scripts/import_sf.py

   1 """ Import tracker data from Sourceforge.NET
   2
   3 This script needs four steps to work:
   4
   5 1. Export the project XML data using the admin web interface at sf.net
   6 2. Run the file fetching (these are not included in the XML):
   7
   8     import_sf.py files <path to XML> <path to files dir>
   9
  10    this will place all the downloaded files in the files dir by file id.
  11 3. Convert the sf.net XML to Roundup "export" format:
  12
  13     import_sf.py import <tracker home> <path to XML> <path to files dir>
  14
  15    this will generate a directory "/tmp/imported" which contains the
  16    data to be imported into a Roundup tracker.
  17 4. Import the data:
  18
  19     roundup-admin -i <tracker home> import /tmp/imported
  20
  21 And you're done!
  22 """
  23
  24 import sys, os, csv, time, urllib2, httplib, mimetypes, urlparse
  25 # Python 2.3 ... 2.6 compatibility:
  26 from roundup.anypy.sets_ import set
  27
  28 try:
  29     import cElementTree as ElementTree
  30 except ImportError:
  31     from elementtree import ElementTree
  32
  33 from roundup import instance, hyperdb, date, support, password
  34
  35 today = date.Date('.')
  36
  37 DL_URL = 'http://sourceforge.net/tracker/download.php?group_id=%(group_id)s&atid=%(atid)s&aid=%(aid)s'
  38
  39 def get_url(aid):
  40     """ so basically we have to jump through hoops, given an artifact id, to
  41     figure what the URL should be to access that artifact, and hence any
  42     attached files."""
  43     # first we hit this URL...
  44     conn = httplib.HTTPConnection("sourceforge.net")
  45     conn.request("GET", "/support/tracker.php?aid=%s"%aid)
  46     response = conn.getresponse()
  47     # which should respond with a redirect to the correct url which has the
  48     # magic "group_id" and "atid" values in it that we need
  49     assert response.status == 302, 'response code was %s'%response.status
  50     location = response.getheader('location')
  51     query = urlparse.urlparse(response.getheader('location'))[-2]
  52     info = dict([param.split('=') for param in query.split('&')])
  53     return DL_URL%info
  54
  55 def fetch_files(xml_file, file_dir):
  56     """ Fetch files referenced in the xml_file into the dir file_dir. """
  57     root = ElementTree.parse(xml_file).getroot()
  58     to_fetch = set()
  59     deleted = set()
  60     for artifact in root.find('artifacts'):
  61         for field in artifact.findall('field'):
  62             if field.get('name') == 'artifact_id':
  63                 aid = field.text
  64         for field in artifact.findall('field'):
  65             if field.get('name') != 'artifact_history': continue
  66             for event in field.findall('history'):
  67                 d = {}
  68                 for field in event.findall('field'):
  69                     d[field.get('name')] = field.text
  70                 if d['field_name'] == 'File Added':
  71                     fid = d['old_value'].split(':')[0]
  72                     to_fetch.add((aid, fid))
  73                 if d['field_name'] == 'File Deleted':
  74                     fid = d['old_value'].split(':')[0]
  75                     deleted.add((aid, fid))
  76     to_fetch = to_fetch - deleted
  77
  78     got = set(os.listdir(file_dir))
  79     to_fetch = to_fetch - got
  80
  81     # load cached urls (sigh)
  82     urls = {}
  83     if os.path.exists(os.path.join(file_dir, 'urls.txt')):
  84         for line in open(os.path.join(file_dir, 'urls.txt')):
  85             aid, url = line.strip().split()
  86             urls[aid] = url
  87
  88     for aid, fid in support.Progress('Fetching files', list(to_fetch)):
  89         if fid in got: continue
  90         if not urls.has_key(aid):
  91             urls[aid] = get_url(aid)
  92             f = open(os.path.join(file_dir, 'urls.txt'), 'a')
  93             f.write('%s %s\n'%(aid, urls[aid]))
  94             f.close()
  95         url = urls[aid] + '&file_id=' + fid
  96         f = urllib2.urlopen(url)
  97         data = f.read()
  98         n = open(os.path.join(file_dir, fid), 'w')
  99         n.write(data)
 100         f.close()
 101         n.close()
 102
 103 def import_xml(tracker_home, xml_file, file_dir):
 104     """ Generate Roundup tracker import files based on the tracker schema,
 105     sf.net xml export and downloaded files from sf.net. """
 106     tracker = instance.open(tracker_home)
 107     db = tracker.open('admin')
 108
 109     resolved = db.status.lookup('resolved')
 110     unread = db.status.lookup('unread')
 111     chatting = db.status.lookup('unread')
 112     critical = db.priority.lookup('critical')
 113     urgent = db.priority.lookup('urgent')
 114     bug = db.priority.lookup('bug')
 115     feature = db.priority.lookup('feature')
 116     wish = db.priority.lookup('wish')
 117     adminuid = db.user.lookup('admin')
 118     anonuid = db.user.lookup('anonymous')
 119
 120     root = ElementTree.parse(xml_file).getroot()
 121
 122     def to_date(ts):
 123         return date.Date(time.gmtime(float(ts)))
 124
 125     # parse out the XML
 126     artifacts = []
 127     categories = set()
 128     users = set()
 129     add_files = set()
 130     remove_files = set()
 131     for artifact in root.find('artifacts'):
 132         d = {}
 133         op = {}
 134         artifacts.append(d)
 135         for field in artifact.findall('field'):
 136             name = field.get('name')
 137             if name == 'artifact_messages':
 138                 for message in field.findall('message'):
 139                     l = d.setdefault('messages', [])
 140                     m = {}
 141                     l.append(m)
 142                     for field in message.findall('field'):
 143                         name = field.get('name')
 144                         if name == 'adddate':
 145                             m[name] = to_date(field.text)
 146                         else:
 147                             m[name] = field.text
 148                         if name == 'user_name': users.add(field.text)
 149             elif name == 'artifact_history':
 150                 for event in field.findall('history'):
 151                     l = d.setdefault('history', [])
 152                     e = {}
 153                     l.append(e)
 154                     for field in event.findall('field'):
 155                         name = field.get('name')
 156                         if name == 'entrydate':
 157                             e[name] = to_date(field.text)
 158                         else:
 159                             e[name] = field.text
 160                         if name == 'mod_by': users.add(field.text)
 161                     if e['field_name'] == 'File Added':
 162                         add_files.add(e['old_value'].split(':')[0])
 163                     elif e['field_name'] == 'File Deleted':
 164                         remove_files.add(e['old_value'].split(':')[0])
 165             elif name == 'details':
 166                 op['body'] = field.text
 167             elif name == 'submitted_by':
 168                 op['user_name'] = field.text
 169                 d[name] = field.text
 170                 users.add(field.text)
 171             elif name == 'open_date':
 172                 thedate = to_date(field.text)
 173                 op['adddate'] = thedate
 174                 d[name] = thedate
 175             else:
 176                 d[name] = field.text
 177
 178         categories.add(d['category'])
 179
 180         if op.has_key('body'):
 181             l = d.setdefault('messages', [])
 182             l.insert(0, op)
 183
 184     add_files -= remove_files
 185
 186     # create users
 187     userd = {'nobody': '2'}
 188     users.remove('nobody')
 189     data = [
 190         {'id': '1', 'username': 'admin', 'password': password.Password('admin'),
 191             'roles': 'Admin', 'address': 'richard@python.org'},
 192         {'id': '2', 'username': 'anonymous', 'roles': 'Anonymous'},
 193     ]
 194     for n, user in enumerate(list(users)):
 195         userd[user] = n+3
 196         data.append({'id': str(n+3), 'username': user, 'roles': 'User',
 197             'address': '%s@users.sourceforge.net'%user})
 198     write_csv(db.user, data)
 199     users=userd
 200
 201     # create categories
 202     categoryd = {'None': None}
 203     categories.remove('None')
 204     data = []
 205     for n, category in enumerate(list(categories)):
 206         categoryd[category] = n
 207         data.append({'id': str(n), 'name': category})
 208     write_csv(db.keyword, data)
 209     categories = categoryd
 210
 211     # create issues
 212     issue_data = []
 213     file_data = []
 214     message_data = []
 215     issue_journal = []
 216     message_id = 0
 217     for artifact in artifacts:
 218         d = {}
 219         d['id'] = artifact['artifact_id']
 220         d['title'] = artifact['summary']
 221         d['assignedto'] = users[artifact['assigned_to']]
 222         if d['assignedto'] == '2':
 223             d['assignedto'] = None
 224         d['creation'] = artifact['open_date']
 225         activity = artifact['open_date']
 226         d['creator'] = users[artifact['submitted_by']]
 227         actor = d['creator']
 228         if categories[artifact['category']]:
 229             d['keyword'] = [categories[artifact['category']]]
 230         issue_journal.append((
 231             d['id'], d['creation'].get_tuple(), d['creator'], "'create'", {}
 232         ))
 233
 234         p = int(artifact['priority'])
 235         if artifact['artifact_type'] == 'Feature Requests':
 236             if p > 3:
 237                 d['priority'] = feature
 238             else:
 239                 d['priority'] = wish
 240         else:
 241             if p > 7:
 242                 d['priority'] = critical
 243             elif p > 5:
 244                 d['priority'] = urgent
 245             elif p > 3:
 246                 d['priority'] = bug
 247             else:
 248                 d['priority'] = feature
 249
 250         s = artifact['status']
 251         if s == 'Closed':
 252             d['status'] = resolved
 253         elif s == 'Deleted':
 254             d['status'] = resolved
 255             d['is retired'] = True
 256         else:
 257             d['status'] = unread
 258
 259         nosy = set()
 260         for message in artifact.get('messages', []):
 261             authid = users[message['user_name']]
 262             if not message['body']: continue
 263             body = convert_message(message['body'], message_id)
 264             if not body: continue
 265             m = {'content': body, 'author': authid,
 266                 'date': message['adddate'],
 267                 'creation': message['adddate'], }
 268             message_data.append(m)
 269             if authid not in (None, '2'):
 270                 nosy.add(authid)
 271             activity = message['adddate']
 272             actor = authid
 273             if d['status'] == unread:
 274                 d['status'] = chatting
 275
 276         # add import message
 277         m = {'content': 'IMPORT FROM SOURCEFORGE', 'author': '1',
 278             'date': today, 'creation': today}
 279         message_data.append(m)
 280
 281         # sort messages and assign ids
 282         d['messages'] = []
 283         message_data.sort(lambda a,b:cmp(a['date'],b['date']))
 284         for message in message_data:
 285             message_id += 1
 286             message['id'] = str(message_id)
 287             d['messages'].append(message_id)
 288
 289         d['nosy'] = list(nosy)
 290
 291         files = []
 292         for event in artifact.get('history', []):
 293             if event['field_name'] == 'File Added':
 294                 fid, name = event['old_value'].split(':', 1)
 295                 if fid in add_files:
 296                     files.append(fid)
 297                     name = name.strip()
 298                     try:
 299                         f = open(os.path.join(file_dir, fid))
 300                         content = f.read()
 301                         f.close()
 302                     except:
 303                         content = 'content missing'
 304                     file_data.append({
 305                         'id': fid,
 306                         'creation': event['entrydate'],
 307                         'creator': users[event['mod_by']],
 308                         'name': name,
 309                         'type': mimetypes.guess_type(name)[0],
 310                         'content': content,
 311                     })
 312                 continue
 313             elif event['field_name'] == 'close_date':
 314                 action = "'set'"
 315                 info = { 'status': unread }
 316             elif event['field_name'] == 'summary':
 317                 action = "'set'"
 318                 info = { 'title': event['old_value'] }
 319             else:
 320                 # not an interesting / translatable event
 321                 continue
 322             row = [ d['id'], event['entrydate'].get_tuple(),
 323                 users[event['mod_by']], action, info ]
 324             if event['entrydate'] > activity:
 325                 activity = event['entrydate']
 326             issue_journal.append(row)
 327         d['files'] = files
 328
 329         d['activity'] = activity
 330         d['actor'] = actor
 331         issue_data.append(d)
 332
 333     write_csv(db.issue, issue_data)
 334     write_csv(db.msg, message_data)
 335     write_csv(db.file, file_data)
 336
 337     f = open('/tmp/imported/issue-journals.csv', 'w')
 338     writer = csv.writer(f, colon_separated)
 339     writer.writerows(issue_journal)
 340     f.close()
 341
 342 def convert_message(content, id):
 343     """ Strip off the useless sf message header crap """
 344     if content[:14] == 'Logged In: YES':
 345         return '\n'.join(content.splitlines()[3:]).strip()
 346     return content
 347
 348 class colon_separated(csv.excel):
 349     delimiter = ':'
 350
 351 def write_csv(klass, data):
 352     props = klass.getprops()
 353     if not os.path.exists('/tmp/imported'):
 354         os.mkdir('/tmp/imported')
 355     f = open('/tmp/imported/%s.csv'%klass.classname, 'w')
 356     writer = csv.writer(f, colon_separated)
 357     propnames = klass.export_propnames()
 358     propnames.append('is retired')
 359     writer.writerow(propnames)
 360     for entry in data:
 361         row = []
 362         for name in propnames:
 363             if name == 'is retired':
 364                 continue
 365             prop = props[name]
 366             if entry.has_key(name):
 367                 if isinstance(prop, hyperdb.Date) or \
 368                         isinstance(prop, hyperdb.Interval):
 369                     row.append(repr(entry[name].get_tuple()))
 370                 elif isinstance(prop, hyperdb.Password):
 371                     row.append(repr(str(entry[name])))
 372                 else:
 373                     row.append(repr(entry[name]))
 374             elif isinstance(prop, hyperdb.Multilink):
 375                 row.append('[]')
 376             elif name in ('creator', 'actor'):
 377                 row.append("'1'")
 378             elif name in ('created', 'activity'):
 379                 row.append(repr(today.get_tuple()))
 380             else:
 381                 row.append('None')
 382         row.append(entry.get('is retired', False))
 383         writer.writerow(row)
 384
 385         if isinstance(klass, hyperdb.FileClass) and entry.get('content'):
 386             fname = klass.exportFilename('/tmp/imported/', entry['id'])
 387             support.ensureParentsExist(fname)
 388             c = open(fname, 'w')
 389             if isinstance(entry['content'], unicode):
 390                 c.write(entry['content'].encode('utf8'))
 391             else:
 392                 c.write(entry['content'])
 393             c.close()
 394
 395     f.close()
 396     f = open('/tmp/imported/%s-journals.csv'%klass.classname, 'w')
 397     f.close()
 398
 399 if __name__ == '__main__':
 400     if sys.argv[1] == 'import':
 401         import_xml(*sys.argv[2:])
 402     elif sys.argv[1] == 'files':
 403         fetch_files(*sys.argv[2:])
 404