gitstats

   1 #!/usr/bin/env python
   2 # Copyright (c) 2007-2009 Heikki Hokkanen <hoxu@users.sf.net> & others (see doc/author.txt)
   3 # GPLv2 / GPLv3
   4 import datetime
   5 import glob
   6 import os
   7 import pickle
   8 import platform
   9 import re
  10 import shutil
  11 import subprocess
  12 import sys
  13 import time
  14 import zlib
  15
  16 GNUPLOT_COMMON = 'set terminal png transparent\nset size 0.5,0.5\n'
  17 MAX_EXT_LENGTH = 10 # maximum file extension length
  18 ON_LINUX = (platform.system() == 'Linux')
  19
  20 exectime_internal = 0.0
  21 exectime_external = 0.0
  22 time_start = time.time()
  23
  24 # By default, gnuplot is searched from path, but can be overridden with the
  25 # environment variable "GNUPLOT"
  26 gnuplot_cmd = 'gnuplot'
  27 if 'GNUPLOT' in os.environ:
  28         gnuplot_cmd = os.environ['GNUPLOT']
  29
  30 def getpipeoutput(cmds, quiet = False):
  31         global exectime_external
  32         start = time.time()
  33         if not quiet and ON_LINUX and os.isatty(1):
  34                 print '>> ' + ' | '.join(cmds),
  35                 sys.stdout.flush()
  36         p0 = subprocess.Popen(cmds[0], stdout = subprocess.PIPE, shell = True)
  37         p = p0
  38         for x in cmds[1:]:
  39                 p = subprocess.Popen(x, stdin = p0.stdout, stdout = subprocess.PIPE, shell = True)
  40                 p0 = p
  41         output = p.communicate()[0]
  42         end = time.time()
  43         if not quiet:
  44                 if ON_LINUX and os.isatty(1):
  45                         print '\r',
  46                 print '[%.5f] >> %s' % (end - start, ' | '.join(cmds))
  47         exectime_external += (end - start)
  48         return output.rstrip('\n')
  49
  50 def getkeyssortedbyvalues(dict):
  51         return map(lambda el : el[1], sorted(map(lambda el : (el[1], el[0]), dict.items())))
  52
  53 # dict['author'] = { 'commits': 512 } - ...key(dict, 'commits')
  54 def getkeyssortedbyvaluekey(d, key):
  55         return map(lambda el : el[1], sorted(map(lambda el : (d[el][key], el), d.keys())))
  56
  57 VERSION = 0
  58 def getversion():
  59         global VERSION
  60         if VERSION == 0:
  61                 VERSION = getpipeoutput(["git rev-parse --short HEAD"]).split('\n')[0]
  62         return VERSION
  63
  64 class DataCollector:
  65         """Manages data collection from a revision control repository."""
  66         def __init__(self):
  67                 self.stamp_created = time.time()
  68                 self.cache = {}
  69
  70         ##
  71         # This should be the main function to extract data from the repository.
  72         def collect(self, dir):
  73                 self.dir = dir
  74                 self.projectname = os.path.basename(os.path.abspath(dir))
  75
  76         ##
  77         # Load cacheable data
  78         def loadCache(self, cachefile):
  79                 if not os.path.exists(cachefile):
  80                         return
  81                 print 'Loading cache...'
  82                 f = open(cachefile)
  83                 try:
  84                         self.cache = pickle.loads(zlib.decompress(f.read()))
  85                 except:
  86                         # temporary hack to upgrade non-compressed caches
  87                         f.seek(0)
  88                         self.cache = pickle.load(f)
  89                 f.close()
  90
  91         ##
  92         # Produce any additional statistics from the extracted data.
  93         def refine(self):
  94                 pass
  95
  96         ##
  97         # : get a dictionary of author
  98         def getAuthorInfo(self, author):
  99                 return None
 100
 101         def getActivityByDayOfWeek(self):
 102                 return {}
 103
 104         def getActivityByHourOfDay(self):
 105                 return {}
 106
 107         ##
 108         # Get a list of authors
 109         def getAuthors(self):
 110                 return []
 111
 112         def getFirstCommitDate(self):
 113                 return datetime.datetime.now()
 114
 115         def getLastCommitDate(self):
 116                 return datetime.datetime.now()
 117
 118         def getStampCreated(self):
 119                 return self.stamp_created
 120
 121         def getTags(self):
 122                 return []
 123
 124         def getTotalAuthors(self):
 125                 return -1
 126
 127         def getTotalCommits(self):
 128                 return -1
 129
 130         def getTotalFiles(self):
 131                 return -1
 132
 133         def getTotalLOC(self):
 134                 return -1
 135
 136         ##
 137         # Save cacheable data
 138         def saveCache(self, filename):
 139                 print 'Saving cache...'
 140                 f = open(cachefile, 'w')
 141                 #pickle.dump(self.cache, f)
 142                 data = zlib.compress(pickle.dumps(self.cache))
 143                 f.write(data)
 144                 f.close()
 145
 146 class GitDataCollector(DataCollector):
 147         def collect(self, dir):
 148                 DataCollector.collect(self, dir)
 149
 150                 try:
 151                         self.total_authors = int(getpipeoutput(['git log', 'git shortlog -s', 'wc -l']))
 152                 except:
 153                         self.total_authors = 0
 154                 #self.total_lines = int(getoutput('git-ls-files -z |xargs -0 cat |wc -l'))
 155
 156                 self.activity_by_hour_of_day = {} # hour -> commits
 157                 self.activity_by_day_of_week = {} # day -> commits
 158                 self.activity_by_month_of_year = {} # month [1-12] -> commits
 159                 self.activity_by_hour_of_week = {} # weekday -> hour -> commits
 160                 self.activity_by_hour_of_day_busiest = 0
 161                 self.activity_by_hour_of_week_busiest = 0
 162                 self.activity_by_year_week = {} # yy_wNN -> commits
 163                 self.activity_by_year_week_peak = 0
 164
 165                 self.authors = {} # name -> {commits, first_commit_stamp, last_commit_stamp, last_active_day, active_days, lines_added, lines_removed}
 166
 167                 # author of the month
 168                 self.author_of_month = {} # month -> author -> commits
 169                 self.author_of_year = {} # year -> author -> commits
 170                 self.commits_by_month = {} # month -> commits
 171                 self.commits_by_year = {} # year -> commits
 172                 self.first_commit_stamp = 0
 173                 self.last_commit_stamp = 0
 174                 self.last_active_day = None
 175                 self.active_days = set()
 176
 177                 # lines
 178                 self.total_lines = 0
 179                 self.total_lines_added = 0
 180                 self.total_lines_removed = 0
 181
 182                 # timezone
 183                 self.commits_by_timezone = {} # timezone -> commits
 184
 185                 # tags
 186                 self.tags = {}
 187                 lines = getpipeoutput(['git show-ref --tags']).split('\n')
 188                 for line in lines:
 189                         if len(line) == 0:
 190                                 continue
 191                         (hash, tag) = line.split(' ')
 192
 193                         tag = tag.replace('refs/tags/', '')
 194                         output = getpipeoutput(['git log "%s" --pretty=format:"%%at %%an" -n 1' % hash])
 195                         if len(output) > 0:
 196                                 parts = output.split(' ')
 197                                 stamp = 0
 198                                 try:
 199                                         stamp = int(parts[0])
 200                                 except ValueError:
 201                                         stamp = 0
 202                                 self.tags[tag] = { 'stamp': stamp, 'hash' : hash, 'date' : datetime.datetime.fromtimestamp(stamp).strftime('%Y-%m-%d'), 'commits': 0, 'authors': {} }
 203
 204                 # collect info on tags, starting from latest
 205                 tags_sorted_by_date_desc = map(lambda el : el[1], reversed(sorted(map(lambda el : (el[1]['date'], el[0]), data.tags.items()))))
 206                 prev = None
 207                 for tag in reversed(tags_sorted_by_date_desc):
 208                         cmd = 'git shortlog -s "%s"' % tag
 209                         if prev != None:
 210                                 cmd += ' "^%s"' % prev
 211                         output = getpipeoutput([cmd])
 212                         if len(output) == 0:
 213                                 continue
 214                         prev = tag
 215                         for line in output.split('\n'):
 216                                 parts = re.split('\s+', line, 2)
 217                                 commits = int(parts[1])
 218                                 author = parts[2]
 219                                 self.tags[tag]['commits'] += commits
 220                                 self.tags[tag]['authors'][author] = commits
 221
 222                 # Collect revision statistics
 223                 # Outputs "<stamp> <author>"
 224                 lines = getpipeoutput(['git rev-list --pretty=format:"%at %ai %an" HEAD', 'grep -v ^commit']).split('\n')
 225                 for line in lines:
 226                         parts = line.split(' ')
 227                         author = ''
 228                         try:
 229                                 stamp = int(parts[0])
 230                         except ValueError:
 231                                 stamp = 0
 232                         timezone = parts[3]
 233                         if len(parts) > 4:
 234                                 author = ' '.join(parts[4:])
 235                         date = datetime.datetime.fromtimestamp(float(stamp))
 236
 237                         # First and last commit stamp
 238                         if self.last_commit_stamp == 0:
 239                                 self.last_commit_stamp = stamp
 240                         self.first_commit_stamp = stamp
 241
 242                         # activity
 243                         # hour
 244                         hour = date.hour
 245                         self.activity_by_hour_of_day[hour] = self.activity_by_hour_of_day.get(hour, 0) + 1
 246                         # most active hour?
 247                         if self.activity_by_hour_of_day[hour] > self.activity_by_hour_of_day_busiest:
 248                                 self.activity_by_hour_of_day_busiest = self.activity_by_hour_of_day[hour]
 249
 250                         # day of week
 251                         day = date.weekday()
 252                         self.activity_by_day_of_week[day] = self.activity_by_day_of_week.get(day, 0) + 1
 253
 254                         # hour of week
 255                         if day not in self.activity_by_hour_of_week:
 256                                 self.activity_by_hour_of_week[day] = {}
 257                         self.activity_by_hour_of_week[day][hour] = self.activity_by_hour_of_week[day].get(hour, 0) + 1
 258                         # most active hour?
 259                         if self.activity_by_hour_of_week[day][hour] > self.activity_by_hour_of_week_busiest:
 260                                 self.activity_by_hour_of_week_busiest = self.activity_by_hour_of_week[day][hour]
 261
 262                         # month of year
 263                         month = date.month
 264                         self.activity_by_month_of_year[month] = self.activity_by_month_of_year.get(month, 0) + 1
 265
 266                         # yearly/weekly activity
 267                         yyw = date.strftime('%Y-%W')
 268                         self.activity_by_year_week[yyw] = self.activity_by_year_week.get(yyw, 0) + 1
 269                         if self.activity_by_year_week_peak < self.activity_by_year_week[yyw]:
 270                                 self.activity_by_year_week_peak = self.activity_by_year_week[yyw]
 271
 272                         # author stats
 273                         if author not in self.authors:
 274                                 self.authors[author] = {}
 275                         # commits
 276                         if 'last_commit_stamp' not in self.authors[author]:
 277                                 self.authors[author]['last_commit_stamp'] = stamp
 278                         self.authors[author]['first_commit_stamp'] = stamp
 279                         self.authors[author]['commits'] = self.authors[author].get('commits', 0) + 1
 280
 281                         # author of the month/year
 282                         yymm = date.strftime('%Y-%m')
 283                         if yymm in self.author_of_month:
 284                                 self.author_of_month[yymm][author] = self.author_of_month[yymm].get(author, 0) + 1
 285                         else:
 286                                 self.author_of_month[yymm] = {}
 287                                 self.author_of_month[yymm][author] = 1
 288                         self.commits_by_month[yymm] = self.commits_by_month.get(yymm, 0) + 1
 289
 290                         yy = date.year
 291                         if yy in self.author_of_year:
 292                                 self.author_of_year[yy][author] = self.author_of_year[yy].get(author, 0) + 1
 293                         else:
 294                                 self.author_of_year[yy] = {}
 295                                 self.author_of_year[yy][author] = 1
 296                         self.commits_by_year[yy] = self.commits_by_year.get(yy, 0) + 1
 297
 298                         # authors: active days
 299                         yymmdd = date.strftime('%Y-%m-%d')
 300                         if 'last_active_day' not in self.authors[author]:
 301                                 self.authors[author]['last_active_day'] = yymmdd
 302                                 self.authors[author]['active_days'] = 1
 303                         elif yymmdd != self.authors[author]['last_active_day']:
 304                                 self.authors[author]['last_active_day'] = yymmdd
 305                                 self.authors[author]['active_days'] += 1
 306
 307                         # project: active days
 308                         if yymmdd != self.last_active_day:
 309                                 self.last_active_day = yymmdd
 310                                 self.active_days.add(yymmdd)
 311
 312                         # timezone
 313                         self.commits_by_timezone[timezone] = self.commits_by_timezone.get(timezone, 0) + 1
 314
 315                 # TODO Optimize this, it's the worst bottleneck
 316                 # outputs "<stamp> <files>" for each revision
 317                 self.files_by_stamp = {} # stamp -> files
 318                 revlines = getpipeoutput(['git rev-list --pretty=format:"%at %T" HEAD', 'grep -v ^commit']).strip().split('\n')
 319                 lines = []
 320                 for revline in revlines:
 321                         time, rev = revline.split(' ')
 322                         linecount = self.getFilesInCommit(rev)
 323                         lines.append('%d %d' % (int(time), linecount))
 324
 325                 self.total_commits = len(lines)
 326                 for line in lines:
 327                         parts = line.split(' ')
 328                         if len(parts) != 2:
 329                                 continue
 330                         (stamp, files) = parts[0:2]
 331                         try:
 332                                 self.files_by_stamp[int(stamp)] = int(files)
 333                         except ValueError:
 334                                 print 'Warning: failed to parse line "%s"' % line
 335
 336                 # extensions
 337                 self.extensions = {} # extension -> files, lines
 338                 lines = getpipeoutput(['git ls-tree -r -z HEAD']).split('\000')
 339                 self.total_files = len(lines)
 340                 for line in lines:
 341                         if len(line) == 0:
 342                                 continue
 343                         parts = re.split('\s+', line, 4)
 344                         sha1 = parts[2]
 345                         filename = parts[3]
 346
 347                         if filename.find('.') == -1 or filename.rfind('.') == 0:
 348                                 ext = ''
 349                         else:
 350                                 ext = filename[(filename.rfind('.') + 1):]
 351                         if len(ext) > MAX_EXT_LENGTH:
 352                                 ext = ''
 353
 354                         if ext not in self.extensions:
 355                                 self.extensions[ext] = {'files': 0, 'lines': 0}
 356
 357                         self.extensions[ext]['files'] += 1
 358                         try:
 359                                 self.extensions[ext]['lines'] += int(getpipeoutput(['git cat-file blob %s' % sha1, 'wc -l']).split()[0])
 360                         except:
 361                                 print 'Warning: Could not count lines for file "%s"' % line
 362
 363                 # line statistics
 364                 # outputs:
 365                 #  N files changed, N insertions (+), N deletions(-)
 366                 # <stamp> <author>
 367                 self.changes_by_date = {} # stamp -> { files, ins, del }
 368                 lines = getpipeoutput(['git log --shortstat --pretty=format:"%at %an"']).split('\n')
 369                 lines.reverse()
 370                 files = 0; inserted = 0; deleted = 0; total_lines = 0
 371                 author = None
 372                 for line in lines:
 373                         if len(line) == 0:
 374                                 continue
 375
 376                         # <stamp> <author>
 377                         if line.find('files changed,') == -1:
 378                                 pos = line.find(' ')
 379                                 if pos != -1:
 380                                         try:
 381                                                 (stamp, author) = (int(line[:pos]), line[pos+1:])
 382                                                 self.changes_by_date[stamp] = { 'files': files, 'ins': inserted, 'del': deleted, 'lines': total_lines }
 383                                                 if author not in self.authors:
 384                                                         self.authors[author] = { 'lines_added' : 0, 'lines_removed' : 0 }
 385                                                 self.authors[author]['lines_added'] = self.authors[author].get('lines_added', 0) + inserted
 386                                                 self.authors[author]['lines_removed'] = self.authors[author].get('lines_removed', 0) + deleted
 387                                         except ValueError:
 388                                                 print 'Warning: unexpected line "%s"' % line
 389                                 else:
 390                                         print 'Warning: unexpected line "%s"' % line
 391                         else:
 392                                 numbers = re.findall('\d+', line)
 393                                 if len(numbers) == 3:
 394                                         (files, inserted, deleted) = map(lambda el : int(el), numbers)
 395                                         total_lines += inserted
 396                                         total_lines -= deleted
 397                                         self.total_lines_added += inserted
 398                                         self.total_lines_removed += deleted
 399                                 else:
 400                                         print 'Warning: failed to handle line "%s"' % line
 401                                         (files, inserted, deleted) = (0, 0, 0)
 402                                 #self.changes_by_date[stamp] = { 'files': files, 'ins': inserted, 'del': deleted }
 403                 self.total_lines = total_lines
 404
 405         def refine(self):
 406                 # authors
 407                 # name -> {place_by_commits, commits_frac, date_first, date_last, timedelta}
 408                 authors_by_commits = getkeyssortedbyvaluekey(self.authors, 'commits')
 409                 authors_by_commits.reverse() # most first
 410                 for i, name in enumerate(authors_by_commits):
 411                         self.authors[name]['place_by_commits'] = i + 1
 412
 413                 for name in self.authors.keys():
 414                         a = self.authors[name]
 415                         a['commits_frac'] = (100 * float(a['commits'])) / self.getTotalCommits()
 416                         date_first = datetime.datetime.fromtimestamp(a['first_commit_stamp'])
 417                         date_last = datetime.datetime.fromtimestamp(a['last_commit_stamp'])
 418                         delta = date_last - date_first
 419                         a['date_first'] = date_first.strftime('%Y-%m-%d')
 420                         a['date_last'] = date_last.strftime('%Y-%m-%d')
 421                         a['timedelta'] = delta
 422
 423         def getActiveDays(self):
 424                 return self.active_days
 425
 426         def getActivityByDayOfWeek(self):
 427                 return self.activity_by_day_of_week
 428
 429         def getActivityByHourOfDay(self):
 430                 return self.activity_by_hour_of_day
 431
 432         def getAuthorInfo(self, author):
 433                 return self.authors[author]
 434
 435         def getAuthors(self):
 436                 return self.authors.keys()
 437
 438         def getCommitDeltaDays(self):
 439                 return (self.last_commit_stamp - self.first_commit_stamp) / 86400
 440
 441         def getFilesInCommit(self, rev):
 442                 try:
 443                         res = self.cache['files_in_tree'][rev]
 444                 except:
 445                         res = int(getpipeoutput(['git ls-tree -r --name-only "%s"' % rev, 'wc -l']).split('\n')[0])
 446                         if 'files_in_tree' not in self.cache:
 447                                 self.cache['files_in_tree'] = {}
 448                         self.cache['files_in_tree'][rev] = res
 449
 450                 return res
 451
 452         def getFirstCommitDate(self):
 453                 return datetime.datetime.fromtimestamp(self.first_commit_stamp)
 454
 455         def getLastCommitDate(self):
 456                 return datetime.datetime.fromtimestamp(self.last_commit_stamp)
 457
 458         def getTags(self):
 459                 lines = getpipeoutput(['git show-ref --tags', 'cut -d/ -f3'])
 460                 return lines.split('\n')
 461
 462         def getTagDate(self, tag):
 463                 return self.revToDate('tags/' + tag)
 464
 465         def getTotalAuthors(self):
 466                 return self.total_authors
 467
 468         def getTotalCommits(self):
 469                 return self.total_commits
 470
 471         def getTotalFiles(self):
 472                 return self.total_files
 473
 474         def getTotalLOC(self):
 475                 return self.total_lines
 476
 477         def revToDate(self, rev):
 478                 stamp = int(getpipeoutput(['git log --pretty=format:%%at "%s" -n 1' % rev]))
 479                 return datetime.datetime.fromtimestamp(stamp).strftime('%Y-%m-%d')
 480
 481 class ReportCreator:
 482         """Creates the actual report based on given data."""
 483         def __init__(self):
 484                 pass
 485
 486         def create(self, data, path):
 487                 self.data = data
 488                 self.path = path
 489
 490 def html_linkify(text):
 491         return text.lower().replace(' ', '_')
 492
 493 def html_header(level, text):
 494         name = html_linkify(text)
 495         return '\n<h%d><a href="#%s" name="%s">%s</a></h%d>\n\n' % (level, name, name, text, level)
 496
 497 class HTMLReportCreator(ReportCreator):
 498         def create(self, data, path):
 499                 ReportCreator.create(self, data, path)
 500                 self.title = data.projectname
 501
 502                 # copy static files. Looks in the binary directory, ../share/gitstats and /usr/share/gitstats
 503                 binarypath = os.path.dirname(os.path.abspath(__file__))
 504                 secondarypath = os.path.join(binarypath, '..', 'share', 'gitstats')
 505                 basedirs = [binarypath, secondarypath, '/usr/share/gitstats']
 506                 for file in ('gitstats.css', 'sortable.js', 'arrow-up.gif', 'arrow-down.gif', 'arrow-none.gif'):
 507                         for base in basedirs:
 508                                 src = base + '/' + file
 509                                 if os.path.exists(src):
 510                                         shutil.copyfile(src, path + '/' + file)
 511                                         break
 512                         else:
 513                                 print 'Warning: "%s" not found, so not copied (searched: %s)' % (file, basedirs)
 514
 515                 f = open(path + "/index.html", 'w')
 516                 format = '%Y-%m-%d %H:%M:%S'
 517                 self.printHeader(f)
 518
 519                 f.write('<h1>GitStats - %s</h1>' % data.projectname)
 520
 521                 self.printNav(f)
 522
 523                 f.write('<dl>')
 524                 f.write('<dt>Project name</dt><dd>%s</dd>' % (data.projectname))
 525                 f.write('<dt>Generated</dt><dd>%s (in %d seconds)</dd>' % (datetime.datetime.now().strftime(format), time.time() - data.getStampCreated()))
 526                 f.write('<dt>Generator</dt><dd><a href="http://gitstats.sourceforge.net/">GitStats</a> (version %s)</dd>' % getversion())
 527                 f.write('<dt>Report Period</dt><dd>%s to %s</dd>' % (data.getFirstCommitDate().strftime(format), data.getLastCommitDate().strftime(format)))
 528                 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())))
 529                 f.write('<dt>Total Files</dt><dd>%s</dd>' % data.getTotalFiles())
 530                 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))
 531                 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()))
 532                 f.write('<dt>Authors</dt><dd>%s</dd>' % data.getTotalAuthors())
 533                 f.write('</dl>')
 534
 535                 f.write('</body>\n</html>')
 536                 f.close()
 537
 538                 ###
 539                 # Activity
 540                 f = open(path + '/activity.html', 'w')
 541                 self.printHeader(f)
 542                 f.write('<h1>Activity</h1>')
 543                 self.printNav(f)
 544
 545                 #f.write('<h2>Last 30 days</h2>')
 546
 547                 #f.write('<h2>Last 12 months</h2>')
 548
 549                 # Weekly activity
 550                 WEEKS = 32
 551                 f.write(html_header(2, 'Weekly activity'))
 552                 f.write('<p>Last %d weeks</p>' % WEEKS)
 553
 554                 # generate weeks to show (previous N weeks from now)
 555                 now = datetime.datetime.now()
 556                 deltaweek = datetime.timedelta(7)
 557                 weeks = []
 558                 stampcur = now
 559                 for i in range(0, WEEKS):
 560                         weeks.insert(0, stampcur.strftime('%Y-%W'))
 561                         stampcur -= deltaweek
 562
 563                 # top row: commits & bar
 564                 f.write('<table class="noborders"><tr>')
 565                 for i in range(0, WEEKS):
 566                         commits = 0
 567                         if weeks[i] in data.activity_by_year_week:
 568                                 commits = data.activity_by_year_week[weeks[i]]
 569
 570                         percentage = 0
 571                         if weeks[i] in data.activity_by_year_week:
 572                                 percentage = float(data.activity_by_year_week[weeks[i]]) / data.activity_by_year_week_peak
 573                         height = max(1, int(200 * percentage))
 574                         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))
 575
 576                 # bottom row: year/week
 577                 f.write('</tr><tr>')
 578                 for i in range(0, WEEKS):
 579                         f.write('<td>%s</td>' % (WEEKS - i))
 580                 f.write('</tr></table>')
 581
 582                 # Hour of Day
 583                 f.write(html_header(2, 'Hour of Day'))
 584                 hour_of_day = data.getActivityByHourOfDay()
 585                 f.write('<table><tr><th>Hour</th>')
 586                 for i in range(0, 24):
 587                         f.write('<th>%d</th>' % i)
 588                 f.write('</tr>\n<tr><th>Commits</th>')
 589                 fp = open(path + '/hour_of_day.dat', 'w')
 590                 for i in range(0, 24):
 591                         if i in hour_of_day:
 592                                 r = 127 + int((float(hour_of_day[i]) / data.activity_by_hour_of_day_busiest) * 128)
 593                                 f.write('<td style="background-color: rgb(%d, 0, 0)">%d</td>' % (r, hour_of_day[i]))
 594                                 fp.write('%d %d\n' % (i, hour_of_day[i]))
 595                         else:
 596                                 f.write('<td>0</td>')
 597                                 fp.write('%d 0\n' % i)
 598                 fp.close()
 599                 f.write('</tr>\n<tr><th>%</th>')
 600                 totalcommits = data.getTotalCommits()
 601                 for i in range(0, 24):
 602                         if i in hour_of_day:
 603                                 r = 127 + int((float(hour_of_day[i]) / data.activity_by_hour_of_day_busiest) * 128)
 604                                 f.write('<td style="background-color: rgb(%d, 0, 0)">%.2f</td>' % (r, (100.0 * hour_of_day[i]) / totalcommits))
 605                         else:
 606                                 f.write('<td>0.00</td>')
 607                 f.write('</tr></table>')
 608                 f.write('<img src="hour_of_day.png" alt="Hour of Day" />')
 609                 fg = open(path + '/hour_of_day.dat', 'w')
 610                 for i in range(0, 24):
 611                         if i in hour_of_day:
 612                                 fg.write('%d %d\n' % (i + 1, hour_of_day[i]))
 613                         else:
 614                                 fg.write('%d 0\n' % (i + 1))
 615                 fg.close()
 616
 617                 # Day of Week
 618                 f.write(html_header(2, 'Day of Week'))
 619                 day_of_week = data.getActivityByDayOfWeek()
 620                 f.write('<div class="vtable"><table>')
 621                 f.write('<tr><th>Day</th><th>Total (%)</th></tr>')
 622                 fp = open(path + '/day_of_week.dat', 'w')
 623                 for d in range(0, 7):
 624                         commits = 0
 625                         if d in day_of_week:
 626                                 commits = day_of_week[d]
 627                         fp.write('%d %d\n' % (d + 1, commits))
 628                         f.write('<tr>')
 629                         f.write('<th>%d</th>' % (d + 1))
 630                         if d in day_of_week:
 631                                 f.write('<td>%d (%.2f%%)</td>' % (day_of_week[d], (100.0 * day_of_week[d]) / totalcommits))
 632                         else:
 633                                 f.write('<td>0</td>')
 634                         f.write('</tr>')
 635                 f.write('</table></div>')
 636                 f.write('<img src="day_of_week.png" alt="Day of Week" />')
 637                 fp.close()
 638
 639                 # Hour of Week
 640                 f.write(html_header(2, 'Hour of Week'))
 641                 f.write('<table>')
 642
 643                 f.write('<tr><th>Weekday</th>')
 644                 for hour in range(0, 24):
 645                         f.write('<th>%d</th>' % (hour))
 646                 f.write('</tr>')
 647
 648                 for weekday in range(0, 7):
 649                         f.write('<tr><th>%d</th>' % (weekday + 1))
 650                         for hour in range(0, 24):
 651                                 try:
 652                                         commits = data.activity_by_hour_of_week[weekday][hour]
 653                                 except KeyError:
 654                                         commits = 0
 655                                 if commits != 0:
 656                                         f.write('<td')
 657                                         r = 127 + int((float(commits) / data.activity_by_hour_of_week_busiest) * 128)
 658                                         f.write(' style="background-color: rgb(%d, 0, 0)"' % r)
 659                                         f.write('>%d</td>' % commits)
 660                                 else:
 661                                         f.write('<td></td>')
 662                         f.write('</tr>')
 663
 664                 f.write('</table>')
 665
 666                 # Month of Year
 667                 f.write(html_header(2, 'Month of Year'))
 668                 f.write('<div class="vtable"><table>')
 669                 f.write('<tr><th>Month</th><th>Commits (%)</th></tr>')
 670                 fp = open (path + '/month_of_year.dat', 'w')
 671                 for mm in range(1, 13):
 672                         commits = 0
 673                         if mm in data.activity_by_month_of_year:
 674                                 commits = data.activity_by_month_of_year[mm]
 675                         f.write('<tr><td>%d</td><td>%d (%.2f %%)</td></tr>' % (mm, commits, (100.0 * commits) / data.getTotalCommits()))
 676                         fp.write('%d %d\n' % (mm, commits))
 677                 fp.close()
 678                 f.write('</table></div>')
 679                 f.write('<img src="month_of_year.png" alt="Month of Year" />')
 680
 681                 # Commits by year/month
 682                 f.write(html_header(2, 'Commits by year/month'))
 683                 f.write('<div class="vtable"><table><tr><th>Month</th><th>Commits</th></tr>')
 684                 for yymm in reversed(sorted(data.commits_by_month.keys())):
 685                         f.write('<tr><td>%s</td><td>%d</td></tr>' % (yymm, data.commits_by_month[yymm]))
 686                 f.write('</table></div>')
 687                 f.write('<img src="commits_by_year_month.png" alt="Commits by year/month" />')
 688                 fg = open(path + '/commits_by_year_month.dat', 'w')
 689                 for yymm in sorted(data.commits_by_month.keys()):
 690                         fg.write('%s %s\n' % (yymm, data.commits_by_month[yymm]))
 691                 fg.close()
 692
 693                 # Commits by year
 694                 f.write(html_header(2, 'Commits by Year'))
 695                 f.write('<div class="vtable"><table><tr><th>Year</th><th>Commits (% of all)</th></tr>')
 696                 for yy in reversed(sorted(data.commits_by_year.keys())):
 697                         f.write('<tr><td>%s</td><td>%d (%.2f%%)</td></tr>' % (yy, data.commits_by_year[yy], (100.0 * data.commits_by_year[yy]) / data.getTotalCommits()))
 698                 f.write('</table></div>')
 699                 f.write('<img src="commits_by_year.png" alt="Commits by Year" />')
 700                 fg = open(path + '/commits_by_year.dat', 'w')
 701                 for yy in sorted(data.commits_by_year.keys()):
 702                         fg.write('%d %d\n' % (yy, data.commits_by_year[yy]))
 703                 fg.close()
 704
 705                 # Commits by timezone
 706                 f.write(html_header(2, 'Commits by Timezone'))
 707                 f.write('<table><tr>')
 708                 f.write('<th>Timezone</th><th>Commits</th>')
 709                 max_commits_on_tz = max(data.commits_by_timezone.values())
 710                 for i in sorted(data.commits_by_timezone.keys(), key = lambda n : int(n)):
 711                         commits = data.commits_by_timezone[i]
 712                         r = 127 + int((float(commits) / max_commits_on_tz) * 128)
 713                         f.write('<tr><th>%s</th><td style="background-color: rgb(%d, 0, 0)">%d</td></tr>' % (i, r, commits))
 714                 f.write('</tr></table>')
 715
 716                 f.write('</body></html>')
 717                 f.close()
 718
 719                 ###
 720                 # Authors
 721                 f = open(path + '/authors.html', 'w')
 722                 self.printHeader(f)
 723
 724                 f.write('<h1>Authors</h1>')
 725                 self.printNav(f)
 726
 727                 # Authors :: List of authors
 728                 f.write(html_header(2, 'List of Authors'))
 729
 730                 f.write('<table class="authors sortable" id="authors">')
 731                 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>')
 732                 for author in sorted(data.getAuthors()):
 733                         info = data.getAuthorInfo(author)
 734                         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['lines_added'], info['lines_removed'], info['commits_frac'], info['date_first'], info['date_last'], info['timedelta'], info['active_days'], info['place_by_commits']))
 735                 f.write('</table>')
 736
 737                 # Authors :: Author of Month
 738                 f.write(html_header(2, 'Author of Month'))
 739                 f.write('<table class="sortable" id="aom">')
 740                 f.write('<tr><th>Month</th><th>Author</th><th>Commits (%)</th><th class="unsortable">Next top 5</th></tr>')
 741                 for yymm in reversed(sorted(data.author_of_month.keys())):
 742                         authordict = data.author_of_month[yymm]
 743                         authors = getkeyssortedbyvalues(authordict)
 744                         authors.reverse()
 745                         commits = data.author_of_month[yymm][authors[0]]
 746                         next = ', '.join(authors[1:5])
 747                         f.write('<tr><td>%s</td><td>%s</td><td>%d (%.2f%% of %d)</td><td>%s</td></tr>' % (yymm, authors[0], commits, (100.0 * commits) / data.commits_by_month[yymm], data.commits_by_month[yymm], next))
 748
 749                 f.write('</table>')
 750
 751                 f.write(html_header(2, 'Author of Year'))
 752                 f.write('<table class="sortable" id="aoy"><tr><th>Year</th><th>Author</th><th>Commits (%)</th><th class="unsortable">Next top 5</th></tr>')
 753                 for yy in reversed(sorted(data.author_of_year.keys())):
 754                         authordict = data.author_of_year[yy]
 755                         authors = getkeyssortedbyvalues(authordict)
 756                         authors.reverse()
 757                         commits = data.author_of_year[yy][authors[0]]
 758                         next = ', '.join(authors[1:5])
 759                         f.write('<tr><td>%s</td><td>%s</td><td>%d (%.2f%% of %d)</td><td>%s</td></tr>' % (yy, authors[0], commits, (100.0 * commits) / data.commits_by_year[yy], data.commits_by_year[yy], next))
 760                 f.write('</table>')
 761
 762                 f.write('</body></html>')
 763                 f.close()
 764
 765                 ###
 766                 # Files
 767                 f = open(path + '/files.html', 'w')
 768                 self.printHeader(f)
 769                 f.write('<h1>Files</h1>')
 770                 self.printNav(f)
 771
 772                 f.write('<dl>\n')
 773                 f.write('<dt>Total files</dt><dd>%d</dd>' % data.getTotalFiles())
 774                 f.write('<dt>Total lines</dt><dd>%d</dd>' % data.getTotalLOC())
 775                 f.write('<dt>Average file size</dt><dd>%.2f bytes</dd>' % ((100.0 * data.getTotalLOC()) / data.getTotalFiles()))
 776                 f.write('</dl>\n')
 777
 778                 # Files :: File count by date
 779                 f.write(html_header(2, 'File count by date'))
 780
 781                 # use set to get rid of duplicate/unnecessary entries
 782                 files_by_date = set()
 783                 for stamp in sorted(data.files_by_stamp.keys()):
 784                         files_by_date.add('%s %d' % (datetime.datetime.fromtimestamp(stamp).strftime('%Y-%m-%d'), data.files_by_stamp[stamp]))
 785
 786                 fg = open(path + '/files_by_date.dat', 'w')
 787                 for line in sorted(list(files_by_date)):
 788                         fg.write('%s\n' % line)
 789                 #for stamp in sorted(data.files_by_stamp.keys()):
 790                 #       fg.write('%s %d\n' % (datetime.datetime.fromtimestamp(stamp).strftime('%Y-%m-%d'), data.files_by_stamp[stamp]))
 791                 fg.close()
 792
 793                 f.write('<img src="files_by_date.png" alt="Files by Date" />')
 794
 795                 #f.write('<h2>Average file size by date</h2>')
 796
 797                 # Files :: Extensions
 798                 f.write(html_header(2, 'Extensions'))
 799                 f.write('<table class="sortable" id="ext"><tr><th>Extension</th><th>Files (%)</th><th>Lines (%)</th><th>Lines/file</th></tr>')
 800                 for ext in sorted(data.extensions.keys()):
 801                         files = data.extensions[ext]['files']
 802                         lines = data.extensions[ext]['lines']
 803                         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, (100.0 * lines) / data.getTotalLOC(), lines / files))
 804                 f.write('</table>')
 805
 806                 f.write('</body></html>')
 807                 f.close()
 808
 809                 ###
 810                 # Lines
 811                 f = open(path + '/lines.html', 'w')
 812                 self.printHeader(f)
 813                 f.write('<h1>Lines</h1>')
 814                 self.printNav(f)
 815
 816                 f.write('<dl>\n')
 817                 f.write('<dt>Total lines</dt><dd>%d</dd>' % data.getTotalLOC())
 818                 f.write('</dl>\n')
 819
 820                 f.write(html_header(2, 'Lines of Code'))
 821                 f.write('<img src="lines_of_code.png" />')
 822
 823                 fg = open(path + '/lines_of_code.dat', 'w')
 824                 for stamp in sorted(data.changes_by_date.keys()):
 825                         fg.write('%d %d\n' % (stamp, data.changes_by_date[stamp]['lines']))
 826                 fg.close()
 827
 828                 f.write('</body></html>')
 829                 f.close()
 830
 831                 ###
 832                 # tags.html
 833                 f = open(path + '/tags.html', 'w')
 834                 self.printHeader(f)
 835                 f.write('<h1>Tags</h1>')
 836                 self.printNav(f)
 837
 838                 f.write('<dl>')
 839                 f.write('<dt>Total tags</dt><dd>%d</dd>' % len(data.tags))
 840                 if len(data.tags) > 0:
 841                         f.write('<dt>Average commits per tag</dt><dd>%.2f</dd>' % (1.0 * data.getTotalCommits() / len(data.tags)))
 842                 f.write('</dl>')
 843
 844                 f.write('<table class="tags">')
 845                 f.write('<tr><th>Name</th><th>Date</th><th>Commits</th><th>Authors</th></tr>')
 846                 # sort the tags by date desc
 847                 tags_sorted_by_date_desc = map(lambda el : el[1], reversed(sorted(map(lambda el : (el[1]['date'], el[0]), data.tags.items()))))
 848                 for tag in tags_sorted_by_date_desc:
 849                         authorinfo = []
 850                         authors_by_commits = getkeyssortedbyvalues(data.tags[tag]['authors'])
 851                         for i in reversed(authors_by_commits):
 852                                 authorinfo.append('%s (%d)' % (i, data.tags[tag]['authors'][i]))
 853                         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)))
 854                 f.write('</table>')
 855
 856                 f.write('</body></html>')
 857                 f.close()
 858
 859                 self.createGraphs(path)
 860
 861         def createGraphs(self, path):
 862                 print 'Generating graphs...'
 863
 864                 # hour of day
 865                 f = open(path + '/hour_of_day.plot', 'w')
 866                 f.write(GNUPLOT_COMMON)
 867                 f.write(
 868 """
 869 set output 'hour_of_day.png'
 870 unset key
 871 set xrange [0.5:24.5]
 872 set xtics 4
 873 set ylabel "Commits"
 874 plot 'hour_of_day.dat' using 1:2:(0.5) w boxes fs solid
 875 """)
 876                 f.close()
 877
 878                 # day of week
 879                 f = open(path + '/day_of_week.plot', 'w')
 880                 f.write(GNUPLOT_COMMON)
 881                 f.write(
 882 """
 883 set output 'day_of_week.png'
 884 unset key
 885 set xrange [0.5:7.5]
 886 set xtics 1
 887 set ylabel "Commits"
 888 plot 'day_of_week.dat' using 1:2:(0.5) w boxes fs solid
 889 """)
 890                 f.close()
 891
 892                 # Month of Year
 893                 f = open(path + '/month_of_year.plot', 'w')
 894                 f.write(GNUPLOT_COMMON)
 895                 f.write(
 896 """
 897 set output 'month_of_year.png'
 898 unset key
 899 set xrange [0.5:12.5]
 900 set xtics 1
 901 set ylabel "Commits"
 902 plot 'month_of_year.dat' using 1:2:(0.5) w boxes fs solid
 903 """)
 904                 f.close()
 905
 906                 # commits_by_year_month
 907                 f = open(path + '/commits_by_year_month.plot', 'w')
 908                 f.write(GNUPLOT_COMMON)
 909                 f.write(
 910 """
 911 set output 'commits_by_year_month.png'
 912 unset key
 913 set xdata time
 914 set timefmt "%Y-%m"
 915 set format x "%Y-%m"
 916 set xtics rotate by 90 15768000
 917 set bmargin 5
 918 set ylabel "Commits"
 919 plot 'commits_by_year_month.dat' using 1:2:(0.5) w boxes fs solid
 920 """)
 921                 f.close()
 922
 923                 # commits_by_year
 924                 f = open(path + '/commits_by_year.plot', 'w')
 925                 f.write(GNUPLOT_COMMON)
 926                 f.write(
 927 """
 928 set output 'commits_by_year.png'
 929 unset key
 930 set xtics 1
 931 set ylabel "Commits"
 932 set yrange [0:]
 933 plot 'commits_by_year.dat' using 1:2:(0.5) w boxes fs solid
 934 """)
 935                 f.close()
 936
 937                 # Files by date
 938                 f = open(path + '/files_by_date.plot', 'w')
 939                 f.write(GNUPLOT_COMMON)
 940                 f.write(
 941 """
 942 set output 'files_by_date.png'
 943 unset key
 944 set xdata time
 945 set timefmt "%Y-%m-%d"
 946 set format x "%Y-%m-%d"
 947 set ylabel "Files"
 948 set xtics rotate by 90
 949 set ytics autofreq
 950 set bmargin 6
 951 plot 'files_by_date.dat' using 1:2 w steps
 952 """)
 953                 f.close()
 954
 955                 # Lines of Code
 956                 f = open(path + '/lines_of_code.plot', 'w')
 957                 f.write(GNUPLOT_COMMON)
 958                 f.write(
 959 """
 960 set output 'lines_of_code.png'
 961 unset key
 962 set xdata time
 963 set timefmt "%s"
 964 set format x "%Y-%m-%d"
 965 set ylabel "Lines"
 966 set xtics rotate by 90
 967 set bmargin 6
 968 plot 'lines_of_code.dat' using 1:2 w lines
 969 """)
 970                 f.close()
 971
 972                 os.chdir(path)
 973                 files = glob.glob(path + '/*.plot')
 974                 for f in files:
 975                         out = getpipeoutput([gnuplot_cmd + ' "%s"' % f])
 976                         if len(out) > 0:
 977                                 print out
 978
 979         def printHeader(self, f, title = ''):
 980                 f.write(
 981 """<?xml version="1.0" encoding="UTF-8"?>
 982 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
 983 <html xmlns="http://www.w3.org/1999/xhtml">
 984 <head>
 985         <title>GitStats - %s</title>
 986         <link rel="stylesheet" href="gitstats.css" type="text/css" />
 987         <meta name="generator" content="GitStats %s" />
 988         <script type="text/javascript" src="sortable.js"></script>
 989 </head>
 990 <body>
 991 """ % (self.title, getversion()))
 992
 993         def printNav(self, f):
 994                 f.write("""
 995 <div class="nav">
 996 <ul>
 997 <li><a href="index.html">General</a></li>
 998 <li><a href="activity.html">Activity</a></li>
 999 <li><a href="authors.html">Authors</a></li>
1000 <li><a href="files.html">Files</a></li>
1001 <li><a href="lines.html">Lines</a></li>
1002 <li><a href="tags.html">Tags</a></li>
1003 </ul>
1004 </div>
1005 """)
1006
1007
1008 usage = """
1009 Usage: gitstats [options] <gitpath> <outputpath>
1010
1011 Options:
1012 """
1013
1014 if len(sys.argv) <  3:
1015         print usage
1016         sys.exit(0)
1017
1018 gitpath = sys.argv[1]
1019 outputpath = os.path.abspath(sys.argv[2])
1020 rundir = os.getcwd()
1021
1022 try:
1023         os.makedirs(outputpath)
1024 except OSError:
1025         pass
1026 if not os.path.isdir(outputpath):
1027         print 'FATAL: Output path is not a directory or does not exist'
1028         sys.exit(1)
1029
1030 print 'Git path: %s' % gitpath
1031 print 'Output path: %s' % outputpath
1032
1033 os.chdir(gitpath)
1034
1035 cachefile = os.path.join(outputpath, 'gitstats.cache')
1036
1037 print 'Collecting data...'
1038 data = GitDataCollector()
1039 data.loadCache(cachefile)
1040 data.collect(gitpath)
1041 print 'Refining data...'
1042 data.saveCache(cachefile)
1043 data.refine()
1044
1045 os.chdir(rundir)
1046
1047 print 'Generating report...'
1048 report = HTMLReportCreator()
1049 report.create(data, outputpath)
1050
1051 time_end = time.time()
1052 exectime_internal = time_end - time_start
1053 print 'Execution time %.5f secs, %.5f secs (%.2f %%) in external commands)' % (exectime_internal, exectime_external, (100.0 * exectime_external) / exectime_internal)