gitstats

   1 #!/usr/bin/env python
   2 # Copyright (c) 2007-2014 Heikki Hokkanen <hoxu@users.sf.net> & others (see doc/AUTHOR)
   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         'processes': 8,
  51 }
  52
  53 def getpipeoutput(cmds, quiet = False):
  54         global exectime_external
  55         start = time.time()
  56         if not quiet and ON_LINUX and os.isatty(1):
  57                 print '>> ' + ' | '.join(cmds),
  58                 sys.stdout.flush()
  59         p0 = subprocess.Popen(cmds[0], stdout = subprocess.PIPE, shell = True)
  60         p = p0
  61         for x in cmds[1:]:
  62                 p = subprocess.Popen(x, stdin = p0.stdout, stdout = subprocess.PIPE, shell = True)
  63                 p0 = p
  64         output = p.communicate()[0]
  65         end = time.time()
  66         if not quiet:
  67                 if ON_LINUX and os.isatty(1):
  68                         print '\r',
  69                 print '[%.5f] >> %s' % (end - start, ' | '.join(cmds))
  70         exectime_external += (end - start)
  71         return output.rstrip('\n')
  72
  73 def getcommitrange(defaultrange = 'HEAD', end_only = False):
  74         if len(conf['commit_end']) > 0:
  75                 if end_only or len(conf['commit_begin']) == 0:
  76                         return conf['commit_end']
  77                 return '%s..%s' % (conf['commit_begin'], conf['commit_end'])
  78         return defaultrange
  79
  80 def getkeyssortedbyvalues(dict):
  81         return map(lambda el : el[1], sorted(map(lambda el : (el[1], el[0]), dict.items())))
  82
  83 # dict['author'] = { 'commits': 512 } - ...key(dict, 'commits')
  84 def getkeyssortedbyvaluekey(d, key):
  85         return map(lambda el : el[1], sorted(map(lambda el : (d[el][key], el), d.keys())))
  86
  87 def getstatsummarycounts(line):
  88         numbers = re.findall('\d+', line)
  89         if   len(numbers) == 1:
  90                 # neither insertions nor deletions: may probably only happen for "0 files changed"
  91                 numbers.append(0);
  92                 numbers.append(0);
  93         elif len(numbers) == 2 and line.find('(+)') != -1:
  94                 numbers.append(0);    # only insertions were printed on line
  95         elif len(numbers) == 2 and line.find('(-)') != -1:
  96                 numbers.insert(1, 0); # only deletions were printed on line
  97         return numbers
  98
  99 VERSION = 0
 100 def getversion():
 101         global VERSION
 102         if VERSION == 0:
 103                 gitstats_repo = os.path.dirname(os.path.abspath(__file__))
 104                 VERSION = getpipeoutput(["git --git-dir=%s/.git --work-tree=%s rev-parse --short %s" %
 105                         (gitstats_repo, gitstats_repo, getcommitrange('HEAD').split('\n')[0])])
 106         return VERSION
 107
 108 def getgitversion():
 109         return getpipeoutput(['git --version']).split('\n')[0]
 110
 111 def getgnuplotversion():
 112         return getpipeoutput(['%s --version' % gnuplot_cmd]).split('\n')[0]
 113
 114 def getnumoffilesfromrev(time_rev):
 115         """
 116         Get number of files changed in commit
 117         """
 118         time, rev = time_rev
 119         return (int(time), rev, int(getpipeoutput(['git ls-tree -r --name-only "%s"' % rev, 'wc -l']).split('\n')[0]))
 120
 121 def getnumoflinesinblob(ext_blob):
 122         """
 123         Get number of lines in blob
 124         """
 125         ext, blob_id = ext_blob
 126         return (ext, blob_id, int(getpipeoutput(['git cat-file blob %s' % blob_id, 'wc -l']).split()[0]))
 127
 128 class DataCollector:
 129         """Manages data collection from a revision control repository."""
 130         def __init__(self):
 131                 self.stamp_created = time.time()
 132                 self.cache = {}
 133                 self.total_authors = 0
 134                 self.activity_by_hour_of_day = {} # hour -> commits
 135                 self.activity_by_day_of_week = {} # day -> commits
 136                 self.activity_by_month_of_year = {} # month [1-12] -> commits
 137                 self.activity_by_hour_of_week = {} # weekday -> hour -> commits
 138                 self.activity_by_hour_of_day_busiest = 0
 139                 self.activity_by_hour_of_week_busiest = 0
 140                 self.activity_by_year_week = {} # yy_wNN -> commits
 141                 self.activity_by_year_week_peak = 0
 142
 143                 self.authors = {} # name -> {commits, first_commit_stamp, last_commit_stamp, last_active_day, active_days, lines_added, lines_removed}
 144
 145                 self.total_commits = 0
 146                 self.total_files = 0
 147                 self.authors_by_commits = 0
 148
 149                 # domains
 150                 self.domains = {} # domain -> commits
 151
 152                 # author of the month
 153                 self.author_of_month = {} # month -> author -> commits
 154                 self.author_of_year = {} # year -> author -> commits
 155                 self.commits_by_month = {} # month -> commits
 156                 self.commits_by_year = {} # year -> commits
 157                 self.lines_added_by_month = {} # month -> lines added
 158                 self.lines_added_by_year = {} # year -> lines added
 159                 self.lines_removed_by_month = {} # month -> lines removed
 160                 self.lines_removed_by_year = {} # year -> lines removed
 161                 self.first_commit_stamp = 0
 162                 self.last_commit_stamp = 0
 163                 self.last_active_day = None
 164                 self.active_days = set()
 165
 166                 # lines
 167                 self.total_lines = 0
 168                 self.total_lines_added = 0
 169                 self.total_lines_removed = 0
 170
 171                 # size
 172                 self.total_size = 0
 173
 174                 # timezone
 175                 self.commits_by_timezone = {} # timezone -> commits
 176
 177                 # tags
 178                 self.tags = {}
 179
 180                 self.files_by_stamp = {} # stamp -> files
 181
 182                 # extensions
 183                 self.extensions = {} # extension -> files, lines
 184
 185                 # line statistics
 186                 self.changes_by_date = {} # stamp -> { files, ins, del }
 187
 188         ##
 189         # This should be the main function to extract data from the repository.
 190         def collect(self, dir):
 191                 self.dir = dir
 192                 if len(conf['project_name']) == 0:
 193                         self.projectname = os.path.basename(os.path.abspath(dir))
 194                 else:
 195                         self.projectname = conf['project_name']
 196
 197         ##
 198         # Load cacheable data
 199         def loadCache(self, cachefile):
 200                 if not os.path.exists(cachefile):
 201                         return
 202                 print 'Loading cache...'
 203                 f = open(cachefile, 'rb')
 204                 try:
 205                         self.cache = pickle.loads(zlib.decompress(f.read()))
 206                 except:
 207                         # temporary hack to upgrade non-compressed caches
 208                         f.seek(0)
 209                         self.cache = pickle.load(f)
 210                 f.close()
 211
 212         ##
 213         # Produce any additional statistics from the extracted data.
 214         def refine(self):
 215                 pass
 216
 217         ##
 218         # : get a dictionary of author
 219         def getAuthorInfo(self, author):
 220                 return None
 221
 222         def getActivityByDayOfWeek(self):
 223                 return {}
 224
 225         def getActivityByHourOfDay(self):
 226                 return {}
 227
 228         # : get a dictionary of domains
 229         def getDomainInfo(self, domain):
 230                 return None
 231
 232         ##
 233         # Get a list of authors
 234         def getAuthors(self):
 235                 return []
 236
 237         def getFirstCommitDate(self):
 238                 return datetime.datetime.now()
 239
 240         def getLastCommitDate(self):
 241                 return datetime.datetime.now()
 242
 243         def getStampCreated(self):
 244                 return self.stamp_created
 245
 246         def getTags(self):
 247                 return []
 248
 249         def getTotalAuthors(self):
 250                 return -1
 251
 252         def getTotalCommits(self):
 253                 return -1
 254
 255         def getTotalFiles(self):
 256                 return -1
 257
 258         def getTotalLOC(self):
 259                 return -1
 260
 261         ##
 262         # Save cacheable data
 263         def saveCache(self, cachefile):
 264                 print 'Saving cache...'
 265                 tempfile = cachefile + '.tmp'
 266                 f = open(tempfile, 'wb')
 267                 #pickle.dump(self.cache, f)
 268                 data = zlib.compress(pickle.dumps(self.cache))
 269                 f.write(data)
 270                 f.close()
 271                 try:
 272                         os.remove(cachefile)
 273                 except OSError:
 274                         pass
 275                 os.rename(tempfile, cachefile)
 276
 277 class GitDataCollector(DataCollector):
 278         def collect(self, dir):
 279                 DataCollector.collect(self, dir)
 280
 281                 self.total_authors += int(getpipeoutput(['git shortlog -s %s' % getcommitrange(), 'wc -l']))
 282                 #self.total_lines = int(getoutput('git-ls-files -z |xargs -0 cat |wc -l'))
 283
 284                 # tags
 285                 lines = getpipeoutput(['git show-ref --tags']).split('\n')
 286                 for line in lines:
 287                         if len(line) == 0:
 288                                 continue
 289                         (hash, tag) = line.split(' ')
 290
 291                         tag = tag.replace('refs/tags/', '')
 292                         output = getpipeoutput(['git log "%s" --pretty=format:"%%at %%aN" -n 1' % hash])
 293                         if len(output) > 0:
 294                                 parts = output.split(' ')
 295                                 stamp = 0
 296                                 try:
 297                                         stamp = int(parts[0])
 298                                 except ValueError:
 299                                         stamp = 0
 300                                 self.tags[tag] = { 'stamp': stamp, 'hash' : hash, 'date' : datetime.datetime.fromtimestamp(stamp).strftime('%Y-%m-%d'), 'commits': 0, 'authors': {} }
 301
 302                 # collect info on tags, starting from latest
 303                 tags_sorted_by_date_desc = map(lambda el : el[1], reversed(sorted(map(lambda el : (el[1]['date'], el[0]), self.tags.items()))))
 304                 prev = None
 305                 for tag in reversed(tags_sorted_by_date_desc):
 306                         cmd = 'git shortlog -s "%s"' % tag
 307                         if prev != None:
 308                                 cmd += ' "^%s"' % prev
 309                         output = getpipeoutput([cmd])
 310                         if len(output) == 0:
 311                                 continue
 312                         prev = tag
 313                         for line in output.split('\n'):
 314                                 parts = re.split('\s+', line, 2)
 315                                 commits = int(parts[1])
 316                                 author = parts[2]
 317                                 if author in conf['merge_authors']:
 318                                         author = conf['merge_authors'][author]
 319                                 self.tags[tag]['commits'] += commits
 320                                 self.tags[tag]['authors'][author] = commits
 321
 322                 # Collect revision statistics
 323                 # Outputs "<stamp> <date> <time> <timezone> <author> '<' <mail> '>'"
 324                 lines = getpipeoutput(['git rev-list --pretty=format:"%%at %%ai %%aN <%%aE>" %s' % getcommitrange('HEAD'), 'grep -v ^commit']).split('\n')
 325                 for line in lines:
 326                         parts = line.split(' ', 4)
 327                         author = ''
 328                         try:
 329                                 stamp = int(parts[0])
 330                         except ValueError:
 331                                 stamp = 0
 332                         timezone = parts[3]
 333                         author, mail = parts[4].split('<', 1)
 334                         author = author.rstrip()
 335                         if author in conf['merge_authors']:
 336                                 author = conf['merge_authors'][author]
 337                         mail = mail.rstrip('>')
 338                         domain = '?'
 339                         if mail.find('@') != -1:
 340                                 domain = mail.rsplit('@', 1)[1]
 341                         date = datetime.datetime.fromtimestamp(float(stamp))
 342
 343                         # First and last commit stamp (may be in any order because of cherry-picking and patches)
 344                         if stamp > self.last_commit_stamp:
 345                                 self.last_commit_stamp = stamp
 346                         if self.first_commit_stamp == 0 or stamp < self.first_commit_stamp:
 347                                 self.first_commit_stamp = stamp
 348
 349                         # activity
 350                         # hour
 351                         hour = date.hour
 352                         self.activity_by_hour_of_day[hour] = self.activity_by_hour_of_day.get(hour, 0) + 1
 353                         # most active hour?
 354                         if self.activity_by_hour_of_day[hour] > self.activity_by_hour_of_day_busiest:
 355                                 self.activity_by_hour_of_day_busiest = self.activity_by_hour_of_day[hour]
 356
 357                         # day of week
 358                         day = date.weekday()
 359                         self.activity_by_day_of_week[day] = self.activity_by_day_of_week.get(day, 0) + 1
 360
 361                         # domain stats
 362                         if domain not in self.domains:
 363                                 self.domains[domain] = {}
 364                         # commits
 365                         self.domains[domain]['commits'] = self.domains[domain].get('commits', 0) + 1
 366
 367                         # hour of week
 368                         if day not in self.activity_by_hour_of_week:
 369                                 self.activity_by_hour_of_week[day] = {}
 370                         self.activity_by_hour_of_week[day][hour] = self.activity_by_hour_of_week[day].get(hour, 0) + 1
 371                         # most active hour?
 372                         if self.activity_by_hour_of_week[day][hour] > self.activity_by_hour_of_week_busiest:
 373                                 self.activity_by_hour_of_week_busiest = self.activity_by_hour_of_week[day][hour]
 374
 375                         # month of year
 376                         month = date.month
 377                         self.activity_by_month_of_year[month] = self.activity_by_month_of_year.get(month, 0) + 1
 378
 379                         # yearly/weekly activity
 380                         yyw = date.strftime('%Y-%W')
 381                         self.activity_by_year_week[yyw] = self.activity_by_year_week.get(yyw, 0) + 1
 382                         if self.activity_by_year_week_peak < self.activity_by_year_week[yyw]:
 383                                 self.activity_by_year_week_peak = self.activity_by_year_week[yyw]
 384
 385                         # author stats
 386                         if author not in self.authors:
 387                                 self.authors[author] = {}
 388                         # commits, note again that commits may be in any date order because of cherry-picking and patches
 389                         if 'last_commit_stamp' not in self.authors[author]:
 390                                 self.authors[author]['last_commit_stamp'] = stamp
 391                         if stamp > self.authors[author]['last_commit_stamp']:
 392                                 self.authors[author]['last_commit_stamp'] = stamp
 393                         if 'first_commit_stamp' not in self.authors[author]:
 394                                 self.authors[author]['first_commit_stamp'] = stamp
 395                         if stamp < self.authors[author]['first_commit_stamp']:
 396                                 self.authors[author]['first_commit_stamp'] = stamp
 397
 398                         # author of the month/year
 399                         yymm = date.strftime('%Y-%m')
 400                         if yymm in self.author_of_month:
 401                                 self.author_of_month[yymm][author] = self.author_of_month[yymm].get(author, 0) + 1
 402                         else:
 403                                 self.author_of_month[yymm] = {}
 404                                 self.author_of_month[yymm][author] = 1
 405                         self.commits_by_month[yymm] = self.commits_by_month.get(yymm, 0) + 1
 406
 407                         yy = date.year
 408                         if yy in self.author_of_year:
 409                                 self.author_of_year[yy][author] = self.author_of_year[yy].get(author, 0) + 1
 410                         else:
 411                                 self.author_of_year[yy] = {}
 412                                 self.author_of_year[yy][author] = 1
 413                         self.commits_by_year[yy] = self.commits_by_year.get(yy, 0) + 1
 414
 415                         # authors: active days
 416                         yymmdd = date.strftime('%Y-%m-%d')
 417                         if 'last_active_day' not in self.authors[author]:
 418                                 self.authors[author]['last_active_day'] = yymmdd
 419                                 self.authors[author]['active_days'] = set([yymmdd])
 420                         elif yymmdd != self.authors[author]['last_active_day']:
 421                                 self.authors[author]['last_active_day'] = yymmdd
 422                                 self.authors[author]['active_days'].add(yymmdd)
 423
 424                         # project: active days
 425                         if yymmdd != self.last_active_day:
 426                                 self.last_active_day = yymmdd
 427                                 self.active_days.add(yymmdd)
 428
 429                         # timezone
 430                         self.commits_by_timezone[timezone] = self.commits_by_timezone.get(timezone, 0) + 1
 431
 432                 # outputs "<stamp> <files>" for each revision
 433                 revlines = getpipeoutput(['git rev-list --pretty=format:"%%at %%T" %s' % getcommitrange('HEAD'), 'grep -v ^commit']).strip().split('\n')
 434                 lines = []
 435                 revs_to_read = []
 436                 time_rev_count = []
 437                 #Look up rev in cache and take info from cache if found
 438                 #If not append rev to list of rev to read from repo
 439                 for revline in revlines:
 440                         time, rev = revline.split(' ')
 441                         #if cache empty then add time and rev to list of new rev's
 442                         #otherwise try to read needed info from cache
 443                         if 'files_in_tree' not in self.cache.keys():
 444                                 revs_to_read.append((time,rev))
 445                                 continue
 446                         if rev in self.cache['files_in_tree'].keys():
 447                                 lines.append('%d %d' % (int(time), self.cache['files_in_tree'][rev]))
 448                         else:
 449                                 revs_to_read.append((time,rev))
 450
 451                 #Read revisions from repo
 452                 time_rev_count = Pool(processes=conf['processes']).map(getnumoffilesfromrev, revs_to_read)
 453
 454                 #Update cache with new revisions and append then to general list
 455                 for (time, rev, count) in time_rev_count:
 456                         if 'files_in_tree' not in self.cache:
 457                                 self.cache['files_in_tree'] = {}
 458                         self.cache['files_in_tree'][rev] = count
 459                         lines.append('%d %d' % (int(time), count))
 460
 461                 self.total_commits += len(lines)
 462                 for line in lines:
 463                         parts = line.split(' ')
 464                         if len(parts) != 2:
 465                                 continue
 466                         (stamp, files) = parts[0:2]
 467                         try:
 468                                 self.files_by_stamp[int(stamp)] = int(files)
 469                         except ValueError:
 470                                 print 'Warning: failed to parse line "%s"' % line
 471
 472                 # extensions and size of files
 473                 lines = getpipeoutput(['git ls-tree -r -l -z %s' % getcommitrange('HEAD', end_only = True)]).split('\000')
 474                 blobs_to_read = []
 475                 for line in lines:
 476                         if len(line) == 0:
 477                                 continue
 478                         parts = re.split('\s+', line, 5)
 479                         if parts[0] == '160000' and parts[3] == '-':
 480                                 # skip submodules
 481                                 continue
 482                         blob_id = parts[2]
 483                         size = int(parts[3])
 484                         fullpath = parts[4]
 485
 486                         self.total_size += size
 487                         self.total_files += 1
 488
 489                         filename = fullpath.split('/')[-1] # strip directories
 490                         if filename.find('.') == -1 or filename.rfind('.') == 0:
 491                                 ext = ''
 492                         else:
 493                                 ext = filename[(filename.rfind('.') + 1):]
 494                         if len(ext) > conf['max_ext_length']:
 495                                 ext = ''
 496                         if ext not in self.extensions:
 497                                 self.extensions[ext] = {'files': 0, 'lines': 0}
 498                         self.extensions[ext]['files'] += 1
 499                         #if cache empty then add ext and blob id to list of new blob's
 500                         #otherwise try to read needed info from cache
 501                         if 'lines_in_blob' not in self.cache.keys():
 502                                 blobs_to_read.append((ext,blob_id))
 503                                 continue
 504                         if blob_id in self.cache['lines_in_blob'].keys():
 505                                 self.extensions[ext]['lines'] += self.cache['lines_in_blob'][blob_id]
 506                         else:
 507                                 blobs_to_read.append((ext,blob_id))
 508
 509                 #Get info abount line count for new blob's that wasn't found in cache
 510                 ext_blob_linecount = Pool(processes=conf['processes']).map(getnumoflinesinblob, blobs_to_read)
 511
 512                 #Update cache and write down info about number of number of lines
 513                 for (ext, blob_id, linecount) in ext_blob_linecount:
 514                         if 'lines_in_blob' not in self.cache:
 515                                 self.cache['lines_in_blob'] = {}
 516                         self.cache['lines_in_blob'][blob_id] = linecount
 517                         self.extensions[ext]['lines'] += self.cache['lines_in_blob'][blob_id]
 518
 519                 # line statistics
 520                 # outputs:
 521                 #  N files changed, N insertions (+), N deletions(-)
 522                 # <stamp> <author>
 523                 self.changes_by_date = {} # stamp -> { files, ins, del }
 524                 # computation of lines of code by date is better done
 525                 # on a linear history.
 526                 extra = ''
 527                 if conf['linear_linestats']:
 528                         extra = '--first-parent -m'
 529                 lines = getpipeoutput(['git log --shortstat %s --pretty=format:"%%at %%aN" %s' % (extra, getcommitrange('HEAD'))]).split('\n')
 530                 lines.reverse()
 531                 files = 0; inserted = 0; deleted = 0; total_lines = 0
 532                 author = None
 533                 for line in lines:
 534                         if len(line) == 0:
 535                                 continue
 536
 537                         # <stamp> <author>
 538                         if re.search('files? changed', line) == None:
 539                                 pos = line.find(' ')
 540                                 if pos != -1:
 541                                         try:
 542                                                 (stamp, author) = (int(line[:pos]), line[pos+1:])
 543                                                 if author in conf['merge_authors']:
 544                                                         author = conf['merge_authors'][author]
 545                                                 self.changes_by_date[stamp] = { 'files': files, 'ins': inserted, 'del': deleted, 'lines': total_lines }
 546
 547                                                 date = datetime.datetime.fromtimestamp(stamp)
 548                                                 yymm = date.strftime('%Y-%m')
 549                                                 self.lines_added_by_month[yymm] = self.lines_added_by_month.get(yymm, 0) + inserted
 550                                                 self.lines_removed_by_month[yymm] = self.lines_removed_by_month.get(yymm, 0) + deleted
 551
 552                                                 yy = date.year
 553                                                 self.lines_added_by_year[yy] = self.lines_added_by_year.get(yy,0) + inserted
 554                                                 self.lines_removed_by_year[yy] = self.lines_removed_by_year.get(yy, 0) + deleted
 555
 556                                                 files, inserted, deleted = 0, 0, 0
 557                                         except ValueError:
 558                                                 print 'Warning: unexpected line "%s"' % line
 559                                 else:
 560                                         print 'Warning: unexpected line "%s"' % line
 561                         else:
 562                                 numbers = getstatsummarycounts(line)
 563
 564                                 if len(numbers) == 3:
 565                                         (files, inserted, deleted) = map(lambda el : int(el), numbers)
 566                                         total_lines += inserted
 567                                         total_lines -= deleted
 568                                         self.total_lines_added += inserted
 569                                         self.total_lines_removed += deleted
 570
 571                                 else:
 572                                         print 'Warning: failed to handle line "%s"' % line
 573                                         (files, inserted, deleted) = (0, 0, 0)
 574                                 #self.changes_by_date[stamp] = { 'files': files, 'ins': inserted, 'del': deleted }
 575                 self.total_lines += total_lines
 576
 577                 # Per-author statistics
 578
 579                 # defined for stamp, author only if author commited at this timestamp.
 580                 self.changes_by_date_by_author = {} # stamp -> author -> lines_added
 581
 582                 # Similar to the above, but never use --first-parent
 583                 # (we need to walk through every commit to know who
 584                 # committed what, not just through mainline)
 585                 lines = getpipeoutput(['git log --shortstat --date-order --pretty=format:"%%at %%aN" %s' % (getcommitrange('HEAD'))]).split('\n')
 586                 lines.reverse()
 587                 files = 0; inserted = 0; deleted = 0
 588                 author = None
 589                 stamp = 0
 590                 for line in lines:
 591                         if len(line) == 0:
 592                                 continue
 593
 594                         # <stamp> <author>
 595                         if re.search('files? changed', line) == None:
 596                                 pos = line.find(' ')
 597                                 if pos != -1:
 598                                         try:
 599                                                 oldstamp = stamp
 600                                                 (stamp, author) = (int(line[:pos]), line[pos+1:])
 601                                                 if author in conf['merge_authors']:
 602                                                         author = conf['merge_authors'][author]
 603                                                 if oldstamp > stamp:
 604                                                         # clock skew, keep old timestamp to avoid having ugly graph
 605                                                         stamp = oldstamp
 606                                                 if author not in self.authors:
 607                                                         self.authors[author] = { 'lines_added' : 0, 'lines_removed' : 0, 'commits' : 0}
 608                                                 self.authors[author]['commits'] = self.authors[author].get('commits', 0) + 1
 609                                                 self.authors[author]['lines_added'] = self.authors[author].get('lines_added', 0) + inserted
 610                                                 self.authors[author]['lines_removed'] = self.authors[author].get('lines_removed', 0) + deleted
 611                                                 if stamp not in self.changes_by_date_by_author:
 612                                                         self.changes_by_date_by_author[stamp] = {}
 613                                                 if author not in self.changes_by_date_by_author[stamp]:
 614                                                         self.changes_by_date_by_author[stamp][author] = {}
 615                                                 self.changes_by_date_by_author[stamp][author]['lines_added'] = self.authors[author]['lines_added']
 616                                                 self.changes_by_date_by_author[stamp][author]['commits'] = self.authors[author]['commits']
 617                                                 files, inserted, deleted = 0, 0, 0
 618                                         except ValueError:
 619                                                 print 'Warning: unexpected line "%s"' % line
 620                                 else:
 621                                         print 'Warning: unexpected line "%s"' % line
 622                         else:
 623                                 numbers = getstatsummarycounts(line);
 624
 625                                 if len(numbers) == 3:
 626                                         (files, inserted, deleted) = map(lambda el : int(el), numbers)
 627                                 else:
 628                                         print 'Warning: failed to handle line "%s"' % line
 629                                         (files, inserted, deleted) = (0, 0, 0)
 630
 631         def refine(self):
 632                 # authors
 633                 # name -> {place_by_commits, commits_frac, date_first, date_last, timedelta}
 634                 self.authors_by_commits = getkeyssortedbyvaluekey(self.authors, 'commits')
 635                 self.authors_by_commits.reverse() # most first
 636                 for i, name in enumerate(self.authors_by_commits):
 637                         self.authors[name]['place_by_commits'] = i + 1
 638
 639                 for name in self.authors.keys():
 640                         a = self.authors[name]
 641                         a['commits_frac'] = (100 * float(a['commits'])) / self.getTotalCommits()
 642                         date_first = datetime.datetime.fromtimestamp(a['first_commit_stamp'])
 643                         date_last = datetime.datetime.fromtimestamp(a['last_commit_stamp'])
 644                         delta = date_last - date_first
 645                         a['date_first'] = date_first.strftime('%Y-%m-%d')
 646                         a['date_last'] = date_last.strftime('%Y-%m-%d')
 647                         a['timedelta'] = delta
 648                         if 'lines_added' not in a: a['lines_added'] = 0
 649                         if 'lines_removed' not in a: a['lines_removed'] = 0
 650
 651         def getActiveDays(self):
 652                 return self.active_days
 653
 654         def getActivityByDayOfWeek(self):
 655                 return self.activity_by_day_of_week
 656
 657         def getActivityByHourOfDay(self):
 658                 return self.activity_by_hour_of_day
 659
 660         def getAuthorInfo(self, author):
 661                 return self.authors[author]
 662
 663         def getAuthors(self, limit = None):
 664                 res = getkeyssortedbyvaluekey(self.authors, 'commits')
 665                 res.reverse()
 666                 return res[:limit]
 667
 668         def getCommitDeltaDays(self):
 669                 return (self.last_commit_stamp / 86400 - self.first_commit_stamp / 86400) + 1
 670
 671         def getDomainInfo(self, domain):
 672                 return self.domains[domain]
 673
 674         def getDomains(self):
 675                 return self.domains.keys()
 676
 677         def getFirstCommitDate(self):
 678                 return datetime.datetime.fromtimestamp(self.first_commit_stamp)
 679
 680         def getLastCommitDate(self):
 681                 return datetime.datetime.fromtimestamp(self.last_commit_stamp)
 682
 683         def getTags(self):
 684                 lines = getpipeoutput(['git show-ref --tags', 'cut -d/ -f3'])
 685                 return lines.split('\n')
 686
 687         def getTagDate(self, tag):
 688                 return self.revToDate('tags/' + tag)
 689
 690         def getTotalAuthors(self):
 691                 return self.total_authors
 692
 693         def getTotalCommits(self):
 694                 return self.total_commits
 695
 696         def getTotalFiles(self):
 697                 return self.total_files
 698
 699         def getTotalLOC(self):
 700                 return self.total_lines
 701
 702         def getTotalSize(self):
 703                 return self.total_size
 704
 705         def revToDate(self, rev):
 706                 stamp = int(getpipeoutput(['git log --pretty=format:%%at "%s" -n 1' % rev]))
 707                 return datetime.datetime.fromtimestamp(stamp).strftime('%Y-%m-%d')
 708
 709 class ReportCreator:
 710         """Creates the actual report based on given data."""
 711         def __init__(self):
 712                 pass
 713
 714         def create(self, data, path):
 715                 self.data = data
 716                 self.path = path
 717
 718 def html_linkify(text):
 719         return text.lower().replace(' ', '_')
 720
 721 def html_header(level, text):
 722         name = html_linkify(text)
 723         return '\n<h%d><a href="#%s" name="%s">%s</a></h%d>\n\n' % (level, name, name, text, level)
 724
 725 class HTMLReportCreator(ReportCreator):
 726         def create(self, data, path):
 727                 ReportCreator.create(self, data, path)
 728                 self.title = data.projectname
 729
 730                 # copy static files. Looks in the binary directory, ../share/gitstats and /usr/share/gitstats
 731                 binarypath = os.path.dirname(os.path.abspath(__file__))
 732                 secondarypath = os.path.join(binarypath, '..', 'share', 'gitstats')
 733                 basedirs = [binarypath, secondarypath, '/usr/share/gitstats']
 734                 for file in ('gitstats.css', 'sortable.js', 'arrow-up.gif', 'arrow-down.gif', 'arrow-none.gif'):
 735                         for base in basedirs:
 736                                 src = base + '/' + file
 737                                 if os.path.exists(src):
 738                                         shutil.copyfile(src, path + '/' + file)
 739                                         break
 740                         else:
 741                                 print 'Warning: "%s" not found, so not copied (searched: %s)' % (file, basedirs)
 742
 743                 f = open(path + "/index.html", 'w')
 744                 format = '%Y-%m-%d %H:%M:%S'
 745                 self.printHeader(f)
 746
 747                 f.write('<h1>GitStats - %s</h1>' % data.projectname)
 748
 749                 self.printNav(f)
 750
 751                 f.write('<dl>')
 752                 f.write('<dt>Project name</dt><dd>%s</dd>' % (data.projectname))
 753                 f.write('<dt>Generated</dt><dd>%s (in %d seconds)</dd>' % (datetime.datetime.now().strftime(format), time.time() - data.getStampCreated()))
 754                 f.write('<dt>Generator</dt><dd><a href="http://gitstats.sourceforge.net/">GitStats</a> (version %s), %s, %s</dd>' % (getversion(), getgitversion(), getgnuplotversion()))
 755                 f.write('<dt>Report Period</dt><dd>%s to %s</dd>' % (data.getFirstCommitDate().strftime(format), data.getLastCommitDate().strftime(format)))
 756                 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())))
 757                 f.write('<dt>Total Files</dt><dd>%s</dd>' % data.getTotalFiles())
 758                 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))
 759                 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()))
 760                 f.write('<dt>Authors</dt><dd>%s (average %.1f commits per author)</dd>' % (data.getTotalAuthors(), (1.0 * data.getTotalCommits()) / data.getTotalAuthors()))
 761                 f.write('</dl>')
 762
 763                 f.write('</body>\n</html>')
 764                 f.close()
 765
 766                 ###
 767                 # Activity
 768                 f = open(path + '/activity.html', 'w')
 769                 self.printHeader(f)
 770                 f.write('<h1>Activity</h1>')
 771                 self.printNav(f)
 772
 773                 #f.write('<h2>Last 30 days</h2>')
 774
 775                 #f.write('<h2>Last 12 months</h2>')
 776
 777                 # Weekly activity
 778                 WEEKS = 32
 779                 f.write(html_header(2, 'Weekly activity'))
 780                 f.write('<p>Last %d weeks</p>' % WEEKS)
 781
 782                 # generate weeks to show (previous N weeks from now)
 783                 now = datetime.datetime.now()
 784                 deltaweek = datetime.timedelta(7)
 785                 weeks = []
 786                 stampcur = now
 787                 for i in range(0, WEEKS):
 788                         weeks.insert(0, stampcur.strftime('%Y-%W'))
 789                         stampcur -= deltaweek
 790
 791                 # top row: commits & bar
 792                 f.write('<table class="noborders"><tr>')
 793                 for i in range(0, WEEKS):
 794                         commits = 0
 795                         if weeks[i] in data.activity_by_year_week:
 796                                 commits = data.activity_by_year_week[weeks[i]]
 797
 798                         percentage = 0
 799                         if weeks[i] in data.activity_by_year_week:
 800                                 percentage = float(data.activity_by_year_week[weeks[i]]) / data.activity_by_year_week_peak
 801                         height = max(1, int(200 * percentage))
 802                         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))
 803
 804                 # bottom row: year/week
 805                 f.write('</tr><tr>')
 806                 for i in range(0, WEEKS):
 807                         f.write('<td>%s</td>' % (WEEKS - i))
 808                 f.write('</tr></table>')
 809
 810                 # Hour of Day
 811                 f.write(html_header(2, 'Hour of Day'))
 812                 hour_of_day = data.getActivityByHourOfDay()
 813                 f.write('<table><tr><th>Hour</th>')
 814                 for i in range(0, 24):
 815                         f.write('<th>%d</th>' % i)
 816                 f.write('</tr>\n<tr><th>Commits</th>')
 817                 fp = open(path + '/hour_of_day.dat', 'w')
 818                 for i in range(0, 24):
 819                         if i in hour_of_day:
 820                                 r = 127 + int((float(hour_of_day[i]) / data.activity_by_hour_of_day_busiest) * 128)
 821                                 f.write('<td style="background-color: rgb(%d, 0, 0)">%d</td>' % (r, hour_of_day[i]))
 822                                 fp.write('%d %d\n' % (i, hour_of_day[i]))
 823                         else:
 824                                 f.write('<td>0</td>')
 825                                 fp.write('%d 0\n' % i)
 826                 fp.close()
 827                 f.write('</tr>\n<tr><th>%</th>')
 828                 totalcommits = data.getTotalCommits()
 829                 for i in range(0, 24):
 830                         if i in hour_of_day:
 831                                 r = 127 + int((float(hour_of_day[i]) / data.activity_by_hour_of_day_busiest) * 128)
 832                                 f.write('<td style="background-color: rgb(%d, 0, 0)">%.2f</td>' % (r, (100.0 * hour_of_day[i]) / totalcommits))
 833                         else:
 834                                 f.write('<td>0.00</td>')
 835                 f.write('</tr></table>')
 836                 f.write('<img src="hour_of_day.png" alt="Hour of Day" />')
 837                 fg = open(path + '/hour_of_day.dat', 'w')
 838                 for i in range(0, 24):
 839                         if i in hour_of_day:
 840                                 fg.write('%d %d\n' % (i + 1, hour_of_day[i]))
 841                         else:
 842                                 fg.write('%d 0\n' % (i + 1))
 843                 fg.close()
 844
 845                 # Day of Week
 846                 f.write(html_header(2, 'Day of Week'))
 847                 day_of_week = data.getActivityByDayOfWeek()
 848                 f.write('<div class="vtable"><table>')
 849                 f.write('<tr><th>Day</th><th>Total (%)</th></tr>')
 850                 fp = open(path + '/day_of_week.dat', 'w')
 851                 for d in range(0, 7):
 852                         commits = 0
 853                         if d in day_of_week:
 854                                 commits = day_of_week[d]
 855                         fp.write('%d %s %d\n' % (d + 1, WEEKDAYS[d], commits))
 856                         f.write('<tr>')
 857                         f.write('<th>%s</th>' % (WEEKDAYS[d]))
 858                         if d in day_of_week:
 859                                 f.write('<td>%d (%.2f%%)</td>' % (day_of_week[d], (100.0 * day_of_week[d]) / totalcommits))
 860                         else:
 861                                 f.write('<td>0</td>')
 862                         f.write('</tr>')
 863                 f.write('</table></div>')
 864                 f.write('<img src="day_of_week.png" alt="Day of Week" />')
 865                 fp.close()
 866
 867                 # Hour of Week
 868                 f.write(html_header(2, 'Hour of Week'))
 869                 f.write('<table>')
 870
 871                 f.write('<tr><th>Weekday</th>')
 872                 for hour in range(0, 24):
 873                         f.write('<th>%d</th>' % (hour))
 874                 f.write('</tr>')
 875
 876                 for weekday in range(0, 7):
 877                         f.write('<tr><th>%s</th>' % (WEEKDAYS[weekday]))
 878                         for hour in range(0, 24):
 879                                 try:
 880                                         commits = data.activity_by_hour_of_week[weekday][hour]
 881                                 except KeyError:
 882                                         commits = 0
 883                                 if commits != 0:
 884                                         f.write('<td')
 885                                         r = 127 + int((float(commits) / data.activity_by_hour_of_week_busiest) * 128)
 886                                         f.write(' style="background-color: rgb(%d, 0, 0)"' % r)
 887                                         f.write('>%d</td>' % commits)
 888                                 else:
 889                                         f.write('<td></td>')
 890                         f.write('</tr>')
 891
 892                 f.write('</table>')
 893
 894                 # Month of Year
 895                 f.write(html_header(2, 'Month of Year'))
 896                 f.write('<div class="vtable"><table>')
 897                 f.write('<tr><th>Month</th><th>Commits (%)</th></tr>')
 898                 fp = open (path + '/month_of_year.dat', 'w')
 899                 for mm in range(1, 13):
 900                         commits = 0
 901                         if mm in data.activity_by_month_of_year:
 902                                 commits = data.activity_by_month_of_year[mm]
 903                         f.write('<tr><td>%d</td><td>%d (%.2f %%)</td></tr>' % (mm, commits, (100.0 * commits) / data.getTotalCommits()))
 904                         fp.write('%d %d\n' % (mm, commits))
 905                 fp.close()
 906                 f.write('</table></div>')
 907                 f.write('<img src="month_of_year.png" alt="Month of Year" />')
 908
 909                 # Commits by year/month
 910                 f.write(html_header(2, 'Commits by year/month'))
 911                 f.write('<div class="vtable"><table><tr><th>Month</th><th>Commits</th><th>Lines added</th><th>Lines removed</th></tr>')
 912                 for yymm in reversed(sorted(data.commits_by_month.keys())):
 913                         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)))
 914                 f.write('</table></div>')
 915                 f.write('<img src="commits_by_year_month.png" alt="Commits by year/month" />')
 916                 fg = open(path + '/commits_by_year_month.dat', 'w')
 917                 for yymm in sorted(data.commits_by_month.keys()):
 918                         fg.write('%s %s\n' % (yymm, data.commits_by_month[yymm]))
 919                 fg.close()
 920
 921                 # Commits by year
 922                 f.write(html_header(2, 'Commits by Year'))
 923                 f.write('<div class="vtable"><table><tr><th>Year</th><th>Commits (% of all)</th><th>Lines added</th><th>Lines removed</th></tr>')
 924                 for yy in reversed(sorted(data.commits_by_year.keys())):
 925                         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)))
 926                 f.write('</table></div>')
 927                 f.write('<img src="commits_by_year.png" alt="Commits by Year" />')
 928                 fg = open(path + '/commits_by_year.dat', 'w')
 929                 for yy in sorted(data.commits_by_year.keys()):
 930                         fg.write('%d %d\n' % (yy, data.commits_by_year[yy]))
 931                 fg.close()
 932
 933                 # Commits by timezone
 934                 f.write(html_header(2, 'Commits by Timezone'))
 935                 f.write('<table><tr>')
 936                 f.write('<th>Timezone</th><th>Commits</th>')
 937                 max_commits_on_tz = max(data.commits_by_timezone.values())
 938                 for i in sorted(data.commits_by_timezone.keys(), key = lambda n : int(n)):
 939                         commits = data.commits_by_timezone[i]
 940                         r = 127 + int((float(commits) / max_commits_on_tz) * 128)
 941                         f.write('<tr><th>%s</th><td style="background-color: rgb(%d, 0, 0)">%d</td></tr>' % (i, r, commits))
 942                 f.write('</tr></table>')
 943
 944                 f.write('</body></html>')
 945                 f.close()
 946
 947                 ###
 948                 # Authors
 949                 f = open(path + '/authors.html', 'w')
 950                 self.printHeader(f)
 951
 952                 f.write('<h1>Authors</h1>')
 953                 self.printNav(f)
 954
 955                 # Authors :: List of authors
 956                 f.write(html_header(2, 'List of Authors'))
 957
 958                 f.write('<table class="authors sortable" id="authors">')
 959                 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>')
 960                 for author in data.getAuthors(conf['max_authors']):
 961                         info = data.getAuthorInfo(author)
 962                         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']))
 963                 f.write('</table>')
 964
 965                 allauthors = data.getAuthors()
 966                 if len(allauthors) > conf['max_authors']:
 967                         rest = allauthors[conf['max_authors']:]
 968                         f.write('<p class="moreauthors">These didn\'t make it to the top: %s</p>' % ', '.join(rest))
 969
 970                 f.write(html_header(2, 'Cumulated Added Lines of Code per Author'))
 971                 f.write('<img src="lines_of_code_by_author.png" alt="Lines of code per Author" />')
 972                 if len(allauthors) > conf['max_authors']:
 973                         f.write('<p class="moreauthors">Only top %d authors shown</p>' % conf['max_authors'])
 974
 975                 f.write(html_header(2, 'Commits per Author'))
 976                 f.write('<img src="commits_by_author.png" alt="Commits per Author" />')
 977                 if len(allauthors) > conf['max_authors']:
 978                         f.write('<p class="moreauthors">Only top %d authors shown</p>' % conf['max_authors'])
 979
 980                 fgl = open(path + '/lines_of_code_by_author.dat', 'w')
 981                 fgc = open(path + '/commits_by_author.dat', 'w')
 982
 983                 lines_by_authors = {} # cumulated added lines by
 984                 # author. to save memory,
 985                 # changes_by_date_by_author[stamp][author] is defined
 986                 # only at points where author commits.
 987                 # lines_by_authors allows us to generate all the
 988                 # points in the .dat file.
 989
 990                 # Don't rely on getAuthors to give the same order each
 991                 # time. Be robust and keep the list in a variable.
 992                 commits_by_authors = {} # cumulated added lines by
 993
 994                 self.authors_to_plot = data.getAuthors(conf['max_authors'])
 995                 for author in self.authors_to_plot:
 996                         lines_by_authors[author] = 0
 997                         commits_by_authors[author] = 0
 998                 for stamp in sorted(data.changes_by_date_by_author.keys()):
 999                         fgl.write('%d' % stamp)
1000                         fgc.write('%d' % stamp)
1001                         for author in self.authors_to_plot:
1002                                 if author in data.changes_by_date_by_author[stamp].keys():
1003                                         lines_by_authors[author] = data.changes_by_date_by_author[stamp][author]['lines_added']
1004                                         commits_by_authors[author] = data.changes_by_date_by_author[stamp][author]['commits']
1005                                 fgl.write(' %d' % lines_by_authors[author])
1006                                 fgc.write(' %d' % commits_by_authors[author])
1007                         fgl.write('\n')
1008                         fgc.write('\n')
1009                 fgl.close()
1010                 fgc.close()
1011
1012                 # Authors :: Author of Month
1013                 f.write(html_header(2, 'Author of Month'))
1014                 f.write('<table class="sortable" id="aom">')
1015                 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'])
1016                 for yymm in reversed(sorted(data.author_of_month.keys())):
1017                         authordict = data.author_of_month[yymm]
1018                         authors = getkeyssortedbyvalues(authordict)
1019                         authors.reverse()
1020                         commits = data.author_of_month[yymm][authors[0]]
1021                         next = ', '.join(authors[1:conf['authors_top']+1])
1022                         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)))
1023
1024                 f.write('</table>')
1025
1026                 f.write(html_header(2, 'Author of Year'))
1027                 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'])
1028                 for yy in reversed(sorted(data.author_of_year.keys())):
1029                         authordict = data.author_of_year[yy]
1030                         authors = getkeyssortedbyvalues(authordict)
1031                         authors.reverse()
1032                         commits = data.author_of_year[yy][authors[0]]
1033                         next = ', '.join(authors[1:conf['authors_top']+1])
1034                         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)))
1035                 f.write('</table>')
1036
1037                 # Domains
1038                 f.write(html_header(2, 'Commits by Domains'))
1039                 domains_by_commits = getkeyssortedbyvaluekey(data.domains, 'commits')
1040                 domains_by_commits.reverse() # most first
1041                 f.write('<div class="vtable"><table>')
1042                 f.write('<tr><th>Domains</th><th>Total (%)</th></tr>')
1043                 fp = open(path + '/domains.dat', 'w')
1044                 n = 0
1045                 for domain in domains_by_commits:
1046                         if n == conf['max_domains']:
1047                                 break
1048                         commits = 0
1049                         n += 1
1050                         info = data.getDomainInfo(domain)
1051                         fp.write('%s %d %d\n' % (domain, n , info['commits']))
1052                         f.write('<tr><th>%s</th><td>%d (%.2f%%)</td></tr>' % (domain, info['commits'], (100.0 * info['commits'] / totalcommits)))
1053                 f.write('</table></div>')
1054                 f.write('<img src="domains.png" alt="Commits by Domains" />')
1055                 fp.close()
1056
1057                 f.write('</body></html>')
1058                 f.close()
1059
1060                 ###
1061                 # Files
1062                 f = open(path + '/files.html', 'w')
1063                 self.printHeader(f)
1064                 f.write('<h1>Files</h1>')
1065                 self.printNav(f)
1066
1067                 f.write('<dl>\n')
1068                 f.write('<dt>Total files</dt><dd>%d</dd>' % data.getTotalFiles())
1069                 f.write('<dt>Total lines</dt><dd>%d</dd>' % data.getTotalLOC())
1070                 try:
1071                         f.write('<dt>Average file size</dt><dd>%.2f bytes</dd>' % (float(data.getTotalSize()) / data.getTotalFiles()))
1072                 except ZeroDivisionError:
1073                         pass
1074                 f.write('</dl>\n')
1075
1076                 # Files :: File count by date
1077                 f.write(html_header(2, 'File count by date'))
1078
1079                 # use set to get rid of duplicate/unnecessary entries
1080                 files_by_date = set()
1081                 for stamp in sorted(data.files_by_stamp.keys()):
1082                         files_by_date.add('%s %d' % (datetime.datetime.fromtimestamp(stamp).strftime('%Y-%m-%d'), data.files_by_stamp[stamp]))
1083
1084                 fg = open(path + '/files_by_date.dat', 'w')
1085                 for line in sorted(list(files_by_date)):
1086                         fg.write('%s\n' % line)
1087                 #for stamp in sorted(data.files_by_stamp.keys()):
1088                 #       fg.write('%s %d\n' % (datetime.datetime.fromtimestamp(stamp).strftime('%Y-%m-%d'), data.files_by_stamp[stamp]))
1089                 fg.close()
1090
1091                 f.write('<img src="files_by_date.png" alt="Files by Date" />')
1092
1093                 #f.write('<h2>Average file size by date</h2>')
1094
1095                 # Files :: Extensions
1096                 f.write(html_header(2, 'Extensions'))
1097                 f.write('<table class="sortable" id="ext"><tr><th>Extension</th><th>Files (%)</th><th>Lines (%)</th><th>Lines/file</th></tr>')
1098                 for ext in sorted(data.extensions.keys()):
1099                         files = data.extensions[ext]['files']
1100                         lines = data.extensions[ext]['lines']
1101                         try:
1102                                 loc_percentage = (100.0 * lines) / data.getTotalLOC()
1103                         except ZeroDivisionError:
1104                                 loc_percentage = 0
1105                         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))
1106                 f.write('</table>')
1107
1108                 f.write('</body></html>')
1109                 f.close()
1110
1111                 ###
1112                 # Lines
1113                 f = open(path + '/lines.html', 'w')
1114                 self.printHeader(f)
1115                 f.write('<h1>Lines</h1>')
1116                 self.printNav(f)
1117
1118                 f.write('<dl>\n')
1119                 f.write('<dt>Total lines</dt><dd>%d</dd>' % data.getTotalLOC())
1120                 f.write('</dl>\n')
1121
1122                 f.write(html_header(2, 'Lines of Code'))
1123                 f.write('<img src="lines_of_code.png" />')
1124
1125                 fg = open(path + '/lines_of_code.dat', 'w')
1126                 for stamp in sorted(data.changes_by_date.keys()):
1127                         fg.write('%d %d\n' % (stamp, data.changes_by_date[stamp]['lines']))
1128                 fg.close()
1129
1130                 f.write('</body></html>')
1131                 f.close()
1132
1133                 ###
1134                 # tags.html
1135                 f = open(path + '/tags.html', 'w')
1136                 self.printHeader(f)
1137                 f.write('<h1>Tags</h1>')
1138                 self.printNav(f)
1139
1140                 f.write('<dl>')
1141                 f.write('<dt>Total tags</dt><dd>%d</dd>' % len(data.tags))
1142                 if len(data.tags) > 0:
1143                         f.write('<dt>Average commits per tag</dt><dd>%.2f</dd>' % (1.0 * data.getTotalCommits() / len(data.tags)))
1144                 f.write('</dl>')
1145
1146                 f.write('<table class="tags">')
1147                 f.write('<tr><th>Name</th><th>Date</th><th>Commits</th><th>Authors</th></tr>')
1148                 # sort the tags by date desc
1149                 tags_sorted_by_date_desc = map(lambda el : el[1], reversed(sorted(map(lambda el : (el[1]['date'], el[0]), data.tags.items()))))
1150                 for tag in tags_sorted_by_date_desc:
1151                         authorinfo = []
1152                         self.authors_by_commits = getkeyssortedbyvalues(data.tags[tag]['authors'])
1153                         for i in reversed(self.authors_by_commits):
1154                                 authorinfo.append('%s (%d)' % (i, data.tags[tag]['authors'][i]))
1155                         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)))
1156                 f.write('</table>')
1157
1158                 f.write('</body></html>')
1159                 f.close()
1160
1161                 self.createGraphs(path)
1162
1163         def createGraphs(self, path):
1164                 print 'Generating graphs...'
1165
1166                 # hour of day
1167                 f = open(path + '/hour_of_day.plot', 'w')
1168                 f.write(GNUPLOT_COMMON)
1169                 f.write(
1170 """
1171 set output 'hour_of_day.png'
1172 unset key
1173 set xrange [0.5:24.5]
1174 set yrange [0:]
1175 set xtics 4
1176 set grid y
1177 set ylabel "Commits"
1178 plot 'hour_of_day.dat' using 1:2:(0.5) w boxes fs solid
1179 """)
1180                 f.close()
1181
1182                 # day of week
1183                 f = open(path + '/day_of_week.plot', 'w')
1184                 f.write(GNUPLOT_COMMON)
1185                 f.write(
1186 """
1187 set output 'day_of_week.png'
1188 unset key
1189 set xrange [0.5:7.5]
1190 set yrange [0:]
1191 set xtics 1
1192 set grid y
1193 set ylabel "Commits"
1194 plot 'day_of_week.dat' using 1:3:(0.5):xtic(2) w boxes fs solid
1195 """)
1196                 f.close()
1197
1198                 # Domains
1199                 f = open(path + '/domains.plot', 'w')
1200                 f.write(GNUPLOT_COMMON)
1201                 f.write(
1202 """
1203 set output 'domains.png'
1204 unset key
1205 unset xtics
1206 set yrange [0:]
1207 set grid y
1208 set ylabel "Commits"
1209 plot 'domains.dat' using 2:3:(0.5) with boxes fs solid, '' using 2:3:1 with labels rotate by 45 offset 0,1
1210 """)
1211                 f.close()
1212
1213                 # Month of Year
1214                 f = open(path + '/month_of_year.plot', 'w')
1215                 f.write(GNUPLOT_COMMON)
1216                 f.write(
1217 """
1218 set output 'month_of_year.png'
1219 unset key
1220 set xrange [0.5:12.5]
1221 set yrange [0:]
1222 set xtics 1
1223 set grid y
1224 set ylabel "Commits"
1225 plot 'month_of_year.dat' using 1:2:(0.5) w boxes fs solid
1226 """)
1227                 f.close()
1228
1229                 # commits_by_year_month
1230                 f = open(path + '/commits_by_year_month.plot', 'w')
1231                 f.write(GNUPLOT_COMMON)
1232                 f.write(
1233 """
1234 set output 'commits_by_year_month.png'
1235 unset key
1236 set yrange [0:]
1237 set xdata time
1238 set timefmt "%Y-%m"
1239 set format x "%Y-%m"
1240 set xtics rotate
1241 set bmargin 5
1242 set grid y
1243 set ylabel "Commits"
1244 plot 'commits_by_year_month.dat' using 1:2:(0.5) w boxes fs solid
1245 """)
1246                 f.close()
1247
1248                 # commits_by_year
1249                 f = open(path + '/commits_by_year.plot', 'w')
1250                 f.write(GNUPLOT_COMMON)
1251                 f.write(
1252 """
1253 set output 'commits_by_year.png'
1254 unset key
1255 set yrange [0:]
1256 set xtics 1 rotate
1257 set grid y
1258 set ylabel "Commits"
1259 set yrange [0:]
1260 plot 'commits_by_year.dat' using 1:2:(0.5) w boxes fs solid
1261 """)
1262                 f.close()
1263
1264                 # Files by date
1265                 f = open(path + '/files_by_date.plot', 'w')
1266                 f.write(GNUPLOT_COMMON)
1267                 f.write(
1268 """
1269 set output 'files_by_date.png'
1270 unset key
1271 set yrange [0:]
1272 set xdata time
1273 set timefmt "%Y-%m-%d"
1274 set format x "%Y-%m-%d"
1275 set grid y
1276 set ylabel "Files"
1277 set xtics rotate
1278 set ytics autofreq
1279 set bmargin 6
1280 plot 'files_by_date.dat' using 1:2 w steps
1281 """)
1282                 f.close()
1283
1284                 # Lines of Code
1285                 f = open(path + '/lines_of_code.plot', 'w')
1286                 f.write(GNUPLOT_COMMON)
1287                 f.write(
1288 """
1289 set output 'lines_of_code.png'
1290 unset key
1291 set yrange [0:]
1292 set xdata time
1293 set timefmt "%s"
1294 set format x "%Y-%m-%d"
1295 set grid y
1296 set ylabel "Lines"
1297 set xtics rotate
1298 set bmargin 6
1299 plot 'lines_of_code.dat' using 1:2 w lines
1300 """)
1301                 f.close()
1302
1303                 # Lines of Code Added per author
1304                 f = open(path + '/lines_of_code_by_author.plot', 'w')
1305                 f.write(GNUPLOT_COMMON)
1306                 f.write(
1307 """
1308 set terminal png transparent size 640,480
1309 set output 'lines_of_code_by_author.png'
1310 set key left top
1311 set yrange [0:]
1312 set xdata time
1313 set timefmt "%s"
1314 set format x "%Y-%m-%d"
1315 set grid y
1316 set ylabel "Lines"
1317 set xtics rotate
1318 set bmargin 6
1319 plot """
1320 )
1321                 i = 1
1322                 plots = []
1323                 for a in self.authors_to_plot:
1324                         i = i + 1
1325                         author = a.replace("\"", "\\\"").replace("`", "")
1326                         plots.append("""'lines_of_code_by_author.dat' using 1:%d title "%s" w lines""" % (i, author))
1327                 f.write(", ".join(plots))
1328                 f.write('\n')
1329
1330                 f.close()
1331
1332                 # Commits per author
1333                 f = open(path + '/commits_by_author.plot', 'w')
1334                 f.write(GNUPLOT_COMMON)
1335                 f.write(
1336 """
1337 set terminal png transparent size 640,480
1338 set output 'commits_by_author.png'
1339 set key left top
1340 set yrange [0:]
1341 set xdata time
1342 set timefmt "%s"
1343 set format x "%Y-%m-%d"
1344 set grid y
1345 set ylabel "Commits"
1346 set xtics rotate
1347 set bmargin 6
1348 plot """
1349 )
1350                 i = 1
1351                 plots = []
1352                 for a in self.authors_to_plot:
1353                         i = i + 1
1354                         author = a.replace("\"", "\\\"").replace("`", "")
1355                         plots.append("""'commits_by_author.dat' using 1:%d title "%s" w lines""" % (i, author))
1356                 f.write(", ".join(plots))
1357                 f.write('\n')
1358
1359                 f.close()
1360
1361                 os.chdir(path)
1362                 files = glob.glob(path + '/*.plot')
1363                 for f in files:
1364                         out = getpipeoutput([gnuplot_cmd + ' "%s"' % f])
1365                         if len(out) > 0:
1366                                 print out
1367
1368         def printHeader(self, f, title = ''):
1369                 f.write(
1370 """<?xml version="1.0" encoding="UTF-8"?>
1371 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
1372 <html xmlns="http://www.w3.org/1999/xhtml">
1373 <head>
1374         <title>GitStats - %s</title>
1375         <link rel="stylesheet" href="%s" type="text/css" />
1376         <meta name="generator" content="GitStats %s" />
1377         <script type="text/javascript" src="sortable.js"></script>
1378 </head>
1379 <body>
1380 """ % (self.title, conf['style'], getversion()))
1381
1382         def printNav(self, f):
1383                 f.write("""
1384 <div class="nav">
1385 <ul>
1386 <li><a href="index.html">General</a></li>
1387 <li><a href="activity.html">Activity</a></li>
1388 <li><a href="authors.html">Authors</a></li>
1389 <li><a href="files.html">Files</a></li>
1390 <li><a href="lines.html">Lines</a></li>
1391 <li><a href="tags.html">Tags</a></li>
1392 </ul>
1393 </div>
1394 """)
1395
1396 def usage():
1397         print """
1398 Usage: gitstats [options] <gitpath..> <outputpath>
1399
1400 Options:
1401 -c key=value     Override configuration value
1402
1403 Default config values:
1404 %s
1405
1406 Please see the manual page for more details.
1407 """ % conf
1408
1409
1410 class GitStats:
1411         def run(self, args_orig):
1412                 optlist, args = getopt.getopt(args_orig, 'hc:', ["help"])
1413                 for o,v in optlist:
1414                         if o == '-c':
1415                                 key, value = v.split('=', 1)
1416                                 if key not in conf:
1417                                         raise KeyError('no such key "%s" in config' % key)
1418                                 if isinstance(conf[key], int):
1419                                         conf[key] = int(value)
1420                                 elif isinstance(conf[key], dict):
1421                                         kk,vv = value.split(',', 1)
1422                                         conf[key][kk] = vv
1423                                 else:
1424                                         conf[key] = value
1425                         elif o in ('-h', '--help'):
1426                                 usage()
1427                                 sys.exit()
1428
1429                 if len(args) < 2:
1430                         usage()
1431                         sys.exit(0)
1432
1433                 outputpath = os.path.abspath(args[-1])
1434                 rundir = os.getcwd()
1435
1436                 try:
1437                         os.makedirs(outputpath)
1438                 except OSError:
1439                         pass
1440                 if not os.path.isdir(outputpath):
1441                         print 'FATAL: Output path is not a directory or does not exist'
1442                         sys.exit(1)
1443
1444                 if not getgnuplotversion():
1445                         print 'gnuplot not found'
1446                         sys.exit(1)
1447
1448                 print 'Output path: %s' % outputpath
1449                 cachefile = os.path.join(outputpath, 'gitstats.cache')
1450
1451                 data = GitDataCollector()
1452                 data.loadCache(cachefile)
1453
1454                 for gitpath in args[0:-1]:
1455                         print 'Git path: %s' % gitpath
1456
1457                         os.chdir(gitpath)
1458
1459                         print 'Collecting data...'
1460                         data.collect(gitpath)
1461
1462                 print 'Refining data...'
1463                 data.saveCache(cachefile)
1464                 data.refine()
1465
1466                 os.chdir(rundir)
1467
1468                 print 'Generating report...'
1469                 report = HTMLReportCreator()
1470                 report.create(data, outputpath)
1471
1472                 time_end = time.time()
1473                 exectime_internal = time_end - time_start
1474                 print 'Execution time %.5f secs, %.5f secs (%.2f %%) in external commands)' % (exectime_internal, exectime_external, (100.0 * exectime_external) / exectime_internal)
1475                 if sys.stdin.isatty():
1476                         print 'You may now run:'
1477                         print
1478                         print '   sensible-browser \'%s\'' % os.path.join(outputpath, 'index.html').replace("'", "'\\''")
1479                         print
1480
1481 if __name__=='__main__':
1482         g = GitStats()
1483         g.run(sys.argv[1:])
1484