X-Git-Url: http://www.git.stargrave.org/?a=blobdiff_plain;f=Documentation%2Ftxt2pre;h=3ecd9100815dc69eb227444d988ed9dffe755ef1;hb=f349c77a421ef1082eb0b22e8f71304df610fa6b;hp=f84f9c0685ece45da6dcdc555ce343cc43ce7e08;hpb=6a414a4087a59ad8c62cbef30984632ea31ced23;p=public-inbox.git

diff --git a/Documentation/txt2pre b/Documentation/txt2pre
index f84f9c06..3ecd9100 100755
--- a/Documentation/txt2pre
+++ b/Documentation/txt2pre
@@ -1,26 +1,179 @@
 #!/usr/bin/env perl
-# Copyright (C) 2014, Eric Wong <normalperson@yhbt.net> and all contributors
-# License: AGPLv3 or later (https://www.gnu.org/licenses/agpl-3.0.txt)
+# Copyright (C) 2014-2021 all contributors <meta@public-inbox.org>
+# License: AGPL-3.0+ <https://www.gnu.org/licenses/agpl-3.0.txt>
 #
 # Stupid script to make HTML from preformatted, utf-8 text versions,
 # only generating links for http(s).  Markdown does too much
 # and requires indentation to output preformatted text.
 use strict;
 use warnings;
-use CGI qw/escapeHTML/;
-use Encode qw/encode/;
-my $str = eval { local $/; <> };
-$str = escapeHTML($str);
-$str = encode('us-ascii', $str, Encode::HTMLCREF);
+use PublicInbox::Linkify;
+use PublicInbox::Hval qw(ascii_html);
+my %xurls;
+for (qw[lei(1)
+	lei-add-external(1)
+	lei-add-watch(1)
+	lei-blob(1)
+	lei-config(1)
+	lei-convert(1)
+	lei-daemon(8)
+	lei-daemon-kill(1)
+	lei-daemon-pid(1)
+	lei-edit-search(1)
+	lei-export-kw(1)
+	lei-forget-external(1)
+	lei-forget-mail-sync(1)
+	lei-forget-search(1)
+	lei-import(1)
+	lei-index(1)
+	lei-init(1)
+	lei-inspect(1)
+	lei-lcat(1)
+	lei-ls-external(1)
+	lei-ls-label(1)
+	lei-ls-mail-source(1)
+	lei-ls-mail-sync(1)
+	lei-ls-search(1)
+	lei-ls-watch(1)
+	lei-mail-diff(1)
+	lei-mail-sync-overview(7)
+	lei-overview(7)
+	lei-p2q(1)
+	lei-q(1)
+	lei-rediff(1)
+	lei-refresh-mail-sync(1)
+	lei-rm(1)
+	lei-rm-watch(1)
+	lei-security(7)
+	lei-store-format(5)
+	lei-tag(1)
+	lei-up(1)
+	public-inbox.cgi(1)
+	public-inbox-clone(1)
+	public-inbox-config(5)
+	public-inbox-config(5)
+	public-inbox-convert(1)
+	public-inbox-daemon(8)
+	public-inbox-edit(1)
+	public-inbox-fetch(1)
+	public-inbox-glossary(7)
+	public-inbox-httpd(1)
+	public-inbox-imapd(1)
+	public-inbox-index(1)
+	public-inbox-init(1)
+	public-inbox-learn(1)
+	public-inbox-mda(1)
+	public-inbox-nntpd(1)
+	public-inbox-overview(7)
+	public-inbox-purge(1)
+	public-inbox-v1-format(5)
+	public-inbox-v2-format(5)
+	public-inbox-watch(1)
+	public-inbox-xcpdb(1)
+]) {
+	my ($n) = (/([\w\-\.]+)/);
+	$xurls{$_} = "$n.html";
+	$xurls{$n} = "$n.html";
+}
+
+for (qw[make(1) flock(2) setrlimit(2) vfork(2) tmpfs(5) inotify(7) unix(7)
+		syslog(3)]) {
+	my ($n, $s) = (/([\w\-]+)\((\d)\)/);
+	$xurls{$_} = "https://www.man7.org/linux/man-pages/man$s/$n.$s.html";
+}
+
+for (qw[git(1)
+	git-am(1)
+	git-apply(1)
+	git-config(1)
+	git-credential(1)
+	git-daemon(1)
+	git-diff(1)
+	git-fast-import(1)
+	git-fetch(1)
+	git-filter-branch(1)
+	git-format-patch(1)
+	git-gc(1)
+	git-http-backend(1)
+	git-imap-send(1)
+	git-init(1)
+	git-send-email(1)
+	gitrepository-layout(5)
+	gitglossary(7)
+]) {
+	my ($n) = (/([\w\-\.]+)/);
+	$xurls{$_} = "https://kernel.org/pub/software/scm/git/docs/$n.html"
+}
+
+for (qw[
+	sd_listen_fds(3)
+	systemd(1)
+	systemd.unit(5)
+	systemd.socket(5)
+]) {
+	my ($n) = (/([\w\-\.]+)/);
+	$xurls{$_} = "https://www.freedesktop.org/software/systemd/man/$n.html";
+}
+
+# favor upstream docs if they exist, use manpages.debian.org if they don't
+$xurls{'netrc(5)'} = 'https://manpages.debian.org/stable/ftp/netrc.5.en.html';
+$xurls{'mbsync(1)'} =
+	'https://manpages.debian.org/stable/isync/mbsync.1.en.html';
+$xurls{'offlineimap(1)'} =
+	'https://manpages.debian.org/stable/offlineimap/offlineimap.1.en.html';
+$xurls{'spamc(1)'} =
+	'https://spamassassin.apache.org/full/3.4.x/doc/spamc.html';
+$xurls{'grok-pull'} =
+	'https://git.kernel.org/pub/scm/utils/grokmirror/grokmirror.git' .
+	'/tree/man/grok-pull.1.rst';
+$xurls{'git-filter-repo(1)'} = 'https://github.com/newren/git-filter-repo'.
+			'/blob/master/Documentation/git-filter-repo.txt';
+$xurls{'ssoma(1)'} = 'https://ssoma.public-inbox.org/ssoma.txt';
+$xurls{'cgitrc(5)'} = 'https://git.zx2c4.com/cgit/tree/cgitrc.5.txt';
+$xurls{'prove(1)'} = 'https://perldoc.perl.org/prove.html';
+$xurls{'mbox(5)'} = 'https://manpages.debian.org/stable/mutt/mbox.5.en.html';
+$xurls{'mmdf(5)'} = 'https://manpages.debian.org/stable/mutt/mmdf.5.en.html';
+$xurls{'mutt(1)'} = 'https://manpages.debian.org/stable/mutt/mutt.1.en.html';
+$xurls{'torsocks(1)'} =
+	'https://manpages.debian.org/stable/torsocks/torsocks.1.en.html';
+$xurls{'curl(1)'} = 'https://manpages.debian.org/stable/curl/curl.1.en.html';
+$xurls{'copydatabase(1)'} =
+ 'https://manpages.debian.org/stable/xapian-tools/copydatabase.1.en.html';
+$xurls{'xapian-compact(1)'} =
+ 'https://manpages.debian.org/stable/xapian-tools/xapian-compact.1.en.html';
+$xurls{'gzip(1)'} = 'https://manpages.debian.org/stable/gzip/gzip.1.en.html';
+$xurls{'chmod(1)'} =
+	'https://manpages.debian.org/stable/coreutils/chmod.1.en.html';
+$xurls{'kqueue(2)'} =
+	'https://www.freebsd.org/cgi/man.cgi?query=kqueue&sektion=2';
+$xurls{'notmuch(1)'} = 'https://notmuchmail.org/manpages/notmuch-1/';
+$xurls{'mairix(1)'} =
+	'https://manpages.debian.org/stable/mairix/mairix.1.en.html';
+
+my $str = do { local $/; <STDIN> };
 my ($title) = ($str =~ /\A([^\n]+)/);
+if ($str =~ /^NAME\n\s+([^\n]+)/sm) {
+	# don't link to ourselves
+	$title = $1;
+	if ($title =~ /([\w\.\-]+)/) {
+		delete $xurls{$1};
+	}
+}
+$title = ascii_html($title);
+my $l = PublicInbox::Linkify->new;
+$str = $l->linkify_1($str);
+$str = ascii_html($str);
+
+# longest matches, first
+my @keys = sort { length($b) <=> length($a) } keys %xurls;
+my $xkeys = join('|', map { quotemeta } @keys);
+$str =~ s,(?<![>\w_])($xkeys)(?!(?:[\w<\-]|\.html)),
+	qq(<a\nhref=").$xurls{$1}.qq(">$1).($2//'').'</a>',sge;
 
-# temporarily swap &gt; for escape so our s!! to add href works.
-# there's probably a way to do this with only a single s!! ...
-$str =~ s!&gt;!\e!g;
-$str =~ s!\b(https?://[\w+\+\&\?\.\%\;/-]+)!<a\nhref="$1"\n>$1</a>!g;
-$str =~ s!\e!&gt;!g; # swap escapes back to &gt;
+$str = $l->linkify_2($str);
 
 print '<html><head>',
-  '<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />',
+  qq(<meta\nhttp-equiv="Content-Type"\ncontent="text/html; charset=utf-8"\n/>),
   "<title>$title</title>",
-  "</head><body>\n<pre>",  $str , '</pre></body></html>';
+  "</head><body><pre>",  $str , '</pre></body></html>';
+STDOUT->flush;