gitstats

   1 #!/usr/bin/env python
   2 # Copyright (c) 2007-2012 Heikki Hokkanen <hoxu@users.sf.net> & others (see doc/author.txt)
   3 # GPLv2 / GPLv3
   4 import datetime
   5 import getopt
   6 import glob
   7 import os
   8 import pickle
   9 import platform
  10 import re
  11 import shutil
  12 import subprocess
  13 import sys
  14 import time
  15 import zlib
  16
  17 os.environ['LC_ALL'] = 'C'
  18
  19 GNUPLOT_COMMON = 'set terminal png transparent size 640,240\nset size 1.0,1.0\n'
  20 ON_LINUX = (platform.system() == 'Linux')
  21 WEEKDAYS = ('Mon', 'Tue', 'Wed', 'Thu', 'Fri', 'Sat', 'Sun')
  22
  23 exectime_internal = 0.0
  24 exectime_external = 0.0
  25 time_start = time.time()
  26
  27 # By default, gnuplot is searched from path, but can be overridden with the
  28 # environment variable "GNUPLOT"
  29 gnuplot_cmd = 'gnuplot'
  30 if 'GNUPLOT' in os.environ:
  31         gnuplot_cmd = os.environ['GNUPLOT']
  32
  33 conf = {
  34         'max_domains': 10,
  35         'max_ext_length': 10,
  36         'style': 'gitstats.css',
  37         'max_authors': 20,
  38         'authors_top': 5,
  39         'commit_begin': '',
  40         'commit_end': 'HEAD',
  41         'linear_linestats': 1,
  42         'project_name': '',
  43 }
  44
  45 def getpipeoutput(cmds, quiet = False):
  46         global exectime_external
  47         start = time.time()
  48         if not quiet and ON_LINUX and os.isatty(1):
  49                 print '>> ' + ' | '.join(cmds),
  50                 sys.stdout.flush()
  51         p0 = subprocess.Popen(cmds[0], stdout = subprocess.PIPE, shell = True)
  52         p = p0
  53         for x in cmds[1:]:
  54                 p = subprocess.Popen(x, stdin = p0.stdout, stdout = subprocess.PIPE, shell = True)
  55                 p0 = p
  56         output = p.communicate()[0]
  57         end = time.time()
  58         if not quiet:
  59                 if ON_LINUX and os.isatty(1):
  60                         print '\r',
  61                 print '[%.5f] >> %s' % (end - start, ' | '.join(cmds))
  62         exectime_external += (end - start)
  63         return output.rstrip('\n')
  64
  65 def getcommitrange(defaultrange = 'HEAD', end_only = False):
  66         if len(conf['commit_end']) > 0:
  67                 if end_only or len(conf['commit_begin']) == 0:
  68                         return conf['commit_end']
  69                 return '%s..%s' % (conf['commit_begin'], conf['commit_end'])
  70         return defaultrange
  71
  72 def getkeyssortedbyvalues(dict):
  73         return map(lambda el : el[1], sorted(map(lambda el : (el[1], el[0]), dict.items())))
  74
  75 # dict['author'] = { 'commits': 512 } - ...key(dict, 'commits')
  76 def getkeyssortedbyvaluekey(d, key):
  77         return map(lambda el : el[1], sorted(map(lambda el : (d[el][key], el), d.keys())))
  78
  79 def getstatsummarycounts(line):
  80         numbers = re.findall('\d+', line)
  81         if   len(numbers) == 1:
  82                 # neither insertions nor deletions: may probably only happen for "0 files changed"
  83                 numbers.append(0);
  84                 numbers.append(0);
  85         elif len(numbers) == 2 and line.find('(+)') != -1:
  86                 numbers.append(0);    # only insertions were printed on line
  87         elif len(numbers) == 2 and line.find('(-)') != -1:
  88                 numbers.insert(1, 0); # only deletions were printed on line
  89         return numbers
  90
  91 VERSION = 0
  92 def getversion():
  93         global VERSION
  94         if VERSION == 0:
  95                 gitstats_repo = os.path.dirname(os.path.abspath(__file__))
  96                 VERSION = getpipeoutput(["git --git-dir=%s/.git --work-tree=%s rev-parse --short %s" %
  97                         (gitstats_repo, gitstats_repo, getcommitrange('HEAD').split('\n')[0])])
  98         return VERSION
  99
 100 def getgitversion():
 101         return getpipeoutput(['git --version']).split('\n')[0]
 102
 103 def getgnuplotversion():
 104         return getpipeoutput(['%s --version' % gnuplot_cmd]).split('\n')[0]
 105
 106 class DataCollector:
 107         """Manages data collection from a revision control repository."""
 108         def __init__(self):
 109                 self.stamp_created = time.time()
 110                 self.cache = {}
 111                 self.total_authors = 0
 112                 self.activity_by_hour_of_day = {} # hour -> commits
 113                 self.activity_by_day_of_week = {} # day -> commits
 114                 self.activity_by_month_of_year = {} # month [1-12] -> commits
 115                 self.activity_by_hour_of_week = {} # weekday -> hour -> commits
 116                 self.activity_by_hour_of_day_busiest = 0
 117                 self.activity_by_hour_of_week_busiest = 0
 118                 self.activity_by_year_week = {} # yy_wNN -> commits
 119                 self.activity_by_year_week_peak = 0
 120
 121                 self.authors = {} # name -> {commits, first_commit_stamp, last_commit_stamp, last_active_day, active_days, lines_added, lines_removed}
 122
 123                 self.total_commits = 0
 124                 self.total_files = 0
 125                 self.authors_by_commits = 0
 126
 127                 # domains
 128                 self.domains = {} # domain -> commits
 129
 130                 # author of the month
 131                 self.author_of_month = {} # month -> author -> commits
 132                 self.author_of_year = {} # year -> author -> commits
 133                 self.commits_by_month = {} # month -> commits
 134                 self.commits_by_year = {} # year -> commits
 135                 self.lines_added_by_month = {} # month -> lines added
 136                 self.lines_added_by_year = {} # year -> lines added
 137                 self.lines_removed_by_month = {} # month -> lines removed
 138                 self.lines_removed_by_year = {} # year -> lines removed
 139                 self.first_commit_stamp = 0
 140                 self.last_commit_stamp = 0
 141                 self.last_active_day = None
 142                 self.active_days = set()
 143
 144                 # lines
 145                 self.total_lines = 0
 146                 self.total_lines_added = 0
 147                 self.total_lines_removed = 0
 148
 149                 # size
 150                 self.total_size = 0
 151
 152                 # timezone
 153                 self.commits_by_timezone = {} # timezone -> commits
 154
 155                 # tags
 156                 self.tags = {}
 157
 158                 self.files_by_stamp = {} # stamp -> files
 159
 160                 # extensions
 161                 self.extensions = {} # extension -> files, lines
 162
 163                 # line statistics
 164                 self.changes_by_date = {} # stamp -> { files, ins, del }
 165
 166         ##
 167         # This should be the main function to extract data from the repository.
 168         def collect(self, dir):
 169                 self.dir = dir
 170                 if len(conf['project_name']) == 0:
 171                         self.projectname = os.path.basename(os.path.abspath(dir))
 172                 else:
 173                         self.projectname = conf['project_name']
 174
 175         ##
 176         # Load cacheable data
 177         def loadCache(self, cachefile):
 178                 if not os.path.exists(cachefile):
 179                         return
 180                 print 'Loading cache...'
 181                 f = open(cachefile, 'rb')
 182                 try:
 183                         self.cache = pickle.loads(zlib.decompress(f.read()))
 184                 except:
 185                         # temporary hack to upgrade non-compressed caches
 186                         f.seek(0)
 187                         self.cache = pickle.load(f)
 188                 f.close()
 189
 190         ##
 191         # Produce any additional statistics from the extracted data.
 192         def refine(self):
 193                 pass
 194
 195         ##
 196         # : get a dictionary of author
 197         def getAuthorInfo(self, author):
 198                 return None
 199
 200         def getActivityByDayOfWeek(self):
 201                 return {}
 202
 203         def getActivityByHourOfDay(self):
 204                 return {}
 205
 206         # : get a dictionary of domains
 207         def getDomainInfo(self, domain):
 208                 return None
 209
 210         ##
 211         # Get a list of authors
 212         def getAuthors(self):
 213                 return []
 214
 215         def getFirstCommitDate(self):
 216                 return datetime.datetime.now()
 217
 218         def getLastCommitDate(self):
 219                 return datetime.datetime.now()
 220
 221         def getStampCreated(self):
 222                 return self.stamp_created
 223
 224         def getTags(self):
 225                 return []
 226
 227         def getTotalAuthors(self):
 228                 return -1
 229
 230         def getTotalCommits(self):
 231                 return -1
 232
 233         def getTotalFiles(self):
 234                 return -1
 235
 236         def getTotalLOC(self):
 237                 return -1
 238
 239         ##
 240         # Save cacheable data
 241         def saveCache(self, cachefile):
 242                 print 'Saving cache...'
 243                 f = open(cachefile, 'wb')
 244                 #pickle.dump(self.cache, f)
 245                 data = zlib.compress(pickle.dumps(self.cache))
 246                 f.write(data)
 247                 f.close()
 248
 249 class GitDataCollector(DataCollector):
 250         def collect(self, dir):
 251                 DataCollector.collect(self, dir)
 252
 253                 self.total_authors += int(getpipeoutput(['git shortlog -s %s' % getcommitrange(), 'wc -l']))
 254                 #self.total_lines = int(getoutput('git-ls-files -z |xargs -0 cat |wc -l'))
 255
 256                 # tags
 257                 lines = getpipeoutput(['git show-ref --tags']).split('\n')
 258                 for line in lines:
 259                         if len(line) == 0:
 260                                 continue
 261                         (hash, tag) = line.split(' ')
 262
 263                         tag = tag.replace('refs/tags/', '')
 264                         output = getpipeoutput(['git log "%s" --pretty=format:"%%at %%aN" -n 1' % hash])
 265                         if len(output) > 0:
 266                                 parts = output.split(' ')
 267                                 stamp = 0
 268                                 try:
 269                                         stamp = int(parts[0])
 270                                 except ValueError:
 271                                         stamp = 0
 272                                 self.tags[tag] = { 'stamp': stamp, 'hash' : hash, 'date' : datetime.datetime.fromtimestamp(stamp).strftime('%Y-%m-%d'), 'commits': 0, 'authors': {} }
 273
 274                 # collect info on tags, starting from latest
 275                 tags_sorted_by_date_desc = map(lambda el : el[1], reversed(sorted(map(lambda el : (el[1]['date'], el[0]), self.tags.items()))))
 276                 prev = None
 277                 for tag in reversed(tags_sorted_by_date_desc):
 278                         cmd = 'git shortlog -s "%s"' % tag
 279                         if prev != None:
 280                                 cmd += ' "^%s"' % prev
 281                         output = getpipeoutput([cmd])
 282                         if len(output) == 0:
 283                                 continue
 284                         prev = tag
 285                         for line in output.split('\n'):
 286                                 parts = re.split('\s+', line, 2)
 287                                 commits = int(parts[1])
 288                                 author = parts[2]
 289                                 self.tags[tag]['commits'] += commits
 290                                 self.tags[tag]['authors'][author] = commits
 291
 292                 # Collect revision statistics
 293                 # Outputs "<stamp> <date> <time> <timezone> <author> '<' <mail> '>'"
 294                 lines = getpipeoutput(['git rev-list --pretty=format:"%%at %%ai %%aN <%%aE>" %s' % getcommitrange('HEAD'), 'grep -v ^commit']).split('\n')
 295                 for line in lines:
 296                         parts = line.split(' ', 4)
 297                         author = ''
 298                         try:
 299                                 stamp = int(parts[0])
 300                         except ValueError:
 301                                 stamp = 0
 302                         timezone = parts[3]
 303                         author, mail = parts[4].split('<', 1)
 304                         author = author.rstrip()
 305                         mail = mail.rstrip('>')
 306                         domain = '?'
 307                         if mail.find('@') != -1:
 308                                 domain = mail.rsplit('@', 1)[1]
 309                         date = datetime.datetime.fromtimestamp(float(stamp))
 310
 311                         # First and last commit stamp (may be in any order because of cherry-picking and patches)
 312                         if stamp > self.last_commit_stamp:
 313                                 self.last_commit_stamp = stamp
 314                         if self.first_commit_stamp == 0 or stamp < self.first_commit_stamp:
 315                                 self.first_commit_stamp = stamp
 316
 317                         # activity
 318                         # hour
 319                         hour = date.hour
 320                         self.activity_by_hour_of_day[hour] = self.activity_by_hour_of_day.get(hour, 0) + 1
 321                         # most active hour?
 322                         if self.activity_by_hour_of_day[hour] > self.activity_by_hour_of_day_busiest:
 323                                 self.activity_by_hour_of_day_busiest = self.activity_by_hour_of_day[hour]
 324
 325                         # day of week
 326                         day = date.weekday()
 327                         self.activity_by_day_of_week[day] = self.activity_by_day_of_week.get(day, 0) + 1
 328
 329                         # domain stats
 330                         if domain not in self.domains:
 331                                 self.domains[domain] = {}
 332                         # commits
 333                         self.domains[domain]['commits'] = self.domains[domain].get('commits', 0) + 1
 334
 335                         # hour of week
 336                         if day not in self.activity_by_hour_of_week:
 337                                 self.activity_by_hour_of_week[day] = {}
 338                         self.activity_by_hour_of_week[day][hour] = self.activity_by_hour_of_week[day].get(hour, 0) + 1
 339                         # most active hour?
 340                         if self.activity_by_hour_of_week[day][hour] > self.activity_by_hour_of_week_busiest:
 341                                 self.activity_by_hour_of_week_busiest = self.activity_by_hour_of_week[day][hour]
 342
 343                         # month of year
 344                         month = date.month
 345                         self.activity_by_month_of_year[month] = self.activity_by_month_of_year.get(month, 0) + 1
 346
 347                         # yearly/weekly activity
 348                         yyw = date.strftime('%Y-%W')
 349                         self.activity_by_year_week[yyw] = self.activity_by_year_week.get(yyw, 0) + 1
 350                         if self.activity_by_year_week_peak < self.activity_by_year_week[yyw]:
 351                                 self.activity_by_year_week_peak = self.activity_by_year_week[yyw]
 352
 353                         # author stats
 354                         if author not in self.authors:
 355                                 self.authors[author] = {}
 356                         # commits, note again that commits may be in any date order because of cherry-picking and patches
 357                         if 'last_commit_stamp' not in self.authors[author]:
 358                                 self.authors[author]['last_commit_stamp'] = stamp
 359                         if stamp > self.authors[author]['last_commit_stamp']:
 360                                 self.authors[author]['last_commit_stamp'] = stamp
 361                         if 'first_commit_stamp' not in self.authors[author]:
 362                                 self.authors[author]['first_commit_stamp'] = stamp
 363                         if stamp < self.authors[author]['first_commit_stamp']:
 364                                 self.authors[author]['first_commit_stamp'] = stamp
 365
 366                         # author of the month/year
 367                         yymm = date.strftime('%Y-%m')
 368                         if yymm in self.author_of_month:
 369                                 self.author_of_month[yymm][author] = self.author_of_month[yymm].get(author, 0) + 1
 370                         else:
 371                                 self.author_of_month[yymm] = {}
 372                                 self.author_of_month[yymm][author] = 1
 373                         self.commits_by_month[yymm] = self.commits_by_month.get(yymm, 0) + 1
 374
 375                         yy = date.year
 376                         if yy in self.author_of_year:
 377                                 self.author_of_year[yy][author] = self.author_of_year[yy].get(author, 0) + 1
 378                         else:
 379                                 self.author_of_year[yy] = {}
 380                                 self.author_of_year[yy][author] = 1
 381                         self.commits_by_year[yy] = self.commits_by_year.get(yy, 0) + 1
 382
 383                         # authors: active days
 384                         yymmdd = date.strftime('%Y-%m-%d')
 385                         if 'last_active_day' not in self.authors[author]:
 386                                 self.authors[author]['last_active_day'] = yymmdd
 387                                 self.authors[author]['active_days'] = set([yymmdd])
 388                         elif yymmdd != self.authors[author]['last_active_day']:
 389                                 self.authors[author]['last_active_day'] = yymmdd
 390                                 self.authors[author]['active_days'].add(yymmdd)
 391
 392                         # project: active days
 393                         if yymmdd != self.last_active_day:
 394                                 self.last_active_day = yymmdd
 395                                 self.active_days.add(yymmdd)
 396
 397                         # timezone
 398                         self.commits_by_timezone[timezone] = self.commits_by_timezone.get(timezone, 0) + 1
 399
 400                 # TODO Optimize this, it's the worst bottleneck
 401                 # outputs "<stamp> <files>" for each revision
 402                 revlines = getpipeoutput(['git rev-list --pretty=format:"%%at %%T" %s' % getcommitrange('HEAD'), 'grep -v ^commit']).strip().split('\n')
 403                 lines = []
 404                 for revline in revlines:
 405                         time, rev = revline.split(' ')
 406                         linecount = self.getFilesInCommit(rev)
 407                         lines.append('%d %d' % (int(time), linecount))
 408
 409                 self.total_commits += len(lines)
 410                 for line in lines:
 411                         parts = line.split(' ')
 412                         if len(parts) != 2:
 413                                 continue
 414                         (stamp, files) = parts[0:2]
 415                         try:
 416                                 self.files_by_stamp[int(stamp)] = int(files)
 417                         except ValueError:
 418                                 print 'Warning: failed to parse line "%s"' % line
 419
 420                 # extensions and size of files
 421                 lines = getpipeoutput(['git ls-tree -r -l -z %s' % getcommitrange('HEAD', end_only = True)]).split('\000')
 422                 for line in lines:
 423                         if len(line) == 0:
 424                                 continue
 425                         parts = re.split('\s+', line, 5)
 426                         if parts[0] == '160000' and parts[3] == '-':
 427                                 # skip submodules
 428                                 continue
 429                         sha1 = parts[2]
 430                         size = int(parts[3])
 431                         fullpath = parts[4]
 432
 433                         self.total_size += size
 434                         self.total_files += 1
 435
 436                         filename = fullpath.split('/')[-1] # strip directories
 437                         if filename.find('.') == -1 or filename.rfind('.') == 0:
 438                                 ext = ''
 439                         else:
 440                                 ext = filename[(filename.rfind('.') + 1):]
 441                         if len(ext) > conf['max_ext_length']:
 442                                 ext = ''
 443
 444                         if ext not in self.extensions:
 445                                 self.extensions[ext] = {'files': 0, 'lines': 0}
 446
 447                         self.extensions[ext]['files'] += 1
 448                         try:
 449                                 self.extensions[ext]['lines'] += self.getLinesInBlob(sha1)
 450                         except:
 451                                 print 'Warning: Could not count lines for file "%s"' % line
 452
 453                 # line statistics
 454                 # outputs:
 455                 #  N files changed, N insertions (+), N deletions(-)
 456                 # <stamp> <author>
 457                 self.changes_by_date = {} # stamp -> { files, ins, del }
 458                 # computation of lines of code by date is better done
 459                 # on a linear history.
 460                 extra = ''
 461                 if conf['linear_linestats']:
 462                         extra = '--first-parent -m'
 463                 lines = getpipeoutput(['git log --shortstat %s --pretty=format:"%%at %%aN" %s' % (extra, getcommitrange('HEAD'))]).split('\n')
 464                 lines.reverse()
 465                 files = 0; inserted = 0; deleted = 0; total_lines = 0
 466                 author = None
 467                 for line in lines:
 468                         if len(line) == 0:
 469                                 continue
 470
 471                         # <stamp> <author>
 472                         if re.search('files? changed', line) == None:
 473                                 pos = line.find(' ')
 474                                 if pos != -1:
 475                                         try:
 476                                                 (stamp, author) = (int(line[:pos]), line[pos+1:])
 477                                                 self.changes_by_date[stamp] = { 'files': files, 'ins': inserted, 'del': deleted, 'lines': total_lines }
 478
 479                                                 date = datetime.datetime.fromtimestamp(stamp)
 480                                                 yymm = date.strftime('%Y-%m')
 481                                                 self.lines_added_by_month[yymm] = self.lines_added_by_month.get(yymm, 0) + inserted
 482                                                 self.lines_removed_by_month[yymm] = self.lines_removed_by_month.get(yymm, 0) + deleted
 483
 484                                                 yy = date.year
 485                                                 self.lines_added_by_year[yy] = self.lines_added_by_year.get(yy,0) + inserted
 486                                                 self.lines_removed_by_year[yy] = self.lines_removed_by_year.get(yy, 0) + deleted
 487
 488                                                 files, inserted, deleted = 0, 0, 0
 489                                         except ValueError:
 490                                                 print 'Warning: unexpected line "%s"' % line
 491                                 else:
 492                                         print 'Warning: unexpected line "%s"' % line
 493                         else:
 494                                 numbers = getstatsummarycounts(line)
 495
 496                                 if len(numbers) == 3:
 497                                         (files, inserted, deleted) = map(lambda el : int(el), numbers)
 498                                         total_lines += inserted
 499                                         total_lines -= deleted
 500                                         self.total_lines_added += inserted
 501                                         self.total_lines_removed += deleted
 502
 503                                 else:
 504                                         print 'Warning: failed to handle line "%s"' % line
 505                                         (files, inserted, deleted) = (0, 0, 0)
 506                                 #self.changes_by_date[stamp] = { 'files': files, 'ins': inserted, 'del': deleted }
 507                 self.total_lines = total_lines
 508
 509                 # Per-author statistics
 510
 511                 # defined for stamp, author only if author commited at this timestamp.
 512                 self.changes_by_date_by_author = {} # stamp -> author -> lines_added
 513
 514                 # Similar to the above, but never use --first-parent
 515                 # (we need to walk through every commit to know who
 516                 # committed what, not just through mainline)
 517                 lines = getpipeoutput(['git log --shortstat --date-order --pretty=format:"%%at %%aN" %s' % (getcommitrange('HEAD'))]).split('\n')
 518                 lines.reverse()
 519                 files = 0; inserted = 0; deleted = 0
 520                 author = None
 521                 stamp = 0
 522                 for line in lines:
 523                         if len(line) == 0:
 524                                 continue
 525
 526                         # <stamp> <author>
 527                         if re.search('files? changed', line) == None:
 528                                 pos = line.find(' ')
 529                                 if pos != -1:
 530                                         try:
 531                                                 oldstamp = stamp
 532                                                 (stamp, author) = (int(line[:pos]), line[pos+1:])
 533                                                 if oldstamp > stamp:
 534                                                         # clock skew, keep old timestamp to avoid having ugly graph
 535                                                         stamp = oldstamp
 536                                                 if author not in self.authors:
 537                                                         self.authors[author] = { 'lines_added' : 0, 'lines_removed' : 0, 'commits' : 0}
 538                                                 self.authors[author]['commits'] = self.authors[author].get('commits', 0) + 1
 539                                                 self.authors[author]['lines_added'] = self.authors[author].get('lines_added', 0) + inserted
 540                                                 self.authors[author]['lines_removed'] = self.authors[author].get('lines_removed', 0) + deleted
 541                                                 if stamp not in self.changes_by_date_by_author:
 542                                                         self.changes_by_date_by_author[stamp] = {}
 543                                                 if author not in self.changes_by_date_by_author[stamp]:
 544                                                         self.changes_by_date_by_author[stamp][author] = {}
 545                                                 self.changes_by_date_by_author[stamp][author]['lines_added'] = self.authors[author]['lines_added']
 546                                                 self.changes_by_date_by_author[stamp][author]['commits'] = self.authors[author]['commits']
 547                                                 files, inserted, deleted = 0, 0, 0
 548                                         except ValueError:
 549                                                 print 'Warning: unexpected line "%s"' % line
 550                                 else:
 551                                         print 'Warning: unexpected line "%s"' % line
 552                         else:
 553                                 numbers = getstatsummarycounts(line);
 554
 555                                 if len(numbers) == 3:
 556                                         (files, inserted, deleted) = map(lambda el : int(el), numbers)
 557                                 else:
 558                                         print 'Warning: failed to handle line "%s"' % line
 559                                         (files, inserted, deleted) = (0, 0, 0)
 560
 561         def refine(self):
 562                 # authors
 563                 # name -> {place_by_commits, commits_frac, date_first, date_last, timedelta}
 564                 self.authors_by_commits = getkeyssortedbyvaluekey(self.authors, 'commits')
 565                 self.authors_by_commits.reverse() # most first
 566                 for i, name in enumerate(self.authors_by_commits):
 567                         self.authors[name]['place_by_commits'] = i + 1
 568
 569                 for name in self.authors.keys():
 570                         a = self.authors[name]
 571                         a['commits_frac'] = (100 * float(a['commits'])) / self.getTotalCommits()
 572                         date_first = datetime.datetime.fromtimestamp(a['first_commit_stamp'])
 573                         date_last = datetime.datetime.fromtimestamp(a['last_commit_stamp'])
 574                         delta = date_last - date_first
 575                         a['date_first'] = date_first.strftime('%Y-%m-%d')
 576                         a['date_last'] = date_last.strftime('%Y-%m-%d')
 577                         a['timedelta'] = delta
 578                         if 'lines_added' not in a: a['lines_added'] = 0
 579                         if 'lines_removed' not in a: a['lines_removed'] = 0
 580
 581         def getActiveDays(self):
 582                 return self.active_days
 583
 584         def getActivityByDayOfWeek(self):
 585                 return self.activity_by_day_of_week
 586
 587         def getActivityByHourOfDay(self):
 588                 return self.activity_by_hour_of_day
 589
 590         def getAuthorInfo(self, author):
 591                 return self.authors[author]
 592
 593         def getAuthors(self, limit = None):
 594                 res = getkeyssortedbyvaluekey(self.authors, 'commits')
 595                 res.reverse()
 596                 return res[:limit]
 597
 598         def getCommitDeltaDays(self):
 599                 return (self.last_commit_stamp / 86400 - self.first_commit_stamp / 86400) + 1
 600
 601         def getDomainInfo(self, domain):
 602                 return self.domains[domain]
 603
 604         def getDomains(self):
 605                 return self.domains.keys()
 606
 607         def getFilesInCommit(self, rev):
 608                 try:
 609                         res = self.cache['files_in_tree'][rev]
 610                 except:
 611                         res = int(getpipeoutput(['git ls-tree -r --name-only "%s"' % rev, 'wc -l']).split('\n')[0])
 612                         if 'files_in_tree' not in self.cache:
 613                                 self.cache['files_in_tree'] = {}
 614                         self.cache['files_in_tree'][rev] = res
 615
 616                 return res
 617
 618         def getFirstCommitDate(self):
 619                 return datetime.datetime.fromtimestamp(self.first_commit_stamp)
 620
 621         def getLastCommitDate(self):
 622                 return datetime.datetime.fromtimestamp(self.last_commit_stamp)
 623
 624         def getLinesInBlob(self, sha1):
 625                 try:
 626                         res = self.cache['lines_in_blob'][sha1]
 627                 except:
 628                         res = int(getpipeoutput(['git cat-file blob %s' % sha1, 'wc -l']).split()[0])
 629                         if 'lines_in_blob' not in self.cache:
 630                                 self.cache['lines_in_blob'] = {}
 631                         self.cache['lines_in_blob'][sha1] = res
 632                 return res
 633
 634         def getTags(self):
 635                 lines = getpipeoutput(['git show-ref --tags', 'cut -d/ -f3'])
 636                 return lines.split('\n')
 637
 638         def getTagDate(self, tag):
 639                 return self.revToDate('tags/' + tag)
 640
 641         def getTotalAuthors(self):
 642                 return self.total_authors
 643
 644         def getTotalCommits(self):
 645                 return self.total_commits
 646
 647         def getTotalFiles(self):
 648                 return self.total_files
 649
 650         def getTotalLOC(self):
 651                 return self.total_lines
 652
 653         def getTotalSize(self):
 654                 return self.total_size
 655
 656         def revToDate(self, rev):
 657                 stamp = int(getpipeoutput(['git log --pretty=format:%%at "%s" -n 1' % rev]))
 658                 return datetime.datetime.fromtimestamp(stamp).strftime('%Y-%m-%d')
 659
 660 class ReportCreator:
 661         """Creates the actual report based on given data."""
 662         def __init__(self):
 663                 pass
 664
 665         def create(self, data, path):
 666                 self.data = data
 667                 self.path = path
 668
 669 def html_linkify(text):
 670         return text.lower().replace(' ', '_')
 671
 672 def html_header(level, text):
 673         name = html_linkify(text)
 674         return '\n<h%d><a href="#%s" name="%s">%s</a></h%d>\n\n' % (level, name, name, text, level)
 675
 676 class HTMLReportCreator(ReportCreator):
 677         def create(self, data, path):
 678                 ReportCreator.create(self, data, path)
 679                 self.title = data.projectname
 680
 681                 # copy static files. Looks in the binary directory, ../share/gitstats and /usr/share/gitstats
 682                 binarypath = os.path.dirname(os.path.abspath(__file__))
 683                 secondarypath = os.path.join(binarypath, '..', 'share', 'gitstats')
 684                 basedirs = [binarypath, secondarypath, '/usr/share/gitstats']
 685                 for file in ('gitstats.css', 'sortable.js', 'arrow-up.gif', 'arrow-down.gif', 'arrow-none.gif'):
 686                         for base in basedirs:
 687                                 src = base + '/' + file
 688                                 if os.path.exists(src):
 689                                         shutil.copyfile(src, path + '/' + file)
 690                                         break
 691                         else:
 692                                 print 'Warning: "%s" not found, so not copied (searched: %s)' % (file, basedirs)
 693
 694                 f = open(path + "/index.html", 'w')
 695                 format = '%Y-%m-%d %H:%M:%S'
 696                 self.printHeader(f)
 697
 698                 f.write('<h1>GitStats - %s</h1>' % data.projectname)
 699
 700                 self.printNav(f)
 701
 702                 f.write('<dl>')
 703                 f.write('<dt>Project name</dt><dd>%s</dd>' % (data.projectname))
 704                 f.write('<dt>Generated</dt><dd>%s (in %d seconds)</dd>' % (datetime.datetime.now().strftime(format), time.time() - data.getStampCreated()))
 705                 f.write('<dt>Generator</dt><dd><a href="http://gitstats.sourceforge.net/">GitStats</a> (version %s), %s, %s</dd>' % (getversion(), getgitversion(), getgnuplotversion()))
 706                 f.write('<dt>Report Period</dt><dd>%s to %s</dd>' % (data.getFirstCommitDate().strftime(format), data.getLastCommitDate().strftime(format)))
 707                 f.write('<dt>Age</dt><dd>%d days, %d active days (%3.2f%%)</dd>' % (data.getCommitDeltaDays(), len(data.getActiveDays()), (100.0 * len(data.getActiveDays()) / data.getCommitDeltaDays())))
 708                 f.write('<dt>Total Files</dt><dd>%s</dd>' % data.getTotalFiles())
 709                 f.write('<dt>Total Lines of Code</dt><dd>%s (%d added, %d removed)</dd>' % (data.getTotalLOC(), data.total_lines_added, data.total_lines_removed))
 710                 f.write('<dt>Total Commits</dt><dd>%s (average %.1f commits per active day, %.1f per all days)</dd>' % (data.getTotalCommits(), float(data.getTotalCommits()) / len(data.getActiveDays()), float(data.getTotalCommits()) / data.getCommitDeltaDays()))
 711                 f.write('<dt>Authors</dt><dd>%s (average %.1f commits per author)</dd>' % (data.getTotalAuthors(), (1.0 * data.getTotalCommits()) / data.getTotalAuthors()))
 712                 f.write('</dl>')
 713
 714                 f.write('</body>\n</html>')
 715                 f.close()
 716
 717                 ###
 718                 # Activity
 719                 f = open(path + '/activity.html', 'w')
 720                 self.printHeader(f)
 721                 f.write('<h1>Activity</h1>')
 722                 self.printNav(f)
 723
 724                 #f.write('<h2>Last 30 days</h2>')
 725
 726                 #f.write('<h2>Last 12 months</h2>')
 727
 728                 # Weekly activity
 729                 WEEKS = 32
 730                 f.write(html_header(2, 'Weekly activity'))
 731                 f.write('<p>Last %d weeks</p>' % WEEKS)
 732
 733                 # generate weeks to show (previous N weeks from now)
 734                 now = datetime.datetime.now()
 735                 deltaweek = datetime.timedelta(7)
 736                 weeks = []
 737                 stampcur = now
 738                 for i in range(0, WEEKS):
 739                         weeks.insert(0, stampcur.strftime('%Y-%W'))
 740                         stampcur -= deltaweek
 741
 742                 # top row: commits & bar
 743                 f.write('<table class="noborders"><tr>')
 744                 for i in range(0, WEEKS):
 745                         commits = 0
 746                         if weeks[i] in data.activity_by_year_week:
 747                                 commits = data.activity_by_year_week[weeks[i]]
 748
 749                         percentage = 0
 750                         if weeks[i] in data.activity_by_year_week:
 751                                 percentage = float(data.activity_by_year_week[weeks[i]]) / data.activity_by_year_week_peak
 752                         height = max(1, int(200 * percentage))
 753                         f.write('<td style="text-align: center; vertical-align: bottom">%d<div style="display: block; background-color: red; width: 20px; height: %dpx"></div></td>' % (commits, height))
 754
 755                 # bottom row: year/week
 756                 f.write('</tr><tr>')
 757                 for i in range(0, WEEKS):
 758                         f.write('<td>%s</td>' % (WEEKS - i))
 759                 f.write('</tr></table>')
 760
 761                 # Hour of Day
 762                 f.write(html_header(2, 'Hour of Day'))
 763                 hour_of_day = data.getActivityByHourOfDay()
 764                 f.write('<table><tr><th>Hour</th>')
 765                 for i in range(0, 24):
 766                         f.write('<th>%d</th>' % i)
 767                 f.write('</tr>\n<tr><th>Commits</th>')
 768                 fp = open(path + '/hour_of_day.dat', 'w')
 769                 for i in range(0, 24):
 770                         if i in hour_of_day:
 771                                 r = 127 + int((float(hour_of_day[i]) / data.activity_by_hour_of_day_busiest) * 128)
 772                                 f.write('<td style="background-color: rgb(%d, 0, 0)">%d</td>' % (r, hour_of_day[i]))
 773                                 fp.write('%d %d\n' % (i, hour_of_day[i]))
 774                         else:
 775                                 f.write('<td>0</td>')
 776                                 fp.write('%d 0\n' % i)
 777                 fp.close()
 778                 f.write('</tr>\n<tr><th>%</th>')
 779                 totalcommits = data.getTotalCommits()
 780                 for i in range(0, 24):
 781                         if i in hour_of_day:
 782                                 r = 127 + int((float(hour_of_day[i]) / data.activity_by_hour_of_day_busiest) * 128)
 783                                 f.write('<td style="background-color: rgb(%d, 0, 0)">%.2f</td>' % (r, (100.0 * hour_of_day[i]) / totalcommits))
 784                         else:
 785                                 f.write('<td>0.00</td>')
 786                 f.write('</tr></table>')
 787                 f.write('<img src="hour_of_day.png" alt="Hour of Day" />')
 788                 fg = open(path + '/hour_of_day.dat', 'w')
 789                 for i in range(0, 24):
 790                         if i in hour_of_day:
 791                                 fg.write('%d %d\n' % (i + 1, hour_of_day[i]))
 792                         else:
 793                                 fg.write('%d 0\n' % (i + 1))
 794                 fg.close()
 795
 796                 # Day of Week
 797                 f.write(html_header(2, 'Day of Week'))
 798                 day_of_week = data.getActivityByDayOfWeek()
 799                 f.write('<div class="vtable"><table>')
 800                 f.write('<tr><th>Day</th><th>Total (%)</th></tr>')
 801                 fp = open(path + '/day_of_week.dat', 'w')
 802                 for d in range(0, 7):
 803                         commits = 0
 804                         if d in day_of_week:
 805                                 commits = day_of_week[d]
 806                         fp.write('%d %s %d\n' % (d + 1, WEEKDAYS[d], commits))
 807                         f.write('<tr>')
 808                         f.write('<th>%s</th>' % (WEEKDAYS[d]))
 809                         if d in day_of_week:
 810                                 f.write('<td>%d (%.2f%%)</td>' % (day_of_week[d], (100.0 * day_of_week[d]) / totalcommits))
 811                         else:
 812                                 f.write('<td>0</td>')
 813                         f.write('</tr>')
 814                 f.write('</table></div>')
 815                 f.write('<img src="day_of_week.png" alt="Day of Week" />')
 816                 fp.close()
 817
 818                 # Hour of Week
 819                 f.write(html_header(2, 'Hour of Week'))
 820                 f.write('<table>')
 821
 822                 f.write('<tr><th>Weekday</th>')
 823                 for hour in range(0, 24):
 824                         f.write('<th>%d</th>' % (hour))
 825                 f.write('</tr>')
 826
 827                 for weekday in range(0, 7):
 828                         f.write('<tr><th>%s</th>' % (WEEKDAYS[weekday]))
 829                         for hour in range(0, 24):
 830                                 try:
 831                                         commits = data.activity_by_hour_of_week[weekday][hour]
 832                                 except KeyError:
 833                                         commits = 0
 834                                 if commits != 0:
 835                                         f.write('<td')
 836                                         r = 127 + int((float(commits) / data.activity_by_hour_of_week_busiest) * 128)
 837                                         f.write(' style="background-color: rgb(%d, 0, 0)"' % r)
 838                                         f.write('>%d</td>' % commits)
 839                                 else:
 840                                         f.write('<td></td>')
 841                         f.write('</tr>')
 842
 843                 f.write('</table>')
 844
 845                 # Month of Year
 846                 f.write(html_header(2, 'Month of Year'))
 847                 f.write('<div class="vtable"><table>')
 848                 f.write('<tr><th>Month</th><th>Commits (%)</th></tr>')
 849                 fp = open (path + '/month_of_year.dat', 'w')
 850                 for mm in range(1, 13):
 851                         commits = 0
 852                         if mm in data.activity_by_month_of_year:
 853                                 commits = data.activity_by_month_of_year[mm]
 854                         f.write('<tr><td>%d</td><td>%d (%.2f %%)</td></tr>' % (mm, commits, (100.0 * commits) / data.getTotalCommits()))
 855                         fp.write('%d %d\n' % (mm, commits))
 856                 fp.close()
 857                 f.write('</table></div>')
 858                 f.write('<img src="month_of_year.png" alt="Month of Year" />')
 859
 860                 # Commits by year/month
 861                 f.write(html_header(2, 'Commits by year/month'))
 862                 f.write('<div class="vtable"><table><tr><th>Month</th><th>Commits</th><th>Lines added</th><th>Lines removed</th></tr>')
 863                 for yymm in reversed(sorted(data.commits_by_month.keys())):
 864                         f.write('<tr><td>%s</td><td>%d</td><td>%d</td><td>%d</td></tr>' % (yymm, data.commits_by_month.get(yymm,0), data.lines_added_by_month.get(yymm,0), data.lines_removed_by_month.get(yymm,0)))
 865                 f.write('</table></div>')
 866                 f.write('<img src="commits_by_year_month.png" alt="Commits by year/month" />')
 867                 fg = open(path + '/commits_by_year_month.dat', 'w')
 868                 for yymm in sorted(data.commits_by_month.keys()):
 869                         fg.write('%s %s\n' % (yymm, data.commits_by_month[yymm]))
 870                 fg.close()
 871
 872                 # Commits by year
 873                 f.write(html_header(2, 'Commits by Year'))
 874                 f.write('<div class="vtable"><table><tr><th>Year</th><th>Commits (% of all)</th><th>Lines added</th><th>Lines removed</th></tr>')
 875                 for yy in reversed(sorted(data.commits_by_year.keys())):
 876                         f.write('<tr><td>%s</td><td>%d (%.2f%%)</td><td>%d</td><td>%d</td></tr>' % (yy, data.commits_by_year.get(yy,0), (100.0 * data.commits_by_year.get(yy,0)) / data.getTotalCommits(), data.lines_added_by_year.get(yy,0), data.lines_removed_by_year.get(yy,0)))
 877                 f.write('</table></div>')
 878                 f.write('<img src="commits_by_year.png" alt="Commits by Year" />')
 879                 fg = open(path + '/commits_by_year.dat', 'w')
 880                 for yy in sorted(data.commits_by_year.keys()):
 881                         fg.write('%d %d\n' % (yy, data.commits_by_year[yy]))
 882                 fg.close()
 883
 884                 # Commits by timezone
 885                 f.write(html_header(2, 'Commits by Timezone'))
 886                 f.write('<table><tr>')
 887                 f.write('<th>Timezone</th><th>Commits</th>')
 888                 max_commits_on_tz = max(data.commits_by_timezone.values())
 889                 for i in sorted(data.commits_by_timezone.keys(), key = lambda n : int(n)):
 890                         commits = data.commits_by_timezone[i]
 891                         r = 127 + int((float(commits) / max_commits_on_tz) * 128)
 892                         f.write('<tr><th>%s</th><td style="background-color: rgb(%d, 0, 0)">%d</td></tr>' % (i, r, commits))
 893                 f.write('</tr></table>')
 894
 895                 f.write('</body></html>')
 896                 f.close()
 897
 898                 ###
 899                 # Authors
 900                 f = open(path + '/authors.html', 'w')
 901                 self.printHeader(f)
 902
 903                 f.write('<h1>Authors</h1>')
 904                 self.printNav(f)
 905
 906                 # Authors :: List of authors
 907                 f.write(html_header(2, 'List of Authors'))
 908
 909                 f.write('<table class="authors sortable" id="authors">')
 910                 f.write('<tr><th>Author</th><th>Commits (%)</th><th>+ lines</th><th>- lines</th><th>First commit</th><th>Last commit</th><th class="unsortable">Age</th><th>Active days</th><th># by commits</th></tr>')
 911                 for author in data.getAuthors(conf['max_authors']):
 912                         info = data.getAuthorInfo(author)
 913                         f.write('<tr><td>%s</td><td>%d (%.2f%%)</td><td>%d</td><td>%d</td><td>%s</td><td>%s</td><td>%s</td><td>%d</td><td>%d</td></tr>' % (author, info['commits'], info['commits_frac'], info['lines_added'], info['lines_removed'], info['date_first'], info['date_last'], info['timedelta'], len(info['active_days']), info['place_by_commits']))
 914                 f.write('</table>')
 915
 916                 allauthors = data.getAuthors()
 917                 if len(allauthors) > conf['max_authors']:
 918                         rest = allauthors[conf['max_authors']:]
 919                         f.write('<p class="moreauthors">These didn\'t make it to the top: %s</p>' % ', '.join(rest))
 920
 921                 f.write(html_header(2, 'Cumulated Added Lines of Code per Author'))
 922                 f.write('<img src="lines_of_code_by_author.png" alt="Lines of code per Author" />')
 923                 if len(allauthors) > conf['max_authors']:
 924                         f.write('<p class="moreauthors">Only top %d authors shown</p>' % conf['max_authors'])
 925
 926                 f.write(html_header(2, 'Commits per Author'))
 927                 f.write('<img src="commits_by_author.png" alt="Commits per Author" />')
 928                 if len(allauthors) > conf['max_authors']:
 929                         f.write('<p class="moreauthors">Only top %d authors shown</p>' % conf['max_authors'])
 930
 931                 fgl = open(path + '/lines_of_code_by_author.dat', 'w')
 932                 fgc = open(path + '/commits_by_author.dat', 'w')
 933
 934                 lines_by_authors = {} # cumulated added lines by
 935                 # author. to save memory,
 936                 # changes_by_date_by_author[stamp][author] is defined
 937                 # only at points where author commits.
 938                 # lines_by_authors allows us to generate all the
 939                 # points in the .dat file.
 940
 941                 # Don't rely on getAuthors to give the same order each
 942                 # time. Be robust and keep the list in a variable.
 943                 commits_by_authors = {} # cumulated added lines by
 944
 945                 self.authors_to_plot = data.getAuthors(conf['max_authors'])
 946                 for author in self.authors_to_plot:
 947                         lines_by_authors[author] = 0
 948                         commits_by_authors[author] = 0
 949                 for stamp in sorted(data.changes_by_date_by_author.keys()):
 950                         fgl.write('%d' % stamp)
 951                         fgc.write('%d' % stamp)
 952                         for author in self.authors_to_plot:
 953                                 if author in data.changes_by_date_by_author[stamp].keys():
 954                                         lines_by_authors[author] = data.changes_by_date_by_author[stamp][author]['lines_added']
 955                                         commits_by_authors[author] = data.changes_by_date_by_author[stamp][author]['commits']
 956                                 fgl.write(' %d' % lines_by_authors[author])
 957                                 fgc.write(' %d' % commits_by_authors[author])
 958                         fgl.write('\n')
 959                         fgc.write('\n')
 960                 fgl.close()
 961                 fgc.close()
 962
 963                 # Authors :: Author of Month
 964                 f.write(html_header(2, 'Author of Month'))
 965                 f.write('<table class="sortable" id="aom">')
 966                 f.write('<tr><th>Month</th><th>Author</th><th>Commits (%%)</th><th class="unsortable">Next top %d</th><th>Number of authors</th></tr>' % conf['authors_top'])
 967                 for yymm in reversed(sorted(data.author_of_month.keys())):
 968                         authordict = data.author_of_month[yymm]
 969                         authors = getkeyssortedbyvalues(authordict)
 970                         authors.reverse()
 971                         commits = data.author_of_month[yymm][authors[0]]
 972                         next = ', '.join(authors[1:conf['authors_top']+1])
 973                         f.write('<tr><td>%s</td><td>%s</td><td>%d (%.2f%% of %d)</td><td>%s</td><td>%d</td></tr>' % (yymm, authors[0], commits, (100.0 * commits) / data.commits_by_month[yymm], data.commits_by_month[yymm], next, len(authors)))
 974
 975                 f.write('</table>')
 976
 977                 f.write(html_header(2, 'Author of Year'))
 978                 f.write('<table class="sortable" id="aoy"><tr><th>Year</th><th>Author</th><th>Commits (%%)</th><th class="unsortable">Next top %d</th><th>Number of authors</th></tr>' % conf['authors_top'])
 979                 for yy in reversed(sorted(data.author_of_year.keys())):
 980                         authordict = data.author_of_year[yy]
 981                         authors = getkeyssortedbyvalues(authordict)
 982                         authors.reverse()
 983                         commits = data.author_of_year[yy][authors[0]]
 984                         next = ', '.join(authors[1:conf['authors_top']+1])
 985                         f.write('<tr><td>%s</td><td>%s</td><td>%d (%.2f%% of %d)</td><td>%s</td><td>%d</td></tr>' % (yy, authors[0], commits, (100.0 * commits) / data.commits_by_year[yy], data.commits_by_year[yy], next, len(authors)))
 986                 f.write('</table>')
 987
 988                 # Domains
 989                 f.write(html_header(2, 'Commits by Domains'))
 990                 domains_by_commits = getkeyssortedbyvaluekey(data.domains, 'commits')
 991                 domains_by_commits.reverse() # most first
 992                 f.write('<div class="vtable"><table>')
 993                 f.write('<tr><th>Domains</th><th>Total (%)</th></tr>')
 994                 fp = open(path + '/domains.dat', 'w')
 995                 n = 0
 996                 for domain in domains_by_commits:
 997                         if n == conf['max_domains']:
 998                                 break
 999                         commits = 0
1000                         n += 1
1001                         info = data.getDomainInfo(domain)
1002                         fp.write('%s %d %d\n' % (domain, n , info['commits']))
1003                         f.write('<tr><th>%s</th><td>%d (%.2f%%)</td></tr>' % (domain, info['commits'], (100.0 * info['commits'] / totalcommits)))
1004                 f.write('</table></div>')
1005                 f.write('<img src="domains.png" alt="Commits by Domains" />')
1006                 fp.close()
1007
1008                 f.write('</body></html>')
1009                 f.close()
1010
1011                 ###
1012                 # Files
1013                 f = open(path + '/files.html', 'w')
1014                 self.printHeader(f)
1015                 f.write('<h1>Files</h1>')
1016                 self.printNav(f)
1017
1018                 f.write('<dl>\n')
1019                 f.write('<dt>Total files</dt><dd>%d</dd>' % data.getTotalFiles())
1020                 f.write('<dt>Total lines</dt><dd>%d</dd>' % data.getTotalLOC())
1021                 try:
1022                         f.write('<dt>Average file size</dt><dd>%.2f bytes</dd>' % (float(data.getTotalSize()) / data.getTotalFiles()))
1023                 except ZeroDivisionError:
1024                         pass
1025                 f.write('</dl>\n')
1026
1027                 # Files :: File count by date
1028                 f.write(html_header(2, 'File count by date'))
1029
1030                 # use set to get rid of duplicate/unnecessary entries
1031                 files_by_date = set()
1032                 for stamp in sorted(data.files_by_stamp.keys()):
1033                         files_by_date.add('%s %d' % (datetime.datetime.fromtimestamp(stamp).strftime('%Y-%m-%d'), data.files_by_stamp[stamp]))
1034
1035                 fg = open(path + '/files_by_date.dat', 'w')
1036                 for line in sorted(list(files_by_date)):
1037                         fg.write('%s\n' % line)
1038                 #for stamp in sorted(data.files_by_stamp.keys()):
1039                 #       fg.write('%s %d\n' % (datetime.datetime.fromtimestamp(stamp).strftime('%Y-%m-%d'), data.files_by_stamp[stamp]))
1040                 fg.close()
1041
1042                 f.write('<img src="files_by_date.png" alt="Files by Date" />')
1043
1044                 #f.write('<h2>Average file size by date</h2>')
1045
1046                 # Files :: Extensions
1047                 f.write(html_header(2, 'Extensions'))
1048                 f.write('<table class="sortable" id="ext"><tr><th>Extension</th><th>Files (%)</th><th>Lines (%)</th><th>Lines/file</th></tr>')
1049                 for ext in sorted(data.extensions.keys()):
1050                         files = data.extensions[ext]['files']
1051                         lines = data.extensions[ext]['lines']
1052                         try:
1053                                 loc_percentage = (100.0 * lines) / data.getTotalLOC()
1054                         except ZeroDivisionError:
1055                                 loc_percentage = 0
1056                         f.write('<tr><td>%s</td><td>%d (%.2f%%)</td><td>%d (%.2f%%)</td><td>%d</td></tr>' % (ext, files, (100.0 * files) / data.getTotalFiles(), lines, loc_percentage, lines / files))
1057                 f.write('</table>')
1058
1059                 f.write('</body></html>')
1060                 f.close()
1061
1062                 ###
1063                 # Lines
1064                 f = open(path + '/lines.html', 'w')
1065                 self.printHeader(f)
1066                 f.write('<h1>Lines</h1>')
1067                 self.printNav(f)
1068
1069                 f.write('<dl>\n')
1070                 f.write('<dt>Total lines</dt><dd>%d</dd>' % data.getTotalLOC())
1071                 f.write('</dl>\n')
1072
1073                 f.write(html_header(2, 'Lines of Code'))
1074                 f.write('<img src="lines_of_code.png" />')
1075
1076                 fg = open(path + '/lines_of_code.dat', 'w')
1077                 for stamp in sorted(data.changes_by_date.keys()):
1078                         fg.write('%d %d\n' % (stamp, data.changes_by_date[stamp]['lines']))
1079                 fg.close()
1080
1081                 f.write('</body></html>')
1082                 f.close()
1083
1084                 ###
1085                 # tags.html
1086                 f = open(path + '/tags.html', 'w')
1087                 self.printHeader(f)
1088                 f.write('<h1>Tags</h1>')
1089                 self.printNav(f)
1090
1091                 f.write('<dl>')
1092                 f.write('<dt>Total tags</dt><dd>%d</dd>' % len(data.tags))
1093                 if len(data.tags) > 0:
1094                         f.write('<dt>Average commits per tag</dt><dd>%.2f</dd>' % (1.0 * data.getTotalCommits() / len(data.tags)))
1095                 f.write('</dl>')
1096
1097                 f.write('<table class="tags">')
1098                 f.write('<tr><th>Name</th><th>Date</th><th>Commits</th><th>Authors</th></tr>')
1099                 # sort the tags by date desc
1100                 tags_sorted_by_date_desc = map(lambda el : el[1], reversed(sorted(map(lambda el : (el[1]['date'], el[0]), data.tags.items()))))
1101                 for tag in tags_sorted_by_date_desc:
1102                         authorinfo = []
1103                         self.authors_by_commits = getkeyssortedbyvalues(data.tags[tag]['authors'])
1104                         for i in reversed(self.authors_by_commits):
1105                                 authorinfo.append('%s (%d)' % (i, data.tags[tag]['authors'][i]))
1106                         f.write('<tr><td>%s</td><td>%s</td><td>%d</td><td>%s</td></tr>' % (tag, data.tags[tag]['date'], data.tags[tag]['commits'], ', '.join(authorinfo)))
1107                 f.write('</table>')
1108
1109                 f.write('</body></html>')
1110                 f.close()
1111
1112                 self.createGraphs(path)
1113
1114         def createGraphs(self, path):
1115                 print 'Generating graphs...'
1116
1117                 # hour of day
1118                 f = open(path + '/hour_of_day.plot', 'w')
1119                 f.write(GNUPLOT_COMMON)
1120                 f.write(
1121 """
1122 set output 'hour_of_day.png'
1123 unset key
1124 set xrange [0.5:24.5]
1125 set xtics 4
1126 set grid y
1127 set ylabel "Commits"
1128 plot 'hour_of_day.dat' using 1:2:(0.5) w boxes fs solid
1129 """)
1130                 f.close()
1131
1132                 # day of week
1133                 f = open(path + '/day_of_week.plot', 'w')
1134                 f.write(GNUPLOT_COMMON)
1135                 f.write(
1136 """
1137 set output 'day_of_week.png'
1138 unset key
1139 set xrange [0.5:7.5]
1140 set xtics 1
1141 set grid y
1142 set ylabel "Commits"
1143 plot 'day_of_week.dat' using 1:3:(0.5):xtic(2) w boxes fs solid
1144 """)
1145                 f.close()
1146
1147                 # Domains
1148                 f = open(path + '/domains.plot', 'w')
1149                 f.write(GNUPLOT_COMMON)
1150                 f.write(
1151 """
1152 set output 'domains.png'
1153 unset key
1154 unset xtics
1155 set yrange [0:]
1156 set grid y
1157 set ylabel "Commits"
1158 plot 'domains.dat' using 2:3:(0.5) with boxes fs solid, '' using 2:3:1 with labels rotate by 45 offset 0,1
1159 """)
1160                 f.close()
1161
1162                 # Month of Year
1163                 f = open(path + '/month_of_year.plot', 'w')
1164                 f.write(GNUPLOT_COMMON)
1165                 f.write(
1166 """
1167 set output 'month_of_year.png'
1168 unset key
1169 set xrange [0.5:12.5]
1170 set xtics 1
1171 set grid y
1172 set ylabel "Commits"
1173 plot 'month_of_year.dat' using 1:2:(0.5) w boxes fs solid
1174 """)
1175                 f.close()
1176
1177                 # commits_by_year_month
1178                 f = open(path + '/commits_by_year_month.plot', 'w')
1179                 f.write(GNUPLOT_COMMON)
1180                 f.write(
1181 """
1182 set output 'commits_by_year_month.png'
1183 unset key
1184 set xdata time
1185 set timefmt "%Y-%m"
1186 set format x "%Y-%m"
1187 set xtics rotate
1188 set bmargin 5
1189 set grid y
1190 set ylabel "Commits"
1191 plot 'commits_by_year_month.dat' using 1:2:(0.5) w boxes fs solid
1192 """)
1193                 f.close()
1194
1195                 # commits_by_year
1196                 f = open(path + '/commits_by_year.plot', 'w')
1197                 f.write(GNUPLOT_COMMON)
1198                 f.write(
1199 """
1200 set output 'commits_by_year.png'
1201 unset key
1202 set xtics 1 rotate
1203 set grid y
1204 set ylabel "Commits"
1205 set yrange [0:]
1206 plot 'commits_by_year.dat' using 1:2:(0.5) w boxes fs solid
1207 """)
1208                 f.close()
1209
1210                 # Files by date
1211                 f = open(path + '/files_by_date.plot', 'w')
1212                 f.write(GNUPLOT_COMMON)
1213                 f.write(
1214 """
1215 set output 'files_by_date.png'
1216 unset key
1217 set xdata time
1218 set timefmt "%Y-%m-%d"
1219 set format x "%Y-%m-%d"
1220 set grid y
1221 set ylabel "Files"
1222 set xtics rotate
1223 set ytics autofreq
1224 set bmargin 6
1225 plot 'files_by_date.dat' using 1:2 w steps
1226 """)
1227                 f.close()
1228
1229                 # Lines of Code
1230                 f = open(path + '/lines_of_code.plot', 'w')
1231                 f.write(GNUPLOT_COMMON)
1232                 f.write(
1233 """
1234 set output 'lines_of_code.png'
1235 unset key
1236 set xdata time
1237 set timefmt "%s"
1238 set format x "%Y-%m-%d"
1239 set grid y
1240 set ylabel "Lines"
1241 set xtics rotate
1242 set bmargin 6
1243 plot 'lines_of_code.dat' using 1:2 w lines
1244 """)
1245                 f.close()
1246
1247                 # Lines of Code Added per author
1248                 f = open(path + '/lines_of_code_by_author.plot', 'w')
1249                 f.write(GNUPLOT_COMMON)
1250                 f.write(
1251 """
1252 set terminal png transparent size 640,480
1253 set output 'lines_of_code_by_author.png'
1254 set key left top
1255 set xdata time
1256 set timefmt "%s"
1257 set format x "%Y-%m-%d"
1258 set grid y
1259 set ylabel "Lines"
1260 set xtics rotate
1261 set bmargin 6
1262 plot """
1263 )
1264                 i = 1
1265                 plots = []
1266                 for a in self.authors_to_plot:
1267                         i = i + 1
1268                         plots.append("""'lines_of_code_by_author.dat' using 1:%d title "%s" w lines""" % (i, a.replace("\"", "\\\"")))
1269                 f.write(", ".join(plots))
1270                 f.write('\n')
1271
1272                 f.close()
1273
1274                 # Commits per author
1275                 f = open(path + '/commits_by_author.plot', 'w')
1276                 f.write(GNUPLOT_COMMON)
1277                 f.write(
1278 """
1279 set terminal png transparent size 640,480
1280 set output 'commits_by_author.png'
1281 set key left top
1282 set xdata time
1283 set timefmt "%s"
1284 set format x "%Y-%m-%d"
1285 set grid y
1286 set ylabel "Commits"
1287 set xtics rotate
1288 set bmargin 6
1289 plot """
1290 )
1291                 i = 1
1292                 plots = []
1293                 for a in self.authors_to_plot:
1294                         i = i + 1
1295                         plots.append("""'commits_by_author.dat' using 1:%d title "%s" w lines""" % (i, a.replace("\"", "\\\"")))
1296                 f.write(", ".join(plots))
1297                 f.write('\n')
1298
1299                 f.close()
1300
1301                 os.chdir(path)
1302                 files = glob.glob(path + '/*.plot')
1303                 for f in files:
1304                         out = getpipeoutput([gnuplot_cmd + ' "%s"' % f])
1305                         if len(out) > 0:
1306                                 print out
1307
1308         def printHeader(self, f, title = ''):
1309                 f.write(
1310 """<?xml version="1.0" encoding="UTF-8"?>
1311 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
1312 <html xmlns="http://www.w3.org/1999/xhtml">
1313 <head>
1314         <title>GitStats - %s</title>
1315         <link rel="stylesheet" href="%s" type="text/css" />
1316         <meta name="generator" content="GitStats %s" />
1317         <script type="text/javascript" src="sortable.js"></script>
1318 </head>
1319 <body>
1320 """ % (self.title, conf['style'], getversion()))
1321
1322         def printNav(self, f):
1323                 f.write("""
1324 <div class="nav">
1325 <ul>
1326 <li><a href="index.html">General</a></li>
1327 <li><a href="activity.html">Activity</a></li>
1328 <li><a href="authors.html">Authors</a></li>
1329 <li><a href="files.html">Files</a></li>
1330 <li><a href="lines.html">Lines</a></li>
1331 <li><a href="tags.html">Tags</a></li>
1332 </ul>
1333 </div>
1334 """)
1335
1336
1337 class GitStats:
1338         def run(self, args_orig):
1339                 optlist, args = getopt.getopt(args_orig, 'c:')
1340                 for o,v in optlist:
1341                         if o == '-c':
1342                                 key, value = v.split('=', 1)
1343                                 if key not in conf:
1344                                         raise KeyError('no such key "%s" in config' % key)
1345                                 if isinstance(conf[key], int):
1346                                         conf[key] = int(value)
1347                                 else:
1348                                         conf[key] = value
1349
1350                 if len(args) < 2:
1351                         print """
1352 Usage: gitstats [options] <gitpath..> <outputpath>
1353
1354 Options:
1355 -c key=value     Override configuration value
1356
1357 Default config values:
1358 %s
1359 """ % conf
1360                         sys.exit(0)
1361
1362                 outputpath = os.path.abspath(args[-1])
1363                 rundir = os.getcwd()
1364
1365                 try:
1366                         os.makedirs(outputpath)
1367                 except OSError:
1368                         pass
1369                 if not os.path.isdir(outputpath):
1370                         print 'FATAL: Output path is not a directory or does not exist'
1371                         sys.exit(1)
1372
1373                 if not getgnuplotversion():
1374                         print 'gnuplot not found'
1375                         sys.exit(1)
1376
1377                 print 'Output path: %s' % outputpath
1378                 cachefile = os.path.join(outputpath, 'gitstats.cache')
1379
1380                 data = GitDataCollector()
1381                 data.loadCache(cachefile)
1382
1383                 for gitpath in args[0:-1]:
1384                         print 'Git path: %s' % gitpath
1385
1386                         os.chdir(gitpath)
1387
1388                         print 'Collecting data...'
1389                         data.collect(gitpath)
1390
1391                 print 'Refining data...'
1392                 data.saveCache(cachefile)
1393                 data.refine()
1394
1395                 os.chdir(rundir)
1396
1397                 print 'Generating report...'
1398                 report = HTMLReportCreator()
1399                 report.create(data, outputpath)
1400
1401                 time_end = time.time()
1402                 exectime_internal = time_end - time_start
1403                 print 'Execution time %.5f secs, %.5f secs (%.2f %%) in external commands)' % (exectime_internal, exectime_external, (100.0 * exectime_external) / exectime_internal)
1404                 if sys.stdin.isatty():
1405                         print 'You may now run:'
1406                         print
1407                         print '   sensible-browser \'%s\'' % os.path.join(outputpath, 'index.html').replace("'", "'\\''")
1408                         print
1409
1410 if __name__=='__main__':
1411         g = GitStats()
1412         g.run(sys.argv[1:])
1413