From 8c1b0d1c11d754250c87df3a54cfd5343b743ac9 Mon Sep 17 00:00:00 2001
From: =?utf8?q?Paul=20H=C3=A4nsch?= <paul@plutz.net>
Date: Wed, 7 Feb 2024 13:56:44 +0100
Subject: [PATCH] prevent double escape in page title

---
 tools.sh | 16 +++++++++++++---
 1 file changed, 13 insertions(+), 3 deletions(-)

diff --git a/tools.sh b/tools.sh
index a85e63c..2ee1818 100755
--- a/tools.sh
+++ b/tools.sh
@@ -193,23 +193,33 @@ page_title() {
     return 1
   fi
   PAGE_TITLE="$(
+    # pick title from %title pragma
     sed -nE '
       s;^%title[ \t]+([[:graph:]][[:print:]]+)\r?$;\1;p; tQ;
       b; :Q q;
     ' "$mdfile"
   )"
   [ ! "${PAGE_TITLE}" ] && PAGE_TITLE="$(
+    # pick title from first h1/h2 headline
     MD_MACROS="" md <"$mdfile" \
     | sed -nE '
-      s;^.*<h1[^>]*>(.*>)?([^<]+)(<.*)?</h1>.*$;\2;p; tQ;
-      s;^.*<h2[^>]*>(.*>)?([^<]+)(<.*)?</h2>.*$;\2;p; tQ;
-      b; :Q q;
+      s;^.*<h1[^>]*>(.*>)?([^<]+)(<.*)?</h1>.*$;\2;; tQ;
+      s;^.*<h2[^>]*>(.*>)?([^<]+)(<.*)?</h2>.*$;\2;; tQ;
+      b; :Q
+      # reverse escapes of cgilite HTML function,
+      # to prevent later double escaping
+      # later escaping must not be omited
+      s/&lt;/</g; s/&gt;/>/g;  s/&quot;/'\"'/g; s/&#x27;/'\''/g;
+      s/&#x5B;/[/g; s/&#x5D;/]/g;  s/&#x0D;/\r/g; s/&amp;/\&/g;
+      p; q;
     '
   )"
   if [ ! "${PAGE_TITLE}" ]; then
+    # use last part of page URL as title
     PAGE_TITLE="${1:-${PATH_INFO}}"
     PAGE_TITLE="${PAGE_TITLE%/*}"
     PAGE_TITLE="${PAGE_TITLE##*/}"
   fi
+  debug "TITLE: $PAGE_TITLE"
   printf %s\\n "$PAGE_TITLE"
 }
-- 
2.39.2