WebKitTools/Scripts/prepare-ChangeLog

#!/usr/bin/perl -w
# -*- Mode: perl; indent-tabs-mode: nil; c-basic-offset: 2  -*-

#
#  Copyright (C) 2000, 2001 Eazel, Inc.
#  Copyright (C) 2002, 2003 Apple Computer, Inc.
#
#  prepare-ChangeLog is free software; you can redistribute it and/or
#  modify it under the terms of the GNU General Public
#  License as published by the Free Software Foundation; either
#  version 2 of the License, or (at your option) any later version.
#
#  prepare-ChangeLog is distributed in the hope that it will be useful,
#  but WITHOUT ANY WARRANTY; without even the implied warranty of
#  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
#  General Public License for more details.
#
#  You should have received a copy of the GNU General Public
#  License along with this program; if not, write to the Free
#  Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
#


# Perl script to create a ChangeLog entry with names of files
# and functions from a cvs diff.
#
# Darin Adler <darin@bentspoon.com>, started 20 April 2000
# Java support added by Maciej Stachowiak <mjs@eazel.com>
# Objective-C, C++ and Objective-C++ support added by Maciej Stachowiak <mjs@apple.com>


#
# TODO:
#   List functions that have been removed too.
#   Decide what a good logical order is for the changed files
#     other than a normal text "sort" (top level first?)
#     (group directories?) (.h before .c?)
#   Handle yacc source files too (other languages?).
#   Help merge when there are ChangeLog conflicts or if there's
#     already a partly written ChangeLog entry.
#   Add command line option to put the ChangeLog into a separate
#     file or just spew it out stdout.
#   Figure out how to allow -z options from .cvsrc to work without
#     letting other bad options work. Currently the -f disables
#     everything from the .cvsrc.
#   Add CVS version numbers for each file too (can't do that until
#     the changes are checked in, though).
#   Work around diff stupidity where deleting a function that starts
#     with a comment makes diff think that the following function
#     has been changed (if the following function starts with a comment
#     with the same first line, such as /**)
#   Work around diff stupidity where deleting an entire function and
#     the blank lines before it makes diff think you've changed the
#     previous function.

use strict;

use Getopt::Long;

my $spewDiff = $ENV{"PREPARE_CHANGELOG_DIFF"};
my $openChangeLogs = 0;
GetOptions("diff|d!" => \$spewDiff,
	   "open|o!" => \$openChangeLogs);

# Find the list of modified files
my @changed_files;
my $changed_files_string;
my %changed_line_ranges;
my %function_lists;
my @conflict_files;

my $CVS = "cvs";

print STDERR "  Running cvs -n update to find changed, added, or removed files.\n";
open UPDATE, "$CVS -qn update 2> /dev/stdout |" or die "The cvs update failed: $!.\n";
while (<UPDATE>)
  {
    if (/^[MA] (.+)$/)
      {
        my $file = $1;
        $function_lists{$file} = "";
        push @changed_files, $file if $file ne "ChangeLog";
      }

    push @conflict_files, $1 if /^[C] (.+)$/;

    $function_lists{$1} = " Added." if /^A (.+)$/;
    $function_lists{$1} = " Removed." if /^R (.+)$/;
    
    print unless /^[A-Z] ./ || /^cvs server: New directory/;
  }
close UPDATE;

if (@conflict_files)
  {
    print STDERR "  The following files have conflicts. Run prepare-ChangeLog again after fixing the conflicts:\n";
    print STDERR join("\n", @conflict_files), "\n";
    exit 1;
  }

if (@changed_files)
  {
    $changed_files_string = "'" . join ("' '", @changed_files) . "'";

    # For each file, build a list of modified lines.
    # Use line numbers from the "after" side of each diff.
    print STDERR "  Running cvs diff to determine which lines changed.\n";
    my $file;
    open DIFF, "$CVS -fq diff -N $changed_files_string |" or die "The cvs diff failed: $!.\n";
    while (<DIFF>)
      {
        $file = $1 if /^Index: (\S+)$/;
        if (defined $file) {
          if (/^\d+(,\d+)?[acd](\d+)(,(\d+))?/)
            { push @{$changed_line_ranges{$file}}, [ $2, $4 || $2 ]; }
	  elsif (/DO_NOT_COMMIT/) {
	    print STDERR "WARNING: file $file contains the string DO_NOT_COMMIT, line $.\n";
	  }
	}
      }
    close DIFF;
  }

# For each source file, convert line range to function list.
if (%changed_line_ranges)
  {
    print STDERR "  Extracting affected function names from source files.\n";
    foreach my $file (keys %changed_line_ranges)
      {
        # Only look for function names in .c files.
        next unless $file =~ /\.(c|cpp|m|mm|h|java)/;
    
        # Find all the functions in the file.
        open SOURCE, $file or next;
        my @function_ranges = get_function_line_ranges(\*SOURCE, $file);
        close SOURCE;
    
        # Find all the modified functions.
        my @functions;
        my %saw_function;
        my @change_ranges = (@{$changed_line_ranges{$file}}, []);
        my @change_range = (0, 0);
        FUNCTION: foreach my $function_range_ref (@function_ranges)
          {
            my @function_range = @$function_range_ref;
    
            # Advance to successive change ranges.
            for (;; @change_range = @{shift @change_ranges})
              {
                last FUNCTION unless @change_range;
    
                # If past this function, move on to the next one.
                next FUNCTION if $change_range[0] > $function_range[1];
    
                # If an overlap with this function range, record the function name.
                if ($change_range[1] >= $function_range[0]
                    and $change_range[0] <= $function_range[1])
                  {
                    if (!$saw_function{$function_range[2]})
                      {
                        $saw_function{$function_range[2]} = 1;
                        push @functions, $function_range[2];
                      }
                    next FUNCTION;
                  }
              }
          }
    
        # Format the list of functions now.

        if (@functions) {
            $function_lists{$file} = "" if !defined $function_lists{$file};
            $function_lists{$file} .= "\n        (" . join("):\n        (", @functions) . "):";
        }
      }
  }

if (!%function_lists)
  {
    print STDERR "  No changes found.\n";
    exit 1;
  }

# Get some parameters for the ChangeLog we are about to write.
my $date = sprintf "%d-%02d-%02d",
  1900 + (localtime $^T)[5], # year
  1 + (localtime $^T)[4], # month
  (localtime $^T)[3]; # day within month
my $name = $ENV{CHANGE_LOG_NAME}
  || $ENV{REAL_NAME}
  || (getpwuid $<)[6]
  || "set REAL_NAME environment variable";
my $email_address = $ENV{CHANGE_LOG_EMAIL_ADDRESS}
  || $ENV{EMAIL_ADDRESS}
  || "set EMAIL_ADDRESS environment variable";

# Remove trailing parenthesized notes from user name (bit of hack).
$name =~ s/\(.*?\)\s*$//g;

# Find the change logs.
my %has_log;
my %files;
foreach my $file (sort keys %function_lists)
  {
    my $prefix = $file;
    my $has_log = 0;
    while ($prefix)
      {
        $prefix =~ s-/[^/]+/?$-/- or $prefix = "";
        $has_log = $has_log{$prefix};
        if (!defined $has_log)
          {
            $has_log = -f "${prefix}ChangeLog";
            $has_log{$prefix} = $has_log;
          }
        last if $has_log;
      }
    if (!$has_log)
      {
        print STDERR "No ChangeLog found for $file.\n";
      }
    else
      {
        push @{$files{$prefix}}, $file;
      }
  }

# Get the latest ChangeLog files from cvs.
my $logs = "";
foreach my $prefix (sort keys %files)
  {
    $logs .= " ${prefix}ChangeLog";
  }
if ($logs)
  {
    print STDERR "  Running cvs update to update ChangeLog files.\n";
    open ERRORS, "$CVS update$logs |" or die "The cvs update of ChangeLog files failed: $!.\n";
    print STDERR "    $_" while <ERRORS>;
    close ERRORS;
  }

# Write out a new ChangeLog file.
foreach my $prefix (sort keys %files)
  {
    print STDERR "  Editing the ${prefix}ChangeLog file.\n";
    open OLD_CHANGE_LOG, "${prefix}ChangeLog" or die "Could not open ${prefix}ChangeLog file: $!.\n";
    # It's less efficient to read the whole thing into memory than it would be
    # to read it while we prepend to it later, but I like doing this part first.
    my @old_change_log = <OLD_CHANGE_LOG>;
    close OLD_CHANGE_LOG;
    open CHANGE_LOG, "> ${prefix}ChangeLog" or die "Could not write ${prefix}ChangeLog\n.";
    print CHANGE_LOG "$date  $name  <$email_address>\n\n";
    print CHANGE_LOG "        Reviewed by NOBODY (OOPS!).\n\n";
    if ($prefix =~ m/WebCore/ || `pwd` =~ m/WebCore/) {
        my $testsDir = "../LayoutTests";
        $testsDir = "$prefix/$testsDir" if length($prefix);
        my $haveNewTests = ((-x "$testsDir/../LayoutTests") && (system("find \"$testsDir/../LayoutTests\" -path \"*/CVS/Entries\" | xargs grep -q \"/0/dummy timestamp//\"") == 0));
        print CHANGE_LOG "        WARNING: NO TEST CASES ADDED\n\n" unless $haveNewTests;
    }

    foreach my $file (sort @{$files{$prefix}})
      {
        my $file_stem = substr $file, length $prefix;
        print CHANGE_LOG "        * $file_stem:$function_lists{$file}\n";
      }
    print CHANGE_LOG "\n", @old_change_log;
    close CHANGE_LOG;
  }

# Write out another diff.
if ($spewDiff && @changed_files)
  {
    print STDERR "  Running cvs diff -u to help you write the ChangeLog entries.\n";
    open DIFF, "$CVS -q diff -uN $changed_files_string |" or die "The cvs diff failed: $!.\n";
    while (<DIFF>) { print; }
    close DIFF;
  }

# Open ChangeLogs.
if ($openChangeLogs && $logs)
  {
    print STDERR "  Opening the edited ChangeLog files.\n";
    my $editor = $ENV{"CHANGE_LOG_EDIT_APPLICATION"};
    if ($editor) {
        system "open -a '$editor'$logs";
    } else {
        system "open -e$logs";
    }
  }

# Done.
exit;

sub get_function_line_ranges
  {
    my ($file_handle, $file_name) = @_;

    if ($file_name =~ /\.(c|cpp|m|mm|h)$/) {
        return get_function_line_ranges_for_c ($file_handle, $file_name);
    } elsif ($file_name =~ /\.java$/) {
        return get_function_line_ranges_for_java ($file_handle, $file_name);
    }
    return ();
  }


sub method_decl_to_selector
  {
    (my $method_decl) = @_;

    $_ = $method_decl;

    if ((my $comment_stripped) = m-([^/]*)(//|/*).*-) 
      {
	$_ = $comment_stripped;
      }

    s/,\s*...//;

    if (/:/) 
      {
        my @components = split /:/;
	pop @components if (scalar @components > 1);
	$_ = (join ':', map {s/.*[^[:word:]]//; scalar $_;} @components) . ':';
      } else {
        s/\s*$//;
	s/.*[^[:word:]]//;
      }

    return $_;
  }



# Read a file and get all the line ranges of the things that look like C functions.
# A function name is the last word before an open parenthesis before the outer
# level open brace. A function starts at the first character after the last close
# brace or semicolon before the function name and ends at the close brace.
# Comment handling is simple-minded but will work for all but pathological cases.
#
# Result is a list of triples: [ start_line, end_line, function_name ].

sub get_function_line_ranges_for_c
  {
    my ($file_handle, $file_name) = @_;

    my @ranges;

    my $in_comment = 0;
    my $in_macro = 0;
    my $in_method_declaration = 0;
    my $in_parentheses = 0;
    my $in_braces = 0;
    my $brace_start = 0;
    my $brace_end = 0;
    my $skip_til_brace_or_semicolon = 0;

    my $word = "";
    my $interface_name = "";

    my $potential_method_char = "";
    my $potential_method_spec = "";

    my $potential_start = 0;
    my $potential_name = "";

    my $start = 0;
    my $name = "";

    my $next_word_could_be_namespace = 0;
    my $potential_namespace = "";
    my @namespaces;

    while (<$file_handle>)
      {
        # Handle continued multi-line comment.
        if ($in_comment)
          {
            next unless s-.*\*/--;
            $in_comment = 0;
          }

        # Handle continued macro.
        if ($in_macro)
          {
            $in_macro = 0 unless /\\$/;
            next;
          }

        # Handle start of macro (or any preprocessor directive).
        if (/^\s*\#/)
          {
            $in_macro = 1 if /^([^\\]|\\.)*\\$/;
            next;
          }

        # Handle comments and quoted text.
        while (m-(/\*|//|\'|\")-) # \' and \" keep emacs perl mode happy
          {
            my $match = $1;
            if ($match eq "/*")
              {
                if (!s-/\*.*?\*/--)
                  {
                    s-/\*.*--;
                    $in_comment = 1;
                  }
              }
            elsif ($match eq "//")
              {
                s-//.*--;
              }
            else # ' or "
              {
                if (!s-$match([^\\]|\\.)*?$match--)
                  {
                    warn "mismatched quotes at line $. in $file_name\n";
                    s-$match.*--;
                  }
              }
          }


        # continued method declaration
        if ($in_method_declaration) 
          {
              my $original = $_;
              my $method_cont = $_;

              chomp $method_cont;
              $method_cont =~ s/[;\{].*//;
              $potential_method_spec = "${potential_method_spec} ${method_cont}";

              $_ = $original;
              if (/;/) 
                {
                  $potential_start = 0;
                  $potential_method_spec = "";
                  $potential_method_char = "";
                  $in_method_declaration = 0;
                  s/^[^;\{]*//;
                } elsif (/{/) {
                  my $selector = method_decl_to_selector ($potential_method_spec);
                  $potential_name = "${potential_method_char}\[${interface_name} ${selector}\]";
                  
                  $potential_method_spec = "";
                  $potential_method_char = "";
                  $in_method_declaration = 0;
  
                  $_ = $original;
                  s/^[^;{]*//;
                } else {
                  next;
                }
          }

        
        # start of method declaration
        if ((my $method_char, my $method_spec) = m&^([-+])([^0-9;][^;]*);?$&)
          {
            my $original = $_;

            if ($interface_name) 
              {
                chomp $method_spec;
                $method_spec =~ s/\{.*//;
            
                $potential_method_char = $method_char;
                $potential_method_spec = $method_spec;
                $potential_start = $.;
                $in_method_declaration = 1;
              } else { 
                warn "declaring a method but don't have interface on line $. in $file_name\n";
              }
            $_ = $original;
            if (/\{/) {
              my $selector = method_decl_to_selector ($potential_method_spec);
              $potential_name = "${potential_method_char}\[${interface_name} ${selector}\]";
              
              $potential_method_spec = "";
              $potential_method_char = "";
              $in_method_declaration = 0;
              $_ = $original;
              s/^[^{]*//;
            } else {
              next;
            }
          }


        # Find function, interface and method names.
        while (m&((?:[[:word:]]+::)*operator(?:[ \t]*\(\)|[^()]*)|[[:word:]:~]+|[(){}:;])|\@(?:implementation|interface|protocol)\s+(\w+)[^{]*&g)
          {
            # interface name
            if ($2) 
              {
                $interface_name = $2;
                next;
              }

            # Open parenthesis.
            if ($1 eq "(")
              {
                $potential_name = $word unless $in_parentheses || $skip_til_brace_or_semicolon;
                $in_parentheses++;
                next;
              }

            # Close parenthesis.
            if ($1 eq ")")
              {
                $in_parentheses--;
                next;
              }

            # C++ constructor initializers
            if ($1 eq ":")
              {
                  $skip_til_brace_or_semicolon = 1 unless ($in_parentheses || $in_braces);
              }

            # Open brace.
            if ($1 eq "{")
              {
                $skip_til_brace_or_semicolon = 0;

                if ($potential_namespace) {
                    push @namespaces, $potential_namespace;
                    $potential_namespace = "";
                    next;
                }

                # Promote potential name to real function name at the
                # start of the outer level set of braces (function body?).
                if (!$in_braces and $potential_start)
                  {
                    $start = $potential_start;
                    $name = $potential_name;
                    if (@namespaces && (length($name) < 2 || substr($name,1,1) ne "[")) {
                        $name = join ('::', @namespaces, $name);
                    }
                  }

                $in_method_declaration = 0;

                $brace_start = $. if (!$in_braces);
                $in_braces++;
                next;
              }

            # Close brace.
            if ($1 eq "}")
              {
                if (!$in_braces && @namespaces) {
                    pop @namespaces;
                    next;
                }

                $in_braces--;
                $brace_end = $. if (!$in_braces);

                # End of an outer level set of braces.
                # This could be a function body.
                if (!$in_braces and $name)
                  {
                    push @ranges, [ $start, $., $name ];
                    $name = "";
                  }

                $potential_start = 0;
                $potential_name = "";
                next;
              }

            # Semicolon.
            if ($1 eq ";")
              {
                $skip_til_brace_or_semicolon = 0;
                $potential_start = 0;
                $potential_name = "";
                $in_method_declaration = 0;
                next;
              }

            # Ignore "const" method qualifier.
            if ($1 eq "const") {
                next;
            }

            if ($1 eq "namespace" || $1 eq "class" || $1 eq "struct") {
                $next_word_could_be_namespace = 1;
                next;
            }

            # Word.
            $word = $1;
            if (!$skip_til_brace_or_semicolon) {
              if ($next_word_could_be_namespace) {
                $potential_namespace = $word;
                $next_word_could_be_namespace = 0;
              } elsif ($potential_namespace) {
                $potential_namespace = "";
              }

              if (!$in_parentheses) {
                $potential_start = 0;
                $potential_name = "";
              }
              if (!$potential_start) {
                $potential_start = $.;
                $potential_name = "";
              }
            }
          }
      }

    warn "missing close braces in $file_name (probable start at $brace_start)\n" if ($in_braces > 0);
    warn "too many close braces in $file_name (probable start at $brace_end)\n" if ($in_braces < 0);

    warn "mismatched parentheses in $file_name\n" if $in_parentheses;

    return @ranges;
  }



# Read a file and get all the line ranges of the things that look like Java
# classes, interfaces and methods.
#
# A class or interface name is the word that immediately follows
# `class' or `interface' when followed by an open curly brace and not
# a semicolon. It can appear at the top level, or inside another class
# or interface block, but not inside a function block
#
# A class or interface starts at the first character after the first close
# brace or after the function name and ends at the close brace.
#
# A function name is the last word before an open parenthesis before
# an open brace rather than a semicolon. It can appear at top level or
# inside a class or interface block, but not inside a function block.
#
# A function starts at the first character after the first close
# brace or after the function name and ends at the close brace.
#
# Comment handling is simple-minded but will work for all but pathological cases.
#
# Result is a list of triples: [ start_line, end_line, function_name ].

sub get_function_line_ranges_for_java
  {
    my ($file_handle, $file_name) = @_;

    my @current_scopes;

    my @ranges;

    my $in_comment = 0;
    my $in_macro = 0;
    my $in_parentheses = 0;
    my $in_braces = 0;
    my $in_non_block_braces = 0;
    my $class_or_interface_just_seen = 0;

    my $word = "";

    my $potential_start = 0;
    my $potential_name = "";
    my $potential_name_is_class_or_interface = 0;

    my $start = 0;
    my $name = "";
    my $current_name_is_class_or_interface = 0;

    while (<$file_handle>)
      {
        # Handle continued multi-line comment.
        if ($in_comment)
          {
            next unless s-.*\*/--;
            $in_comment = 0;
          }

        # Handle continued macro.
        if ($in_macro)
          {
            $in_macro = 0 unless /\\$/;
            next;
          }

        # Handle start of macro (or any preprocessor directive).
        if (/^\s*\#/)
          {
            $in_macro = 1 if /^([^\\]|\\.)*\\$/;
            next;
          }

        # Handle comments and quoted text.
        while (m-(/\*|//|\'|\")-) # \' and \" keep emacs perl mode happy
          {
            my $match = $1;
            if ($match eq "/*")
              {
                if (!s-/\*.*?\*/--)
                  {
                    s-/\*.*--;
                    $in_comment = 1;
                  }
              }
            elsif ($match eq "//")
              {
                s-//.*--;
              }
            else # ' or "
              {
                if (!s-$match([^\\]|\\.)*?$match--)
                  {
                    warn "mismatched quotes at line $. in $file_name\n";
                    s-$match.*--;
                  }
              }
          }

        # Find function names.
        while (m-(\w+|[(){};])-g)
          {
            # Open parenthesis.
            if ($1 eq "(")
              {
                if (!$in_parentheses) {
                    $potential_name = $word;
                    $potential_name_is_class_or_interface = 0;
                }
                $in_parentheses++;
                next;
              }

            # Close parenthesis.
            if ($1 eq ")")
              {
                $in_parentheses--;
                next;
              }

            # Open brace.
            if ($1 eq "{")
              {
                # Promote potential name to real function name at the
                # start of the outer level set of braces (function/class/interface body?).
                if (!$in_non_block_braces
                    and (!$in_braces or $current_name_is_class_or_interface)
                    and $potential_start)
                  {
                    if ($name)
                      {
                          push @ranges, [ $start, ($. - 1),
                                          join ('.', @current_scopes) ];
                      }


                    $current_name_is_class_or_interface = $potential_name_is_class_or_interface;

                    $start = $potential_start;
                    $name = $potential_name;

                    push (@current_scopes, $name);
                  } else {
                      $in_non_block_braces++;
                  }

                $potential_name = "";
                $potential_start = 0;

                $in_braces++;
                next;
              }

            # Close brace.
            if ($1 eq "}")
              {
                $in_braces--;

                # End of an outer level set of braces.
                # This could be a function body.
                if (!$in_non_block_braces)
                  {
                    if ($name)
                      {
                        push @ranges, [ $start, $.,
                                        join ('.', @current_scopes) ];

                        pop (@current_scopes);

                        if (@current_scopes)
                          {
                            $current_name_is_class_or_interface = 1;

                            $start = $. + 1;
                            $name =  $current_scopes[$#current_scopes-1];
                          }
                        else
                          {
                            $current_name_is_class_or_interface = 0;
                            $start = 0;
                            $name =  "";
                          }
                    }
                  }
                else
                  {
                    $in_non_block_braces-- if $in_non_block_braces;
                  }

                $potential_start = 0;
                $potential_name = "";
                next;
              }

            # Semicolon.
            if ($1 eq ";")
              {
                $potential_start = 0;
                $potential_name = "";
                next;
              }

            if ($1 eq "class" or $1 eq "interface")
              {
                $class_or_interface_just_seen = 1;
                next;
              }

            # Word.
            $word = $1;
            if (!$in_parentheses)
              {
                if ($class_or_interface_just_seen) {
                    $potential_name = $word;
                    $potential_start = $.;
                    $class_or_interface_just_seen = 0;
                    $potential_name_is_class_or_interface = 1;
                    next;
                }
              }
            if (!$potential_start)
              {
                $potential_start = $.;
                $potential_name = "";
              }
            $class_or_interface_just_seen = 0;
          }
      }

    warn "mismatched braces in $file_name\n" if $in_braces;
    warn "mismatched parentheses in $file_name\n" if $in_parentheses;

    return @ranges;
  }