branch master updated: * tp/Texinfo/Convert/Converter.pm (determine_files_and_directory), tp/Texinfo/Convert/Text.pm (output), tp/Texinfo/Convert/HTML.pm (_prepare_css): use a different variable for file name string bytes and file name character string.

[Patrice Dumas]

This is an automated email from the git hooks/post-receive script.

pertusus pushed a commit to branch master
in repository texinfo.

The following commit(s) were added to refs/heads/master by this push:
     new 6b19a72c46 * tp/Texinfo/Convert/Converter.pm 
(determine_files_and_directory), tp/Texinfo/Convert/Text.pm (output), 
tp/Texinfo/Convert/HTML.pm (_prepare_css): use a different variable for file 
name string bytes and file name character string.
6b19a72c46 is described below

commit 6b19a72c4683d5f4d32075ff17efb0bb12642899
Author: Patrice Dumas <pertusus@free.fr>
AuthorDate: Fri Apr 5 22:10:29 2024 +0200

    * tp/Texinfo/Convert/Converter.pm (determine_files_and_directory),
    tp/Texinfo/Convert/Text.pm (output), tp/Texinfo/Convert/HTML.pm
    (_prepare_css): use a different variable for file name string bytes
    and file name character string.
    
    * tp/Texinfo/Convert/HTML.pm (_prepare_css): rename variables.
    
    * tp/t/test_utils.pl: change in comments.
---
 ChangeLog                       | 11 +++++++++++
 tp/Texinfo/Convert/Converter.pm |  7 +++++--
 tp/Texinfo/Convert/HTML.pm      | 28 +++++++++++++++-------------
 tp/Texinfo/Convert/Text.pm      |  7 +++++--
 tp/t/test_utils.pl              | 29 +++++++++++++++++------------
 5 files changed, 53 insertions(+), 29 deletions(-)

diff --git a/ChangeLog b/ChangeLog
index f2e70a6025..35de89a96d 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,14 @@
+2024-04-05  Patrice Dumas  <pertusus@free.fr>
+
+       * tp/Texinfo/Convert/Converter.pm (determine_files_and_directory),
+       tp/Texinfo/Convert/Text.pm (output), tp/Texinfo/Convert/HTML.pm
+       (_prepare_css): use a different variable for file name string bytes
+       and file name character string.
+
+       * tp/Texinfo/Convert/HTML.pm (_prepare_css): rename variables.
+
+       * tp/t/test_utils.pl: change in comments.
+
 2024-04-05  Gavin Smith <gavinsmith0123@gmail.com>
 
        * tp/Texinfo/XS/parsetexi/Parsetexi.pm (parse_texi_line):
diff --git a/tp/Texinfo/Convert/Converter.pm b/tp/Texinfo/Convert/Converter.pm
index 2dbfb6c3f4..dfa882c1eb 100644
--- a/tp/Texinfo/Convert/Converter.pm
+++ b/tp/Texinfo/Convert/Converter.pm
@@ -667,12 +667,15 @@ sub determine_files_and_directory($$)
     # file which is not decoded either.  We want to return only
     # decoded character strings such that they can easily be mixed
     # with other character strings, so we decode here.
-    my $input_file_name = $document_info->{'input_file_name'};
+    my $input_file_name_bytes = $document_info->{'input_file_name'};
     my $encoding = $self->get_conf('COMMAND_LINE_ENCODING');
+    my $input_file_name;
     if (defined($encoding)) {
-      $input_file_name = decode($encoding, $input_file_name, sub { '?' });
+      $input_file_name = decode($encoding, $input_file_name_bytes, sub { '?' 
});
       # use '?' as replacement character rather than U+FFFD in case it
       # is re-encoded to an encoding without this character
+    } else {
+      $input_file_name = $input_file_name_bytes;
     }
     my ($directories, $suffix);
     ($input_basefile, $directories, $suffix) = fileparse($input_file_name);
diff --git a/tp/Texinfo/Convert/HTML.pm b/tp/Texinfo/Convert/HTML.pm
index 7f58948676..47f4e9f579 100644
--- a/tp/Texinfo/Convert/HTML.pm
+++ b/tp/Texinfo/Convert/HTML.pm
@@ -9297,26 +9297,28 @@ sub _prepare_css($)
   my @css_rule_lines;
 
   my $css_files = $self->get_conf('CSS_FILES');
-  foreach my $file (@$css_files) {
+  foreach my $css_file (@$css_files) {
     my $css_file_fh;
-    my $css_file;
-    if ($file eq '-') {
+    my $css_file_path;
+    if ($css_file eq '-') {
       $css_file_fh = \*STDIN;
-      $css_file = '-';
+      $css_file_path = '-';
     } else {
-      $css_file = $self->Texinfo::Common::locate_include_file($file);
-      unless (defined($css_file)) {
-        my $input_file_name = $file;
+      $css_file_path = $self->Texinfo::Common::locate_include_file($css_file);
+      unless (defined($css_file_path)) {
+        my $css_input_file_name;
         my $encoding = $self->get_conf('COMMAND_LINE_ENCODING');
         if (defined($encoding)) {
-          $input_file_name = decode($encoding, $input_file_name);
+          $css_input_file_name = decode($encoding, $css_file);
+        } else {
+          $css_input_file_name = $css_file;
         }
         $self->converter_document_warn(sprintf(
-               __("CSS file %s not found"), $input_file_name));
+               __("CSS file %s not found"), $css_input_file_name));
         next;
       }
-      unless (open (CSSFILE, $css_file)) {
-        my $css_file_name = $css_file;
+      unless (open (CSSFILE, $css_file_path)) {
+        my $css_file_name = $css_file_path;
         my $encoding = $self->get_conf('COMMAND_LINE_ENCODING');
         if (defined($encoding)) {
           $css_file_name = decode($encoding, $css_file_name);
@@ -9333,9 +9335,9 @@ sub _prepare_css($)
     # binmode($css_file_fh)
     my ($import_lines, $rules_lines);
     ($import_lines, $rules_lines)
-      = $self->_process_css_file($css_file_fh, $css_file);
+      = $self->_process_css_file($css_file_fh, $css_file_path);
     if (!close($css_file_fh)) {
-      my $css_file_name = $css_file;
+      my $css_file_name = $css_file_path;
       my $encoding = $self->get_conf('COMMAND_LINE_ENCODING');
       if (defined($encoding)) {
         $css_file_name = decode($encoding, $css_file_name);
diff --git a/tp/Texinfo/Convert/Text.pm b/tp/Texinfo/Convert/Text.pm
index 2ca74f4586..060771b877 100644
--- a/tp/Texinfo/Convert/Text.pm
+++ b/tp/Texinfo/Convert/Text.pm
@@ -897,10 +897,13 @@ sub output($$)
   #print STDERR "OUTPUT\n";
   my $input_basename;
   if ($document_info and defined($document_info->{'input_file_name'})) {
-    my $input_file_name = $document_info->{'input_file_name'};
+    my $input_file_name_bytes = $document_info->{'input_file_name'};
     my $encoding = $self->{'COMMAND_LINE_ENCODING'};
+    my $input_file_name;
     if (defined($encoding)) {
-      $input_file_name = decode($encoding, $input_file_name);
+      $input_file_name = decode($encoding, $input_file_name_bytes);
+    } else {
+      $input_file_name = $input_file_name_bytes;
     }
     my ($directories, $suffix);
     ($input_basename, $directories, $suffix) = fileparse($input_file_name);
diff --git a/tp/t/test_utils.pl b/tp/t/test_utils.pl
index 743b566193..21592df39b 100644
--- a/tp/t/test_utils.pl
+++ b/tp/t/test_utils.pl
@@ -969,10 +969,9 @@ sub test($$)
   Texinfo::Config::GNUT_reinitialize_init_files();
   my $init_files_options = {};
   my $init_file_directories = [$srcdir.'init/', $srcdir.'t/init/'];
-  # the init file names are supposed to be binary strings.  Since they
-  # are not encoded anywhere, probably only ascii file names should
-  # be used.
-  # FIXME what if srcdir is non ascii (srcdir is truly a binary string).
+  # the init file names should be binary strings.  Since they
+  # are not encoded here, ascii file names should be used or they
+  # should be encoded in test specification files.
   if ($parser_options and $parser_options->{'init_files'}) {
     $symbols_before_init_file = {};
     foreach my $symbol (keys(%Texinfo::Config::)) {
@@ -1022,8 +1021,12 @@ sub test($$)
       }
     }
     if (defined($test_input_file_name)) {
-      # FIXME should we need to encode or do we assume that
-      # $test_input_file_name is already bytes?
+      # argument should be byte strings.  In most if not all cases,
+      # 'test_input_file_name' is based on $test_name.  $test_name should
+      # only consist of ascii characters as it is used both as a
+      # character string and a byte string (see the comment below on that),
+      # so we do not encode, but we could if needed.  If we encode, an
+      # output encoding should be determined.
       $document->set_document_global_info('input_file_name',
                                           $test_input_file_name);
     }
@@ -1436,11 +1439,12 @@ sub test($$)
     print OUT 'use utf8;'."\n\n";
 
     #print STDERR "Generate: ".Data::Dumper->Dump([$tree], ['$res']);
-    # NOTE $test_name is in general used for directories and
-    # file names, here it is used as a text string.  If non ascii, it
+    # NOTE $test_name is in general used for directories, file names,
+    # and in messages.  Here it is used as a text string.  If non ascii, it
     # should be a character string in internal perl codepoints as OUT
-    # is encoded as utf8.  It should also be encoded to be used as file name
-    # in that case.
+    # is encoded as utf8.  In that case, it should be encoded to be
+    # used as a file name for the above cases.  Since this is not the case,
+    # $test_name should consist of ascii characters only.
     my $out_result;
     {
       local $Data::Dumper::Sortkeys = \&filter_tree_keys;
@@ -1577,8 +1581,9 @@ sub test($$)
     ok (Data::Compare::Compare($indices_sorted_sort_strings,
                                $result_indices_sort_strings{$test_name}),
         $test_name.' indices sort');
-    # FIXME use PlainTexinfo converter to test the XS converter until
-    # convert_to_texinfo goes through XS.
+    # NOTE either a PlainTexinfo converter or a direct call to
+    # convert_to_texinfo can be used to test conversion to raw text,
+    # both for pure Perl and XS.
     my $converter_to_texinfo = Texinfo::Convert::PlainTexinfo->converter();
     my $texi_result = $converter_to_texinfo->convert($document);
     #my $texi_result = Texinfo::Convert::Texinfo::convert_to_texinfo($tree);