Changeset 15152 for gsdl/trunk/bin/script/gsConvert.pl

Timestamp:

2008-03-29T17:53:51+13:00 (16 years ago)

Author:

ak19

Message:

Regular expression in make_links_to_assocdir_relative is corrected and a utility subroutine is added

File:

• gsdl/trunk/bin/script/gsConvert.pl (modified) (8 diffs)

gsdl/trunk/bin/script/gsConvert.pl

@@ -614 +614 @@
         # Inserted this code to remove the images directory if it was still empty after 
         # the html was generated (in case there were no images in the word document)
-        if(is_dir_empty($assoc_dir)) {
-        print STDERR "***gsConvert.pl: Image dir $assoc_dir is empty, removing***\n";
+        if(&is_dir_empty($assoc_dir)) {
+        #print STDERR "***gsConvert.pl: Image dir $assoc_dir is empty, removing***\n";
         &util::rm_r($assoc_dir);
         } else { # there was an image folder (it was generated)
@@ -621 +621 @@
         # If the folder contains images
         # Replace them with relative links instead, so it can be moved elsewhere
-        make_links_to_assocdir_relative($toppath, $docname, "$output_filestem.html", $assoc_dir, $docname."_files");    
+        &make_links_to_assocdir_relative($toppath, $docname, "$output_filestem.html", $assoc_dir, $docname."_files");   
         }
         return 1;
@@ -640 +640 @@
     
     return 0;
+}
+
+
+# A method to check if a directory is empty (note that an empty directory still has non-zero size!!!) 
+# Code is from http://episteme.arstechnica.com/eve/forums/a/tpc/f/6330927813/m/436007700831
+sub is_dir_empty
+{
+    my ($path) = @_;
+    opendir DIR, $path;
+    while(my $entry = readdir DIR) {
+        next if($entry =~ /^\.\.?$/);
+        closedir DIR;
+        return 0;
+    }
+    closedir DIR;
+    return 1;
 }
 
@@ -661 +677 @@
     unless(open(FIN, "<$html_file")) { 
     print STDERR "gsConvert.pl: Unable to open $html_file for reading absolute urls...ERROR\n";
-    return;
+    return 0;
     }
     # From http://perl.plover.com/local.html
@@ -667 +683 @@
     # (Some people call this slurping the file.) Perl has a special feature to support this: 
     # If the $/ variable is undefined, the <...> operator will read the entire file all at once"
-    $/ = undef;                 # Read entire file at once
-    my $html_contents = <FIN>;  # Now file is read in as one single 'line'
+    my $html_contents;
+    {
+    local $/ = undef;        # Read entire file at once
+    $html_contents = <FIN>;  # Now file is read in as one single 'line'
+    }
     close(FIN); # close the file
-    print STDERR $html_contents;
+    #print STDERR $html_contents;
    
     # 2. Replace (substitute) *all* ocurrences of the assoc_dir_path in a hrefs and img src
@@ -677 +696 @@
     # all new lines as a regular space. This interacts with g to consider all the lines 
     # together as a single line so that multi-occurrences can be replaced.
-    
+
+    # we can't just replace $assoc_dir_path with $assoc_dir
+    # $assoc_dir_path represents a regular expression that needs to be replaced
+    # if it contains ., -, [ or ] -- which all have special meaning in Perl regular expressions --
+    # we need to escape these first
+    my $safe_reg_expression = $assoc_dir_path;
+    $safe_reg_expression =~ s/\./\\./g;
+    $safe_reg_expression =~ s/\-/\\-/g;
+    $safe_reg_expression =~ s/\[/\\[/g;
+    $safe_reg_expression =~ s/\]/\\]/g;
+    $safe_reg_expression =~ s/ /%20/g; # wvWare put %20 in place of space, so we need to change our prefix to match
+
     # The following regular expression substitution looks for <a or <image, followed by any other 
     # attributes and values until it comes to the FIRST (indicated by ?) href= or src= 
@@ -689 +719 @@
     # and performs a global replace (g) meaning that all occurrences that match in that single line
     # are substituted.
-    $html_contents =~ s/(<(a|img).*?(href|src)=(\"|\')?)$assoc_dir_path(.*?(\"|\')?.*?>)/$1$assoc_dirname$5/sg;
-    #$html_contents =~ s/$assoc_dir_path/$assoc_dirname/gs; # this works, used as fall-back
-
+    $html_contents =~ s/(<(a|img).*?(href|src)=(\"|\')?)$safe_reg_expression(.*?(\"|\')?.*?>)/$1$assoc_dirname$5/sg;
+               #$html_contents =~ s/$safe_reg_expression/$assoc_dirname/gs; # this works, used as fall-back
+    # now replace any %20 chars in filenames of href or src attributes to use literal space ' '. Calls a function for this
+    $html_contents =~ s/(<(a|img).*?(href|src)=(\"|\')?)(.*)(.*?(\"|\')?.*?>)/&percent_twenty_to_space($1, $5, $6)/sge;
+   
+    #print STDERR "assoc_dirname: ****$assoc_dirname***\n";
+    #print STDERR "safe_reg_expression: ****$safe_reg_expression***\n";
+   
     # delete the original file and recreate it
     my $copy_of_filename = $html_file;
@@ -699 +734 @@
     unless(open(FOUT, ">$html_file")) {  # open it as a new file for writing
     print STDERR "gsConvert.pl: Unable to open $html_file for writing relative links...ERROR\n";
-    return;
+    return 0;
     }
     # write out the updated contents and close the file
     print FOUT $html_contents;
     close(FOUT);
-   
-}
-
-# A method to check if directory is empty (note that an empty directory still has non-zero size!!!) 
-# Code is from http://episteme.arstechnica.com/eve/forums/a/tpc/f/6330927813/m/436007700831
-sub is_dir_empty
+    return 1;
+}
+
+# Utility routine to convert all %20 introduced by wvWare in link pathnames into space again
+sub percent_twenty_to_space
 {
-    my ($path) = @_;
-    opendir DIR, $path;
-    while(my $entry = readdir DIR) {
-        next if($entry =~ /^\.\.?$/);
-        closedir DIR;
-        return 0;
-    }
-    closedir DIR;
-    return 1;
-}
-
+    my ($pre, $text, $post) = @_;
+
+    $text =~ s/%20/ /g;
+
+    return "$pre$text$post";
+}
 
 # Attempt to convert a word document to html with the word2html scripting program