I need to register many thousands of users in ejabberd from csv file. For this, I wrote a simple script.
#!/bin/sh
OLDIFS=$IFS
IFS=','
[ ! -f $INPUT ] && { echo "$INPUT file not found"; exit 99; }
while read username domain pass p1 p2 p3 p4
do
echo "ejabberdctl register $username $domain $pass"
ejabberdctl register $username $domain $pass
done < users.csv
IFS=$OLDIFS
But in the end, the answer is: Error: cannot_register
If I just run the line copied from the output, everything is ok. The user is created normally.
This is just a little trick for later: once you get your loop working, if you consider ejabberdctl is too slow, you can try using the ReST API. That should be a lot faster when doing many requests.
Configure temporarily something like this (remember to remove this when you finished):
listen:
-
port: 5280
module: ejabberd_http
tls: false
request_handlers:
/api: mod_http_api
api_permissions:
"console commands":
from:
- ejabberd_ctl
- mod_http_api
who: all
what: "*"
modules:
mod_http_api: {}
Then execute this in a shell to register an account:
curl 'localhost:5280/api/register?user=user2&host=localhost&password=somepass123'
I'm trying to add some text to the end of a few files.
I have made a file, where I have 5 servername. Each servername corresponds to a separate config file. (The path of these config files is not known).
I am using below code to get the file path,
MyCode:
#!/bin/bash
for i in $(cat serverlist-file | while read f; do find . -name "*$f*"; done);
do
echo $i
done
Output:
/data/servers/customer01/server01.cfg
/data/servers/customer01/server02.cfg
/data/servers/customer02/server03.cfg
/data/servers/customer03/server04.cfg
/data/servers/customer03/server05.cfg
I am using below code to get the list of servers,
MyCode:
#!/bin/bash
for j in $(cat serverlist-file);
do
echo $j
done
Output:
server01
server02
server03
server04
server05
Now I want to edit those config files and add text to it.
I am using below code to add the required text:
#!/bin/bash
for i in $(cat serverlist-file | while read f; do find . -name "*$f*"; done);
do
for j in $(cat serverlist-file);
do
sed -i -e "\$a\
this\ is\ a\ config\ file\nfor\ $j" $i
done
done
Expected Output:
/data/servers/customer01/server01.cfg
this is a config file
for server01
/data/servers/customer01/server02.cfg
this is a config file
for server02
/data/servers/customer02/server03.cfg
this is a config file
for server03
/data/servers/customer03/server04.cfg
this is a config file
for server04
/data/servers/customer03/server05.cfg
this is a config file
for server05
Edit for a reply to #ShawnMilo:
I am trying to bulk add some config to some nagios config files, but not to all server config files.
So, searching with find . -name '*.config' isn't going to work, because then all the config files will get edited.
I only want specific files to get edited, just the servers from the serverlist-file.
Nagios configs need to have the hostname of the server in them, like:
define service {
use generic-service
host_name server01
service_description SSH
contact_groups linux
check_command check_something
}
Seems like an odd requirement. What are you actually trying to do?
In any case, this will do what was requested:
$ find . -name '*.config' | while read x; do echo $x; cat $x; echo; done
./data/servers/customer02/server03.config
default stuff here
./data/servers/customer03/server05.config
default stuff here
./data/servers/customer03/server04.config
default stuff here
./data/servers/customer01/server01.config
default stuff here
./data/servers/customer01/server02.config
default stuff here
$ find . -name '*.config' | while read x; do name=$(basename $x); echo -e "this is a config file\nfor ${name%%.*}" >> $x; done
$ find . -name '*.config' | while read x; do echo $x; cat $x; echo; done
./data/servers/customer02/server03.config
default stuff here
this is a config file
for server03
./data/servers/customer03/server05.config
default stuff here
this is a config file
for server05
./data/servers/customer03/server04.config
default stuff here
this is a config file
for server04
./data/servers/customer01/server01.config
default stuff here
this is a config file
for server01
./data/servers/customer01/server02.config
default stuff here
this is a config file
for server02
In a script I'm writing right now, I create many background processes in attempts to run my script on multiple devices in parallel. This functionality works, but it would appear I have no control of it. The simple wait command does not get me the results I need.
Abridged code:
#!/bin/bash
echo ""
date
echo ""
echo "Displaying devices to be configured:"
./adb devices | sed "1d ; $ d"
echo ""
echo "###########################"
echo "# #"
echo "# Starting configuration! #"
echo "# #"
echo "###########################"
echo ""
# All commands ran through this function
DeviceConfig () {
...
# Large list of commands
...
}
# This is the loop that spawns all the processes. Note the ampersand I'm using.
for usb in $(./adb devices -l | awk '/ device usb:/{print $3}'); do ( DeviceConfig & ) ; done
echo ""
echo "###########################"
echo "# #"
echo "# Configuration complete! #"
echo "# #"
echo "###########################"
While this will successfully run all my commands in parallel, my output is not as intended.
Actual output:
Wed Oct 5 13:11:26 EDT 2016
Displaying devices to be configured:
3100c2759da2a200 device
3100c2ddbbafa200 device
###########################
# #
# Starting configuration! #
# #
###########################
###########################
# #
# Configuration complete! #
# #
###########################
Starting: Intent { cmp=com.android.settings/.Settings }
Warning: Activity not started, its current task has been brought to the front
Starting: Intent { cmp=com.android.settings/.Settings }
Warning: Activity not started, its current task has been brought to the front
...
(The ... is to imply more output from the script.)
Putting a wait in the loop does not solve the issue. Putting a wait after the loop does not solve the issue. How do I write this loop so the configurations happen in between the Starting configuration! and Configuration complete! output?
You can ask wait to wait on multiple processes, e.g.:
pids=()
for usb in $(./adb devices -l | awk '/ device usb:/{print $3}'); do DeviceConfig & pids+=($!); done
wait "${pids[#]}"
i tried to setup squid3 with multiple auth_param. Basically, the first choice should be basic_ldap_auth and if this doesnt return OK it should try basic_ncsa_auth with the same values. As far as i know squid doesnt support it however there is the possibility to use "external" ACL
auth_param basic program /usr/lib/squid3/basic_fake_auth
external_acl_type MultAuth %SRC %LOGIN %{Proxy-Authorization} /etc/squid3/multAuth.pl
acl extAuth external MultAuth
my "multAuth.pl"
use URI::Escape;
use MIME::Base64;
$|=1;
while (<>) {
($ip,$user,$auth) = split();
# Retrieve the password from the authentication header
$auth = uri_unescape($auth);
($type,$authData) = split(/ /, $auth);
$authString = decode_base64($authData);
($username,$password) = split(/:/, $authString);
# do the authentication and pass results back to Squid.
$ldap = `/bin/bash auth/ldap.sh`;
if ($ldap == "OK") {
print "OK";
}
$ncsa = `/bin/bash auth/ncsa.sh`;
if ($ncsa == "OK") {
print "OK";
} else {
print "ERR";
}
}
now i am trying to run with ncsa.sh and ldap.sh the "normal" shell command for these auth methods.
./basic_ldap_auth -R -b "dc=domain,dc=de" -D "CN=Administrator,CN=Users,DC=domain,DC=de" -w "password" -f sAMAccountName=%s -h domain.de
user password
and
./basic_ncsa_auth /etc/squid3/users
user password
Therefor i ran:
auth/ncsa.sh
#!/usr/bin/expect
eval spawn [lrange $argv 0 end]
expect ""
send [lindex $argv 1]
send '\r'
expect {
"OK" {
echo OK
exp_continue
}
"ERR" {
echo ERR
exp_continue
}
interact
with
./ncsa.sh "/usr/lib/squid3/basic_ncsa_auth /etc/squid3/users" "user password"
and i generate the following error:
couldn't execute "/usr/lib/squid3/basic_ncsa_auth /etc/squid3/users": no such file or directory
while executing
"spawn {/usr/lib/squid3/basic_ncsa_auth /etc/squid3/users} {user password}"
("eval" body line 1)
invoked from within
"eval spawn [lrange $argv 0 end]"
(file "./ncsa.sh" line 2)
Besides this error, i am not sure how to pass the variables (username & password) forward and i am also not sure how to answer the shell questions like for example the user & pw input for basic_ldap_auth .
Is there a nice way how to solve that? or any other good plan ?
thanks!
FWIW, the following script helped me transition from passwd based to LDAP based authentication.
Contrary to your requirements, my script acts the other way around: It first checks passwd, then LDAP.
#!/usr/bin/env bash
# multiple Squid basic auth checks
# originally posted here: https://github.com/HackerHarry/mSquidAuth
#
# credits
# https://stackoverflow.com/questions/24147067/verify-user-and-password-against-a-file-created-by-htpasswd/40131483
# https://stackoverflow.com/questions/38710483/how-to-stop-ldapsearch1-from-base64-encoding-userpassword-and-other-attributes
#
# requires ldap-utils, openssl and perl
# tested with Squid 4 using a "auth_param basic program /usr/lib/squid/mSquidAuth.sh" line
# authenticate first against squid password file
# if this fails, try LDAP (Active Directory) and also check group membership
# variables
# sLOGFILE=/var/log/squid/mSquidAuth.log
sPWDFILE="/etc/squid/passwd"
sLDAPHOST="ldaps://dc.domain.local:636"
sBASE="DC=domain,DC=local"
sLDS_OPTIONS="-o ldif-wrap=no -o nettimeout=7 -LLL -P3 -x "
sBINDDN="CN=LDAP-read-user,OU=Users,DC=domain,DC=local"
sBINDPW="read-user-password"
sGROUP="Proxy-Users"
# functions
function _grantAccess {
# echo "access granted - $sUSER" >>$sLOGFILE
echo "OK"
}
function _denyAccess {
# echo "access denied - $sUSER" >>$sLOGFILE
echo "ERR"
}
function _setUserAndPass {
local sAuth="$1"
local sOldIFS=$IFS
IFS=' '
set -- $sAuth
IFS=$sOldIFS
# set it globally
sUSER="$1"
sPASS="$2"
}
# loop
while (true); do
read -r sAUTH
sUSER=""
sPASS=""
sSALT=""
sUSERENTRY=""
sHASHEDPW=""
sUSERDN=""
iDNCOUNT=0
if [ -z "$sAUTH" ]; then
# echo "exiting" >>$sLOGFILE
exit 0
fi
_setUserAndPass "$sAUTH"
sUSERENTRY=$(grep -E "^${sUSER}:" "$sPWDFILE")
if [ -n "$sUSERENTRY" ]; then
sSALT=$(echo "$sUSERENTRY" | cut -d$ -f3)
if [ -n "$sSALT" ]; then
sHASHEDPW=$(openssl passwd -apr1 -salt "$sSALT" "$sPASS")
if [ "$sUSERENTRY" = "${sUSER}:${sHASHEDPW}" ]; then
_grantAccess
continue
fi
fi
fi
# LDAP is next
iDNCOUNT=$(ldapsearch $sLDS_OPTIONS -H "$sLDAPHOST" -D "$sBINDDN" -w "$sBINDPW" -b "$sBASE" "(|(sAMAccountName=${sUSER})(userPrincipalName=${sUSER}))" dn 2>/dev/null | grep -cE 'dn::? ')
if [ $iDNCOUNT != 1 ]; then
# user needs a unique account
_denyAccess
continue
fi
# get user's DN
# we need the extra grep in case we get lines back starting with "# refldap" :/
sUSERDN=$(ldapsearch $sLDS_OPTIONS -H "$sLDAPHOST" -D "$sBINDDN" -w "$sBINDPW" -b "$sBASE" "(|(sAMAccountName=${sUSER})(userPrincipalName=${sUSER}))" dn 2>/dev/null | perl -MMIME::Base64 -n -00 -e 's/\n +//g;s/(?<=:: )(\S+)/decode_base64($1)/eg;print' | grep -E 'dn::? ' | sed -r 's/dn::? //')
# try and bind using that DN to check password validity
# also test if that user is member of a particular group
# backslash in DN needs special treatment
if ldapsearch $sLDS_OPTIONS -H "$sLDAPHOST" -D "$sUSERDN" -w "$sPASS" -b "$sBASE" "name=${sGROUP}" member 2>/dev/null | perl -MMIME::Base64 -n -00 -e 's/\n +//g;s/(?<=:: )(\S+)/decode_base64($1)/eg;print' | grep -q "${sUSERDN/\\/\\\\}"; then
_grantAccess
continue
fi
_denyAccess
done
I'm trying to set up a minimal web server using netcat (nc). When the browser calls up localhost:1500, for instance, it should show the result of a function (date in the example below, but eventually it'll be a python or c program that yields some data).
My little netcat web server needs to be a while true loop in bash, possibly as simple as this:
while true ; do echo -e "HTTP/1.1 200 OK\n\n $(date)" | nc -l -p 1500 ; done
When I try this the browser shows the currently available data during the moment when nc starts. I want the browser displays the data during the moment the browser requests it, though. How can I achieve this?
Try this:
while true ; do nc -l -p 1500 -c 'echo -e "HTTP/1.1 200 OK\n\n $(date)"'; done
The -cmakes netcat execute the given command in a shell, so you can use echo. If you don't need echo, use -e. For further information on this, try man nc. Note, that when using echo there is no way for your program (the date-replacement) to get the browser request. So you probably finally want to do something like this:
while true ; do nc -l -p 1500 -e /path/to/yourprogram ; done
Where yourprogram must do the protocol stuff like handling GET, sending HTTP 200 etc.
I had the problem where I wanted to return the result of executing a bash command:
$ while true; do { echo -e 'HTTP/1.1 200 OK\r\n'; sh test; } | nc -l 8080; done
NOTE:
This command was taken from: http://www.razvantudorica.com/08/web-server-in-one-line-of-bash
This executes a bash script and returns the result to a browser client connecting to the server running this command on port 8080.
My script does this:
$ nano test
#!/bin/bash
echo "************PRINT SOME TEXT***************\n"
echo "Hello World!!!"
echo "\n"
echo "Resources:"
vmstat -S M
echo "\n"
echo "Addresses:"
echo "$(ifconfig)"
echo "\n"
echo "$(gpio readall)"
and my web browser is showing
************PRINT SOME TEXT***************
Hello World!!!
Resources:
procs -----------memory---------- ---swap-- -----io---- -system-- ----cpu----
r b swpd free buff cache si so bi bo in cs us sy id wa
0 0 0 314 18 78 0 0 2 1 306 31 0 0 100 0
Addresses:
eth0 Link encap:Ethernet HWaddr b8:27:eb:86:e8:c5
inet addr:192.168.1.83 Bcast:192.168.1.255 Mask:255.255.255.0
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:27734 errors:0 dropped:0 overruns:0 frame:0
TX packets:26393 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:1924720 (1.8 MiB) TX bytes:3841998 (3.6 MiB)
lo Link encap:Local Loopback
inet addr:127.0.0.1 Mask:255.0.0.0
UP LOOPBACK RUNNING MTU:65536 Metric:1
RX packets:0 errors:0 dropped:0 overruns:0 frame:0
TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:0
RX bytes:0 (0.0 B) TX bytes:0 (0.0 B)
GPIOs:
+----------+-Rev2-+------+--------+------+-------+
| wiringPi | GPIO | Phys | Name | Mode | Value |
+----------+------+------+--------+------+-------+
| 0 | 17 | 11 | GPIO 0 | IN | Low |
| 1 | 18 | 12 | GPIO 1 | IN | Low |
| 2 | 27 | 13 | GPIO 2 | IN | Low |
| 3 | 22 | 15 | GPIO 3 | IN | Low |
| 4 | 23 | 16 | GPIO 4 | IN | Low |
| 5 | 24 | 18 | GPIO 5 | IN | Low |
| 6 | 25 | 22 | GPIO 6 | IN | Low |
| 7 | 4 | 7 | GPIO 7 | IN | Low |
| 8 | 2 | 3 | SDA | IN | High |
| 9 | 3 | 5 | SCL | IN | High |
| 10 | 8 | 24 | CE0 | IN | Low |
| 11 | 7 | 26 | CE1 | IN | Low |
| 12 | 10 | 19 | MOSI | IN | Low |
| 13 | 9 | 21 | MISO | IN | Low |
| 14 | 11 | 23 | SCLK | IN | Low |
| 15 | 14 | 8 | TxD | ALT0 | High |
| 16 | 15 | 10 | RxD | ALT0 | High |
| 17 | 28 | 3 | GPIO 8 | ALT2 | Low |
| 18 | 29 | 4 | GPIO 9 | ALT2 | Low |
| 19 | 30 | 5 | GPIO10 | ALT2 | Low |
| 20 | 31 | 6 | GPIO11 | ALT2 | Low |
+----------+------+------+--------+------+-------+
Add -q 1 to the netcat command line:
while true; do
echo -e "HTTP/1.1 200 OK\n\n $(date)" | nc -l -p 1500 -q 1
done
The problem you are facing is that nc does not know when the web client is done with its request so it can respond to the request.
A web session should go something like this.
TCP session is established.
Browser Request Header: GET / HTTP/1.1
Browser Request Header: Host: www.google.com
Browser Request Header: \n #Note: Browser is telling Webserver that the request header is complete.
Server Response Header: HTTP/1.1 200 OK
Server Response Header: Content-Type: text/html
Server Response Header: Content-Length: 24
Server Response Header: \n #Note: Webserver is telling browser that response header is complete
Server Message Body: <html>sample html</html>
Server Message Body: \n #Note: Webserver is telling the browser that the requested resource is finished.
The server closes the TCP session.
Lines that begin with "\n" are simply empty lines without even a space and contain nothing more than a new line character.
I have my bash httpd launched by xinetd, xinetd tutorial. It also logs date, time, browser IP address, and the entire browser request to a log file, and calculates Content-Length for the Server header response.
user#machine:/usr/local/bin# cat ./bash_httpd
#!/bin/bash
x=0;
Log=$( echo -n "["$(date "+%F %T %Z")"] $REMOTE_HOST ")$(
while read I[$x] && [ ${#I[$x]} -gt 1 ];do
echo -n '"'${I[$x]} | sed -e's,.$,",'; let "x = $x + 1";
done ;
); echo $Log >> /var/log/bash_httpd
Message_Body=$(echo -en '<html>Sample html</html>')
echo -en "HTTP/1.0 200 OK\nContent-Type: text/html\nContent-Length: ${#Message_Body}\n\n$Message_Body"
To add more functionality, you could incorporate.
METHOD=$(echo ${I[0]} |cut -d" " -f1)
REQUEST=$(echo ${I[0]} |cut -d" " -f2)
HTTP_VERSION=$(echo ${I[0]} |cut -d" " -f3)
If METHOD = "GET" ]; then
case "$REQUEST" in
"/") Message_Body="HTML formatted home page stuff"
;;
/who) Message_Body="HTML formatted results of who"
;;
/ps) Message_Body="HTML formatted results of ps"
;;
*) Message_Body= "Error Page not found header and content"
;;
esac
fi
Happy bashing!
Another way to do this
while true; do (echo -e 'HTTP/1.1 200 OK\r\n'; echo -e "\n\tMy website has date function" ; echo -e "\t$(date)\n") | nc -lp 8080; done
Let's test it with 2 HTTP request using curl
In this example, 172.16.2.6 is the server IP Address.
Server Side
admin#server:~$ while true; do (echo -e 'HTTP/1.1 200 OK\r\n'; echo -e "\n\tMy website has date function" ; echo -e "\t$(date)\n") | nc -lp 8080; done
GET / HTTP/1.1 Host: 172.16.2.6:8080 User-Agent: curl/7.48.0 Accept:
*/*
GET / HTTP/1.1 Host: 172.16.2.6:8080 User-Agent: curl/7.48.0 Accept:
*/*
Client Side
user#client:~$ curl 172.16.2.6:8080
My website has date function
Tue Jun 13 18:00:19 UTC 2017
user#client:~$ curl 172.16.2.6:8080
My website has date function
Tue Jun 13 18:00:24 UTC 2017
user#client:~$
If you want to execute another command, feel free to replace $(date).
I had the same need/problem but nothing here worked for me (or I didn't understand everything), so this is my solution.
I post my minimal_http_server.sh (working with my /bin/bash (4.3.11) but not /bin/sh because of the redirection):
rm -f out
mkfifo out
trap "rm -f out" EXIT
while true
do
cat out | nc -l 1500 > >( # parse the netcat output, to build the answer redirected to the pipe "out".
export REQUEST=
while read -r line
do
line=$(echo "$line" | tr -d '\r\n')
if echo "$line" | grep -qE '^GET /' # if line starts with "GET /"
then
REQUEST=$(echo "$line" | cut -d ' ' -f2) # extract the request
elif [ -z "$line" ] # empty line / end of request
then
# call a script here
# Note: REQUEST is exported, so the script can parse it (to answer 200/403/404 status code + content)
./a_script.sh > out
fi
done
)
done
And my a_script.sh (with your need):
#!/bin/bash
echo -e "HTTP/1.1 200 OK\r"
echo "Content-type: text/html"
echo
date
mkfifo pipe;
while true ;
do
#use read line from pipe to make it blocks before request comes in,
#this is the key.
{ read line<pipe;echo -e "HTTP/1.1 200 OK\r\n";echo $(date);
} | nc -l -q 0 -p 8080 > pipe;
done
Here is a beauty of a little bash webserver, I found it online and forked a copy and spruced it up a bit - it uses socat or netcat I have tested it with socat -- it is self-contained in one-script and generates its own configuration file and favicon.
By default it will start up as a web enabled file browser yet is easily configured by the configuration file for any logic. For files it streams images and music (mp3's), video (mp4's, avi, etc) -- I have tested streaming various file types to Linux,Windows and Android devices including a smartwatch!
I think it streams better than VLC actually. I have found it useful for transferring files to remote clients who have no access beyond a web browser e.g. Android smartwatch without needing to worry about physically connecting to a USB port.
If you want to try it out just copy and paste it to a file named bashttpd, then start it up on the host with $> bashttpd -s
Then you can go to any other computer (presuming the firewall is not blocking inbound tcp connections to port 8080 -- the default port, you can change the port to whatever you want using the global variables at the top of the script). http://bashttpd_server_ip:8080
#!/usr/bin/env bash
#############################################################################
###########################################################################
### bashttpd v 1.12
###
### Original author: Avleen Vig, 2012
### Reworked by: Josh Cartwright, 2012
### Modified by: A.M.Danischewski, 2015
### Issues: If you find any issues leave me a comment at
### http://scriptsandoneliners.blogspot.com/2015/04/bashttpd-self-contained-bash-webserver.html
###
### This is a simple Bash based webserver. By default it will browse files and allows for
### retrieving binary files.
###
### It has been tested successfully to view and stream files including images, mp3s,
### mp4s and downloading files of any type including binary and compressed files via
### any web browser.
###
### Successfully tested on various browsers on Windows, Linux and Android devices (including the
### Android Smartwatch ZGPAX S8).
###
### It handles favicon requests by hardcoded favicon image -- by default a marathon
### runner; change it to whatever you want! By base64 encoding your favorit favicon
### and changing the global variable below this header.
###
### Make sure if you have a firewall it allows connections to the port you plan to
### listen on (8080 by default).
###
### By default this program will allow for the browsing of files from the
### computer where it is run.
###
### Make sure you are allowed connections to the port you plan to listen on
### (8080 by default). Then just drop it on a host machine (that has bash)
### and start it up like this:
###
### $192.168.1.101> bashttpd -s
###
### On the remote machine you should be able to browse and download files from the host
### server via any web browser by visiting:
###
### http://192.168.1.101:8080
###
#### This program requires (to work to full capacity) by default:
### socat or netcat (w/ '-e' option - on Ubuntu netcat-traditional)
### tree - useful for pretty directory listings
### If you are using socat, you can type: bashttpd -s
###
### to start listening on the LISTEN_PORT (default is 8080), you can change
### the port below.
### E.g. nc -lp 8080 -e ./bashttpd ## <-- If your nc has the -e option.
### E.g. nc.traditional -lp 8080 -e ./bashttpd
### E.g. bashttpd -s -or- socat TCP4-LISTEN:8080,fork EXEC:bashttpd
###
### Copyright (C) 2012, Avleen Vig <avleen#gmail.com>
###
### Permission is hereby granted, free of charge, to any person obtaining a copy of
### this software and associated documentation files (the "Software"), to deal in
### the Software without restriction, including without limitation the rights to
### use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of
### the Software, and to permit persons to whom the Software is furnished to do so,
### subject to the following conditions:
###
### The above copyright notice and this permission notice shall be included in all
### copies or substantial portions of the Software.
###
### THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
### IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS
### FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR
### COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER
### IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
### CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
###
###########################################################################
#############################################################################
### CHANGE THIS TO WHERE YOU WANT THE CONFIGURATION FILE TO RESIDE
declare -r BASHTTPD_CONF="/tmp/bashttpd.conf"
### CHANGE THIS IF YOU WOULD LIKE TO LISTEN ON A DIFFERENT PORT
declare -i LISTEN_PORT=8080
## If you are on AIX, IRIX, Solaris, or a hardened system redirecting
## to /dev/random will probably break, you can change it to /dev/null.
declare -a DUMP_DEV="/dev/random"
## Just base64 encode your favorite favicon and change this to whatever you want.
declare -r FAVICON="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"
declare -i DEBUG=1
declare -i VERBOSE=0
declare -a REQUEST_HEADERS
declare REQUEST_URI=""
declare -a HTTP_RESPONSE=(
[200]="OK"
[400]="Bad Request"
[403]="Forbidden"
[404]="Not Found"
[405]="Method Not Allowed"
[500]="Internal Server Error")
declare DATE=$(date +"%a, %d %b %Y %H:%M:%S %Z")
declare -a RESPONSE_HEADERS=(
"Date: $DATE"
"Expires: $DATE"
"Server: Slash Bin Slash Bash"
)
function warn() { ((${VERBOSE})) && echo "WARNING: $#" >&2; }
function chk_conf_file() {
[ -r "${BASHTTPD_CONF}" ] || {
cat >"${BASHTTPD_CONF}" <<'EOF'
#
# bashttpd.conf - configuration for bashttpd
#
# The behavior of bashttpd is dictated by the evaluation
# of rules specified in this configuration file. Each rule
# is evaluated until one is matched. If no rule is matched,
# bashttpd will serve a 500 Internal Server Error.
#
# The format of the rules are:
# on_uri_match REGEX command [args]
# unconditionally command [args]
#
# on_uri_match:
# On an incoming request, the URI is checked against the specified
# (bash-supported extended) regular expression, and if encounters a match the
# specified command is executed with the specified arguments.
#
# For additional flexibility, on_uri_match will also pass the results of the
# regular expression match, ${BASH_REMATCH[#]} as additional arguments to the
# command.
#
# unconditionally:
# Always serve via the specified command. Useful for catchall rules.
#
# The following commands are available for use:
#
# serve_file FILE
# Statically serves a single file.
#
# serve_dir_with_tree DIRECTORY
# Statically serves the specified directory using 'tree'. It must be
# installed and in the PATH.
#
# serve_dir_with_ls DIRECTORY
# Statically serves the specified directory using 'ls -al'.
#
# serve_dir DIRECTORY
# Statically serves a single directory listing. Will use 'tree' if it is
# installed and in the PATH, otherwise, 'ls -al'
#
# serve_dir_or_file_from DIRECTORY
# Serves either a directory listing (using serve_dir) or a file (using
# serve_file). Constructs local path by appending the specified root
# directory, and the URI portion of the client request.
#
# serve_static_string STRING
# Serves the specified static string with Content-Type text/plain.
#
# Examples of rules:
#
# on_uri_match '^/issue$' serve_file "/etc/issue"
#
# When a client's requested URI matches the string '/issue', serve them the
# contents of /etc/issue
#
# on_uri_match 'root' serve_dir /
#
# When a client's requested URI has the word 'root' in it, serve up
# a directory listing of /
#
# DOCROOT=/var/www/html
# on_uri_match '/(.*)' serve_dir_or_file_from "$DOCROOT"
# When any URI request is made, attempt to serve a directory listing
# or file content based on the request URI, by mapping URI's to local
# paths relative to the specified "$DOCROOT"
#
#unconditionally serve_static_string 'Hello, world! You can configure bashttpd by modifying bashttpd.conf.'
DOCROOT=/
on_uri_match '/(.*)' serve_dir_or_file_from
# More about commands:
#
# It is possible to somewhat easily write your own commands. An example
# may help. The following example will serve "Hello, $x!" whenever
# a client sends a request with the URI /say_hello_to/$x:
#
# serve_hello() {
# add_response_header "Content-Type" "text/plain"
# send_response_ok_exit <<< "Hello, $2!"
# }
# on_uri_match '^/say_hello_to/(.*)$' serve_hello
#
# Like mentioned before, the contents of ${BASH_REMATCH[#]} are passed
# to your command, so its possible to use regular expression groups
# to pull out info.
#
# With this example, when the requested URI is /say_hello_to/Josh, serve_hello
# is invoked with the arguments '/say_hello_to/Josh' 'Josh',
# (${BASH_REMATCH[0]} is always the full match)
EOF
warn "Created bashttpd.conf using defaults. Please review and configure bashttpd.conf before running bashttpd again."
# exit 1
}
}
function recv() { ((${VERBOSE})) && echo "< $#" >&2; }
function send() { ((${VERBOSE})) && echo "> $#" >&2; echo "$*"; }
function add_response_header() { RESPONSE_HEADERS+=("$1: $2"); }
function send_response_binary() {
local code="$1"
local file="${2}"
local transfer_stats=""
local tmp_stat_file="/tmp/_send_response_$$_"
send "HTTP/1.0 $1 ${HTTP_RESPONSE[$1]}"
for i in "${RESPONSE_HEADERS[#]}"; do
send "$i"
done
send
if ((${VERBOSE})); then
## Use dd since it handles null bytes
dd 2>"${tmp_stat_file}" < "${file}"
transfer_stats=$(<"${tmp_stat_file}")
echo -en ">> Transferred: ${file}\n>> $(awk '/copied/{print}' <<< "${transfer_stats}")\n" >&2
rm "${tmp_stat_file}"
else
## Use dd since it handles null bytes
dd 2>"${DUMP_DEV}" < "${file}"
fi
}
function send_response() {
local code="$1"
send "HTTP/1.0 $1 ${HTTP_RESPONSE[$1]}"
for i in "${RESPONSE_HEADERS[#]}"; do
send "$i"
done
send
while IFS= read -r line; do
send "${line}"
done
}
function send_response_ok_exit() { send_response 200; exit 0; }
function send_response_ok_exit_binary() { send_response_binary 200 "${1}"; exit 0; }
function fail_with() { send_response "$1" <<< "$1 ${HTTP_RESPONSE[$1]}"; exit 1; }
function serve_file() {
local file="$1"
local CONTENT_TYPE=""
case "${file}" in
*\.css)
CONTENT_TYPE="text/css"
;;
*\.js)
CONTENT_TYPE="text/javascript"
;;
*)
CONTENT_TYPE=$(file -b --mime-type "${file}")
;;
esac
add_response_header "Content-Type" "${CONTENT_TYPE}"
CONTENT_LENGTH=$(stat -c'%s' "${file}")
add_response_header "Content-Length" "${CONTENT_LENGTH}"
## Use binary safe transfer method since text doesn't break.
send_response_ok_exit_binary "${file}"
}
function serve_dir_with_tree() {
local dir="$1" tree_vers tree_opts basehref x
## HTML 5 compatible way to avoid tree html from generating favicon
## requests in certain browsers, such as browsers in android smartwatches. =)
local no_favicon=" <link href=\"data:image/x-icon;base64,${FAVICON}\" rel=\"icon\" type=\"image/x-icon\" />"
local tree_page=""
local base_server_path="/${2%/}"
[ "$base_server_path" = "/" ] && base_server_path=".."
local tree_opts="--du -h -a --dirsfirst"
add_response_header "Content-Type" "text/html"
# The --du option was added in 1.6.0. "/${2%/*}"
read _ tree_vers x < <(tree --version)
tree_page=$(tree -H "$base_server_path" -L 1 "${tree_opts}" -D "${dir}")
tree_page=$(sed "5 i ${no_favicon}" <<< "${tree_page}")
[[ "${tree_vers}" == v1.6* ]]
send_response_ok_exit <<< "${tree_page}"
}
function serve_dir_with_ls() {
local dir="$1"
add_response_header "Content-Type" "text/plain"
send_response_ok_exit < \
<(ls -la "${dir}")
}
function serve_dir() {
local dir="$1"
# If `tree` is installed, use that for pretty output.
which tree &>"${DUMP_DEV}" && \
serve_dir_with_tree "$#"
serve_dir_with_ls "$#"
fail_with 500
}
function urldecode() { [ "${1%/}" = "" ] && echo "/" || echo -e "$(sed 's/%\([[:xdigit:]]\{2\}\)/\\\x\1/g' <<< "${1%/}")"; }
function serve_dir_or_file_from() {
local URL_PATH="${1}/${3}"
shift
URL_PATH=$(urldecode "${URL_PATH}")
[[ $URL_PATH == *..* ]] && fail_with 400
# Serve index file if exists in requested directory
[[ -d "${URL_PATH}" && -f "${URL_PATH}/index.html" && -r "${URL_PATH}/index.html" ]] && \
URL_PATH="${URL_PATH}/index.html"
if [[ -f "${URL_PATH}" ]]; then
[[ -r "${URL_PATH}" ]] && \
serve_file "${URL_PATH}" "$#" || fail_with 403
elif [[ -d "${URL_PATH}" ]]; then
[[ -x "${URL_PATH}" ]] && \
serve_dir "${URL_PATH}" "$#" || fail_with 403
fi
fail_with 404
}
function serve_static_string() {
add_response_header "Content-Type" "text/plain"
send_response_ok_exit <<< "$1"
}
function on_uri_match() {
local regex="$1"
shift
[[ "${REQUEST_URI}" =~ $regex ]] && \
"$#" "${BASH_REMATCH[#]}"
}
function unconditionally() { "$#" "$REQUEST_URI"; }
function main() {
local recv=""
local line=""
local REQUEST_METHOD=""
local REQUEST_HTTP_VERSION=""
chk_conf_file
[[ ${UID} = 0 ]] && warn "It is not recommended to run bashttpd as root."
# Request-Line HTTP RFC 2616 $5.1
read -r line || fail_with 400
line=${line%%$'\r'}
recv "${line}"
read -r REQUEST_METHOD REQUEST_URI REQUEST_HTTP_VERSION <<< "${line}"
[ -n "${REQUEST_METHOD}" ] && [ -n "${REQUEST_URI}" ] && \
[ -n "${REQUEST_HTTP_VERSION}" ] || fail_with 400
# Only GET is supported at this time
[ "${REQUEST_METHOD}" = "GET" ] || fail_with 405
while IFS= read -r line; do
line=${line%%$'\r'}
recv "${line}"
# If we've reached the end of the headers, break.
[ -z "${line}" ] && break
REQUEST_HEADERS+=("${line}")
done
}
if [[ ! -z "{$1}" ]] && [ "${1}" = "-s" ]; then
socat TCP4-LISTEN:${LISTEN_PORT},fork EXEC:"${0}"
else
main
source "${BASHTTPD_CONF}"
fail_with 500
fi
LOL, a super lame hack, but at least curl and firefox accepts it:
while true ; do (dd if=/dev/zero count=10000;echo -e "HTTP/1.1\n\n $(date)") | nc -l 1500 ; done
You better replace it soon with something proper!
Ah yes, my nc were not exactly the same as yours, it did not like the -p option.
If you're using Apline Linux, the BusyBox netcat is slightly different:
while true; do nc -l -p 8080 -e sh -c 'echo -e "HTTP/1.1 200 OK\n\n$(date)"'; done
And another way using printf:
while true; do nc -l -p 8080 -e sh -c "printf 'HTTP/1.1 200 OK\n\n%s' \"$(date)\""; done
while true; do (echo -e 'HTTP/1.1 200 OK\r\nConnection: close\r\n';) | timeout 1 nc -lp 8080 ; done
Closes connection after 1 sec, so curl doesn't hang on it.
Type in nc -h and see if You have -e option available. If yes, You can create a script, for example:
script.sh
echo -e "HTTP/1.1 200 OK\n\n $(date)"
and run it like this:
while true ; do nc -l -p 1500 -e script.sh; done
Note that -e option needs to be enabled at compilation to be available.
I think the problem that all the solution listed doesn't work, is intrinsic in the nature of http service, the every request established is with a different client and the response need to be processed in a different context, every request must fork a new instance of response...
The current solution I think is the -e of netcat but I don't know why doesn't work... maybe is my nc version that I test on openwrt...
with socat it works....
I try this https://github.com/avleen/bashttpd
and it works, but I must run the shell script with this command.
socat tcp-l:80,reuseaddr,fork EXEC:bashttpd &
The socat and netcat samples on github doesn't works for me, but the socat that I used works.
Actually, the best way to close gracefully the connection is to send the Content-Length header like following. Client (like curl will close the connection after receiving the data.
DATA="Date: $(date)";
LENGTH=$(echo $DATA | wc -c);
echo -e "HTTP/1.1 200 OK\nContent-Length: ${LENGTH}\n\n${DATA}" | nc -l -p 8000;
On OSX you can use :
while true; do echo -e "HTTP/1.1 200 OK\n\n $(date)" | nc -l localhost 1500 ; done