jirabot.py

   1 #!/usr/bin/python
   2 import calendar
   3 import feedparser
   4 import formatter
   5 import htmlentitydefs
   6 import htmllib
   7 import mechanize
   8 import os
   9 import random
  10 import string
  11 import StringIO
  12 import time
  13 import traceback
  14 import urlparse
  15 import zephyr
  16
  17 zephyr_sender = 'jira'
  18 zephyr_class = 'andersk-test'
  19 time_file = 'jirabot.time'
  20
  21 class UnicodeHTMLParser(htmllib.HTMLParser):
  22     entitydefs = dict((k, unichr(v)) for (k, v) in htmlentitydefs.name2codepoint.items())
  23
  24     def convert_charref(self, name):
  25         try:
  26             n = int(name)
  27         except ValueError:
  28             return
  29         return self.convert_codepoint(n)
  30
  31     def convert_codepoint(self, codepoint):
  32         return unichr(codepoint)
  33
  34 def jira_init():
  35     b = mechanize.Browser()
  36     b.set_handle_robots(False)
  37     b.add_client_certificate("https://idp.mit.edu:9443", "cert.pem", "cert.pem")
  38     b.addheaders = [("Accept-Language", "en-us,en;q=0.5"),]
  39     return b
  40
  41 def jira_login(b):
  42     b.open("https://jira.mit.edu/jira/secure/Dashboard.jspa")
  43     try:
  44         b.follow_link(text="MIT Touchstone")
  45     except mechanize.LinkNotFoundError:
  46         return
  47     if (urlparse.urlparse(b.geturl())[1] == "jira.mit.edu"):
  48         return
  49     b.select_form("wayfForm1")
  50     b.submit()
  51     b.select_form(predicate=lambda f: any(c.name == 'login_certificate'
  52                                           for c in f.controls))
  53     b.submit()
  54     b.select_form(nr=0)
  55     b.submit()
  56
  57 def feed_to_zephyrs(thing, rss, parse):
  58     zephyrs = []
  59     try:
  60         feed = feedparser.parse(rss)
  61         for e in feed.entries:
  62             global old_time, new_time
  63             t = int(calendar.timegm(e.date_parsed))
  64             if t <= old_time:
  65                 continue
  66             if t > new_time:
  67                 new_time = t
  68             try:
  69                 z = parse(e)
  70             except:
  71                 z = zerror("Error parsing " + thing + ":\n" + e.id + "\n" + traceback.format_exc())
  72             zephyrs.append((t, z))
  73     except:
  74         zephyrs.append((0, zerror("Error parsing " + thing + "s feed:\n" + traceback.format_exc())))
  75     return zephyrs
  76
  77 def parse_issue(e):
  78     issue = urlparse.urlparse(e.id)[2].rsplit('/', 1)[1]
  79     url = e.id
  80     msg = e.id + "\nThis issue was updated."
  81
  82     return zephyr.ZNotice(
  83         sender=zephyr_sender,
  84         auth=False,
  85         opcode='auto',
  86         cls=zephyr_class,
  87         instance=issue,
  88         fields=[e.title, msg],
  89     )
  90
  91 def parse_comment(e):
  92     url = urlparse.urlunsplit(urlparse.urlparse(e.id)[0:3] + (None,None))
  93     issue = url.rsplit('/', 1)[1]
  94
  95     s = StringIO.StringIO()
  96     parser = UnicodeHTMLParser(formatter.AbstractFormatter(formatter.DumbWriter(s)))
  97     parser.feed(e.summary.rsplit('<table>', 1)[0])
  98     parser.close()
  99     comment = s.getvalue()
 100
 101     msg = e.author + " added a comment:\n" + comment.rstrip()
 102
 103     return zephyr.ZNotice(
 104         sender=zephyr_sender,
 105         auth=False,
 106         opcode='auto',
 107         cls=zephyr_class,
 108         instance=issue,
 109         fields=[e.title, msg],
 110     )
 111
 112 def zerror(msg):
 113     return zephyr.ZNotice(
 114         sender=zephyr_sender,
 115         auth=False,
 116         opcode='auto',
 117         cls=zephyr_class,
 118         instance='jira-error',
 119         fields=['Jira bot error', msg]
 120     )
 121
 122 b = jira_init()
 123 zephyr.init()
 124
 125 count = 0
 126
 127 while True:
 128     time_file_new = time_file + '.' + ''.join(random.sample(string.letters, 8))
 129
 130     try:
 131         os.rename(time_file, time_file_new)
 132     except OSError:
 133         print "warning: could not acquire timestamp lock"
 134         time.sleep(17)
 135         continue
 136
 137     if (count >= 200):
 138         b = jira_init()
 139         count = 0
 140     count += 1
 141
 142     jira_login(b)
 143     b.open("https://jira.mit.edu/jira/sr/jira.issueviews:searchrequest-rss/temp/SearchRequest.xml?&pid=10185&updated%3Aprevious=-1w&sorter/field=updated&sorter/order=DESC&tempMax=1000")
 144     issues_rss = b.response().read()
 145     b.open("https://jira.mit.edu/jira/sr/jira.issueviews:searchrequest-comments-rss/temp/SearchRequest.xml?&pid=10185&updated%3Aprevious=-1w&sorter/field=updated&sorter/order=DESC&tempMax=1000")
 146     comments_rss = b.response().read()
 147     b.clear_history()
 148
 149     old_time = int(open(time_file_new).read())
 150     new_time = old_time
 151
 152     zephyrs = (feed_to_zephyrs('issue', issues_rss, parse_issue) +
 153                feed_to_zephyrs('comment', comments_rss, parse_comment))
 154     zephyrs.sort(key=lambda tz: tz[0])
 155     for (t, z) in zephyrs:
 156         z.send()
 157
 158     open(time_file_new, 'w').write(str(new_time))
 159     os.rename(time_file_new, time_file)
 160
 161     time.sleep(20)