From 2d8ce103893a1ef8371c3c33588469e55b6d42b4 Mon Sep 17 00:00:00 2001
From: Uriel <u@berlinblue.org>
Date: Tue, 28 Jul 2009 20:28:35 +0200
Subject: [PATCH] New context var 'local_file' that will point to the
 on-disk-file that represents the main content of a given page request (if
 such exists, otherwise it is '' or ()).

Move title extracting functions from sitemap.tpl to werclib.rc.
Rewrite .md title extraction code to only match first lines when 'underlines' with =.

Use the changes above to generate better page titles when possible.
---
 bin/corehandlers.rc | 25 +++++++++++++++++--------
 bin/werc.rc         | 17 ++++++++++++++---
 bin/werclib.rc      | 30 ++++++++++++++++++++++++++++++
 tpl/sitemap.tpl     | 29 -----------------------------
 4 files changed, 61 insertions(+), 40 deletions(-)

diff --git a/bin/corehandlers.rc b/bin/corehandlers.rc
index e6ea61d..f1052ac 100644
--- a/bin/corehandlers.rc
+++ b/bin/corehandlers.rc
@@ -85,17 +85,26 @@ fn notices_handler {
 
 fn setup_handlers {
 
-    if(test -f $local_path.md)
-        handler_body_main=(md_handler $local_path.md)
-    if not if(test -f $local_path.tpl)
-        handler_body_main=(tpl_handler $local_path.tpl)
-    if not if(test -f $local_path.html)
-        handler_body_main=(html_handler $local_path.html)
+    if(test -f $local_path.md) {
+        local_file=$local_path.md
+        handler_body_main=(md_handler $local_file)
+    }
+    if not if(test -f $local_path.tpl) {
+        local_file=$local_path.tpl
+        handler_body_main=(tpl_handler $local_file)
+    }
+    if not if(test -f $local_path.html) {
+        local_file=$local_path.html
+        handler_body_main=(html_handler $local_file)
+    }
     # Global tpl (eg sitemap.tpl), should take precedence over txt handler!
     if not if(test -f tpl^$req_path^.tpl)
+        # XXX Should we set $local_file for global .tpls?
         handler_body_main=(tpl_handler tpl^$req_path^.tpl)
-    if not if(test -f $local_path.txt)
-        handler_body_main=(txt_handler $local_path.txt)
+    if not if(test -f $local_path.txt) {
+        local_file=$local_path.txt
+        handler_body_main=(txt_handler $local_file)
+    }
 
     # XXX Should check that $enabled_apps exist in $werc_apps?
     # XXX Should split init of apps that provide main handler (eg., blog) and apps that don't (eg., comments)?
diff --git a/bin/werc.rc b/bin/werc.rc
index 2c6aa88..28f63b6 100755
--- a/bin/werc.rc
+++ b/bin/werc.rc
@@ -46,6 +46,7 @@ fn werc_exec_request {
     req_path=`{echo -n $REQUEST_URI | sed 's/\?.*//; s!//+!/!g; s/%5[Ff]/_/g; s/'^$forbidden_uri_chars^'//g; s/\.\.*/./g; 1q'}
     req_url=$base_url^$req_path
     local_path=$sitedir$req_path
+    local_file=''
     ifs='/' { args=`{echo -n $req_path} }
 
     # Preload post args for templates where cgi's stdin is not accessible
@@ -103,14 +104,24 @@ fn werc_exec_request {
             perm_redirect $t
     }
 
+    setup_handlers
+
+
     # Set Page title
-    if(~ $"pageTitle '')
+    if(! ~ $local_file '') {
+        t=`{get_file_title $local_file}
+        if(! ~ $"t '')
+            pageTitle=$t
+    }
+
+    dprint X^$"pageTitle^X
+
+    # XXX Is this never true? because we set pageTitle earlier based on url.
+    if(~ $"pageTitle '') 
         pageTitle=$"siteTitle' '$"siteSubTitle
     if not
         pageTitle=$"pageTitle' | '$"siteTitle' '$"siteSubTitle
 
-    setup_handlers
-
     for(h in $extraHttpHeaders)
         echo $h
     echo 'Content-Type: '^$http_content_type
diff --git a/bin/werclib.rc b/bin/werclib.rc
index 2367dbe..0c43417 100644
--- a/bin/werclib.rc
+++ b/bin/werclib.rc
@@ -90,6 +90,36 @@ fn get_md_file_attr {
     sed -n '/^\* '$2': /p; /^\* '$2': /q; /^$/q' < $1
 }
 
+# File title extraction
+
+fn get_md_title {
+    #sed 's/^(................................................................[^ ]*).*$/\1/g; 1q' < $1 
+    sed -n -e '1N; /^.*\n===*$/N; /.*\n===*\n$/!b' -e 's/\n==*\n//p' < $1
+}
+
+fn get_html_title {
+    # H1 is not reliable because htmlroff doesn't use it :(
+    #desc=`{cat $1 | sed 32q | grep '<[Hh]1>' | sed 's/<[Hh]1>(.*)(<\/[Hh]1>|$)/\1/;s/<[^>]*>//g;1q'}
+    # Pick the first line of body  instead
+    desc=`{sed -n '/<[Bb][Oo][Dd][Yy]/,/./s/(<[^>]*>|$)//gp' < $1}
+    if(~ $#desc 0)
+        desc=`{sed 's/<[^>]*>//g; 1q' < $1}
+}
+
+fn get_file_title {
+    if (~ $1 *.md)
+        get_md_title $1
+    if not if(~ $1 *.html)
+        get_html_title $1
+    if not if(~ $1 */) {
+        if(test -f $1/index.md)
+            get_md_title $1/index.md
+        if not if(test -f $1/index.html)
+            get_html_title $1/index.html
+    }
+}
+
+
 ##########################################################################
 ##########################################################################
 #app_blog_methods = ( _post index.rss )
diff --git a/tpl/sitemap.tpl b/tpl/sitemap.tpl
index 8f0feeb..334f7e5 100644
--- a/tpl/sitemap.tpl
+++ b/tpl/sitemap.tpl
@@ -4,35 +4,6 @@
 tmpfile=/tmp/werc_sitemap_$pid.txt
 saveddf=$dirfilter
 
-fn get_md_title {
-    sed 's/^(................................................................[^ ]*).*$/\1/g; 1q' < $1 
-}
-
-fn get_html_title {
-    # H1 is not reliable because htmlroff doesn't use it :(
-    #desc=`{cat $1 | sed 32q | grep '<[Hh]1>' | sed 's/<[Hh]1>(.*)(<\/[Hh]1>|$)/\1/;s/<[^>]*>//g;1q'}
-    # Pick the first line of body  instead
-    desc=`{sed -n '/<[Bb][Oo][Dd][Yy]/,/./s/(<[^>]*>|$)//gp' < $1}
-    if(~ $#desc 0)
-        desc=`{sed 's/<[^>]*>//g; 1q' < $1}
-}
-
-fn get_file_title {
-        
-    if(~ $1 */) {
-        if(test -f $1/index.md)
-            get_md_title $1/index.md
-        if not if(test -f $1/index.html)
-            get_html_title $1/index.html
-    }
-    if not if(~ $1 *.md)
-        get_md_title $1
-    if not if(~ $1 *.html)
-        get_html_title $1
-    if not
-        echo ''
-}
-
 fn listDir {
     d=$1
     dirfilter=$saveddf