\n'
 364                 elif len(self.text) > 2 \
 365                     and self.text[-1] != u'\n' \
 366                     and self.text[-2] != u'\n':
 367                     self.text = self.text + u'\n\n'
 368
 369         if tag_thats_done in ["h1", "h2", "h3", "h4", "h5", "h6"]:
 370             underline = u''
 371             underlinechar = u'='
 372             headingtext = " ".join(self.curdata.split())
 373             seperator = u'\n' + u' '*self.indentlevel
 374             headingtext = seperator.join( \
 375                 textwrap.wrap( \
 376                     headingtext, \
 377                     self.textwidth - self.indentlevel \
 378                     ) \
 379                 )
 380
 381             if tag_thats_done == u'h2':
 382                 underlinechar = u'-'
 383             elif tag_thats_done != u'h1':
 384                 underlinechar = u'~'
 385
 386             if u'\n' in headingtext:
 387                 underline = u' ' * self.indentlevel \
 388                     + underlinechar * (self.textwidth - self.indentlevel)
 389             else:
 390                 underline = u' ' * self.indentlevel \
 391                     + underlinechar * len(headingtext)
 392             self.text = self.text \
 393                 + headingtext + u'\n' \
 394                 + underline
 395         elif tag_thats_done in [u'p', u'div']:
 396             paragraph = unicode( \
 397                 " ".join(self.curdata.strip().encode("utf-8").split()), \
 398                 "utf-8")
 399             seperator = u'\n' + u' ' * self.indentlevel
 400             self.text = self.text \
 401                 + u' ' * self.indentlevel \
 402                 + seperator.join( \
 403                     textwrap.wrap( \
 404                         paragraph, self.textwidth - self.indentlevel))
 405         elif tag_thats_done == "pre":
 406             self.text = self.text + unicode( \
 407                 self.curdata.encode("utf-8"), "utf-8")
 408         elif tag_thats_done == u'blockquote':
 409             quote = unicode( \
 410                 " ".join(self.curdata.encode("utf-8").strip().split()), \
 411                 "utf-8")
 412             seperator = u'\n' + u' ' * self.indentlevel + u'    '
 413             if len(self.text) > 0 and self.text[-1] != u'\n':
 414                 self.text = self.text + u'\n'
 415             self.text = self.text \
 416                 + u'    ' \
 417                 + seperator.join( \
 418                     textwrap.wrap( \
 419                         quote, \
 420                         self.textwidth - self.indentlevel - 2 \
 421                     )
 422                 )
 423             self.curdata = u''
 424         elif tag_thats_done == "li":
 425             item = unicode(self.curdata.encode("utf-8").strip(), "utf-8")
 426             if len(self.text) > 0 and self.text[-1] != u'\n':
 427                 self.text = self.text + u'\n'
 428             # work out if we're in an ol rather than a ul
 429             latesttags = self.opentags[-4:]
 430             latesttags.reverse()
 431             isul = None
 432             for thing in latesttags:
 433                 if thing == 'ul':
 434                     isul = True
 435                     break
 436                 elif thing == 'ol':
 437                     isul = False
 438                     break
 439
 440             listindent = 3
 441             if not isul:
 442                 listindent = 4
 443
 444             listmarker = u' * '
 445             if isul == False:
 446                 listmarker = u' %2d. ' %(self.listcount[-1])
 447                 self.listcount[-1] = self.listcount[-1] + 1
 448
 449             seperator = u'\n' \
 450                 + u' ' * self.indentlevel \
 451                 + u' ' * listindent
 452             self.text = self.text \
 453                 + u' ' * self.indentlevel \
 454                 + listmarker \
 455                 + seperator.join( \
 456                     textwrap.wrap( \
 457                         item, \
 458                         self.textwidth - self.indentlevel - listindent \
 459                     ) \
 460                 )
 461             self.curdata = u''
 462         elif tag_thats_done == u'dt':
 463             definition = unicode(" ".join( \
 464                     self.curdata.encode("utf-8").strip().split()), \
 465                 "utf-8")
 466             if len(self.text) > 0 and self.text[-1] != u'\n':
 467                 self.text = self.text + u'\n\n'
 468             elif len(self.text) > 1 and self.text[-2] != u'\n':
 469                 self.text = self.text + u'\n'
 470             definition = u' ' * (self.indentlevel - 4) + definition + "::"
 471             indentstring = u'\n' + u' ' * (self.indentlevel - 3)
 472             self.text = self.text \
 473                 + indentstring.join(
 474                     textwrap.wrap(definition, \
 475                         self.textwidth - self.indentlevel - 4))
 476             self.curdata = u''
 477         elif tag_thats_done == u'dd':
 478             definition = unicode(" ".join( \
 479                     self.curdata.encode("utf-8").strip().split()),
 480                 "utf-8")
 481             if len(definition) > 0:
 482                 if len(self.text) > 0 and self.text[-1] != u'\n':
 483                     self.text = self.text + u'\n'
 484                 indentstring = u'\n' + u' ' * self.indentlevel
 485                 self.text = self.text \
 486                     + indentstring \
 487                     + indentstring.join( \
 488                         textwrap.wrap( \
 489                             definition, \
 490                             self.textwidth - self.indentlevel \
 491                             ) \
 492                         )
 493                 self.curdata = u''
 494         elif tag_thats_done == u'a':
 495             self.curdata = self.curdata + u'`__'
 496             pass
 497         elif tag_thats_done in self.liststarttags:
 498             pass
 499
 500         if tag_thats_done in self.blockleveltags:
 501             self.curdata = u''
 502
 503         self.ignorenodata = False
 504
 505     def handle_endtag(self, tag):
 506         self.ignorenodata = False
 507         if tag == "span":
 508             return
 509
 510         try:
 511             tagindex = self.opentags.index(tag)
 512         except:
 513             return
 514         tag = tag.lower()
 515
 516         if tag in [u'br', u'img']:
 517             return
 518
 519         if tag == u'dl':
 520             self.indentlevel = self.indentlevel - 4
 521
 522         if tag in self.liststarttags:
 523             if tag in [u'ol', u'dl', u'ul', u'dd']:
 524                 self.handle_curdata()
 525                 # find if there was a previous list level
 526                 smalllist = self.opentags[:-1]
 527                 smalllist.reverse()
 528                 for prev_listtag in smalllist:
 529                     if prev_listtag in [u'ol', u'dl']:
 530                         self.indentlevel = self.indentlevel - 4
 531                         break
 532                     elif prev_listtag == u'ul':
 533                         self.indentlevel = self.indentlevel - 3
 534                         break
 535
 536         if tag == u'ol':
 537             self.listcount = self.listcount[:-1]
 538
 539         while tagindex < len(self.opentags) \
 540             and tag in self.opentags[tagindex+1:]:
 541             try:
 542                 tagindex = self.opentags.index(tag, tagindex+1)
 543             except:
 544                 # well, we don't want to do that then
 545                 pass
 546         if tagindex != len(self.opentags) - 1:
 547             # Assuming the data was for the last opened tag first
 548             self.handle_curdata()
 549             # Now kill the list to be a slice before this tag was opened
 550             self.opentags = self.opentags[:tagindex + 1]
 551         else:
 552             self.handle_curdata()
 553             if self.opentags[-1] == tag:
 554                 self.opentags.pop()
 555
 556     def handle_data(self, data):
 557         if len(self.opentags) == 0:
 558             self.opentags.append(u'p')
 559         self.curdata = self.curdata + data.decode("utf-8")
 560
 561     def handle_charref(self, name):
 562         try:
 563             entity = unichr(int(name))
 564         except:
 565             if name[0] == 'x':
 566                 try:
 567                     entity = unichr(int('0%s' %(name,), 16))
 568                 except:
 569                     entity = u'#%s' %(name,)
 570             else:
 571                 entity = u'#%s' %(name,)
 572         self.curdata = self.curdata + unicode(entity.encode('utf-8'), \
 573             "utf-8")
 574
 575     def handle_entityref(self, name):
 576         entity = name
 577         if HTML2Text.entities.has_key(name):
 578             entity = HTML2Text.entities[name]
 579         else:
 580             entity = "&" + name + ";"
 581
 582         self.curdata = self.curdata + unicode(entity.encode('utf-8'), \
 583             "utf-8")
 584
 585     def gettext(self):
 586         self.handle_curdata()
 587         if len(self.text) == 0 or self.text[-1] != u'\n':
 588             self.text = self.text + u'\n'
 589         self.opentags = []
 590         if len(self.text) > 0:
 591             while len(self.text) > 1 and self.text[-1] == u'\n':
 592                 self.text = self.text[:-1]
 593             self.text = self.text + u'\n'
 594         if len(self.urls) > 0:
 595             self.text = self.text + u'\n__ ' + u'\n__ '.join(self.urls) + u'\n'
 596             self.urls = []
 597         if len(self.images.keys()) > 0:
 598             self.text = self.text + u'\n.. ' \
 599                 + u'\n.. '.join( \
 600                     ["|%s| image:: %s" %(a, self.images[a]["url"]) \
 601                 for a in self.images.keys()]) + u'\n'
 602             self.images = {}
 603         return self.text
 604
 605 def open_url(method, url):
 606     redirectcount = 0
 607     while redirectcount < 3:
 608         (type, rest) = urllib.splittype(url)
 609         (host, path) = urllib.splithost(rest)
 610         (host, port) = urllib.splitport(host)
 611         if port == None:
 612             port = 80
 613         try:
 614             conn = httplib.HTTPConnection("%s:%s" %(host, port))
 615             conn.request(method, path)
 616             response = conn.getresponse()
 617             if response.status in [301, 302, 303, 307]:
 618                 headers = response.getheaders()
 619                 for header in headers:
 620                     if header[0] == "location":
 621                         url = header[1]
 622             elif response.status == 200:
 623                 return response
 624         except:
 625             pass
 626         redirectcount = redirectcount + 1
 627     return None
 628
 629 def parse_and_deliver(maildir, url, statedir):
 630     feedhandle = None
 631     headers = None
 632     # first check if we know about this feed already
 633     feeddb = dbm.open(os.path.join(statedir, "feeds"), "c")
 634     if feeddb.has_key(url):
 635         data = feeddb[url]
 636         data = cgi.parse_qs(data)
 637         response = open_url("HEAD", url)
 638         headers = None
 639         if response:
 640             headers = response.getheaders()
 641         ischanged = False
 642         try:
 643             for header in headers:
 644                 if header[0] == "content-length":
 645                     if header[1] != data["content-length"][0]:
 646                         ischanged = True
 647                 elif header[0] == "etag":
 648                     if header[1] != data["etag"][0]:
 649                         ischanged = True
 650                 elif header[0] == "last-modified":
 651                     if header[1] != data["last-modified"][0]:
 652                         ischanged = True
 653                 elif header[0] == "content-md5":
 654                     if header[1] != data["content-md5"][0]:
 655                         ischanged = True
 656         except:
 657             ischanged = True
 658         if ischanged:
 659             response = open_url("GET", url)
 660             if response != None:
 661                 headers = response.getheaders()
 662                 feedhandle = response
 663             else:
 664                 sys.stderr.write("Failed to fetch feed: %s\n" %(url))
 665                 return
 666         else:
 667             return # don't need to do anything, nothings changed.
 668     else:
 669         response = open_url("GET", url)
 670         if response != None:
 671             headers = response.getheaders()
 672             feedhandle = response
 673         else:
 674             sys.stderr.write("Failed to fetch feed: %s\n" %(url))
 675             return
 676
 677     fp = feedparser.parse(feedhandle)
 678     db = dbm.open(os.path.join(statedir, "seen"), "c")
 679     for item in fp["items"]:
 680         # have we seen it before?
 681         # need to work out what the content is first...
 682
 683         if item.has_key("content"):
 684             content = item["content"][0]["value"]
 685         else:
 686             if item.has_key("description"):
 687                 content = item["description"]
 688             else:
 689                 content = u''
 690
 691         md5sum = md5.md5(content.encode("utf-8")).hexdigest()
 692
 693         prevmessageid = None
 694
 695         db_guid_key = None
 696         db_link_key = (url + u'|' + item["link"]).encode("utf-8")
 697
 698         # check if there's a guid too - if that exists and we match the md5,
 699         # return
 700         if item.has_key("guid"):
 701             db_guid_key = (url + u'|' + item["guid"]).encode("utf-8")
 702             if db.has_key(db_guid_key):
 703                 data = db[db_guid_key]
 704                 data = cgi.parse_qs(data)
 705                 if data["contentmd5"][0] == md5sum:
 706                     continue
 707
 708         if db.has_key(db_link_key):
 709             data = db[db_link_key]
 710             data = cgi.parse_qs(data)
 711             if data.has_key("message-id"):
 712                 prevmessageid = data["message-id"][0]
 713             if data["contentmd5"][0] == md5sum:
 714                 continue
 715
 716         try:
 717             author = item["author"]
 718         except:
 719             author = url
 720
 721         # create a basic email message
 722         msg = MIMEMultipart("alternative")
 723         messageid = "<" \
 724             + datetime.datetime.now().strftime("%Y%m%d%H%M") \
 725             + "." \
 726             + "".join( \
 727                 [random.choice( \
 728                     string.ascii_letters + string.digits \
 729                     ) for a in range(0,6) \
 730                 ]) + "@" + socket.gethostname() + ">"
 731         msg.add_header("Message-ID", messageid)
 732         msg.set_unixfrom("\"%s\" <rss2maildir@localhost>" %(url))
 733         msg.add_header("From", "\"%s\" <rss2maildir@localhost>" %(author))
 734         msg.add_header("To", "\"%s\" <rss2maildir@localhost>" %(url))
 735         if prevmessageid:
 736             msg.add_header("References", prevmessageid)
 737         createddate = datetime.datetime.now() \
 738             .strftime("%a, %e %b %Y %T -0000")
 739         try:
 740             createddate = datetime.datetime(*item["updated_parsed"][0:6]) \
 741                 .strftime("%a, %e %b %Y %T -0000")
 742         except:
 743             pass
 744         msg.add_header("Date", createddate)
 745         subj_gen = HTML2Text()
 746         title = item["title"]
 747         title = re.sub(u'<', u'&lt;', title)
 748         title = re.sub(u'>', u'&gt;', title)
 749         subj_gen.feed(title.encode("utf-8"))
 750         msg.add_header("Subject", subj_gen.gettext())
 751         msg.set_default_type("text/plain")
 752
 753         htmlcontent = content.encode("utf-8")
 754         htmlcontent = "%s\n\n<p>Item URL: <a href='%s'>%s</a></p>" %( \
 755             content, \
 756             item["link"], \
 757             item["link"] )
 758         htmlpart = MIMEText(htmlcontent.encode("utf-8"), "html", "utf-8")
 759         textparser = HTML2Text()
 760         textparser.feed(content.encode("utf-8"))
 761         textcontent = textparser.gettext()
 762         textcontent = "%s\n\nItem URL: %s" %( \
 763             textcontent, \
 764             item["link"] )
 765         textpart = MIMEText(textcontent.encode("utf-8"), "plain", "utf-8")
 766         msg.attach(textpart)
 767         msg.attach(htmlpart)
 768
 769         # start by working out the filename we should be writting to, we do
 770         # this following the normal maildir style rules
 771         fname = str(os.getpid()) \
 772             + "." + socket.gethostname() \
 773             + "." + "".join( \
 774                 [random.choice( \
 775                     string.ascii_letters + string.digits \
 776                     ) for a in range(0,10) \
 777                 ]) + "." \
 778             + datetime.datetime.now().strftime('%s')
 779         fn = os.path.join(maildir, "tmp", fname)
 780         fh = open(fn, "w")
 781         fh.write(msg.as_string())
 782         fh.close()
 783         # now move it in to the new directory
 784         newfn = os.path.join(maildir, "new", fname)
 785         os.link(fn, newfn)
 786         os.unlink(fn)
 787
 788         # now add to the database about the item
 789         if prevmessageid:
 790             messageid = prevmessageid + " " + messageid
 791         if item.has_key("guid") and item["guid"] != item["link"]:
 792             data = urllib.urlencode(( \
 793                 ("message-id", messageid), \
 794                 ("created", createddate), \
 795                 ("contentmd5", md5sum) \
 796                 ))
 797             db[db_guid_key] = data
 798             try:
 799                 data = db[db_link_key]
 800                 data = cgi.parse_qs(data)
 801                 newdata = urllib.urlencode(( \
 802                     ("message-id", messageid), \
 803                     ("created", data["created"][0]), \
 804                     ("contentmd5", data["contentmd5"][0]) \
 805                     ))
 806                 db[db_link_key] = newdata
 807             except:
 808                 db[db_link_key] = data
 809         else:
 810             data = urllib.urlencode(( \
 811                 ("message-id", messageid), \
 812                 ("created", createddate), \
 813                 ("contentmd5", md5sum) \
 814                 ))
 815             db[db_link_key] = data
 816
 817     if headers:
 818         data = []
 819         for header in headers:
 820             if header[0] in \
 821                 ["content-md5", "etag", "last-modified", "content-length"]:
 822                 data.append((header[0], header[1]))
 823         if len(data) > 0:
 824             data = urllib.urlencode(data)
 825             feeddb[url] = data
 826
 827     db.close()
 828     feeddb.close()
 829
 830 if __name__ == "__main__":
 831     # This only gets executed if we really called the program
 832     # first off, parse the command line arguments
 833
 834     oparser = OptionParser()
 835     oparser.add_option(
 836         "-c", "--conf", dest="conf",
 837         help="location of config file"
 838         )
 839     oparser.add_option(
 840         "-s", "--statedir", dest="statedir",
 841         help="location of directory to store state in"
 842         )
 843
 844     (options, args) = oparser.parse_args()
 845
 846     # check for the configfile
 847
 848     configfile = None
 849
 850     if options.conf != None:
 851         # does the file exist?
 852         try:
 853             os.stat(options.conf)
 854             configfile = options.conf
 855         except:
 856             # should exit here as the specified file doesn't exist
 857             sys.stderr.write( \
 858                 "Config file %s does not exist. Exiting.\n" %(options.conf,))
 859             sys.exit(2)
 860     else:
 861         # check through the default locations
 862         try:
 863             os.stat("%s/.rss2maildir.conf" %(os.environ["HOME"],))
 864             configfile = "%s/.rss2maildir.conf" %(os.environ["HOME"],)
 865         except:
 866             try:
 867                 os.stat("/etc/rss2maildir.conf")
 868                 configfile = "/etc/rss2maildir.conf"
 869             except:
 870                 sys.stderr.write("No config file found. Exiting.\n")
 871                 sys.exit(2)
 872
 873     # Right - if we've got this far, we've got a config file, now for the hard
 874     # bits...
 875
 876     scp = SafeConfigParser()
 877     scp.read(configfile)
 878
 879     maildir_root = "RSSMaildir"
 880     state_dir = "state"
 881
 882     if options.statedir != None:
 883         state_dir = options.statedir
 884         try:
 885             mode = os.stat(state_dir)[stat.ST_MODE]
 886             if not stat.S_ISDIR(mode):
 887                 sys.stderr.write( \
 888                     "State directory (%s) is not a directory\n" %(state_dir))
 889                 sys.exit(1)
 890         except:
 891             # try to make the directory
 892             try:
 893                 os.mkdir(state_dir)
 894             except:
 895                 sys.stderr.write("Couldn't create statedir %s" %(state_dir))
 896                 sys.exit(1)
 897     elif scp.has_option("general", "state_dir"):
 898         new_state_dir = scp.get("general", "state_dir")
 899         try:
 900             mode = os.stat(new_state_dir)[stat.ST_MODE]
 901             if not stat.S_ISDIR(mode):
 902                 sys.stderr.write( \
 903                     "State directory (%s) is not a directory\n" %(state_dir))
 904                 sys.exit(1)
 905             else:
 906                 state_dir = new_state_dir
 907         except:
 908             # try to create it
 909             try:
 910                 os.mkdir(new_state_dir)
 911                 state_dir = new_state_dir
 912             except:
 913                 sys.stderr.write( \
 914                     "Couldn't create state directory %s\n" %(new_state_dir))
 915                 sys.exit(1)
 916     else:
 917         try:
 918             mode = os.stat(state_dir)[stat.ST_MODE]
 919             if not stat.S_ISDIR(mode):
 920                 sys.stderr.write( \
 921                     "State directory %s is not a directory\n" %(state_dir))
 922                 sys.exit(1)
 923         except:
 924             try:
 925                 os.mkdir(state_dir)
 926             except:
 927                 sys.stderr.write( \
 928                     "State directory %s could not be created\n" %(state_dir))
 929                 sys.exit(1)
 930
 931     if scp.has_option("general", "maildir_root"):
 932         maildir_root = scp.get("general", "maildir_root")
 933
 934     try:
 935         mode = os.stat(maildir_root)[stat.ST_MODE]
 936         if not stat.S_ISDIR(mode):
 937             sys.stderr.write( \
 938                 "Maildir Root %s is not a directory\n" \
 939                 %(maildir_root))
 940             sys.exit(1)
 941     except:
 942         try:
 943             os.mkdir(maildir_root)
 944         except:
 945             sys.stderr.write("Couldn't create Maildir Root %s\n" \
 946                 %(maildir_root))
 947             sys.exit(1)
 948
 949     feeds = scp.sections()
 950     try:
 951         feeds.remove("general")
 952     except:
 953         pass
 954
 955     for section in feeds:
 956         # check if the directory exists
 957         maildir = None
 958         try:
 959             maildir = scp.get(section, "maildir")
 960         except:
 961             maildir = section
 962
 963         maildir = urllib.urlencode(((section, maildir),)).split("=")[1]
 964         maildir = os.path.join(maildir_root, maildir)
 965
 966         try:
 967             exists = os.stat(maildir)
 968             if stat.S_ISDIR(exists[stat.ST_MODE]):
 969                 # check if there's a new, cur and tmp directory
 970                 try:
 971                     mode = os.stat(os.path.join(maildir, "cur"))[stat.ST_MODE]
 972                 except:
 973                     os.mkdir(os.path.join(maildir, "cur"))
 974                     if not stat.S_ISDIR(mode):
 975                         sys.stderr.write("Broken maildir: %s\n" %(maildir))
 976                 try:
 977                     mode = os.stat(os.path.join(maildir, "tmp"))[stat.ST_MODE]
 978                 except:
 979                     os.mkdir(os.path.join(maildir, "tmp"))
 980                     if not stat.S_ISDIR(mode):
 981                         sys.stderr.write("Broken maildir: %s\n" %(maildir))
 982                 try:
 983                     mode = os.stat(os.path.join(maildir, "new"))[stat.ST_MODE]
 984                     if not stat.S_ISDIR(mode):
 985                         sys.stderr.write("Broken maildir: %s\n" %(maildir))
 986                 except:
 987                     os.mkdir(os.path.join(maildir, "new"))
 988             else:
 989                 sys.stderr.write("Broken maildir: %s\n" %(maildir))
 990         except:
 991             try:
 992                 os.mkdir(maildir)
 993             except:
 994                 sys.stderr.write("Couldn't create root maildir %s\n" \
 995                     %(maildir))
 996                 sys.exit(1)
 997             try:
 998                 os.mkdir(os.path.join(maildir, "new"))
 999                 os.mkdir(os.path.join(maildir, "cur"))
1000                 os.mkdir(os.path.join(maildir, "tmp"))
1001             except:
1002                 sys.stderr.write( \
1003                     "Couldn't create required maildir directories for %s\n" \
1004                     %(section,))
1005                 sys.exit(1)
1006
1007         # right - we've got the directories, we've got the section, we know the
1008         # url... lets play!
1009
1010         parse_and_deliver(maildir, section, state_dir)