werc/bin/werc.rc

243 lines
6.9 KiB
Bash
Executable File

#!/usr/local/plan9/bin/rc
. ./cgilib.rc
cd ..
forbidden_uri_chars='[^a-zA-Z0-9_+\-\/\.]'
response_format=html
fn get_lib_file {
if(test -f $sitedir/_werc/lib/$1)
echo -n $sitedir/_werc/lib/$1
if not if(! ~ $#masterSite 0 && test -f $sitesdir/$masterSite/_werc/lib/$1)
echo -n $sitesdir/$masterSite/_werc/lib/$1
if not if(test -f lib/$1)
echo -n lib/$1
}
# Don't change var name or trailing ';', dirfilter might be changed from _werc/config!
# ls -F style input of the form: $sitedir/path/to/files/
# <ls -F+x><symlink hack><Useless?><hiden files >
dirfilter='s/\*$//; s,/+\./+,/,g; s,^\./,,; /\/[._][^\/]/d; /'^$forbidden_uri_chars^'/d; /^\/(robots|sitemap)\.txt$|\/index\.(md|html|txt|tpl)$/d; /_werc\/?$/d; '
dirclean=' s/\.(md|html|txt)$//; '
# To be used from config files
fn hide_paths {
for(i in $*)
dirfilter=$dirfilter^'/^'$i'$/d; '
}
# Sidebar
fn gensidebar {
# Ignore stderr, last path element might be a file that doesn't exist (eg., foo for foo.md)
# /./ to deal with p9p's ls failure to follow dir symlinks otherwise
ls -F $sitedir/./$req_paths_list >[2]/dev/null \
| sed 's!^'$sitedir'!!; '^$dirfilter^'/\/[^_.\/][^\/]*(\.(md|txt|html)|\/)$/!d; '^$dirclean \
| sort -u | awk -F/ '
function p(x, y, s) { for(i=0; i < x-y; i+=1) print s }
{
d = ""
if(match($0, "/$"))
d = "/"
sub("/$", "") # Strip trailing / for dirs so NF is consistent
p(NF, lNF, "<ul class=\"side-bar\">")
p(lNF, NF, "</ul>")
lNF = NF
bname = $NF d
path = $0 d
gsub("_", " ", bname)
if(index(ENVIRON["req_path"] "/", path) == 1)
print "<li><a href=\"" path "\" class=\"thisPage\">&raquo;<i> " bname "</i></a>"
else
print "<li><a href=\"" path "\">&rsaquo; " bname "</a></li>"
}
END { p(lNF, 0, "</ul>") }'
}
# Handlers
fn set_handler { handler=$* }
fn md_handler { cat $* | $formatter }
fn tpl_handler { template $1 }
fn html_handler {
# body states: 0 = no <body> found, 2 = after <body>, 1 = after <body></body>, -1 = after </body>
awk 'gsub(".*<[Bb][Oo][Dd][Yy][^>]*>", "") > 0 {body=2}
gsub("</ *[Bb][Oo][Dd][Yy][^>]*>.*", "") > 0 {print; body=body-1}
body==2 {print}
body==0 {buf=buf "\n" $0}
END {if(body<=0) {print buf}}' < $1
}
fn txt_handler {
# Note: Words are not broken, even if they are way beyond 82 chars long
echo '<pre>' `{ sed 's/</\&lt;/g; s/>/\&gt;/g' < $1 | fmt -l 82 -j } '</pre>'
}
fn dir_listing_handler {
d=`{basename -d $1}
echo $d|sed 's,.*//,,g; s,/$,,; s,/, / ,g; s,.*,<h1 class="dir-list-head">&</h1> <ul class="dir-list">,'
# Symlinks suck: '/.' forces ls to list the linked dir if $d is a symlink.
ls -F $d/. | sed $dirfilter$dirclean' s,.*/([^/]+/?)$,<li><a href="\1">\1</a></li>,'
echo '</ul>'
}
fn select_handler {
if(test -f $local_path.md)
set_handler md_handler $local_path.md
if not if(test -f $local_path.tpl)
set_handler tpl_handler $local_path.tpl
if not if(test -f $local_path.html)
set_handler html_handler $local_path.html
# Global tpl (eg sitemap.tpl), should take precedence over txt handler!
if not if(test -f lib^$req_path^.tpl)
set_handler tpl_handler lib^$req_path^.tpl
if not if(test -f $local_path.txt)
set_handler txt_handler $local_path.txt
# Apps
if not if(! ~ $#enabled_apps 0 && select_apps)
set_handler app_handler
# Dir listing
if not if(~ $local_path */index)
set_handler dir_listing_handler $local_path
# Explicit .html urls, unfortunately usually the web server will handle this as static files
if not if(~ $local_path *.html && test -f $local_path)
perm_redirect `{ echo $req_path|sed 's/.html$//' }
# Fallback static file handler
if not if(test -f $local_path) {
m='text/plain'
if(~ $req_path *.css)
m='text/css'
if not if(~ $req_path *.ico)
m='image/x-icon'
if not if(~ $req_path *.png)
m='image/png'
if not if(~ $req_path *.jpg *.jpeg)
m='image/jpeg'
if not if(~ $req_path *.gif)
m='image/gif'
if not if(~ $req_path *.pdf)
m='application/pdf'
echo 'Content-Type: '^$m
echo
cat $local_path
exit
}
# File not found
if not {
set_handler tpl_handler `{get_lib_file 404.tpl}
dprint 'NOT FOUND: '$SERVER_NAME^$"REQUEST_URI^' - '^$"HTTP_REFERER^' - '^$"HTTP_USER_AGENT
echo 'Status: 404 Not Found'
}
}
fn genbody { $handler(1) $handler(2-) }
# Careful, the proper p9p path might not be set until initrc.local is sourced
path=(. $PLAN9/bin ./bin/ /bin/ /usr/bin)
headers=lib/headers.tpl
master_template=default_master.tpl
sidebar=sidebar
werc_root=`{pwd}
for(i in siteTitle siteSubTitle pageTitle extraHeaders)
$i = ''
# TODO: Per-req variables should move after initrc loading.
site=$SERVER_NAME
base_url=http://$site/
sitesdir=sites
sitedir=$sitesdir/$site
current_date_time=`{date}
. ./etc/initrc
if(test -f etc/initrc.local)
. ./etc/initrc.local
# Parse request URL
# NOTE: $REQUEST_URI is not officially in CGI 1.1, but seems to be de-facto
req_path=`{echo -n $REQUEST_URI | sed 's/\?.*//; s/'^$forbidden_uri_chars^'//g; s/\.\.*/./g; 1q'}
local_path=$sitedir$req_path
ifs='/' { args=`{echo -n $req_path} }
# Hack: preload post data so we can access it from templates where cgi's stdin is not accesible
if(~ $REQUEST_METHOD POST) {
get_post_args
login_user
}
if(! ~ $#args 0)
pageTitle=`{ echo $args|sed -e 's/ / - /g' -e 's/_/ /g' }
if(~ $req_path */index)
perm_redirect `{echo $req_path | sed 's,/index$,/,'}
if(~ $local_path */)
local_path=$local_path^'index'
if not if(test -d $local_path)
perm_redirect $req_path^'/'
p=()
cd $sitedir
if(test -f _werc/config)
. _werc/config
for(i in ('' $args)) {
p=($"p^/^$i)
req_paths_list=($req_paths_list $p)
if(test -d $i) {
cd $i
if(test -f _werc/config)
. _werc/config
}
}
cd $werc_root
# Redirections and other preprocessing
if(~ $#redirectPermanent 1) {
perm_redirect $"redirectPermanent
}
if not if(~ $#redirectPermanent 2) {
from='http://'^$SERVER_NAME^$req_path
to=`{echo $from|sed 's@'^$redirectPermanent(1)^'@'^$redirectPermanent(2)^'@'}
if(! ~ $to $from)
perm_redirect $to
}
# Set Page title
if(~ $pageTitle '')
pageTitle=$siteTitle^' '^$siteSubTitle
if not
pageTitle=$"pageTitle^' | '^$"siteTitle^' '^$"siteSubTitle
select_handler
if(! ~ $#debug 0)
dprint ' '$"SERVER_NAME^$"REQUEST_URI' - '$"HTTP_USER_AGENT' - '$"REQUEST_METHOD' - '$"handler
# Template/body selection
master_template=`{get_lib_file $master_template}
if(~ $response_format html) {
template $headers $master_template | awk_buffer
echo '</body></html>'
}
if not if(~ $response_format raw)
template < $master_template