5 # script to search for url's in lyxfiles
6 # and testing their validity.
8 # Syntax: search_url.pl [(filesToScan|(ignored|reverted|extra|selected)URLS)={path_to_control]*
9 # Param value is a path to a file containing list of xxx:
10 # filesToScan={xxx = lyx-file-names to be scanned for}
11 # ignoredURLS={xxx = urls that are discarded from test}
12 # revertedURLS={xxx = urls that should fail, to test the test with invalid urls}
13 # extraURLS={xxx = urls which should be also checked}
15 # This file is free software; you can redistribute it and/or
16 # modify it under the terms of the GNU General Public
17 # License as published by the Free Software Foundation; either
18 # version 2 of the License, or (at your option) any later version.
20 # This software is distributed in the hope that it will be useful,
21 # but WITHOUT ANY WARRANTY; without even the implied warranty of
22 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
23 # General Public License for more details.
25 # You should have received a copy of the GNU General Public
26 # License along with this software; if not, write to the Free Software
27 # Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
29 # Copyright (c) 2013 Kornel Benko <kornel@lyx.org>
30 # (c) 2013 Scott Kostyshak <skotysh@lyx.org>
36 my $p = File::Spec->rel2abs(__FILE__);
37 $p =~ s/[\/\\]?[^\/\\]+$//;
44 use POSIX qw(locale_h);
46 setlocale(LC_CTYPE, "");
47 setlocale(LC_MESSAGES, "en_US.UTF-8");
50 sub printNotUsedURLS($\%);
58 my %revertedURLS = ();
60 my %selectedURLS = ();
62 my $checkSelectedOnly = 0;
64 die("Bad argument \"$arg\"") if ($arg !~ /=/);
65 my ($type,$val) = split("=", $arg);
66 if ($type eq "filesToScan") {
67 #The file should be a list of files to search in
68 if (open(FLIST, $val)) {
69 while (my $l = <FLIST>) {
76 elsif ($type eq "ignoredURLS") {
77 readUrls($val, %ignoredURLS);
79 elsif ($type eq "revertedURLS") {
80 readUrls($val, %revertedURLS);
82 elsif ($type eq "extraURLS") {
83 readUrls($val, %extraURLS);
85 elsif ($type eq "selectedURLS") {
86 $checkSelectedOnly = 1;
87 readUrls($val, %selectedURLS);
90 die("Invalid argument \"$arg\"");
94 my @urls = sort keys %URLS, keys %extraURLS;
100 if (defined($selectedURLS{$u})) {
101 ${selectedURLS}{$u}->{count} += 1;
103 if (defined($ignoredURLS{$u})) {
104 $ignoredURLS{$u}->{count} += 1;
107 next if ($checkSelectedOnly && ! defined($selectedURLS{$u}));
109 print "Checking '$u': ";
112 $res = check_url($u);
121 $prnt = "Failed, caught error: $_";
125 my $printSourceFiles = 0;
126 my $err_txt = "Error url:";
128 if ($res || $checkSelectedOnly) {
129 $printSourceFiles = 1;
131 if ($res && defined($revertedURLS{$u})) {
132 $err_txt = "Failed url:";
134 $res = ! $res if (defined($revertedURLS{$u}));
135 if ($res || $checkSelectedOnly) {
136 print "$err_txt \"$u\"\n";
138 if ($printSourceFiles) {
139 if (defined($URLS{$u})) {
140 for my $f(sort keys %{$URLS{$u}}) {
141 my $lines = ":" . join(',', @{$URLS{$u}->{$f}});
152 printNotUsedURLS("Ignored", %ignoredURLS);
153 printNotUsedURLS("Selected", %selectedURLS);
154 printNotUsedURLS("KnownInvalid", %extraURLS);
157 print "\n$errorcount URL-tests failed out of $URLScount\n\n";
160 ###############################################################################
162 sub printNotUsedURLS($\%)
164 my ($txt, $rURLS) = @_;
166 for my $u ( sort keys %{$rURLS}) {
167 if ($rURLS->{$u}->{count} < 2) {
169 for my $f (sort keys %{$rURLS->{$u}}) {
170 next if ($f eq "count");
171 push(@submsg, "$f:" . $rURLS->{$u}->{$f});
173 push(@msg, "\n $u\n " . join("\n ", @submsg) . "\n");
177 print "\n$txt URLs not found in sources: " . join(' ',@msg) . "\n";
183 my ($file, $rUrls) = @_;
185 die("Could not read file $file") if (! open(ULIST, $file));
187 while (my $l = <ULIST>) {
189 $l =~ s/[\r\n]+$//; # remove eol
190 $l =~ s/\s*\#.*$//; # remove comment
192 if (! defined($rUrls->{$l} )) {
193 $rUrls->{$l} = {$file => $line, count => 1};
202 my $status = "out"; # outside of URL/href
204 return if ($f =~ /\/attic\//);
207 while(my $l = <FI>) {
209 $l =~ s/[\r\n]+$//; # Simulate chomp
210 if ($status eq "out") {
211 # searching for "\begin_inset Flex URL"
212 if($l =~ /^\s*\\begin_inset\s+Flex\s+URL\s*$/) {
213 $status = "inUrlInset";
215 elsif ($l =~ /^\s*\\begin_inset\s+CommandInset\s+href\s*$/) {
216 $status = "inHrefInset";
219 # Outside of url, check also
220 if ($l =~ /"((ftp|http|https):\/\/[^ ]+)"/) {
222 handle_url($url, $f, "x$line");
227 if($l =~ /^\s*\\end_(layout|inset)\s*$/) {
230 elsif ($status eq "inUrlInset") {
231 if ($l =~ /\s*([a-z]+:\/\/.+)\s*$/) {
234 handle_url($url, $f, "u$line");
237 elsif ($status eq "inHrefInset") {
238 if ($l =~ /^target\s+"([a-z]+:\/\/[^ ]+)"$/) {
241 handle_url($url, $f, "h$line");
252 my($url, $f, $line) = @_;
254 if(!defined($URLS{$url})) {
257 if(!defined($URLS{$url}->{$f})) {
258 $URLS{$url}->{$f} = [];
260 push(@{$URLS{$url}->{$f}}, $line);