#!/usr/bin/perl -w

# Copyright (C) 2005, 2006 Apple Computer, Inc.  All rights reserved.
#
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions
# are met:
#
# 1.  Redistributions of source code must retain the above copyright
#     notice, this list of conditions and the following disclaimer. 
# 2.  Redistributions in binary form must reproduce the above copyright
#     notice, this list of conditions and the following disclaimer in the
#     documentation and/or other materials provided with the distribution. 
# 3.  Neither the name of Apple Computer, Inc. ("Apple") nor the names of
#     its contributors may be used to endorse or promote products derived
#     from this software without specific prior written permission. 
#
# THIS SOFTWARE IS PROVIDED BY APPLE AND ITS CONTRIBUTORS "AS IS" AND ANY
# EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
# WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
# DISCLAIMED. IN NO EVENT SHALL APPLE OR ITS CONTRIBUTORS BE LIABLE FOR ANY
# DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
# (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
# LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
# ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
# THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

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

# Differences from standard "svn diff":
#
#   Uses the real diff, not svn's built-in diff.
#   Always passes "-p" to diff so it will try to include function names.
#   Handles binary files (encoded as a base64 chunk of text).
#   Sorts the diffs alphabetically by text files, then binary files.
#   Handles copied and moved files.
#
# Missing features:
#
#   Handle copied and moved directories.

use strict;
use warnings;

use Config;
use Cwd;
use File::Basename;
use File::Spec;
use File::stat;
use Getopt::Long;
use MIME::Base64;
use POSIX qw(:errno_h);
use Time::gmtime;

sub binarycmp($$);
sub canonicalizePath($);
sub findBaseUrl($);
sub findMimeType($;$);
sub findModificationType($);
sub findSourceFileAndRevision($);
sub fixChangeLogPatch($);
sub generateDiff($);
sub generateFileList($\%);
sub isBinaryMimeType($);
sub manufacturePatchForAdditionWithHistory($);
sub numericcmp($$);
sub outputBinaryContent($);
sub patchpathcmp($$);
sub pathcmp($$);
sub processPaths(\@);
sub splitpath($);
sub testfilecmp($$);

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

my $showHelp;

my $result = GetOptions(
    "help"       => \$showHelp,
);
if (!$result || $showHelp) {
    print STDERR basename($0) . " [-h|--help] [svndir1 [svndir2 ...]]\n";
    exit 1;
}

# Sort the diffs for easier reviewing.
my %paths = processPaths(@ARGV);

# Generate a list of files requiring diffs.
my %diffFiles;
for my $path (keys %paths) {
    generateFileList($path, %diffFiles);
}

# Generate the diffs, in a order chosen for easy reviewing.
for my $path (sort patchpathcmp values %diffFiles) {
    generateDiff($path);
}

exit 0;

# Overall sort, considering multiple criteria.
sub patchpathcmp($$)
{
    my ($a, $b) = @_;

    # All binary files come after all non-binary files.
    my $result = binarycmp($a, $b);
    return $result if $result;

    # All test files come after all non-test files.
    $result = testfilecmp($a, $b);
    return $result if $result;

    # Final sort is a "smart" sort by directory and file name.
    return pathcmp($a, $b);
}

# Sort so text files appear before binary files.
sub binarycmp($$)
{
    my ($fileDataA, $fileDataB) = @_;
    return $fileDataA->{isBinary} <=> $fileDataB->{isBinary};
}

sub canonicalizePath($)
{
    my ($file) = @_;

    # Remove extra slashes and '.' directories in path
    $file = File::Spec->canonpath($file);

    # Remove '..' directories in path
    my @dirs = ();
    foreach my $dir (File::Spec->splitdir($file)) {
        if ($dir eq '..' && $#dirs >= 0 && $dirs[$#dirs] ne '..') {
            pop(@dirs);
        } else {
            push(@dirs, $dir);
        }
    }
    return ($#dirs >= 0) ? File::Spec->catdir(@dirs) : ".";
}

sub findBaseUrl($)
{
    my ($infoPath) = @_;
    my $baseUrl;
    open INFO, "svn info '$infoPath' |" or die;
    while (<INFO>) {
        if (/^URL: (.+)/) {
            $baseUrl = $1;
            last;
        }
    }
    close INFO;
    return $baseUrl;
}

sub findMimeType($;$)
{
    my ($file, $revision) = @_;
    my $args = $revision ? "--revision $revision" : "";
    open PROPGET, "svn propget svn:mime-type $args '$file' |" or die;
    my $mimeType = <PROPGET>;
    close PROPGET;
    chomp $mimeType if $mimeType;
    return $mimeType;
}

sub findModificationType($)
{
    my ($stat) = @_;
    my $fileStat = substr($stat, 0, 1);
    my $propertyStat = substr($stat, 1, 1);
    if ($fileStat eq "A") {
        my $additionWithHistory = substr($stat, 3, 1);
        return $additionWithHistory eq "+" ? "additionWithHistory" : "addition";
    }
    return "modification" if ($fileStat eq "M" || $propertyStat eq "M");
    return "deletion" if ($fileStat eq "D");
    return undef;
}

sub findSourceFileAndRevision($)
{
    my ($file) = @_;
    my $baseUrl = findBaseUrl(".");
    my $sourceFile;
    my $sourceRevision;
    open INFO, "svn info '$file' |" or die;
    while (<INFO>) {
        if (/^Copied From URL: (.+)/) {
            $sourceFile = File::Spec->abs2rel($1, $baseUrl);
        } elsif (/^Copied From Rev: ([0-9]+)/) {
            $sourceRevision = $1;
        }
    }
    close INFO;
    return ($sourceFile, $sourceRevision);
}

sub fixChangeLogPatch($)
{
    my $patch = shift;
    my $contextLineCount = 3;

    return $patch if $patch !~ /\n@@ -1,(\d+) \+1,(\d+) @@\n( .*\n)+(\+.*\n)+( .*\n){$contextLineCount}$/m;
    my ($oldLineCount, $newLineCount) = ($1, $2);
    return $patch if $oldLineCount <= $contextLineCount;

    # The diff(1) command is greedy when matching lines, so a new ChangeLog entry will
    # have lines of context at the top of a patch when the existing entry has the same
    # date and author as the new entry.  This nifty loop alters a ChangeLog patch so
    # that the added lines ("+") in the patch always start at the beginning of the
    # patch and there are no initial lines of context.
    my $newPatch;
    my $lineCountInState = 0;
    my $oldContentLineCountReduction = $oldLineCount - $contextLineCount;
    my $newContentLineCountWithoutContext = $newLineCount - $oldLineCount - $oldContentLineCountReduction;
    my ($stateHeader, $statePreContext, $stateNewChanges, $statePostContext) = (1..4);
    my $state = $stateHeader;
    foreach my $line (split(/\n/, $patch)) {
        $lineCountInState++;
        if ($state == $stateHeader && $line =~ /^@@ -1,$oldLineCount \+1,$newLineCount @\@$/) {
            $line = "@@ -1,$contextLineCount +1," . ($newLineCount - $oldContentLineCountReduction) . " @@";
            $lineCountInState = 0;
            $state = $statePreContext;
        } elsif ($state == $statePreContext && substr($line, 0, 1) eq " ") {
            $line = "+" . substr($line, 1);
            if ($lineCountInState == $oldContentLineCountReduction) {
                $lineCountInState = 0;
                $state = $stateNewChanges;
            }
        } elsif ($state == $stateNewChanges && substr($line, 0, 1) eq "+") {
            # No changes to these lines
            if ($lineCountInState == $newContentLineCountWithoutContext) {
                $lineCountInState = 0;
                $state = $statePostContext;
            }
        } elsif ($state == $statePostContext) {
            if (substr($line, 0, 1) eq "+" && $lineCountInState <= $oldContentLineCountReduction) {
                $line = " " . substr($line, 1);
            } elsif ($lineCountInState > $contextLineCount && substr($line, 0, 1) eq " ") {
                next; # Discard
            }
        }
        $newPatch .= $line . "\n";
    }

    return $newPatch;
}

sub generateDiff($)
{
    my ($fileData) = @_;
    my $file = $fileData->{path};
    my $patch;
    if ($fileData->{modificationType} eq "additionWithHistory") {
        manufacturePatchForAdditionWithHistory($fileData);
    }
    open DIFF, "svn diff --diff-cmd diff -x -uaNp '$file' |" or die;
    while (<DIFF>) {
        $patch .= $_;
    }
    close DIFF;
    $patch = fixChangeLogPatch($patch) if basename($file) eq "ChangeLog";
    print $patch if $patch;
    if ($fileData->{isBinary}) {
        print "\n" if ($patch && $patch =~ m/\n\S+$/m);
        outputBinaryContent($file);
    }
}

sub generateFileList($\%)
{
    my ($statPath, $diffFiles) = @_;
    my %testDirectories = map { $_ => 1 } qw(LayoutTests);
    open STAT, "svn stat '$statPath' |" or die;
    while (my $line = <STAT>) {
        chomp $line;
        my $stat = substr($line, 0, 7);
        my $path = substr($line, 7);
        next if -d $path;
        my $modificationType = findModificationType($stat);
        if ($modificationType) {
            $diffFiles->{$path}->{path} = $path;
            $diffFiles->{$path}->{modificationType} = $modificationType;
            $diffFiles->{$path}->{isBinary} = isBinaryMimeType($path);
            $diffFiles->{$path}->{isTestFile} = exists $testDirectories{(File::Spec->splitdir($path))[0]} ? 1 : 0;
            if ($modificationType eq "additionWithHistory") {
                my ($sourceFile, $sourceRevision) = findSourceFileAndRevision($path);
                $diffFiles->{$path}->{sourceFile} = $sourceFile;
                $diffFiles->{$path}->{sourceRevision} = $sourceRevision;
            }
        } else {
            print STDERR $line, "\n";
        }
    }
    close STAT;
}

sub isBinaryMimeType($)
{
    my ($file) = @_;
    my $mimeType = findMimeType($file);
    return 0 if (!$mimeType || substr($mimeType, 0, 5) eq "text/");
    return 1;
}

sub manufacturePatchForAdditionWithHistory($)
{
    my ($fileData) = @_;
    my $file = $fileData->{path};
    print "Index: ${file}\n";
    print "=" x 67, "\n";
    my $sourceFile = $fileData->{sourceFile};
    my $sourceRevision = $fileData->{sourceRevision};
    print "--- ${file}\t(revision ${sourceRevision})\t(from ${sourceFile}:${sourceRevision})\n";
    print "+++ ${file}\t(working copy)\n";
    if ($fileData->{isBinary}) {
        print "\nCannot display: file marked as a binary type.\n";
        my $mimeType = findMimeType($file, $sourceRevision);
        print "svn:mime-type = ${mimeType}\n\n";
    } else {
        print `svn cat ${sourceFile} | diff -u /dev/null - | tail -n +3`;
    }
}

# Sort numeric parts of strings as numbers, other parts as strings.
# Makes 1.33 come after 1.3, which is cool.
sub numericcmp($$)
{
    my ($aa, $bb) = @_;

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

    # Compare one chunk at a time.
    # Each chunk is either all numeric digits, or all not numeric digits.
    while (@a && @b) {
        my $a = shift @a;
        my $b = shift @b;
        
        # Use numeric comparison if chunks are non-equal numbers.
        return $a <=> $b if $a =~ /^\d/ && $b =~ /^\d/ && $a != $b;

        # Use string comparison if chunks are any other kind of non-equal string.
        return $a cmp $b if $a ne $b;
    }
    
    # One of the two is now empty; compare lengths for result in this case.
    return @a <=> @b;
}

sub outputBinaryContent($)
{
    my ($path) = @_;
    # Deletion
    return if (! -e $path);
    # Addition or Modification
    my $buffer;
    open BINARY, $path  or die;
    while (read(BINARY, $buffer, 60*57)) {
        print encode_base64($buffer);
    }
    close BINARY;
    print "\n";
}

# Sort first by directory, then by file, so all paths in one directory are grouped
# rather than being interspersed with items from subdirectories.
# Use numericcmp to sort directory and filenames to make order logical.
# Also include a special case for ChangeLog, which comes first in any directory.
sub pathcmp($$)
{
    my ($fileDataA, $fileDataB) = @_;

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

    return numericcmp($dira, $dirb) if $dira ne $dirb;
    return -1 if $namea eq "ChangeLog" && $nameb ne "ChangeLog";
    return +1 if $namea ne "ChangeLog" && $nameb eq "ChangeLog";
    return numericcmp($namea, $nameb);
}

sub processPaths(\@)
{
    my ($paths) = @_;
    return ("." => 1) if (!@{$paths});

    my %result = ();

    for my $file (@{$paths}) {
        die "can't handle absolute paths like \"$file\"\n" if File::Spec->file_name_is_absolute($file);
        die "can't handle empty string path\n" if $file eq "";
        die "can't handle path with single quote in the name like \"$file\"\n" if $file =~ /'/; # ' (keep Xcode syntax highlighting happy)

        my $untouchedFile = $file;

        $file = canonicalizePath($file);

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

        $result{$file} = 1;
    }

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

    # Remove any paths that also have a parent listed.
    for my $path (keys %result) {
        for (my $parent = dirname($path); $parent ne '.'; $parent = dirname($parent)) {
            if ($result{$parent}) {
                delete $result{$path};
                last;
            }
        }
    }

    return %result;
}

# Break up a path into the directory (with slash) and base name.
sub splitpath($)
{
    my ($path) = @_;

    my $pathSeparator = "/";
    my $dirname = dirname($path) . $pathSeparator;
    $dirname = "" if $dirname eq "." . $pathSeparator;

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

# Sort so source code files appear before test files.
sub testfilecmp($$)
{
    my ($fileDataA, $fileDataB) = @_;
    return $fileDataA->{isTestFile} <=> $fileDataB->{isTestFile};
}