+ advanced URL routing

README.md

@@ -1,11 +1,9 @@
 # HTTP.sh
 Node.js, but `| sed s/Node/HTTP/;s/js/sh/`.
 
-Launch with `./http.sh`. Does not need root priviliges ~~- in fact, DO NOT run it as superuser~~ unless you're running it on ports lower than 1024. If you're running on 80 and 443, superuser is more or less mandatory, but THIS MAY BE UNSAFE.
+HTTP.sh is (by far) the most extensible attempt at creating a web framework in Bash, and (AFAIK) the only one that's actively maintained. Although I strive for code quality, this is still rather experimental and may contain bugs.
 
-To prevent running malicious scripts, by default only scripts with extension `.shs` can be run by the server, but this can be changed in the config. Also, cfg[index] ignores this. SHS stands for Shell Server.
+Originally made for Junction Stupidhack 2020; Created by [sdomi](https://sakamoto.pl/), [ptrcnull](https://ptrcnull.me/) and [selfisekai](https://selfisekai.rocks/).
-
-Originally made for Junction Stupidhack 2020; Created by [sdomi](https://sakamoto.pl/), [selfisekai](https://selfisekai.rocks/) and [ptrcnull](https://ptrcnull.me/).
 
 ## Quick Start
 
@@ -25,8 +23,9 @@ git clone https://git.sakamoto.pl/laudom/ocw/ app # example repo :P
 
 ## Dependencies
 
-- Bash or 100% compatible shell
+- Bash (4.x should work, but we'll need 5.0 soon)
-- [Ncat](https://nmap.org/ncat)
+- [Ncat](https://nmap.org/ncat), not openbsd-nc, not netcat, not nc
+- socat (because the above is slightly broken)
 - pkill
 - mktemp
 - jq (probably not needed just yet, but it will be in 1.0)
@@ -38,17 +37,19 @@ git clone https://git.sakamoto.pl/laudom/ocw/ app # example repo :P
 
 - can't change the HTTP status code from Shell Server scripts. This could theoretically be done with custom vhost configs and some `if` statements, but this would be a rather nasty solution to that problem.
 - `$post_multipart` doesn't keep original names - could be fixed by parsing individual headers from the multipart request instead of skipping them all
+- it won't ever throw a 500, thus it fails silently
 
-## Directory structure (incomplete)
+## Directory structure
-- app (or any other namespace name! - check cfg[namespace])
+- ${cfg[namespace]} (`app` by default)
-	- webroot/ - place your files **here**
+	- ${cfg[root]} (`webroot` by default) - public application root
-	- workers/ - workers live here
+	- workers/ - scripts that execute periodically live there (see examples)
-	- config.sh
+	- views/ - for use with HTTP.sh router
+	- config.sh - application-level config file
 - config
-	- master.sh - main config file, loaded with every request
+	- master.sh - main server config file - loaded on boot and with every request
-	- localhost:1337 - example vhost file, loaded if `Host: ...` equals its name
+	- host:port - if a file matching the Host header is found, HTTP.sh will load it request-wide
 - src
-	- server source files and modules (e.g. `ws.sh`)
+	- server source files and modules
 	- response
 		- files corresponding to specific HTTP status codes
 		- listing.sh (code 210) is actually HTTP 200, but triggered in a directory with autoindex turned on and without a valid `index.shs` file
@@ -60,8 +61,38 @@ git clone https://git.sakamoto.pl/laudom/ocw/ app # example repo :P
 
 ![](https://f.sakamoto.pl/d6584c01-1c48-42b9-935b-d9a89af4e071file_101.jpg)
 
-- $post_data - array, contains data from urlencoded POSTs
+- get_data - holds data from GET parameters
-- $post_multipart - array, contains URIs to uploaded files from multipart/form-data POSTs
+	- /?test=asdf -> `${get_data[test]}` == `"asdf"`
-- $get_data - array, contains data from GETs
+- params - holds parsed data from URL router
-- $cfg - array, contains config values (from master.sh and vhost configs)
+	- /profile/test (assuming profile/:name) -> `${params[name]}` == `"test"` 
-- $r - array, contains data generated from the request - URI, URL and that kinda stuff.
+- post_data - same as above, but for urlencoded POST params
+	- test=asdf -> `${post_data[test]}` == `"asdf"`
+- post_multipart - contains paths to uploaded files from multipart/form-data POST requests. **WARNING**: it doesn't hold field names yet, it relies on upload order for identification
+	- first file (in upload order) -> `cat ${post_multipart[0]}`
+	- second file -> `cat ${post_multipart[1]}`
+- r - misc request data
+	- authorization
+	- content_boundary
+	- content_boundary
+	- content_length
+	- content_type
+	- headers
+	- host
+	- host_portless
+	- ip
+	- post
+	- proto
+	- status
+	- uri
+	- url
+	- user_agent
+	- view
+	- websocket_key
+- cfg - server and app config - see `config/master.sh` for more details
+	
+## Fun stuff
+
+- To prevent running malicious scripts, by default only scripts with extension `.shs` can be run by the server, but this can be changed in the config.
+- ${cfg[index]} ignores the above - see config/master.sh
+- Trans rights!
+- SHS stands for Shell Server.

http.sh

@@ -13,7 +13,6 @@ if [[ ! -f "$(pwd)/http.sh" ]]; then
 		exit 1
 fi	
 
-echo "-= HTTP.sh =-"
 
 for i in $(cat src/dependencies.required); do
 	which $i > /dev/null 2>&1
@@ -33,14 +32,13 @@ if [[ $error == true ]]; then
 fi
 
 if [[ $1 == "init" ]]; then # will get replaced with proper parameter parsing in 1.0
-	mkdir -p "${cfg[namespace]}/${cfg[root]}" "${cfg[namespace]}/workers/example"
+	mkdir -p "${cfg[namespace]}/${cfg[root]}" "${cfg[namespace]}/workers/example" "${cfg[namespace]}/views"
 	touch "${cfg[namespace]}/config.sh" "${cfg[namespace]}/workers/example/control"
 	cat <<LauraIsCute > "${cfg[namespace]}/config.sh"
-# app config - loaded on server bootup
+## app config
-# your application-specific config goes here!
+## your application-specific config goes here!
+
 
-# worker_add <worker> <interval>
-# ---
 # worker_add example 5
 LauraIsCute
 
@@ -60,14 +58,41 @@ echo "<h1>Hello from HTTP.sh!</h1><br>To get started with your app, check out $(
 	 <li>$(pwd)/src/ - HTTP.sh src, feel free to poke around :P</li></ul>
 	 &copy; sdomi, selfisekai, ptrcnull - 2020"
 LauraIsCute
+	cat <<PtrcIsCute > "${cfg[namespace]}/routes.sh"
+## routes - application-specific routes
+##
+## HTTP.sh supports both serving files using a directory structure (webroot),
+## and using routes. The latter may come in handy if you want to create nicer
+## paths, e.g.
+##
+## (webroot) https://example.com/profile.shs?name=ptrcnull
+## ... may become ...
+## (routes)  https://example.com/profile/ptrcnull
+##
+## To set up routes, define rules in this file (see below for examples)
+
+# router "/test" "app/views/test.shs"
+# router "/profile/:user" "app/views/user.shs"
+PtrcIsCute
+
 	chmod +x "${cfg[namespace]}/workers/example/worker.sh"
 	
 	echo -e "Success..?\nTry running ./http.sh now"
 	exit 0
 fi
 
+cat <<PtrcIsCute >&2
+ _    _ _______ _______ _____  ______ _    _ 
+| |  | |_______|_______|  _  \/  ___/| |  | |
+| |__| |  | |     | |  | |_| | |___  | |__| |
+| |__| |  | |     | |  |  ___/\___ \ | |__| |
+| |  | |  | |     | |  | |     ___\ \| |  | |
+|_|  |_|  |_|     |_|  |_|  □ /_____/|_|  |_|
+PtrcIsCute
+
 if [[ $1 == "debug" ]]; then
 	cfg[dbg]=true
+	echo "[DEBUG] Activated debug mode - stderr will be shown"
 fi
 
 source src/worker.sh

src/response/200.sh

@@ -1,23 +1,33 @@
 printf "HTTP/1.0 200 OK
 ${cfg[extra_headers]}\r\n"
-get_mime "${r[uri]}"
-[[ "$mimetype" != '' ]] && printf "content-type: $mimetype\r\n"
 
-if [[ ${cfg[php_enabled]} == true && ${r[uri]} =~ ".php" ]]; then
+if [[ ${r[status]} == 200 ]]; then
+	get_mime "${r[uri]}"
+	[[ "$mimetype" != '' ]] && printf "content-type: $mimetype\r\n"
+fi
+
+if [[ ${r[status]} == 212 ]]; then
+	temp=$(mktemp)
+	source "${r[view]}" > $temp
+	[[ "${r[headers]}" != '' ]] && printf "${r[headers]}\r\n\r\n" || printf "\r\n"
+	cat $temp
+	rm $temp
+	
+elif [[ "${cfg[php_enabled]}" == true && "${r[uri]}" =~ ".php" ]]; then
 	temp=$(mktemp)
 	php "${r[uri]}" "$(get_dump)" "$(post_dump)" > $temp
-	[[ ${r[headers]} != '' ]] && printf "${r[headers]}\r\n\r\n" || printf "\r\n"
+	[[ "${r[headers]}" != '' ]] && printf "${r[headers]}\r\n\r\n" || printf "\r\n"
 	cat $temp
 	rm $temp
 
-elif [[ ${cfg[python_enabled]} == true && ${r[uri]} =~ ".py" ]]; then
+elif [[ "${cfg[python_enabled]}" == true && "${r[uri]}" =~ ".py" ]]; then
 	temp=$(mktemp)
 	python "${r[uri]}" "$(get_dump)" "$(post_dump)" > $temp
-	[[ ${r[headers]} != '' ]] && printf "${r[headers]}\r\n\r\n" || printf "\r\n"
+	[[ "${r[headers]}" != '' ]] && printf "${r[headers]}\r\n\r\n" || printf "\r\n"
 	cat $temp
 	rm $temp
 
-elif [[ ${r[uri]} =~ \.${cfg[extension]}$ ]]; then
+elif [[ "${r[uri]}" =~ \.${cfg[extension]}$ ]]; then
 	temp=$(mktemp)
 	source "${r[uri]}" > $temp
 	[[ "${r[headers]}" != '' ]] && printf "${r[headers]}\r\n\r\n" || printf "\r\n"

src/route.sh

@@ -0,0 +1,8 @@
+#!/bin/bash
+
+# router(uri, path)
+function router() {
+	route+=("$1")
+	route+=("$(sed -E 's/:[A-Za-z0-9]+/[A-Za-z0-9.,%:-_]+/g' <<< "$1")")
+	route+=("$2")
+}

src/server.sh

@@ -4,12 +4,16 @@ source src/mime.sh
 source src/misc.sh
 source src/account.sh
 source src/mail.sh
+source src/route.sh
 [[ -f "${cfg[namespace]}/config.sh" ]] && source "${cfg[namespace]}/config.sh"
-
+[[ -f "${cfg[namespace]}/routes.sh" ]] && source "${cfg[namespace]}/routes.sh"
 
 declare -A r # current request / response
 declare -A meta # metadata for templates
 declare -A cookies # cookies!
+declare -A get_data # all GET params
+declare -A post_data # all POST params
+declare -A params # parsed router data
 
 r[status]=210 # Mommy always said that I was special
 post_length=0
@@ -73,7 +77,6 @@ while read param; do
 		data="$(echo ${r[url]} | sed -E 's/^(.*)\?//;s/\&/ /g')"
 		if [[ "$data" != "${r[url]}" ]]; then
 			data="$(echo ${r[url]} | sed -E 's/^(.*)\?//')"
-			declare -A get_data
 			IFS='&'
 			for i in $data; do
 				name="$(echo $i | sed -E 's/\=(.*)$//')"
@@ -89,7 +92,6 @@ while read param; do
 		data="$(sed -E 's/^(.*)\?//;s/\&/ /g' <<< "${r[url]}")"
 		if [[ "$data" != "${r[url]}" ]]; then
 			data="$(sed -E 's/^(.*)\?//' <<< "${r[url]}")"
-			declare -A get_data
 			IFS='&'
 			for i in $data; do
 				name="$(echo $i | sed -E 's/\=(.*)$//')"
@@ -114,24 +116,45 @@ fi
 echo "$(date) - IP: ${r[ip]}, PROTO: ${r[proto]}, URL: ${r[url]}, GET_data: ${get_data[@]}, POST_data: ${post_data[@]}, POST_multipart: ${post_multipart[@]}" >> "${cfg[namespace]}/${cfg[log]}"
 
 if [[ ${r[status]} != 101 ]]; then
-	if [[ -a ${r[uri]} && ! -r ${r[uri]} ]]; then
+	for (( i=0; i<${#route[@]}; i=i+3 )); do
-		r[status]=403
+		if [[ "$(grep -Poh "${route[$((i+1))]}$" <<< "${r[url]}")" != "" ]]; then
-	elif [[ "$(echo -n ${r[uri]})" != "$(realpath "${cfg[namespace]}/${cfg[root]}")"* ]]; then
+			r[status]=212
-		r[status]=403
+			r[view]="${route[$((i+2))]}"
-	elif [[ -f ${r[uri]} ]]; then
+			IFS='/'
-		r[status]=200
+			url=(${route[$i]})
-	elif [[ -d ${r[uri]} ]]; then
+			url_=(${r[url]})
-		for name in ${cfg[index]}; do
+			unset IFS
-			if [[ -f "${r[uri]}/$name" ]]; then
+			for (( j=0; j<${#url[@]}; j++ )); do
-				r[uri]="${r[uri]}/$name"
+				if [[ ${url_[$j]} != '' ]]; then
-				r[status]=200
+					params[$(sed 's/://' <<< "${url[$j]}")]="${url_[$j]}"
-			fi
+				fi
-		done
+			done
-	else
+			break
-		r[status]=404
+		fi
+	done
+	unset IFS
+	if [[ ${r[status]} != 212 ]]; then
+		if [[ -a "${r[uri]}" && ! -r "${r[uri]}" ]]; then
+			r[status]=403
+		elif [[ "$(echo -n "${r[uri]}")" != "$(realpath "${cfg[namespace]}/${cfg[root]}")"* ]]; then
+			r[status]=403
+		elif [[ -f "${r[uri]}" ]]; then
+			r[status]=200
+		elif [[ -d "${r[uri]}" ]]; then
+			for name in ${cfg[index]}; do
+				if [[ -f "${r[uri]}/$name" ]]; then
+					r[uri]="${r[uri]}/$name"
+					r[status]=200
+				fi
+			done
+		else
+			r[status]=404
+		fi
 	fi
 fi
 
+echo "${r[url]}" >&2
+
 if [[ ${cfg[auth_required]} == true && ${r[authorized]} != true ]]; then
 	echo "Auth failed." >> ${cfg[log_misc]}
 	r[status]=401
@@ -173,8 +196,7 @@ if [[ ${r[post]} == true && ${r[status]} == 200 ]]; then
 		rm $tmpfile
 	else
 		read -N "${r[content_length]}" data
-		declare -A post_data
+		
-
 		IFS='&'
 		for i in $(tr -d '\n' <<< "$data"); do
 			name="$(sed -E 's/\=(.*)$//' <<< "$i")"
@@ -189,7 +211,7 @@ if [[ ${r[status]} == 210 && ${cfg[autoindex]} == true ]]; then
 	source "src/response/listing.sh"
 elif [[ ${r[status]} == 211 ]]; then
 	source "src/response/proxy.sh"
-elif [[ ${r[status]} == 200 ]]; then
+elif [[ ${r[status]} == 200 || ${r[status]} == 212 ]]; then
 	source "src/response/200.sh"
 elif [[ ${r[status]} == 401 ]]; then
 	source "src/response/401.sh"