From: Nigel Metheringham <nigel.metheringham@dev.intechnology.co.uk>
Date: Fri, 24 Sep 2010 15:49:15 +0000 (+0100)
Subject: Initial script to convert docs to text
X-Git-Tag: exim-4_89_1~88
X-Git-Url: https://git.exim.org/exim-website.git/commitdiff_plain/c6d3a19107d0dc596f04ba74ff18a6988b000441

Initial script to convert docs to text
---

diff --git a/script/eximhtml2txt.pl b/script/eximhtml2txt.pl
new file mode 100755
index 0000000..02a5b49
--- /dev/null
+++ b/script/eximhtml2txt.pl
@@ -0,0 +1,29 @@
+#!/usr/bin/perl
+#
+use strict;
+use warnings;
+
+use HTML::FormatText;
+use HTML::TreeBuilder;
+
+sub process_chapter {
+    my $fn = shift;
+
+    my $tree = HTML::TreeBuilder->new->parse_file($fn);
+    my ($chapter) = $tree->look_down( "_tag", "div", "class", "chapter", );
+    return '' unless ($chapter);
+    my $formatter = HTML::FormatText->new( leftmargin => 0, rightmargin => 72 );
+
+    my $text = $formatter->format($chapter);
+    $tree->delete;
+    return $text;
+}
+
+my $dir = shift;
+foreach my $fn ( glob("$dir/ch*.html") ) {
+    print "=" x 72, "\n";
+    print $fn, "\n";
+    print "=" x 72, "\n";
+    print process_chapter($fn);
+    print "-" x 72, "\n";
+}