gitstats

   1 #!/usr/bin/env python
   2 # Copyright (c) 2007-2013 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 if sys.version_info < (2, 6):
  18        print >> sys.stderr, "Python 2.6 or higher is required for gitstats"
  19        sys.exit(1)
  20
  21 from multiprocessing import Pool
  22
  23 os.environ['LC_ALL'] = 'C'
  24
  25 GNUPLOT_COMMON = 'set terminal png transparent size 640,240\nset size 1.0,1.0\n'
  26 ON_LINUX = (platform.system() == 'Linux')
  27 WEEKDAYS = ('Mon', 'Tue', 'Wed', 'Thu', 'Fri', 'Sat', 'Sun')
  28
  29 exectime_internal = 0.0
  30 exectime_external = 0.0
  31 time_start = time.time()
  32
  33 # By default, gnuplot is searched from path, but can be overridden with the
  34 # environment variable "GNUPLOT"
  35 gnuplot_cmd = 'gnuplot'
  36 if 'GNUPLOT' in os.environ:
  37         gnuplot_cmd = os.environ['GNUPLOT']
  38
  39 conf = {
  40         'max_domains': 10,
  41         'max_ext_length': 10,
  42         'style': 'gitstats.css',
  43         'max_authors': 20,
  44         'authors_top': 5,
  45         'commit_begin': '',
  46         'commit_end': 'HEAD',
  47         'linear_linestats': 1,
  48         'project_name': '',
  49         'merge_authors': {}
  50 }
  51
  52 def getpipeoutput(cmds, quiet = False):
  53         global exectime_external
  54         start = time.time()
  55         if not quiet and ON_LINUX and os.isatty(1):
  56                 print '>> ' + ' | '.join(cmds),
  57                 sys.stdout.flush()
  58         p0 = subprocess.Popen(cmds[0], stdout = subprocess.PIPE, shell = True)
  59         p = p0
  60         for x in cmds[1:]:
  61                 p = subprocess.Popen(x, stdin = p0.stdout, stdout = subprocess.PIPE, shell = True)
  62                 p0 = p
  63         output = p.communicate()[0]
  64         end = time.time()
  65         if not quiet:
  66                 if ON_LINUX and os.isatty(1):
  67                         print '\r',
  68                 print '[%.5f] >> %s' % (end - start, ' | '.join(cmds))
  69         exectime_external += (end - start)
  70         return output.rstrip('\n')
  71
  72 def getcommitrange(defaultrange = 'HEAD', end_only = False):
  73         if len(conf['commit_end']) > 0:
  74                 if end_only or len(conf['commit_begin']) == 0:
  75                         return conf['commit_end']
  76                 return '%s..%s' % (conf['commit_begin'], conf['commit_end'])
  77         return defaultrange
  78
  79 def getkeyssortedbyvalues(dict):
  80         return map(lambda el : el[1], sorted(map(lambda el : (el[1], el[0]), dict.items())))
  81
  82 # dict['author'] = { 'commits': 512 } - ...key(dict, 'commits')
  83 def getkeyssortedbyvaluekey(d, key):
  84         return map(lambda el : el[1], sorted(map(lambda el : (d[el][key], el), d.keys())))
  85
  86 def getstatsummarycounts(line):
  87         numbers = re.findall('\d+', line)
  88         if   len(numbers) == 1:
  89                 # neither insertions nor deletions: may probably only happen for "0 files changed"
  90                 numbers.append(0);
  91                 numbers.append(0);
  92         elif len(numbers) == 2 and line.find('(+)') != -1:
  93                 numbers.append(0);    # only insertions were printed on line
  94         elif len(numbers) == 2 and line.find('(-)') != -1:
  95                 numbers.insert(1, 0); # only deletions were printed on line
  96         return numbers
  97
  98 VERSION = 0
  99 def getversion():
 100         global VERSION
 101         if VERSION == 0:
 102                 gitstats_repo = os.path.dirname(os.path.abspath(__file__))
 103                 VERSION = getpipeoutput(["git --git-dir=%s/.git --work-tree=%s rev-parse --short %s" %
 104                         (gitstats_repo, gitstats_repo, getcommitrange('HEAD').split('\n')[0])])
 105         return VERSION
 106
 107 def getgitversion():
 108         return getpipeoutput(['git --version']).split('\n')[0]
 109
 110 def getgnuplotversion():
 111         return getpipeoutput(['%s --version' % gnuplot_cmd]).split('\n')[0]
 112
 113 def getnumoffilesfromrev(time_rev):
 114         """
 115         Get number of files changed in commit
 116         """
 117         time, rev = time_rev
 118         return (int(time), rev, int(getpipeoutput(['git ls-tree -r --name-only "%s"' % rev, 'wc -l']).split('\n')[0]))
 119
 120 def getnumoflinesinblob(ext_blob):
 121         """
 122         Get number of lines in blob
 123         """
 124         ext, blob_id = ext_blob
 125         return (ext, blob_id, int(getpipeoutput(['git cat-file blob %s' % blob_id, 'wc -l']).split()[0]))
 126
 127 class DataCollector:
 128         """Manages data collection from a revision control repository."""
 129         def __init__(self):
 130                 self.stamp_created = time.time()
 131                 self.cache = {}
 132                 self.total_authors = 0
 133                 self.activity_by_hour_of_day = {} # hour -> commits
 134                 self.activity_by_day_of_week = {} # day -> commits
 135                 self.activity_by_month_of_year = {} # month [1-12] -> commits
 136                 self.activity_by_hour_of_week = {} # weekday -> hour -> commits
 137                 self.activity_by_hour_of_day_busiest = 0
 138                 self.activity_by_hour_of_week_busiest = 0
 139                 self.activity_by_year_week = {} # yy_wNN -> commits
 140                 self.activity_by_year_week_peak = 0
 141
 142                 self.authors = {} # name -> {commits, first_commit_stamp, last_commit_stamp, last_active_day, active_days, lines_added, lines_removed}
 143
 144                 self.total_commits = 0
 145                 self.total_files = 0
 146                 self.authors_by_commits = 0
 147
 148                 # domains
 149                 self.domains = {} # domain -> commits
 150
 151                 # author of the month
 152                 self.author_of_month = {} # month -> author -> commits
 153                 self.author_of_year = {} # year -> author -> commits
 154                 self.commits_by_month = {} # month -> commits
 155                 self.commits_by_year = {} # year -> commits
 156                 self.lines_added_by_month = {} # month -> lines added
 157                 self.lines_added_by_year = {} # year -> lines added
 158                 self.lines_removed_by_month = {} # month -> lines removed
 159                 self.lines_removed_by_year = {} # year -> lines removed
 160                 self.first_commit_stamp = 0
 161                 self.last_commit_stamp = 0
 162                 self.last_active_day = None
 163                 self.active_days = set()
 164
 165                 # lines
 166                 self.total_lines = 0
 167                 self.total_lines_added = 0
 168                 self.total_lines_removed = 0
 169
 170                 # size
 171                 self.total_size = 0
 172
 173                 # timezone
 174                 self.commits_by_timezone = {} # timezone -> commits
 175
 176                 # tags
 177                 self.tags = {}
 178
 179                 self.files_by_stamp = {} # stamp -> files
 180
 181                 # extensions
 182                 self.extensions = {} # extension -> files, lines
 183
 184                 # line statistics
 185                 self.changes_by_date = {} # stamp -> { files, ins, del }
 186
 187         ##
 188         # This should be the main function to extract data from the repository.
 189         def collect(self, dir):
 190                 self.dir = dir
 191                 if len(conf['project_name']) == 0:
 192                         self.projectname = os.path.basename(os.path.abspath(dir))
 193                 else:
 194                         self.projectname = conf['project_name']
 195
 196         ##
 197         # Load cacheable data
 198         def loadCache(self, cachefile):
 199                 if not os.path.exists(cachefile):
 200                         return
 201                 print 'Loading cache...'
 202                 f = open(cachefile, 'rb')
 203                 try:
 204                         self.cache = pickle.loads(zlib.decompress(f.read()))
 205                 except:
 206                         # temporary hack to upgrade non-compressed caches
 207                         f.seek(0)
 208                         self.cache = pickle.load(f)
 209                 f.close()
 210
 211         ##
 212         # Produce any additional statistics from the extracted data.
 213         def refine(self):
 214                 pass
 215
 216         ##
 217         # : get a dictionary of author
 218         def getAuthorInfo(self, author):
 219                 return None
 220
 221         def getActivityByDayOfWeek(self):
 222                 return {}
 223
 224         def getActivityByHourOfDay(self):
 225                 return {}
 226
 227         # : get a dictionary of domains
 228         def getDomainInfo(self, domain):
 229                 return None
 230
 231         ##
 232         # Get a list of authors
 233         def getAuthors(self):
 234                 return []
 235
 236         def getFirstCommitDate(self):
 237                 return datetime.datetime.now()
 238
 239         def getLastCommitDate(self):
 240                 return datetime.datetime.now()
 241
 242         def getStampCreated(self):
 243                 return self.stamp_created
 244
 245         def getTags(self):
 246                 return []
 247
 248         def getTotalAuthors(self):
 249                 return -1
 250
 251         def getTotalCommits(self):
 252                 return -1
 253
 254         def getTotalFiles(self):
 255                 return -1
 256
 257         def getTotalLOC(self):
 258                 return -1
 259
 260         ##
 261         # Save cacheable data
 262         def saveCache(self, cachefile):
 263                 print 'Saving cache...'
 264                 tempfile = cachefile + '.tmp'
 265                 f = open(tempfile, 'wb')
 266                 #pickle.dump(self.cache, f)
 267                 data = zlib.compress(pickle.dumps(self.cache))
 268                 f.write(data)
 269                 f.close()
 270                 try:
 271                         os.remove(cachefile)
 272                 except OSError:
 273                         pass
 274                 os.rename(tempfile, cachefile)
 275
 276 class GitDataCollector(DataCollector):
 277         def collect(self, dir):
 278                 DataCollector.collect(self, dir)
 279
 280                 self.total_authors += int(getpipeoutput(['git shortlog -s %s' % getcommitrange(), 'wc -l']))
 281                 #self.total_lines = int(getoutput('git-ls-files -z |xargs -0 cat |wc -l'))
 282
 283                 # tags
 284                 lines = getpipeoutput(['git show-ref --tags']).split('\n')
 285                 for line in lines:
 286                         if len(line) == 0:
 287                                 continue
 288                         (hash, tag) = line.split(' ')
 289
 290                         tag = tag.replace('refs/tags/', '')
 291                         output = getpipeoutput(['git log "%s" --pretty=format:"%%at %%aN" -n 1' % hash])
 292                         if len(output) > 0:
 293                                 parts = output.split(' ')
 294                                 stamp = 0
 295                                 try:
 296                                         stamp = int(parts[0])
 297                                 except ValueError:
 298                                         stamp = 0
 299                                 self.tags[tag] = { 'stamp': stamp, 'hash' : hash, 'date' : datetime.datetime.fromtimestamp(stamp).strftime('%Y-%m-%d'), 'commits': 0, 'authors': {} }
 300
 301                 # collect info on tags, starting from latest
 302                 tags_sorted_by_date_desc = map(lambda el : el[1], reversed(sorted(map(lambda el : (el[1]['date'], el[0]), self.tags.items()))))
 303                 prev = None
 304                 for tag in reversed(tags_sorted_by_date_desc):
 305                         cmd = 'git shortlog -s "%s"' % tag
 306                         if prev != None:
 307                                 cmd += ' "^%s"' % prev
 308                         output = getpipeoutput([cmd])
 309                         if len(output) == 0:
 310                                 continue
 311                         prev = tag
 312                         for line in output.split('\n'):
 313                                 parts = re.split('\s+', line, 2)
 314                                 commits = int(parts[1])
 315                                 author = parts[2]
 316                                 if author in conf['merge_authors']:
 317                                         author = conf['merge_authors'][author]
 318                                 self.tags[tag]['commits'] += commits
 319                                 self.tags[tag]['authors'][author] = commits
 320
 321                 # Collect revision statistics
 322                 # Outputs "<stamp> <date> <time> <timezone> <author> '<' <mail> '>'"
 323                 lines = getpipeoutput(['git rev-list --pretty=format:"%%at %%ai %%aN <%%aE>" %s' % getcommitrange('HEAD'), 'grep -v ^commit']).split('\n')
 324                 for line in lines:
 325                         parts = line.split(' ', 4)
 326                         author = ''
 327                         try:
 328                                 stamp = int(parts[0])
 329                         except ValueError:
 330                                 stamp = 0
 331                         timezone = parts[3]
 332                         author, mail = parts[4].split('<', 1)
 333                         author = author.rstrip()
 334                         if author in conf['merge_authors']:
 335                                 author = conf['merge_authors'][author]
 336                         mail = mail.rstrip('>')
 337                         domain = '?'
 338                         if mail.find('@') != -1:
 339                                 domain = mail.rsplit('@', 1)[1]
 340                         date = datetime.datetime.fromtimestamp(float(stamp))
 341
 342                         # First and last commit stamp (may be in any order because of cherry-picking and patches)
 343                         if stamp > self.last_commit_stamp:
 344                                 self.last_commit_stamp = stamp
 345                         if self.first_commit_stamp == 0 or stamp < self.first_commit_stamp:
 346                                 self.first_commit_stamp = stamp
 347
 348                         # activity
 349                         # hour
 350                         hour = date.hour
 351                         self.activity_by_hour_of_day[hour] = self.activity_by_hour_of_day.get(hour, 0) + 1
 352                         # most active hour?
 353                         if self.activity_by_hour_of_day[hour] > self.activity_by_hour_of_day_busiest:
 354                                 self.activity_by_hour_of_day_busiest = self.activity_by_hour_of_day[hour]
 355
 356                         # day of week
 357                         day = date.weekday()
 358                         self.activity_by_day_of_week[day] = self.activity_by_day_of_week.get(day, 0) + 1
 359
 360                         # domain stats
 361                         if domain not in self.domains:
 362                                 self.domains[domain] = {}
 363                         # commits
 364                         self.domains[domain]['commits'] = self.domains[domain].get('commits', 0) + 1
 365
 366                         # hour of week
 367                         if day not in self.activity_by_hour_of_week:
 368                                 self.activity_by_hour_of_week[day] = {}
 369                         self.activity_by_hour_of_week[day][hour] = self.activity_by_hour_of_week[day].get(hour, 0) + 1
 370                         # most active hour?
 371                         if self.activity_by_hour_of_week[day][hour] > self.activity_by_hour_of_week_busiest:
 372                                 self.activity_by_hour_of_week_busiest = self.activity_by_hour_of_week[day][hour]
 373
 374                         # month of year
 375                         month = date.month
 376                         self.activity_by_month_of_year[month] = self.activity_by_month_of_year.get(month, 0) + 1
 377
 378                         # yearly/weekly activity
 379                         yyw = date.strftime('%Y-%W')
 380                         self.activity_by_year_week[yyw] = self.activity_by_year_week.get(yyw, 0) + 1
 381                         if self.activity_by_year_week_peak < self.activity_by_year_week[yyw]:
 382                                 self.activity_by_year_week_peak = self.activity_by_year_week[yyw]
 383
 384                         # author stats
 385                         if author not in self.authors:
 386                                 self.authors[author] = {}
 387                         # commits, note again that commits may be in any date order because of cherry-picking and patches
 388                         if 'last_commit_stamp' not in self.authors[author]:
 389                                 self.authors[author]['last_commit_stamp'] = stamp
 390                         if stamp > self.authors[author]['last_commit_stamp']:
 391                                 self.authors[author]['last_commit_stamp'] = stamp
 392                         if 'first_commit_stamp' not in self.authors[author]:
 393                                 self.authors[author]['first_commit_stamp'] = stamp
 394                         if stamp < self.authors[author]['first_commit_stamp']:
 395                                 self.authors[author]['first_commit_stamp'] = stamp
 396
 397                         # author of the month/year
 398                         yymm = date.strftime('%Y-%m')
 399                         if yymm in self.author_of_month:
 400                                 self.author_of_month[yymm][author] = self.author_of_month[yymm].get(author, 0) + 1
 401                         else:
 402                                 self.author_of_month[yymm] = {}
 403                                 self.author_of_month[yymm][author] = 1
 404                         self.commits_by_month[yymm] = self.commits_by_month.get(yymm, 0) + 1
 405
 406                         yy = date.year
 407                         if yy in self.author_of_year:
 408                                 self.author_of_year[yy][author] = self.author_of_year[yy].get(author, 0) + 1
 409                         else:
 410                                 self.author_of_year[yy] = {}
 411                                 self.author_of_year[yy][author] = 1
 412                         self.commits_by_year[yy] = self.commits_by_year.get(yy, 0) + 1
 413
 414                         # authors: active days
 415                         yymmdd = date.strftime('%Y-%m-%d')
 416                         if 'last_active_day' not in self.authors[author]:
 417                                 self.authors[author]['last_active_day'] = yymmdd
 418                                 self.authors[author]['active_days'] = set([yymmdd])
 419                         elif yymmdd != self.authors[author]['last_active_day']:
 420                                 self.authors[author]['last_active_day'] = yymmdd
 421                                 self.authors[author]['active_days'].add(yymmdd)
 422
 423                         # project: active days
 424                         if yymmdd != self.last_active_day:
 425                                 self.last_active_day = yymmdd
 426                                 self.active_days.add(yymmdd)
 427
 428                         # timezone
 429                         self.commits_by_timezone[timezone] = self.commits_by_timezone.get(timezone, 0) + 1
 430
 431                 # outputs "<stamp> <files>" for each revision
 432                 revlines = getpipeoutput(['git rev-list --pretty=format:"%%at %%T" %s' % getcommitrange('HEAD'), 'grep -v ^commit']).strip().split('\n')
 433                 lines = []
 434                 revs_to_read = []
 435                 time_rev_count = []
 436                 #Look up rev in cache and take info from cache if found
 437                 #If not append rev to list of rev to read from repo
 438                 for revline in revlines:
 439                         time, rev = revline.split(' ')
 440                         #if cache empty then add time and rev to list of new rev's
 441                         #otherwise try to read needed info from cache
 442                         if 'files_in_tree' not in self.cache.keys():
 443                                 revs_to_read.append((time,rev))
 444                                 continue
 445                         if rev in self.cache['files_in_tree'].keys():
 446                                 lines.append('%d %d' % (int(time), self.cache['files_in_tree'][rev]))
 447                         else:
 448                                 revs_to_read.append((time,rev))
 449
 450                 #Read revisions from repo
 451                 time_rev_count = Pool(processes=24).map(getnumoffilesfromrev, revs_to_read)
 452
 453                 #Update cache with new revisions and append then to general list
 454                 for (time, rev, count) in time_rev_count:
 455                         if 'files_in_tree' not in self.cache:
 456                                 self.cache['files_in_tree'] = {}
 457                         self.cache['files_in_tree'][rev] = count
 458                         lines.append('%d %d' % (int(time), count))
 459
 460                 self.total_commits += len(lines)
 461                 for line in lines:
 462                         parts = line.split(' ')
 463                         if len(parts) != 2:
 464                                 continue
 465                         (stamp, files) = parts[0:2]
 466                         try:
 467                                 self.files_by_stamp[int(stamp)] = int(files)
 468                         except ValueError:
 469                                 print 'Warning: failed to parse line "%s"' % line
 470
 471                 # extensions and size of files
 472                 lines = getpipeoutput(['git ls-tree -r -l -z %s' % getcommitrange('HEAD', end_only = True)]).split('\000')
 473                 blobs_to_read = []
 474                 for line in lines:
 475                         if len(line) == 0:
 476                                 continue
 477                         parts = re.split('\s+', line, 5)
 478                         if parts[0] == '160000' and parts[3] == '-':
 479                                 # skip submodules
 480                                 continue
 481                         blob_id = parts[2]
 482                         size = int(parts[3])
 483                         fullpath = parts[4]
 484
 485                         self.total_size += size
 486                         self.total_files += 1
 487
 488                         filename = fullpath.split('/')[-1] # strip directories
 489                         if filename.find('.') == -1 or filename.rfind('.') == 0:
 490                                 ext = ''
 491                         else:
 492                                 ext = filename[(filename.rfind('.') + 1):]
 493                         if len(ext) > conf['max_ext_length']:
 494                                 ext = ''
 495                         if ext not in self.extensions:
 496                                 self.extensions[ext] = {'files': 0, 'lines': 0}
 497                         self.extensions[ext]['files'] += 1
 498                         #if cache empty then add ext and blob id to list of new blob's
 499                         #otherwise try to read needed info from cache
 500                         if 'lines_in_blob' not in self.cache.keys():
 501                                 blobs_to_read.append((ext,blob_id))
 502                                 continue
 503                         if blob_id in self.cache['lines_in_blob'].keys():
 504                                 self.extensions[ext]['lines'] += self.cache['lines_in_blob'][blob_id]
 505                         else:
 506                                 blobs_to_read.append((ext,blob_id))
 507
 508                 #Get info abount line count for new blob's that wasn't found in cache
 509                 ext_blob_linecount = Pool(processes=24).map(getnumoflinesinblob, blobs_to_read)
 510
 511                 #Update cache and write down info about number of number of lines
 512                 for (ext, blob_id, linecount) in ext_blob_linecount:
 513                         if 'lines_in_blob' not in self.cache:
 514                                 self.cache['lines_in_blob'] = {}
 515                         self.cache['lines_in_blob'][blob_id] = linecount
 516                         self.extensions[ext]['lines'] += self.cache['lines_in_blob'][blob_id]
 517
 518                 # line statistics
 519                 # outputs:
 520                 #  N files changed, N insertions (+), N deletions(-)
 521                 # <stamp> <author>
 522                 self.changes_by_date = {} # stamp -> { files, ins, del }
 523                 # computation of lines of code by date is better done
 524                 # on a linear history.
 525                 extra = ''
 526                 if conf['linear_linestats']:
 527                         extra = '--first-parent -m'
 528                 lines = getpipeoutput(['git log --shortstat %s --pretty=format:"%%at %%aN" %s' % (extra, getcommitrange('HEAD'))]).split('\n')
 529                 lines.reverse()
 530                 files = 0; inserted = 0; deleted = 0; total_lines = 0
 531                 author = None
 532                 for line in lines:
 533                         if len(line) == 0:
 534                                 continue
 535
 536                         # <stamp> <author>
 537                         if re.search('files? changed', line) == None:
 538                                 pos = line.find(' ')
 539                                 if pos != -1:
 540                                         try:
 541                                                 (stamp, author) = (int(line[:pos]), line[pos+1:])
 542                                                 if author in conf['merge_authors']:
 543                                                         author = conf['merge_authors'][author]
 544                                                 self.changes_by_date[stamp] = { 'files': files, 'ins': inserted, 'del': deleted, 'lines': total_lines }
 545
 546                                                 date = datetime.datetime.fromtimestamp(stamp)
 547                                                 yymm = date.strftime('%Y-%m')
 548                                                 self.lines_added_by_month[yymm] = self.lines_added_by_month.get(yymm, 0) + inserted
 549                                                 self.lines_removed_by_month[yymm] = self.lines_removed_by_month.get(yymm, 0) + deleted
 550
 551                                                 yy = date.year
 552                                                 self.lines_added_by_year[yy] = self.lines_added_by_year.get(yy,0) + inserted
 553                                                 self.lines_removed_by_year[yy] = self.lines_removed_by_year.get(yy, 0) + deleted
 554
 555                                                 files, inserted, deleted = 0, 0, 0
 556                                         except ValueError:
 557                                                 print 'Warning: unexpected line "%s"' % line
 558                                 else:
 559                                         print 'Warning: unexpected line "%s"' % line
 560                         else:
 561                                 numbers = getstatsummarycounts(line)
 562
 563                                 if len(numbers) == 3:
 564                                         (files, inserted, deleted) = map(lambda el : int(el), numbers)
 565                                         total_lines += inserted
 566                                         total_lines -= deleted
 567                                         self.total_lines_added += inserted
 568                                         self.total_lines_removed += deleted
 569
 570                                 else:
 571                                         print 'Warning: failed to handle line "%s"' % line
 572                                         (files, inserted, deleted) = (0, 0, 0)
 573                                 #self.changes_by_date[stamp] = { 'files': files, 'ins': inserted, 'del': deleted }
 574                 self.total_lines += total_lines
 575
 576                 # Per-author statistics
 577
 578                 # defined for stamp, author only if author commited at this timestamp.
 579                 self.changes_by_date_by_author = {} # stamp -> author -> lines_added
 580
 581                 # Similar to the above, but never use --first-parent
 582                 # (we need to walk through every commit to know who
 583                 # committed what, not just through mainline)
 584                 lines = getpipeoutput(['git log --shortstat --date-order --pretty=format:"%%at %%aN" %s' % (getcommitrange('HEAD'))]).split('\n')
 585                 lines.reverse()
 586                 files = 0; inserted = 0; deleted = 0
 587                 author = None
 588                 stamp = 0
 589                 for line in lines:
 590                         if len(line) == 0:
 591                                 continue
 592
 593                         # <stamp> <author>
 594                         if re.search('files? changed', line) == None:
 595                                 pos = line.find(' ')
 596                                 if pos != -1:
 597                                         try:
 598                                                 oldstamp = stamp
 599                                                 (stamp, author) = (int(line[:pos]), line[pos+1:])
 600                                                 if author in conf['merge_authors']:
 601                                                         author = conf['merge_authors'][author]
 602                                                 if oldstamp > stamp:
 603                                                         # clock skew, keep old timestamp to avoid having ugly graph
 604                                                         stamp = oldstamp
 605                                                 if author not in self.authors:
 606                                                         self.authors[author] = { 'lines_added' : 0, 'lines_removed' : 0, 'commits' : 0}
 607                                                 self.authors[author]['commits'] = self.authors[author].get('commits', 0) + 1
 608                                                 self.authors[author]['lines_added'] = self.authors[author].get('lines_added', 0) + inserted
 609                                                 self.authors[author]['lines_removed'] = self.authors[author].get('lines_removed', 0) + deleted
 610                                                 if stamp not in self.changes_by_date_by_author:
 611                                                         self.changes_by_date_by_author[stamp] = {}
 612                                                 if author not in self.changes_by_date_by_author[stamp]:
 613                                                         self.changes_by_date_by_author[stamp][author] = {}
 614                                                 self.changes_by_date_by_author[stamp][author]['lines_added'] = self.authors[author]['lines_added']
 615                                                 self.changes_by_date_by_author[stamp][author]['commits'] = self.authors[author]['commits']
 616                                                 files, inserted, deleted = 0, 0, 0
 617                                         except ValueError:
 618                                                 print 'Warning: unexpected line "%s"' % line
 619                                 else:
 620                                         print 'Warning: unexpected line "%s"' % line
 621                         else:
 622                                 numbers = getstatsummarycounts(line);
 623
 624                                 if len(numbers) == 3:
 625                                         (files, inserted, deleted) = map(lambda el : int(el), numbers)
 626                                 else:
 627                                         print 'Warning: failed to handle line "%s"' % line
 628                                         (files, inserted, deleted) = (0, 0, 0)
 629
 630         def refine(self):
 631                 # authors
 632                 # name -> {place_by_commits, commits_frac, date_first, date_last, timedelta}
 633                 self.authors_by_commits = getkeyssortedbyvaluekey(self.authors, 'commits')
 634                 self.authors_by_commits.reverse() # most first
 635                 for i, name in enumerate(self.authors_by_commits):
 636                         self.authors[name]['place_by_commits'] = i + 1
 637
 638                 for name in self.authors.keys():
 639                         a = self.authors[name]
 640                         a['commits_frac'] = (100 * float(a['commits'])) / self.getTotalCommits()
 641                         date_first = datetime.datetime.fromtimestamp(a['first_commit_stamp'])
 642                         date_last = datetime.datetime.fromtimestamp(a['last_commit_stamp'])
 643                         delta = date_last - date_first
 644                         a['date_first'] = date_first.strftime('%Y-%m-%d')
 645                         a['date_last'] = date_last.strftime('%Y-%m-%d')
 646                         a['timedelta'] = delta
 647                         if 'lines_added' not in a: a['lines_added'] = 0
 648                         if 'lines_removed' not in a: a['lines_removed'] = 0
 649
 650         def getActiveDays(self):
 651                 return self.active_days
 652
 653         def getActivityByDayOfWeek(self):
 654                 return self.activity_by_day_of_week
 655
 656         def getActivityByHourOfDay(self):
 657                 return self.activity_by_hour_of_day
 658
 659         def getAuthorInfo(self, author):
 660                 return self.authors[author]
 661
 662         def getAuthors(self, limit = None):
 663                 res = getkeyssortedbyvaluekey(self.authors, 'commits')
 664                 res.reverse()
 665                 return res[:limit]
 666
 667         def getCommitDeltaDays(self):
 668                 return (self.last_commit_stamp / 86400 - self.first_commit_stamp / 86400) + 1
 669
 670         def getDomainInfo(self, domain):
 671                 return self.domains[domain]
 672
 673         def getDomains(self):
 674                 return self.domains.keys()
 675
 676         def getFirstCommitDate(self):
 677                 return datetime.datetime.fromtimestamp(self.first_commit_stamp)
 678
 679         def getLastCommitDate(self):
 680                 return datetime.datetime.fromtimestamp(self.last_commit_stamp)
 681
 682         def getTags(self):
 683                 lines = getpipeoutput(['git show-ref --tags', 'cut -d/ -f3'])
 684                 return lines.split('\n')
 685
 686         def getTagDate(self, tag):
 687                 return self.revToDate('tags/' + tag)
 688
 689         def getTotalAuthors(self):
 690                 return self.total_authors
 691
 692         def getTotalCommits(self):
 693                 return self.total_commits
 694
 695         def getTotalFiles(self):
 696                 return self.total_files
 697
 698         def getTotalLOC(self):
 699                 return self.total_lines
 700
 701         def getTotalSize(self):
 702                 return self.total_size
 703
 704         def revToDate(self, rev):
 705                 stamp = int(getpipeoutput(['git log --pretty=format:%%at "%s" -n 1' % rev]))
 706                 return datetime.datetime.fromtimestamp(stamp).strftime('%Y-%m-%d')
 707
 708 class ReportCreator:
 709         """Creates the actual report based on given data."""
 710         def __init__(self):
 711                 pass
 712
 713         def create(self, data, path):
 714                 self.data = data
 715                 self.path = path
 716
 717 def html_linkify(text):
 718         return text.lower().replace(' ', '_')
 719
 720 def html_header(level, text):
 721         name = html_linkify(text)
 722         return '\n<h%d><a href="#%s" name="%s">%s</a></h%d>\n\n' % (level, name, name, text, level)
 723
 724 class HTMLReportCreator(ReportCreator):
 725         def create(self, data, path):
 726                 ReportCreator.create(self, data, path)
 727                 self.title = data.projectname
 728
 729                 # copy static files. Looks in the binary directory, ../share/gitstats and /usr/share/gitstats
 730                 binarypath = os.path.dirname(os.path.abspath(__file__))
 731                 secondarypath = os.path.join(binarypath, '..', 'share', 'gitstats')
 732                 basedirs = [binarypath, secondarypath, '/usr/share/gitstats']
 733                 for file in ('gitstats.css', 'sortable.js', 'arrow-up.gif', 'arrow-down.gif', 'arrow-none.gif'):
 734                         for base in basedirs:
 735                                 src = base + '/' + file
 736                                 if os.path.exists(src):
 737                                         shutil.copyfile(src, path + '/' + file)
 738                                         break
 739                         else:
 740                                 print 'Warning: "%s" not found, so not copied (searched: %s)' % (file, basedirs)
 741
 742                 f = open(path + "/index.html", 'w')
 743                 format = '%Y-%m-%d %H:%M:%S'
 744                 self.printHeader(f)
 745
 746                 f.write('<h1>GitStats - %s</h1>' % data.projectname)
 747
 748                 self.printNav(f)
 749
 750                 f.write('<dl>')
 751                 f.write('<dt>Project name</dt><dd>%s</dd>' % (data.projectname))
 752                 f.write('<dt>Generated</dt><dd>%s (in %d seconds)</dd>' % (datetime.datetime.now().strftime(format), time.time() - data.getStampCreated()))
 753                 f.write('<dt>Generator</dt><dd><a href="http://gitstats.sourceforge.net/">GitStats</a> (version %s), %s, %s</dd>' % (getversion(), getgitversion(), getgnuplotversion()))
 754                 f.write('<dt>Report Period</dt><dd>%s to %s</dd>' % (data.getFirstCommitDate().strftime(format), data.getLastCommitDate().strftime(format)))
 755                 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())))
 756                 f.write('<dt>Total Files</dt><dd>%s</dd>' % data.getTotalFiles())
 757                 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))
 758                 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()))
 759                 f.write('<dt>Authors</dt><dd>%s (average %.1f commits per author)</dd>' % (data.getTotalAuthors(), (1.0 * data.getTotalCommits()) / data.getTotalAuthors()))
 760                 f.write('</dl>')
 761
 762                 f.write('</body>\n</html>')
 763                 f.close()
 764
 765                 ###
 766                 # Activity
 767                 f = open(path + '/activity.html', 'w')
 768                 self.printHeader(f)
 769                 f.write('<h1>Activity</h1>')
 770                 self.printNav(f)
 771
 772                 #f.write('<h2>Last 30 days</h2>')
 773
 774                 #f.write('<h2>Last 12 months</h2>')
 775
 776                 # Weekly activity
 777                 WEEKS = 32
 778                 f.write(html_header(2, 'Weekly activity'))
 779                 f.write('<p>Last %d weeks</p>' % WEEKS)
 780
 781                 # generate weeks to show (previous N weeks from now)
 782                 now = datetime.datetime.now()
 783                 deltaweek = datetime.timedelta(7)
 784                 weeks = []
 785                 stampcur = now
 786                 for i in range(0, WEEKS):
 787                         weeks.insert(0, stampcur.strftime('%Y-%W'))
 788                         stampcur -= deltaweek
 789
 790                 # top row: commits & bar
 791                 f.write('<table class="noborders"><tr>')
 792                 for i in range(0, WEEKS):
 793                         commits = 0
 794                         if weeks[i] in data.activity_by_year_week:
 795                                 commits = data.activity_by_year_week[weeks[i]]
 796
 797                         percentage = 0
 798                         if weeks[i] in data.activity_by_year_week:
 799                                 percentage = float(data.activity_by_year_week[weeks[i]]) / data.activity_by_year_week_peak
 800                         height = max(1, int(200 * percentage))
 801                         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))
 802
 803                 # bottom row: year/week
 804                 f.write('</tr><tr>')
 805                 for i in range(0, WEEKS):
 806                         f.write('<td>%s</td>' % (WEEKS - i))
 807                 f.write('</tr></table>')
 808
 809                 # Hour of Day
 810                 f.write(html_header(2, 'Hour of Day'))
 811                 hour_of_day = data.getActivityByHourOfDay()
 812                 f.write('<table><tr><th>Hour</th>')
 813                 for i in range(0, 24):
 814                         f.write('<th>%d</th>' % i)
 815                 f.write('</tr>\n<tr><th>Commits</th>')
 816                 fp = open(path + '/hour_of_day.dat', 'w')
 817                 for i in range(0, 24):
 818                         if i in hour_of_day:
 819                                 r = 127 + int((float(hour_of_day[i]) / data.activity_by_hour_of_day_busiest) * 128)
 820                                 f.write('<td style="background-color: rgb(%d, 0, 0)">%d</td>' % (r, hour_of_day[i]))
 821                                 fp.write('%d %d\n' % (i, hour_of_day[i]))
 822                         else:
 823                                 f.write('<td>0</td>')
 824                                 fp.write('%d 0\n' % i)
 825                 fp.close()
 826                 f.write('</tr>\n<tr><th>%</th>')
 827                 totalcommits = data.getTotalCommits()
 828                 for i in range(0, 24):
 829                         if i in hour_of_day:
 830                                 r = 127 + int((float(hour_of_day[i]) / data.activity_by_hour_of_day_busiest) * 128)
 831                                 f.write('<td style="background-color: rgb(%d, 0, 0)">%.2f</td>' % (r, (100.0 * hour_of_day[i]) / totalcommits))
 832                         else:
 833                                 f.write('<td>0.00</td>')
 834                 f.write('</tr></table>')
 835                 f.write('<img src="hour_of_day.png" alt="Hour of Day" />')
 836                 fg = open(path + '/hour_of_day.dat', 'w')
 837                 for i in range(0, 24):
 838                         if i in hour_of_day:
 839                                 fg.write('%d %d\n' % (i + 1, hour_of_day[i]))
 840                         else:
 841                                 fg.write('%d 0\n' % (i + 1))
 842                 fg.close()
 843
 844                 # Day of Week
 845                 f.write(html_header(2, 'Day of Week'))
 846                 day_of_week = data.getActivityByDayOfWeek()
 847                 f.write('<div class="vtable"><table>')
 848                 f.write('<tr><th>Day</th><th>Total (%)</th></tr>')
 849                 fp = open(path + '/day_of_week.dat', 'w')
 850                 for d in range(0, 7):
 851                         commits = 0
 852                         if d in day_of_week:
 853                                 commits = day_of_week[d]
 854                         fp.write('%d %s %d\n' % (d + 1, WEEKDAYS[d], commits))
 855                         f.write('<tr>')
 856                         f.write('<th>%s</th>' % (WEEKDAYS[d]))
 857                         if d in day_of_week:
 858                                 f.write('<td>%d (%.2f%%)</td>' % (day_of_week[d], (100.0 * day_of_week[d]) / totalcommits))
 859                         else:
 860                                 f.write('<td>0</td>')
 861                         f.write('</tr>')
 862                 f.write('</table></div>')
 863                 f.write('<img src="day_of_week.png" alt="Day of Week" />')
 864                 fp.close()
 865
 866                 # Hour of Week
 867                 f.write(html_header(2, 'Hour of Week'))
 868                 f.write('<table>')
 869
 870                 f.write('<tr><th>Weekday</th>')
 871                 for hour in range(0, 24):
 872                         f.write('<th>%d</th>' % (hour))
 873                 f.write('</tr>')
 874
 875                 for weekday in range(0, 7):
 876                         f.write('<tr><th>%s</th>' % (WEEKDAYS[weekday]))
 877                         for hour in range(0, 24):
 878                                 try:
 879                                         commits = data.activity_by_hour_of_week[weekday][hour]
 880                                 except KeyError:
 881                                         commits = 0
 882                                 if commits != 0:
 883                                         f.write('<td')
 884                                         r = 127 + int((float(commits) / data.activity_by_hour_of_week_busiest) * 128)
 885                                         f.write(' style="background-color: rgb(%d, 0, 0)"' % r)
 886                                         f.write('>%d</td>' % commits)
 887                                 else:
 888                                         f.write('<td></td>')
 889                         f.write('</tr>')
 890
 891                 f.write('</table>')
 892
 893                 # Month of Year
 894                 f.write(html_header(2, 'Month of Year'))
 895                 f.write('<div class="vtable"><table>')
 896                 f.write('<tr><th>Month</th><th>Commits (%)</th></tr>')
 897                 fp = open (path + '/month_of_year.dat', 'w')
 898                 for mm in range(1, 13):
 899                         commits = 0
 900                         if mm in data.activity_by_month_of_year:
 901                                 commits = data.activity_by_month_of_year[mm]
 902                         f.write('<tr><td>%d</td><td>%d (%.2f %%)</td></tr>' % (mm, commits, (100.0 * commits) / data.getTotalCommits()))
 903                         fp.write('%d %d\n' % (mm, commits))
 904                 fp.close()
 905                 f.write('</table></div>')
 906                 f.write('<img src="month_of_year.png" alt="Month of Year" />')
 907
 908                 # Commits by year/month
 909                 f.write(html_header(2, 'Commits by year/month'))
 910                 f.write('<div class="vtable"><table><tr><th>Month</th><th>Commits</th><th>Lines added</th><th>Lines removed</th></tr>')
 911                 for yymm in reversed(sorted(data.commits_by_month.keys())):
 912                         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)))
 913                 f.write('</table></div>')
 914                 f.write('<img src="commits_by_year_month.png" alt="Commits by year/month" />')
 915                 fg = open(path + '/commits_by_year_month.dat', 'w')
 916                 for yymm in sorted(data.commits_by_month.keys()):
 917                         fg.write('%s %s\n' % (yymm, data.commits_by_month[yymm]))
 918                 fg.close()
 919
 920                 # Commits by year
 921                 f.write(html_header(2, 'Commits by Year'))
 922                 f.write('<div class="vtable"><table><tr><th>Year</th><th>Commits (% of all)</th><th>Lines added</th><th>Lines removed</th></tr>')
 923                 for yy in reversed(sorted(data.commits_by_year.keys())):
 924                         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)))
 925                 f.write('</table></div>')
 926                 f.write('<img src="commits_by_year.png" alt="Commits by Year" />')
 927                 fg = open(path + '/commits_by_year.dat', 'w')
 928                 for yy in sorted(data.commits_by_year.keys()):
 929                         fg.write('%d %d\n' % (yy, data.commits_by_year[yy]))
 930                 fg.close()
 931
 932                 # Commits by timezone
 933                 f.write(html_header(2, 'Commits by Timezone'))
 934                 f.write('<table><tr>')
 935                 f.write('<th>Timezone</th><th>Commits</th>')
 936                 max_commits_on_tz = max(data.commits_by_timezone.values())
 937                 for i in sorted(data.commits_by_timezone.keys(), key = lambda n : int(n)):
 938                         commits = data.commits_by_timezone[i]
 939                         r = 127 + int((float(commits) / max_commits_on_tz) * 128)
 940                         f.write('<tr><th>%s</th><td style="background-color: rgb(%d, 0, 0)">%d</td></tr>' % (i, r, commits))
 941                 f.write('</tr></table>')
 942
 943                 f.write('</body></html>')
 944                 f.close()
 945
 946                 ###
 947                 # Authors
 948                 f = open(path + '/authors.html', 'w')
 949                 self.printHeader(f)
 950
 951                 f.write('<h1>Authors</h1>')
 952                 self.printNav(f)
 953
 954                 # Authors :: List of authors
 955                 f.write(html_header(2, 'List of Authors'))
 956
 957                 f.write('<table class="authors sortable" id="authors">')
 958                 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>')
 959                 for author in data.getAuthors(conf['max_authors']):
 960                         info = data.getAuthorInfo(author)
 961                         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']))
 962                 f.write('</table>')
 963
 964                 allauthors = data.getAuthors()
 965                 if len(allauthors) > conf['max_authors']:
 966                         rest = allauthors[conf['max_authors']:]
 967                         f.write('<p class="moreauthors">These didn\'t make it to the top: %s</p>' % ', '.join(rest))
 968
 969                 f.write(html_header(2, 'Cumulated Added Lines of Code per Author'))
 970                 f.write('<img src="lines_of_code_by_author.png" alt="Lines of code per Author" />')
 971                 if len(allauthors) > conf['max_authors']:
 972                         f.write('<p class="moreauthors">Only top %d authors shown</p>' % conf['max_authors'])
 973
 974                 f.write(html_header(2, 'Commits per Author'))
 975                 f.write('<img src="commits_by_author.png" alt="Commits per Author" />')
 976                 if len(allauthors) > conf['max_authors']:
 977                         f.write('<p class="moreauthors">Only top %d authors shown</p>' % conf['max_authors'])
 978
 979                 fgl = open(path + '/lines_of_code_by_author.dat', 'w')
 980                 fgc = open(path + '/commits_by_author.dat', 'w')
 981
 982                 lines_by_authors = {} # cumulated added lines by
 983                 # author. to save memory,
 984                 # changes_by_date_by_author[stamp][author] is defined
 985                 # only at points where author commits.
 986                 # lines_by_authors allows us to generate all the
 987                 # points in the .dat file.
 988
 989                 # Don't rely on getAuthors to give the same order each
 990                 # time. Be robust and keep the list in a variable.
 991                 commits_by_authors = {} # cumulated added lines by
 992
 993                 self.authors_to_plot = data.getAuthors(conf['max_authors'])
 994                 for author in self.authors_to_plot:
 995                         lines_by_authors[author] = 0
 996                         commits_by_authors[author] = 0
 997                 for stamp in sorted(data.changes_by_date_by_author.keys()):
 998                         fgl.write('%d' % stamp)
 999                         fgc.write('%d' % stamp)
1000                         for author in self.authors_to_plot:
1001                                 if author in data.changes_by_date_by_author[stamp].keys():
1002                                         lines_by_authors[author] = data.changes_by_date_by_author[stamp][author]['lines_added']
1003                                         commits_by_authors[author] = data.changes_by_date_by_author[stamp][author]['commits']
1004                                 fgl.write(' %d' % lines_by_authors[author])
1005                                 fgc.write(' %d' % commits_by_authors[author])
1006                         fgl.write('\n')
1007                         fgc.write('\n')
1008                 fgl.close()
1009                 fgc.close()
1010
1011                 # Authors :: Author of Month
1012                 f.write(html_header(2, 'Author of Month'))
1013                 f.write('<table class="sortable" id="aom">')
1014                 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'])
1015                 for yymm in reversed(sorted(data.author_of_month.keys())):
1016                         authordict = data.author_of_month[yymm]
1017                         authors = getkeyssortedbyvalues(authordict)
1018                         authors.reverse()
1019                         commits = data.author_of_month[yymm][authors[0]]
1020                         next = ', '.join(authors[1:conf['authors_top']+1])
1021                         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)))
1022
1023                 f.write('</table>')
1024
1025                 f.write(html_header(2, 'Author of Year'))
1026                 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'])
1027                 for yy in reversed(sorted(data.author_of_year.keys())):
1028                         authordict = data.author_of_year[yy]
1029                         authors = getkeyssortedbyvalues(authordict)
1030                         authors.reverse()
1031                         commits = data.author_of_year[yy][authors[0]]
1032                         next = ', '.join(authors[1:conf['authors_top']+1])
1033                         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)))
1034                 f.write('</table>')
1035
1036                 # Domains
1037                 f.write(html_header(2, 'Commits by Domains'))
1038                 domains_by_commits = getkeyssortedbyvaluekey(data.domains, 'commits')
1039                 domains_by_commits.reverse() # most first
1040                 f.write('<div class="vtable"><table>')
1041                 f.write('<tr><th>Domains</th><th>Total (%)</th></tr>')
1042                 fp = open(path + '/domains.dat', 'w')
1043                 n = 0
1044                 for domain in domains_by_commits:
1045                         if n == conf['max_domains']:
1046                                 break
1047                         commits = 0
1048                         n += 1
1049                         info = data.getDomainInfo(domain)
1050                         fp.write('%s %d %d\n' % (domain, n , info['commits']))
1051                         f.write('<tr><th>%s</th><td>%d (%.2f%%)</td></tr>' % (domain, info['commits'], (100.0 * info['commits'] / totalcommits)))
1052                 f.write('</table></div>')
1053                 f.write('<img src="domains.png" alt="Commits by Domains" />')
1054                 fp.close()
1055
1056                 f.write('</body></html>')
1057                 f.close()
1058
1059                 ###
1060                 # Files
1061                 f = open(path + '/files.html', 'w')
1062                 self.printHeader(f)
1063                 f.write('<h1>Files</h1>')
1064                 self.printNav(f)
1065
1066                 f.write('<dl>\n')
1067                 f.write('<dt>Total files</dt><dd>%d</dd>' % data.getTotalFiles())
1068                 f.write('<dt>Total lines</dt><dd>%d</dd>' % data.getTotalLOC())
1069                 try:
1070                         f.write('<dt>Average file size</dt><dd>%.2f bytes</dd>' % (float(data.getTotalSize()) / data.getTotalFiles()))
1071                 except ZeroDivisionError:
1072                         pass
1073                 f.write('</dl>\n')
1074
1075                 # Files :: File count by date
1076                 f.write(html_header(2, 'File count by date'))
1077
1078                 # use set to get rid of duplicate/unnecessary entries
1079                 files_by_date = set()
1080                 for stamp in sorted(data.files_by_stamp.keys()):
1081                         files_by_date.add('%s %d' % (datetime.datetime.fromtimestamp(stamp).strftime('%Y-%m-%d'), data.files_by_stamp[stamp]))
1082
1083                 fg = open(path + '/files_by_date.dat', 'w')
1084                 for line in sorted(list(files_by_date)):
1085                         fg.write('%s\n' % line)
1086                 #for stamp in sorted(data.files_by_stamp.keys()):
1087                 #       fg.write('%s %d\n' % (datetime.datetime.fromtimestamp(stamp).strftime('%Y-%m-%d'), data.files_by_stamp[stamp]))
1088                 fg.close()
1089
1090                 f.write('<img src="files_by_date.png" alt="Files by Date" />')
1091
1092                 #f.write('<h2>Average file size by date</h2>')
1093
1094                 # Files :: Extensions
1095                 f.write(html_header(2, 'Extensions'))
1096                 f.write('<table class="sortable" id="ext"><tr><th>Extension</th><th>Files (%)</th><th>Lines (%)</th><th>Lines/file</th></tr>')
1097                 for ext in sorted(data.extensions.keys()):
1098                         files = data.extensions[ext]['files']
1099                         lines = data.extensions[ext]['lines']
1100                         try:
1101                                 loc_percentage = (100.0 * lines) / data.getTotalLOC()
1102                         except ZeroDivisionError:
1103                                 loc_percentage = 0
1104                         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))
1105                 f.write('</table>')
1106
1107                 f.write('</body></html>')
1108                 f.close()
1109
1110                 ###
1111                 # Lines
1112                 f = open(path + '/lines.html', 'w')
1113                 self.printHeader(f)
1114                 f.write('<h1>Lines</h1>')
1115                 self.printNav(f)
1116
1117                 f.write('<dl>\n')
1118                 f.write('<dt>Total lines</dt><dd>%d</dd>' % data.getTotalLOC())
1119                 f.write('</dl>\n')
1120
1121                 f.write(html_header(2, 'Lines of Code'))
1122                 f.write('<img src="lines_of_code.png" />')
1123
1124                 fg = open(path + '/lines_of_code.dat', 'w')
1125                 for stamp in sorted(data.changes_by_date.keys()):
1126                         fg.write('%d %d\n' % (stamp, data.changes_by_date[stamp]['lines']))
1127                 fg.close()
1128
1129                 f.write('</body></html>')
1130                 f.close()
1131
1132                 ###
1133                 # tags.html
1134                 f = open(path + '/tags.html', 'w')
1135                 self.printHeader(f)
1136                 f.write('<h1>Tags</h1>')
1137                 self.printNav(f)
1138
1139                 f.write('<dl>')
1140                 f.write('<dt>Total tags</dt><dd>%d</dd>' % len(data.tags))
1141                 if len(data.tags) > 0:
1142                         f.write('<dt>Average commits per tag</dt><dd>%.2f</dd>' % (1.0 * data.getTotalCommits() / len(data.tags)))
1143                 f.write('</dl>')
1144
1145                 f.write('<table class="tags">')
1146                 f.write('<tr><th>Name</th><th>Date</th><th>Commits</th><th>Authors</th></tr>')
1147                 # sort the tags by date desc
1148                 tags_sorted_by_date_desc = map(lambda el : el[1], reversed(sorted(map(lambda el : (el[1]['date'], el[0]), data.tags.items()))))
1149                 for tag in tags_sorted_by_date_desc:
1150                         authorinfo = []
1151                         self.authors_by_commits = getkeyssortedbyvalues(data.tags[tag]['authors'])
1152                         for i in reversed(self.authors_by_commits):
1153                                 authorinfo.append('%s (%d)' % (i, data.tags[tag]['authors'][i]))
1154                         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)))
1155                 f.write('</table>')
1156
1157                 f.write('</body></html>')
1158                 f.close()
1159
1160                 self.createGraphs(path)
1161
1162         def createGraphs(self, path):
1163                 print 'Generating graphs...'
1164
1165                 # hour of day
1166                 f = open(path + '/hour_of_day.plot', 'w')
1167                 f.write(GNUPLOT_COMMON)
1168                 f.write(
1169 """
1170 set output 'hour_of_day.png'
1171 unset key
1172 set xrange [0.5:24.5]
1173 set xtics 4
1174 set grid y
1175 set ylabel "Commits"
1176 plot 'hour_of_day.dat' using 1:2:(0.5) w boxes fs solid
1177 """)
1178                 f.close()
1179
1180                 # day of week
1181                 f = open(path + '/day_of_week.plot', 'w')
1182                 f.write(GNUPLOT_COMMON)
1183                 f.write(
1184 """
1185 set output 'day_of_week.png'
1186 unset key
1187 set xrange [0.5:7.5]
1188 set xtics 1
1189 set grid y
1190 set ylabel "Commits"
1191 plot 'day_of_week.dat' using 1:3:(0.5):xtic(2) w boxes fs solid
1192 """)
1193                 f.close()
1194
1195                 # Domains
1196                 f = open(path + '/domains.plot', 'w')
1197                 f.write(GNUPLOT_COMMON)
1198                 f.write(
1199 """
1200 set output 'domains.png'
1201 unset key
1202 unset xtics
1203 set yrange [0:]
1204 set grid y
1205 set ylabel "Commits"
1206 plot 'domains.dat' using 2:3:(0.5) with boxes fs solid, '' using 2:3:1 with labels rotate by 45 offset 0,1
1207 """)
1208                 f.close()
1209
1210                 # Month of Year
1211                 f = open(path + '/month_of_year.plot', 'w')
1212                 f.write(GNUPLOT_COMMON)
1213                 f.write(
1214 """
1215 set output 'month_of_year.png'
1216 unset key
1217 set xrange [0.5:12.5]
1218 set xtics 1
1219 set grid y
1220 set ylabel "Commits"
1221 plot 'month_of_year.dat' using 1:2:(0.5) w boxes fs solid
1222 """)
1223                 f.close()
1224
1225                 # commits_by_year_month
1226                 f = open(path + '/commits_by_year_month.plot', 'w')
1227                 f.write(GNUPLOT_COMMON)
1228                 f.write(
1229 """
1230 set output 'commits_by_year_month.png'
1231 unset key
1232 set xdata time
1233 set timefmt "%Y-%m"
1234 set format x "%Y-%m"
1235 set xtics rotate
1236 set bmargin 5
1237 set grid y
1238 set ylabel "Commits"
1239 plot 'commits_by_year_month.dat' using 1:2:(0.5) w boxes fs solid
1240 """)
1241                 f.close()
1242
1243                 # commits_by_year
1244                 f = open(path + '/commits_by_year.plot', 'w')
1245                 f.write(GNUPLOT_COMMON)
1246                 f.write(
1247 """
1248 set output 'commits_by_year.png'
1249 unset key
1250 set xtics 1 rotate
1251 set grid y
1252 set ylabel "Commits"
1253 set yrange [0:]
1254 plot 'commits_by_year.dat' using 1:2:(0.5) w boxes fs solid
1255 """)
1256                 f.close()
1257
1258                 # Files by date
1259                 f = open(path + '/files_by_date.plot', 'w')
1260                 f.write(GNUPLOT_COMMON)
1261                 f.write(
1262 """
1263 set output 'files_by_date.png'
1264 unset key
1265 set xdata time
1266 set timefmt "%Y-%m-%d"
1267 set format x "%Y-%m-%d"
1268 set grid y
1269 set ylabel "Files"
1270 set xtics rotate
1271 set ytics autofreq
1272 set bmargin 6
1273 plot 'files_by_date.dat' using 1:2 w steps
1274 """)
1275                 f.close()
1276
1277                 # Lines of Code
1278                 f = open(path + '/lines_of_code.plot', 'w')
1279                 f.write(GNUPLOT_COMMON)
1280                 f.write(
1281 """
1282 set output 'lines_of_code.png'
1283 unset key
1284 set xdata time
1285 set timefmt "%s"
1286 set format x "%Y-%m-%d"
1287 set grid y
1288 set ylabel "Lines"
1289 set xtics rotate
1290 set bmargin 6
1291 plot 'lines_of_code.dat' using 1:2 w lines
1292 """)
1293                 f.close()
1294
1295                 # Lines of Code Added per author
1296                 f = open(path + '/lines_of_code_by_author.plot', 'w')
1297                 f.write(GNUPLOT_COMMON)
1298                 f.write(
1299 """
1300 set terminal png transparent size 640,480
1301 set output 'lines_of_code_by_author.png'
1302 set key left top
1303 set xdata time
1304 set timefmt "%s"
1305 set format x "%Y-%m-%d"
1306 set grid y
1307 set ylabel "Lines"
1308 set xtics rotate
1309 set bmargin 6
1310 plot """
1311 )
1312                 i = 1
1313                 plots = []
1314                 for a in self.authors_to_plot:
1315                         i = i + 1
1316                         plots.append("""'lines_of_code_by_author.dat' using 1:%d title "%s" w lines""" % (i, a.replace("\"", "\\\"")))
1317                 f.write(", ".join(plots))
1318                 f.write('\n')
1319
1320                 f.close()
1321
1322                 # Commits per author
1323                 f = open(path + '/commits_by_author.plot', 'w')
1324                 f.write(GNUPLOT_COMMON)
1325                 f.write(
1326 """
1327 set terminal png transparent size 640,480
1328 set output 'commits_by_author.png'
1329 set key left top
1330 set xdata time
1331 set timefmt "%s"
1332 set format x "%Y-%m-%d"
1333 set grid y
1334 set ylabel "Commits"
1335 set xtics rotate
1336 set bmargin 6
1337 plot """
1338 )
1339                 i = 1
1340                 plots = []
1341                 for a in self.authors_to_plot:
1342                         i = i + 1
1343                         plots.append("""'commits_by_author.dat' using 1:%d title "%s" w lines""" % (i, a.replace("\"", "\\\"")))
1344                 f.write(", ".join(plots))
1345                 f.write('\n')
1346
1347                 f.close()
1348
1349                 os.chdir(path)
1350                 files = glob.glob(path + '/*.plot')
1351                 for f in files:
1352                         out = getpipeoutput([gnuplot_cmd + ' "%s"' % f])
1353                         if len(out) > 0:
1354                                 print out
1355
1356         def printHeader(self, f, title = ''):
1357                 f.write(
1358 """<?xml version="1.0" encoding="UTF-8"?>
1359 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
1360 <html xmlns="http://www.w3.org/1999/xhtml">
1361 <head>
1362         <title>GitStats - %s</title>
1363         <link rel="stylesheet" href="%s" type="text/css" />
1364         <meta name="generator" content="GitStats %s" />
1365         <script type="text/javascript" src="sortable.js"></script>
1366 </head>
1367 <body>
1368 """ % (self.title, conf['style'], getversion()))
1369
1370         def printNav(self, f):
1371                 f.write("""
1372 <div class="nav">
1373 <ul>
1374 <li><a href="index.html">General</a></li>
1375 <li><a href="activity.html">Activity</a></li>
1376 <li><a href="authors.html">Authors</a></li>
1377 <li><a href="files.html">Files</a></li>
1378 <li><a href="lines.html">Lines</a></li>
1379 <li><a href="tags.html">Tags</a></li>
1380 </ul>
1381 </div>
1382 """)
1383
1384 def usage():
1385         print """
1386 Usage: gitstats [options] <gitpath..> <outputpath>
1387
1388 Options:
1389 -c key=value     Override configuration value
1390
1391 Default config values:
1392 %s
1393
1394 Please see the manual page for more details.
1395 """ % conf
1396
1397
1398 class GitStats:
1399         def run(self, args_orig):
1400                 optlist, args = getopt.getopt(args_orig, 'hc:', ["help"])
1401                 for o,v in optlist:
1402                         if o == '-c':
1403                                 key, value = v.split('=', 1)
1404                                 if key not in conf:
1405                                         raise KeyError('no such key "%s" in config' % key)
1406                                 if isinstance(conf[key], int):
1407                                         conf[key] = int(value)
1408                                 elif isinstance(conf[key], dict):
1409                                         kk,vv = value.split(',', 1)
1410                                         conf[key][kk] = vv
1411                                 else:
1412                                         conf[key] = value
1413                         elif o in ('-h', '--help'):
1414                                 usage()
1415                                 sys.exit()
1416
1417                 if len(args) < 2:
1418                         usage()
1419                         sys.exit(0)
1420
1421                 outputpath = os.path.abspath(args[-1])
1422                 rundir = os.getcwd()
1423
1424                 try:
1425                         os.makedirs(outputpath)
1426                 except OSError:
1427                         pass
1428                 if not os.path.isdir(outputpath):
1429                         print 'FATAL: Output path is not a directory or does not exist'
1430                         sys.exit(1)
1431
1432                 if not getgnuplotversion():
1433                         print 'gnuplot not found'
1434                         sys.exit(1)
1435
1436                 print 'Output path: %s' % outputpath
1437                 cachefile = os.path.join(outputpath, 'gitstats.cache')
1438
1439                 data = GitDataCollector()
1440                 data.loadCache(cachefile)
1441
1442                 for gitpath in args[0:-1]:
1443                         print 'Git path: %s' % gitpath
1444
1445                         os.chdir(gitpath)
1446
1447                         print 'Collecting data...'
1448                         data.collect(gitpath)
1449
1450                 print 'Refining data...'
1451                 data.saveCache(cachefile)
1452                 data.refine()
1453
1454                 os.chdir(rundir)
1455
1456                 print 'Generating report...'
1457                 report = HTMLReportCreator()
1458                 report.create(data, outputpath)
1459
1460                 time_end = time.time()
1461                 exectime_internal = time_end - time_start
1462                 print 'Execution time %.5f secs, %.5f secs (%.2f %%) in external commands)' % (exectime_internal, exectime_external, (100.0 * exectime_external) / exectime_internal)
1463                 if sys.stdin.isatty():
1464                         print 'You may now run:'
1465                         print
1466                         print '   sensible-browser \'%s\'' % os.path.join(outputpath, 'index.html').replace("'", "'\\''")
1467                         print
1468
1469 if __name__=='__main__':
1470         g = GitStats()
1471         g.run(sys.argv[1:])
1472