I am able to upload a local file to a remote server using the command
curl -T abc.pom -uusername:password URL
But I am not able to upload a remote file to that URL. The command I am using is this
curl -T remote-file-url -uusername:password URL
Is it not possible to do this? Is downloading it and then uploading it again the only option here?
My approach:
TF=/tmp/temp && curl <REMOTE_FILE_URL> -o $TF && curl -T $TF <UPLOAD_URL> && rm -f $TF
It might be possible to pipe the content of file from 1st to 2nd cURL but then the second one has to prepare the HTML form-encoded body by itself. The -T is a shorthand for this - it creates the form and populates it directly:
curl <REMOTE_FILE_URL> | curl -i -X POST -H "Content-Type: multipart/form-data" -d #- <UPLOAD_URL>
You may curl directly from the remote host by sending a command through ssh to the destination host :
ssh user#destination_host 'curl -o destination_file_path remote_file_url'
Related
My bash code is simply this. I am trying to learn docker but also a newbie with bash scripting. I type in something simple like google.com for the read command but it gives me curl: (3) URL using bad/illegal format or missing URL. Anyone know what I am doing wrong?
docker exec -it missingDependencies sh -c "echo 'Input Website:'; read website; echo 'Searching..'; sleep 1; curl http://$website;"
Curl will give that warning when invoked like this (without a domain):
curl http://
let's define an image that has curl.
$ cat Dockerfile
FROM ubuntu:latest
RUN apt-get update
RUN apt-get install -y curl
and assemble it:
docker build . -t foobar
So now we can run your script.
$ docker run -it --rm foobar /bin/sh -c \
"set -x; echo 'Input Website:'; read website; echo 'Searching..'; curl https://$website ;"
+ echo Input Website:
Input Website:
+ read website
google.com
+ echo Searching..
Searching..
+ curl https://
Solution
docker run -it --rm foobar /bin/sh -c \
"set -x; read -p 'which website?: ' website; echo 'Searching..'; curl https://\$website;"
+ read -p which website?: website
which website?: google.com
+ echo Searching..
Searching..
+ curl https://google.com
<HTML><HEAD><meta http-equiv="content-type" content="text/html;charset=utf-8">
<TITLE>301 Moved</TITLE></HEAD><BODY>
<H1>301 Moved</H1>
The document has moved
here.
</BODY></HTML>
The problem is that when you run bash -c "echo whatever $website", the $website variable will be taken from your current environment and not from the docker environment (from read website). To counteract that the $website variable is interpolated, you could use single quotes like sh -c 'read foo; echo $foo', or escape the dollar sign: sh -c "read foo; echo \$foo"
I use the following line to create database:
curl -X POST 'http://10.1.1.1:8086/db?u=root&p=root' -d '{name: test1}
if i try to do it from shell script:
ip=10.1.1.1
curl -X POST 'http://$ip:8086/db?u=root&p=root' -d '{name: test1}'
i have a problem with shell variable substitution within single quotas, if i try to use them within double quotas:
curl -X POST "http://$ip:8086/db?u=root&p=root" -d '{name: test1}'
variable is expanded to the right value, printing in terminal
curl -X POST "http://10.1.21.1:8086/db?u=root&p=root" -d '{name: test1}': **No such file or directory**
What would be the right solution to this problem?
Try this:
ip=10.1.1.1
curl -X POST 'http://'"$ip"':8086/db?u=root&p=root' -d '{name: test1}'
I am using cURL to post a XML file on Linux as follows:
curl -X POST --header "Content-Type: text/xml" -d #test.xml "https://www.example.com"
How can I check the status of the cURL command to see if the file was posted or not?
Thanks for any help.
I'm not sure if I got you but you can basically check the return value of the curl command. The return value of the last command is stored in the variable $?.
Example:
curl -X POST --header "Content-Type: text/xml" -d #test.xml "https://www.example.com"
ret=$? # store return value for later usage in the error message
if [ $ret != 0 ] ; then
echo "POST failed with exit code $ret"
fi
This list of possible error codes can be found at the bottom of the man page. They are very helpful for debugging.
You can get the response status of your operation using -w %{http_code}
curl -s -o out.txt -w %{http_code} http://www.example.com/
In this example -s means silent mode, and -o out.txt means to save the response(usually html) into a file.
For this above command, you'l have output 200 when its success.
This works:
curl -k -d 'VARTEST=vartest2&TEXTPARAM=text1%0Atext2&COMMENT=abc22' http://localhost:8080/job/TEST1/buildWithParameters?delay=0sec
curl -X POST -k 'http://localhost:8080/job/TEST1/buildWithParameters?delay=0sec&VARTEST=vartest2&TEXTPARAM=text1%0Atext2&COMMENT=abc22'
I want to be able to write the parameters to -d into a file and run the command line like this
curl -k -d #persargfile 'http://localhost:8080/job/TEST1/buildWithParameters?delay=0sec' -o abc.html
OR like this
curl -k --data-urlencode #persargfile 'http://localhost:8080/job/TEST1/buildWithParameters?delay=0sec' -o abc.html
Question:
What should be the format of the persargfile?
As the man page states, there are several formats available:
ascii (--data-ascii)
urlencoded (--data-urlencoded)
binary (--data-binary)
It looks like for your application the ascii option is required. Meaning your file should contain:
VARTEST=vartest2&TEXTPARAM=text1%0Atext2&COMMENT=abc22
Ho to insert cookies value in curl? from firebug request headers I can see in the following
Cookie: PHPSESSID=gg792c2ktu6sch6n8q0udd94o0; was=1; uncheck2=1; uncheck3=1; uncheck4=1; uncheck5=0; hd=1; uncheck1=1"
I have tried the following:
curl http://site.com/ -s -L -b cookie.c -c cookie.c -d "was=1; uncheck2=1; uncheck3=1; uncheck4=1; uncheck5=0; hd=1; uncheck1=1" > comic
and the only thing i see in cookie.c is
PHPSESSID=gg792c2ktu6sch6n8q0udd94o0; was=1;
To pass keys/values to cURL cookie, you need the -b switch, not -d.
For the forms -d, the data will be separated by & and not by ; in your curl command.
So :
curl http://site.com/ \
-s \
-L \
-b cookie.c \
-c cookie.c \
-b "was=1; uncheck2=1; uncheck3=1; uncheck4=1; uncheck5=0; hd=1; uncheck1=1"
> comic
If you need to know the names of the forms to be POSTed, you can run the following command :
mech-dump --forms http://site.com/
It comes with libwww-mechanize-perl package with debian or derivated.