1 #!/usr/bin/perl -w

     2 

     3 # Copyright (C) 2005, 2006 Apple Computer, Inc.  All rights reserved.

     4 #

     5 # Redistribution and use in source and binary forms, with or without

     6 # modification, are permitted provided that the following conditions

     7 # are met:

     8 #

     9 # 1.  Redistributions of source code must retain the above copyright

    10 #     notice, this list of conditions and the following disclaimer. 

    11 # 2.  Redistributions in binary form must reproduce the above copyright

    12 #     notice, this list of conditions and the following disclaimer in the

    13 #     documentation and/or other materials provided with the distribution. 

    14 # 3.  Neither the name of Apple Computer, Inc. ("Apple") nor the names of

    15 #     its contributors may be used to endorse or promote products derived

    16 #     from this software without specific prior written permission. 

    17 #

    18 # THIS SOFTWARE IS PROVIDED BY APPLE AND ITS CONTRIBUTORS "AS IS" AND ANY

    19 # EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED

    20 # WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE

    21 # DISCLAIMED. IN NO EVENT SHALL APPLE OR ITS CONTRIBUTORS BE LIABLE FOR ANY

    22 # DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES

    23 # (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;

    24 # LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND

    25 # ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT

    26 # (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF

    27 # THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

    28 

    29 # Extended "svn diff" script for WebKit Open Source Project, used to make patches.

    30 

    31 # Differences from standard "svn diff":

    32 #

    33 #   Uses the real diff, not svn's built-in diff.

    34 #   Always passes "-p" to diff so it will try to include function names.

    35 #   Handles binary files (encoded as a base64 chunk of text).

    36 #   Sorts the diffs alphabetically by text files, then binary files.

    37 #   Handles copied and moved files.

    38 #

    39 # Missing features:

    40 #

    41 #   Handle copied and moved directories.

    42 

    43 use strict;

    44 use warnings;

    45 

    46 use Config;

    47 use File::Basename;

    48 use File::Spec;

    49 use File::stat;

    50 use FindBin;

    51 use Getopt::Long;

    52 use lib $FindBin::Bin;

    53 use MIME::Base64;

    54 use POSIX qw(:errno_h);

    55 use Time::gmtime;

    56 use VCSUtils;

    57 

    58 sub binarycmp($$);

    59 sub diffOptionsForFile($);

    60 sub findBaseUrl($);

    61 sub findMimeType($;$);

    62 sub findModificationType($);

    63 sub findSourceFileAndRevision($);

    64 sub generateDiff($$);

    65 sub generateFileList($\%);

    66 sub hunkHeaderLineRegExForFile($);

    67 sub isBinaryMimeType($);

    68 sub manufacturePatchForAdditionWithHistory($);

    69 sub numericcmp($$);

    70 sub outputBinaryContent($);

    71 sub patchpathcmp($$);

    72 sub pathcmp($$);

    73 sub processPaths(\@);

    74 sub splitpath($);

    75 sub testfilecmp($$);

    76 

    77 $ENV{'LC_ALL'} = 'C';

    78 

    79 my $showHelp;

    80 my $ignoreChangelogs = 0;

    81 my $devNull = File::Spec->devnull();

    82 

    83 my $result = GetOptions(

    84     "help"       => \$showHelp,

    85     "ignore-changelogs"    => \$ignoreChangelogs

    86 );

    87 if (!$result || $showHelp) {

    88     print STDERR basename($0) . " [-h|--help] [--ignore-changelogs] [svndir1 [svndir2 ...]]\n";

    89     exit 1;

    90 }

    91 

    92 # Sort the diffs for easier reviewing.

    93 my %paths = processPaths(@ARGV);

    94 

    95 # Generate a list of files requiring diffs.

    96 my %diffFiles;

    97 for my $path (keys %paths) {

    98     generateFileList($path, %diffFiles);

    99 }

   100 

   101 my $svnRoot = determineSVNRoot();

   102 my $prefix = chdirReturningRelativePath($svnRoot);

   103 

   104 my $patchSize = 0;

   105 

   106 # Generate the diffs, in a order chosen for easy reviewing.

   107 for my $path (sort patchpathcmp values %diffFiles) {

   108     $patchSize += generateDiff($path, $prefix);

   109 }

   110 

   111 if ($patchSize > 20480) {

   112     print STDERR "WARNING: Patch's size is " . int($patchSize/1024) . " kbytes.\n";

   113     print STDERR "Patches 20k or smaller are more likely to be reviewed. Larger patches may sit unreviewed for a long time.\n";

   114 }

   115 

   116 exit 0;

   117 

   118 # Overall sort, considering multiple criteria.

   119 sub patchpathcmp($$)

   120 {

   121     my ($a, $b) = @_;

   122 

   123     # All binary files come after all non-binary files.

   124     my $result = binarycmp($a, $b);

   125     return $result if $result;

   126 

   127     # All test files come after all non-test files.

   128     $result = testfilecmp($a, $b);

   129     return $result if $result;

   130 

   131     # Final sort is a "smart" sort by directory and file name.

   132     return pathcmp($a, $b);

   133 }

   134 

   135 # Sort so text files appear before binary files.

   136 sub binarycmp($$)

   137 {

   138     my ($fileDataA, $fileDataB) = @_;

   139     return $fileDataA->{isBinary} <=> $fileDataB->{isBinary};

   140 }

   141 

   142 sub diffOptionsForFile($)

   143 {

   144     my ($file) = @_;

   145 

   146     my $options = "uaNp";

   147 

   148     if (my $hunkHeaderLineRegEx = hunkHeaderLineRegExForFile($file)) {

   149         $options .= "F'$hunkHeaderLineRegEx'";

   150     }

   151 

   152     return $options;

   153 }

   154 

   155 sub findBaseUrl($)

   156 {

   157     my ($infoPath) = @_;

   158     my $baseUrl;

   159     open INFO, "svn info '$infoPath' |" or die;

   160     while (<INFO>) {

   161         if (/^URL: (.+?)[\r\n]*$/) {

   162             $baseUrl = $1;

   163         }

   164     }

   165     close INFO;

   166     return $baseUrl;

   167 }

   168 

   169 sub findMimeType($;$)

   170 {

   171     my ($file, $revision) = @_;

   172     my $args = $revision ? "--revision $revision" : "";

   173     open PROPGET, "svn propget svn:mime-type $args '$file' |" or die;

   174     my $mimeType = <PROPGET>;

   175     close PROPGET;

   176     # svn may output a different EOL sequence than $/, so avoid chomp.

   177     if ($mimeType) {

   178         $mimeType =~ s/[\r\n]+$//g;

   179     }

   180     return $mimeType;

   181 }

   182 

   183 sub findModificationType($)

   184 {

   185     my ($stat) = @_;

   186     my $fileStat = substr($stat, 0, 1);

   187     my $propertyStat = substr($stat, 1, 1);

   188     if ($fileStat eq "A" || $fileStat eq "R") {

   189         my $additionWithHistory = substr($stat, 3, 1);

   190         return $additionWithHistory eq "+" ? "additionWithHistory" : "addition";

   191     }

   192     return "modification" if ($fileStat eq "M" || $propertyStat eq "M");

   193     return "deletion" if ($fileStat eq "D");

   194     return undef;

   195 }

   196 

   197 sub findSourceFileAndRevision($)

   198 {

   199     my ($file) = @_;

   200     my $baseUrl = findBaseUrl(".");

   201     my $sourceFile;

   202     my $sourceRevision;

   203     open INFO, "svn info '$file' |" or die;

   204     while (<INFO>) {

   205         if (/^Copied From URL: (.+?)[\r\n]*$/) {

   206             $sourceFile = File::Spec->abs2rel($1, $baseUrl);

   207         } elsif (/^Copied From Rev: ([0-9]+)/) {

   208             $sourceRevision = $1;

   209         }

   210     }

   211     close INFO;

   212     return ($sourceFile, $sourceRevision);

   213 }

   214 

   215 sub generateDiff($$)

   216 {

   217     my ($fileData, $prefix) = @_;

   218     my $file = File::Spec->catdir($prefix, $fileData->{path});

   219     

   220     if ($ignoreChangelogs && basename($file) eq "ChangeLog") {

   221         return 0;

   222     }

   223     

   224     my $patch = "";

   225     if ($fileData->{modificationType} eq "additionWithHistory") {

   226         manufacturePatchForAdditionWithHistory($fileData);

   227     }

   228 

   229     my $diffOptions = diffOptionsForFile($file);

   230     open DIFF, "svn diff --diff-cmd diff -x -$diffOptions '$file' |" or die;

   231     while (<DIFF>) {

   232         $patch .= $_;

   233     }

   234     close DIFF;

   235     $patch = fixChangeLogPatch($patch) if basename($file) eq "ChangeLog";

   236     print $patch;

   237     if ($fileData->{isBinary}) {

   238         print "\n" if ($patch && $patch =~ m/\n\S+$/m);

   239         outputBinaryContent($file);

   240     }

   241     return length($patch);

   242 }

   243 

   244 sub generateFileList($\%)

   245 {

   246     my ($statPath, $diffFiles) = @_;

   247     my %testDirectories = map { $_ => 1 } qw(LayoutTests);

   248     open STAT, "svn stat '$statPath' |" or die;

   249     while (my $line = <STAT>) {

   250         # svn may output a different EOL sequence than $/, so avoid chomp.

   251         $line =~ s/[\r\n]+$//g;

   252         my $stat;

   253         my $path;

   254         if (isSVNVersion16OrNewer()) {

   255             $stat = substr($line, 0, 8);

   256             $path = substr($line, 8);

   257         } else {

   258             $stat = substr($line, 0, 7);

   259             $path = substr($line, 7);

   260         }

   261         next if -d $path;

   262         my $modificationType = findModificationType($stat);

   263         if ($modificationType) {

   264             $diffFiles->{$path}->{path} = $path;

   265             $diffFiles->{$path}->{modificationType} = $modificationType;

   266             $diffFiles->{$path}->{isBinary} = isBinaryMimeType($path);

   267             $diffFiles->{$path}->{isTestFile} = exists $testDirectories{(File::Spec->splitdir($path))[0]} ? 1 : 0;

   268             if ($modificationType eq "additionWithHistory") {

   269                 my ($sourceFile, $sourceRevision) = findSourceFileAndRevision($path);

   270                 $diffFiles->{$path}->{sourceFile} = $sourceFile;

   271                 $diffFiles->{$path}->{sourceRevision} = $sourceRevision;

   272             }

   273         } else {

   274             print STDERR $line, "\n";

   275         }

   276     }

   277     close STAT;

   278 }

   279 

   280 sub hunkHeaderLineRegExForFile($)

   281 {

   282     my ($file) = @_;

   283 

   284     my $startOfObjCInterfaceRegEx = "@(implementation\\|interface\\|protocol)";

   285     return "^[-+]\\|$startOfObjCInterfaceRegEx" if $file =~ /\.mm?$/;

   286     return "^$startOfObjCInterfaceRegEx" if $file =~ /^(.*\/)?(mac|objc)\// && $file =~ /\.h$/;

   287 }

   288 

   289 sub isBinaryMimeType($)

   290 {

   291     my ($file) = @_;

   292     my $mimeType = findMimeType($file);

   293     return 0 if (!$mimeType || substr($mimeType, 0, 5) eq "text/");

   294     return 1;

   295 }

   296 

   297 sub manufacturePatchForAdditionWithHistory($)

   298 {

   299     my ($fileData) = @_;

   300     my $file = $fileData->{path};

   301     print "Index: ${file}\n";

   302     print "=" x 67, "\n";

   303     my $sourceFile = $fileData->{sourceFile};

   304     my $sourceRevision = $fileData->{sourceRevision};

   305     print "--- ${file}\t(revision ${sourceRevision})\t(from ${sourceFile}:${sourceRevision})\n";

   306     print "+++ ${file}\t(working copy)\n";

   307     if ($fileData->{isBinary}) {

   308         print "\nCannot display: file marked as a binary type.\n";

   309         my $mimeType = findMimeType($file, $sourceRevision);

   310         print "svn:mime-type = ${mimeType}\n\n";

   311     } else {

   312         print `svn cat ${sourceFile} | diff -u $devNull - | tail -n +3`;

   313     }

   314 }

   315 

   316 # Sort numeric parts of strings as numbers, other parts as strings.

   317 # Makes 1.33 come after 1.3, which is cool.

   318 sub numericcmp($$)

   319 {

   320     my ($aa, $bb) = @_;

   321 

   322     my @a = split /(\d+)/, $aa;

   323     my @b = split /(\d+)/, $bb;

   324 

   325     # Compare one chunk at a time.

   326     # Each chunk is either all numeric digits, or all not numeric digits.

   327     while (@a && @b) {

   328         my $a = shift @a;

   329         my $b = shift @b;

   330         

   331         # Use numeric comparison if chunks are non-equal numbers.

   332         return $a <=> $b if $a =~ /^\d/ && $b =~ /^\d/ && $a != $b;

   333 

   334         # Use string comparison if chunks are any other kind of non-equal string.

   335         return $a cmp $b if $a ne $b;

   336     }

   337     

   338     # One of the two is now empty; compare lengths for result in this case.

   339     return @a <=> @b;

   340 }

   341 

   342 sub outputBinaryContent($)

   343 {

   344     my ($path) = @_;

   345     # Deletion

   346     return if (! -e $path);

   347     # Addition or Modification

   348     my $buffer;

   349     open BINARY, $path  or die;

   350     while (read(BINARY, $buffer, 60*57)) {

   351         print encode_base64($buffer);

   352     }

   353     close BINARY;

   354     print "\n";

   355 }

   356 

   357 # Sort first by directory, then by file, so all paths in one directory are grouped

   358 # rather than being interspersed with items from subdirectories.

   359 # Use numericcmp to sort directory and filenames to make order logical.

   360 # Also include a special case for ChangeLog, which comes first in any directory.

   361 sub pathcmp($$)

   362 {

   363     my ($fileDataA, $fileDataB) = @_;

   364 

   365     my ($dira, $namea) = splitpath($fileDataA->{path});

   366     my ($dirb, $nameb) = splitpath($fileDataB->{path});

   367 

   368     return numericcmp($dira, $dirb) if $dira ne $dirb;

   369     return -1 if $namea eq "ChangeLog" && $nameb ne "ChangeLog";

   370     return +1 if $namea ne "ChangeLog" && $nameb eq "ChangeLog";

   371     return numericcmp($namea, $nameb);

   372 }

   373 

   374 sub processPaths(\@)

   375 {

   376     my ($paths) = @_;

   377     return ("." => 1) if (!@{$paths});

   378 

   379     my %result = ();

   380 

   381     for my $file (@{$paths}) {

   382         die "can't handle absolute paths like \"$file\"\n" if File::Spec->file_name_is_absolute($file);

   383         die "can't handle empty string path\n" if $file eq "";

   384         die "can't handle path with single quote in the name like \"$file\"\n" if $file =~ /'/; # ' (keep Xcode syntax highlighting happy)

   385 

   386         my $untouchedFile = $file;

   387 

   388         $file = canonicalizePath($file);

   389 

   390         die "can't handle paths with .. like \"$untouchedFile\"\n" if $file =~ m|/\.\./|;

   391 

   392         $result{$file} = 1;

   393     }

   394 

   395     return ("." => 1) if ($result{"."});

   396 

   397     # Remove any paths that also have a parent listed.

   398     for my $path (keys %result) {

   399         for (my $parent = dirname($path); $parent ne '.'; $parent = dirname($parent)) {

   400             if ($result{$parent}) {

   401                 delete $result{$path};

   402                 last;

   403             }

   404         }

   405     }

   406 

   407     return %result;

   408 }

   409 

   410 # Break up a path into the directory (with slash) and base name.

   411 sub splitpath($)

   412 {

   413     my ($path) = @_;

   414 

   415     my $pathSeparator = "/";

   416     my $dirname = dirname($path) . $pathSeparator;

   417     $dirname = "" if $dirname eq "." . $pathSeparator;

   418 

   419     return ($dirname, basename($path));

   420 }

   421 

   422 # Sort so source code files appear before test files.

   423 sub testfilecmp($$)

   424 {

   425     my ($fileDataA, $fileDataB) = @_;

   426     return $fileDataA->{isTestFile} <=> $fileDataB->{isTestFile};

   427 }

   428