Minor tidy to test script, and update the README for some omitted

[exim.git] / doc / doc-docbook / TidyHTML-spec

diff --git a/doc/doc-docbook/TidyHTML-spec b/doc/doc-docbook/TidyHTML-spec
index 8459bcf17757eea43d2172df6d65acf58e08b998..4e7b8a2cab43a4a4cb619acf216501b357f19975 100755 (executable)
--- a/doc/doc-docbook/TidyHTML-spec
+++ b/doc/doc-docbook/TidyHTML-spec
@@ -1,6 +1,6 @@
 #! /usr/bin/perl
 
-# $Cambridge: exim/doc/doc-docbook/TidyHTML-spec,v 1.1 2005/06/16 10:32:31 ph10 Exp $
+# $Cambridge: exim/doc/doc-docbook/TidyHTML-spec,v 1.4 2006/04/04 14:03:49 ph10 Exp $
 
 # Script to tidy up the spec HTML files that are generated by xmlto. The
 # following changes are made:
@@ -9,8 +9,13 @@
 # 2. Create reverse links from chapter and section titles back to the TOC.
 # 3. Tidy the ix01.html file - the actual index - by splitting long lines.
 # 4. Insert links from the letter divisions to the top of the Index.
+# 5. Turn <div class="literallayout"><p> into <div class="literallayout"> and
+#    a matching </p></div> into </div> to get rid of unwanted vertical white
+#    space.
+# 6. Before each occurrence of </td> insert &nbsp; so that the table's cell
+#    is a little bit wider than the text itself.
 
-chdir "spec.html";
+chdir "spec_html";
 
 $tocref = 1;
 
@@ -20,11 +25,23 @@ open(IN, "index.html") || die "Failed to open index.html for reading: $!\n";
 @toc = <IN>;
 close(IN);
 
-# Insert a newline after every > because the whole toc is generated as one
-# humungous line that is hard to check. Then split the lines so that each one
-# is a separate element in the vector.
+# Insert a newline after every > except when it is preceded by 'class="quote"',
+# because the whole toc is generated as one humungous line that is hard to
+# check. We have to avoid it in the quote case because that puts a space into
+# the output, and similarly for the </span> the comes afterwards. Easy way out
+# is just not to do it for all </span> occurrences. Unfortunately, Perl does
+# not implement lookbehinds where the alternatives are of different lengths, so
+# we have to take two passes.
+
+
+foreach $line (@toc)
+  {
+  $line =~ s/(?<!class="quote")>\s*/>\n/g;
+  $line =~ s/<\/span>\n/<\/span>/g;
+  }
+
+# Split the lines so that each one is a separate element in the vector.
 
-foreach $line (@toc) { $line =~ s/>\s*/>\n/g; }
 for ($i = 0; $i < scalar(@toc); $i++)
   { splice @toc, $i, 1, (split /(?<=\n)/, $toc[$i]); }
 
@@ -52,7 +69,10 @@ open (OUT, ">index.html") || die "Failed to open index.html for writing: $!\n";
 print OUT @toc;
 close(OUT);
 
-# Now scan each of the other page files and insert the reverse links.
+# Now scan each of the other page files and insert the reverse links. While
+# we are at it, we tidy up <div class="literallayout"> by removing unwanted
+# paragraph marks, which generate unwanted vertical space. We also insert
+# &nbsp; before </td> to push table cells apart from each other.
 
 foreach $file (@chlist)
   {
@@ -60,9 +80,26 @@ foreach $file (@chlist)
   @text = <IN>;
   close(IN);
 
+  # Insert a newline after certain elements, and split the lines so that each
+  # one is a separate element in the vector. This makes it easier to recognize
+  # these elements.
+
   foreach $line (@text)
     {
-    if ($line =~ /^(.*?)<a( xmlns="[^"]+")? id="([^"]+)"><\/a>(.+?)<\/h(.*)$/)
+    $line =~ s/<p>\s*(?!\n)/<p>\n/g;
+    $line =~ s/<\/p>\s*(?!\n)/<\/p>\n/g;
+    $line =~ s/<\/div>\s*(?!\n)/<\/div>\n/g;
+    $line =~ s/<div([^>]*)>(?!\n)/<div$1>\n/g;
+    }
+
+  for ($i = 0; $i < scalar(@text); $i++)
+    { splice @text, $i, 1, (split /(?<=\n)/, $text[$i]); }
+
+  $thisdiv = 0;
+
+  for ($i = 0; $i < scalar(@text); $i++)
+    {
+    if ($text[$i] =~ /^(.*?)<a( xmlns="[^"]+")? id="([^"]+)"><\/a>(.+?)<\/h(.*)$/)
       {
       my($pre, $opt, $id, $title, $post) = ($1, $2, $3, $4, $5);
 
@@ -73,7 +110,22 @@ foreach $file (@chlist)
       $ref = $backref{"$file"} if !defined $ref;
 
       # Adjust the line
-      $line = "$pre<a$opt href=\"index.html#$ref\" id=\"$id\">$title</a></h$post";
+      $text[$i]= "$pre<a$opt href=\"index.html#$ref\" id=\"$id\">$title</a></h$post";
+      }
+
+    elsif ($text[$i] =~ /^<div [^>]*?class="literallayout">$/ && $text[$i+1] eq "<p>\n")
+      {
+      $text[++$i] = "";
+      $thisdiv = 1;
+      }
+    elsif ($thisdiv && $text[$i] eq "</p>\n" && $text[$i+1] eq "</div>\n")
+      {
+      $text[$i] = "";
+      $thisdiv = 0;
+      }
+    elsif ($text[$i] =~ /^\s*<\/td>/)
+      {
+      $text[$i] = "&nbsp;$text[$i]";
       }
     }