Add heuristic for guessing the source file associated with assembler errors
Generalise text in error: pasting "xxx" and "yyy" does not give a valid preprocessing token
Ignore "No such file or directory", as these are unlikely to be gcc specific.
#! perl
# Copyright (c) 2010 Symbian Foundation Ltd
# This component and the accompanying materials are made available
# under the terms of the License "Eclipse Public License v1.0"
# which accompanies this distribution, and is available
# at the URL "http://www.eclipse.org/legal/epl-v10.html".
#
# Initial Contributors:
# Symbian Foundation Ltd - initial contribution.
#
# Contributors:
#
# Description:
# Perl script to summarise GCC logs
use strict;
use Getopt::Long;
sub Usage($)
{
my ($msg) = @_;
print "$msg\n\n" if ($msg ne "");
print <<'EOF';
summarise_gcc_errors.pl - simple script for analysing gcc error logs
This script will read a collection of GCC output logs and summarise
the error messages in various useful ways.
Options:
-warnings process warnings as well as errors
-verbose list the files associated with each error
EOF
exit (1);
}
my $warnings = 0;
my $verbose = 0;
# Analyse the rest of command-line parameters
if (!GetOptions(
"w|warnings" => \$warnings,
"v|verbose" => \$verbose,
))
{
Usage("Invalid argument");
}
my $current_package = "";
my $saved_filename = "";
my %files;
my %errors_by_file;
my %error_count_by_file;
my %errors;
my %message_ids;
my %files_by_message_id;
my %messages_by_id;
my %unique_message_counts;
my %all_message_counts;
my $next_message_id = 1;
my %packages_by_file;
my %package_count_by_file;
my $line;
while ($line = <>)
{
# </pre>os/usb, usb_CompilerCompatibility.005, SF_builds/usb/builds/CompilerCompatibility/usb_CompilerCompatibility.005/html/os_usb_failures.html
if ($line =~/\/html\/([^\/]+)_failures.html/)
{
$current_package = $1;
$current_package =~ s/_/\//;
next;
}
# M:/epoc32/include/elements/nm_interfaces.h:255: warning: dereferencing type-punned pointer will break strict-aliasing rules
# M:/epoc32/include/f32file.h:2169: warning: invalid access to non-static data member 'TVolFormatParam::iUId' of NULL object
# M:/epoc32/include/f32file.h:2169: warning: (perhaps the 'offsetof' macro was used incorrectly)
# M:/epoc32/include/comms-infras/ss_nodemessages.h:301: error: wrong number of template arguments (3, should be 4)
# M:/epoc32/include/elements/nm_signatures.h:496: error: provided for 'template<class TSIGNATURE, int PAYLOADATTRIBOFFSET, class TATTRIBUTECREATIONPOLICY, int PAYLOADBUFFERMAXLEN> class Messages::TSignatureWithPolymorphicPayloadMetaType'
# M:/epoc32/include/comms-infras/ss_nodemessages.h:301: error: invalid type in declaration before ';' token
if ($line =~ /(^...*):(\d+): ([^:]+): (.*)$/)
{
my $filename = $1;
my $lineno = $2;
my $messagetype = $3;
my $message = $4;
# Heuristic for guessing the problem file for assembler issues
#
if ($filename =~ /\\TEMP\\/i)
{
$filename = $saved_filename;
$lineno = 0;
}
else
{
$saved_filename = $filename;
}
if ($messagetype eq "note")
{
next; # ignore notes
}
if ($messagetype eq "warning" && !$warnings)
{
next; # ignore warnings
}
if ($message =~ /.*: No such file/ && !$warnings)
{
next; # ignore "no such file", as these aren't likely to be GCC-specific
}
$filename =~ s/^.://; # remove drive letter
$message =~ s/&/&/g;
$message =~ s/>/>/g;
$message =~ s/</</g;
$message =~ s/'/'/g;
my $generic_message = "$messagetype: $message";
$generic_message =~ s/'offsetof'/"offsetof"/;
$generic_message =~ s/'([^a-zA-Z])'/"\1"/g; # don't catch ';' in next substitution
$generic_message =~ s/['`][^']+'/XX/g; # suppress quoted bits of the actual instance
$generic_message =~ s/pasting ""(.*)"" and ""(.*)""/pasting XX and YY/g; # suppress detail of "pasting" error
my $message_id = $next_message_id;
if (!defined $message_ids{$generic_message})
{
$next_message_id++;
$message_ids{$generic_message} = $message_id;
$messages_by_id{$message_id} = $generic_message;
$all_message_counts{$message_id} = 1;
}
else
{
$message_id = $message_ids{$generic_message};
$all_message_counts{$message_id} += 1;
}
my $instance = sprintf("%s:%d: %s-#%d", $filename, $lineno, $messagetype, $message_id);
my $packages = $packages_by_file{$filename};
if (!defined $packages)
{
$packages_by_file{$filename} = "\t$current_package\t";
$package_count_by_file{$filename} = 1;
}
else
{
if (index($packages,"\t$current_package\t") < 0)
{
$packages_by_file{$filename} .= "\t$current_package\t";
$package_count_by_file{$filename} += 1;
}
}
if (defined $files{$instance})
{
# already seen this one
next;
}
if (!defined $unique_message_counts{$message_id})
{
$unique_message_counts{$message_id} = 1;
}
else
{
$unique_message_counts{$message_id} += 1;
}
$files{$instance} = $message;
if (!defined $files_by_message_id{$message_id})
{
$files_by_message_id{$message_id} = $filename;
}
else
{
$files_by_message_id{$message_id} .= "\n$filename";
}
my $error = sprintf "%-5d: %s: %s", $lineno, $messagetype, $message;
if (!defined $errors_by_file{$filename})
{
$errors_by_file{$filename} = $error;
$error_count_by_file{$filename} = 1;
}
else
{
$errors_by_file{$filename} .= "\n$error";
$error_count_by_file{$filename} += 1;
}
next;
}
}
# clean up the file lists
my %filecount_by_message_id;
foreach my $id (keys %files_by_message_id)
{
my @longlist = split /\n/, $files_by_message_id{$id};
my %uniq;
foreach my $file (@longlist)
{
$uniq{$file} = 1;
}
my $uniqlist = join( "\n\t", sort keys %uniq);
$files_by_message_id{$id} = $uniqlist;
$filecount_by_message_id{$id} = scalar keys %uniq;
}
print "\n\n====Occurrences of messages (distinct, all)\n";
foreach my $id ( sort {$unique_message_counts{$b} <=> $unique_message_counts{$a}} keys %unique_message_counts)
{
printf "%-6d\t%-6d\t%s\n", $unique_message_counts{$id}, $all_message_counts{$id}, $messages_by_id{$id};
}
print "\n\n====Files affected per message\n";
foreach my $id ( sort {$filecount_by_message_id{$b} <=> $filecount_by_message_id{$a}} keys %filecount_by_message_id)
{
printf "%-6d\t%s\n", $filecount_by_message_id{$id}, $messages_by_id{$id};
if ($verbose)
{
print "\t", $files_by_message_id{$id};
}
}
print "\n\n====Packages impacted (if > 1)\n";
foreach my $file ( sort {$package_count_by_file{$b} <=> $package_count_by_file{$a}} keys %package_count_by_file)
{
if ($package_count_by_file{$file} < 2)
{
next;
}
my ($empty,@packages) = split /\t+/, $packages_by_file{$file};
printf "%-6d\t%s\n\t(%s)\n",$package_count_by_file{$file}, $file, join(", ", @packages);
}
print "\n\n====Affected files by package\n";
my $current_package = "";
my @currentfiles;
foreach my $file (sort keys %error_count_by_file)
{
my ($root, $sf, $layer, $packagename, @rest) = split /[\/\\]/, $file;
my $package = "$sf/$layer/$packagename";
if ($layer eq "include")
{
$package = "$sf/$layer";
}
if ($package ne $current_package)
{
if ($current_package ne "")
{
printf "%-6d\t%s\n", scalar @currentfiles, $current_package;
print join("\n",@currentfiles);
print "\n";
}
$current_package = $package;
@currentfiles = ();
}
my $filedetails = sprintf "\t%-6d\t%s", $error_count_by_file{$file}, $file;
push @currentfiles, $filedetails;
}
printf "%-6d\t%s\n", scalar @currentfiles, $current_package;
print join("\n",@currentfiles);
print "\n";
print "\n\n====Messages by file\n";
foreach my $file ( sort keys %error_count_by_file)
{
my @details = split "\n", $errors_by_file{$file};
printf "%-6d\t%s\n\t", $error_count_by_file{$file}, $file;
print join("\n\t", @details);
print "\n";
}