diff --git a/bin/controller.rc b/bin/controller.rc index d43a654..70cd87c 100755 --- a/bin/controller.rc +++ b/bin/controller.rc @@ -2,6 +2,9 @@ cd .. +forbidden_uri_chars='[^a-zA-Z0-9_+\-\/\.]' +response_format=html + # Useful functions fn dprint { echo $* >[1=2] @@ -52,7 +55,7 @@ fn gentitle { } # Don't change var name or trailing ';', some dirs change the filter! -dirfilter = '/\/[._]/d; /\/robots.\txt$/d; /\/sitemap\.txt$/d; s,^\./,,; s,\.md$,,; s,\.html,,; s,\.txt,,; ' +dirfilter = '/\/[._]/d; /'^$forbidden_uri_chars^'/d; /\/robots.\txt$/d; /\/sitemap\.txt$/d; s,^\./,,; s,\.md$,,; s,\.html,,; s,\.txt,,; ' # To be used from config files fn hide_paths { @@ -197,7 +200,7 @@ fn debug_handler { fn select_handler { -dprint $body + if (test -f $body.md) { if (! ~ $#inBlog 0) set_handler blog_post_handler $body.md @@ -212,14 +215,34 @@ dprint $body if not if (test -f $body.html) set_handler html_handler $body.html - # Handle eplicit .html urls, this should not happen (the web server will usually handle this anyway) + # Handle explicit .html urls, this should not happen (the web server will usually handle this anyway) # XXX We probably should setup a permanent redirect to $body|sed 's/.html$//' here if not if (~ $body *.html && test -f $body) set_handler html_handler $body - # This should probably be merged with the blog_dir_handler - if not if (~ $body */[bB]log/index */[bB]log//index && ~ $#blogDirs 0) - blogDirs = `{basename -d $body} + # Blog handler + if not if (~ $body */[bB]log/index */[bB]log//index || ! ~ $#blogDirs 0) { + if (~ $#blogDirs 0) + blogDirs = `{basename -d $body} + + rssuri=`{basename -d $uri} + rssuri=$baseuri`{cleanname $"rssuri^/index.rss} + extraHeaders=$"extraHeaders ^ \ + ' +' + + set_handler blog_dir_handler $blogDirs + } + + # Rss feeds. TODO: we should check that the request is for a real blog dir + if not if (~ $REQUEST_URI */index.rss) { + uri = `{echo $uri | sed 's/index.rss$//'} + if(~ $blogDirs 0) + blogDirs = $sitedir^$uri + uri=$baseuri$"uri + master_template=feeds/rss20.tpl + response_format=raw + } # Global tpl (eg sitemap.tpl) if not if (test -f lib/^$uri^.tpl) @@ -229,18 +252,16 @@ dprint $body set_handler txt_handler $body.txt # Dir listing - if not if(~ $body */index && ~ $#blogDirs 0) + if not if(~ $body */index) set_handler dir_listing_handler $body # File not found - if not if(~ $#blogDirs 0) { + if not { set_handler 404_handler $body dprint 'NOT FOUND: '$SERVER_NAME^$REQUEST_URI^' - '^$"HTTP_REFERER^' - '^$"HTTP_USER_AGENT echo 'Status: 404 Not Found' } - if(! ~ $#blogDirs 0) - set_handler blog_dir_handler $blogDirs } @@ -268,7 +289,7 @@ if(test -f etc/initrc.local) # Parse request URL -uri = `{echo -n $REQUEST_URI | sed -e 's/\?.*//; s/[^a-zA-Z0-9_+\-\/\.]//g; s/\.\.*/./g;' -e '1q'} +uri = `{echo -n $REQUEST_URI | sed -e 's/\?.*//; s/'^$forbidden_uri_chars^'//g; s/\.\.*/./g;' -e '1q'} ifs='/' { args = `{echo -n $uri} } @@ -325,18 +346,10 @@ if(~ $pageTitle '') if not pageTitle=$"pageTitle^' | '^$"siteTitle^' '^$"siteSubTitle -# Template/body selection -master_template= `{get_lib_file $master_template } - -# DEPRECATED -if(~ $master_template 0) - master_template=$sitedir/_default.tpl - body=$sitedir/$body rssuri=$uri - if (test -d $body) { if(! ~ $body */) perm_redirect $REQUEST_URI^'/' @@ -344,19 +357,18 @@ if (test -d $body) { rssuri=$rssuri/ } -if(! ~ $#blogDirs 0 || ! ~ $#inBlog 0) { - rssuri=`{basename -d $uri} - rssuri=$baseuri`{cleanname $"rssuri^/index.rss} - extraHeaders=$"extraHeaders ^ \ - ' -' -} - - select_handler -fn template { - template.awk | rc $rcargs | +# Template/body selection +master_template= `{get_lib_file $master_template} + +if(~ $master_template 0) { + dprint DEPRECATION WARNING: using $sitedir/_default.tpl, should move to _werc/lib/XXX + master_template=$sitedir/_default.tpl +} + +# Is this really useful? +fn postfil_awk_buf { awk '{ buf = buf $0"\n" if(length(buf) > 8192) { @@ -367,40 +379,20 @@ fn template { END{ printf "%s", buf }' } -if(! ~ $REQUEST_URI */index.rss) { +post_filter=postfil_awk_buf + +# This probably should merge with tpl_handler +fn template { + if(~ $#post_filter 0) + template.awk | rc $rcargs + if not + template.awk | rc $rcargs | $post_filter +} + +if(~ $response_format html) { cat $headers $master_template | template echo '' - exit } - -# RSS -if (~ $body */[bB]log/index.rss */[bB]log//index.rss && ~ $#blogDirs 0) - blogDirs = `{basename -d $body} - - -uri = `{echo $uri | sed 's/index.rss$//'} -uri=$baseuri$"uri - -fn statpost { - f = $1 - uri = `{echo $f | sed 's,^'$sitedir',,'} - title=`{basename $f | sed 's/^[0-9\-]*_(.*)\.md$/\1/; s/_/ /g' } - date=`{/bin/date -Rd `{basename $f |sed 's/(^[0-9\-]*).*/\1/; s/-[0-9]$//'}} - # TODO: use mtime(1) and ls(1) instead of lunix's stat(1) - stat=`{stat -c '%Y %U' $f} - #mdate=`{/bin/date -Rd $stat(1)} # Not used because it is unreliable - uri=$baseuri^`{cleanname `{echo -n $uri | sed 's/\.(md|tpl)//g'}} - by=$stat(2) - ifs=() { - summary=`{awk -v max'='1024 '{ - nc += 1 + length; - if(nc > max) { - print substr($0, 1, nc - max) "..." - exit - } - print - }' $f |fmt -j| sed 's/\]\]>/Fucking goddamn XML garbage/g'} - } +if not if (~ $response_format raw) { + template < $master_template } - -cat lib/feeds/rss20.tpl | template diff --git a/lib/feeds/rss20.tpl b/lib/feeds/rss20.tpl index dde92aa..9c27885 100644 --- a/lib/feeds/rss20.tpl +++ b/lib/feeds/rss20.tpl @@ -1,5 +1,30 @@ Content-Type: text/xml; charset=utf-8 +%{ +fn statpost { + f = $1 + uri = `{echo $f | sed 's,^'$sitedir',,'} + title=`{basename $f | sed 's/^[0-9\-]*_(.*)\.md$/\1/; s/_/ /g' } + date=`{/bin/date -Rd `{basename $f |sed 's/(^[0-9\-]*).*/\1/; s/-[0-9]$//'}} + # TODO: use mtime(1) and ls(1) instead of lunix's stat(1) + stat=`{stat -c '%Y %U' $f} + #mdate=`{/bin/date -Rd $stat(1)} # Not used because it is unreliable + uri=$baseuri^`{cleanname `{echo -n $uri | sed 's/\.(md|tpl)//g'}} + by=$stat(2) + ifs=() { + summary=`{awk -v max'='1024 '{ + nc += 1 + length; + if(nc > max) { + print substr($0, 1, nc - max) "..." + exit + } + print + }' $f |fmt -j| sed 's/\]\]>/Fucking goddamn XML garbage/g'} + } +} + +%} +