req is an opinionated HTTP scripting language. It is designed for easily making HTTP requests, and working with their responses. Below is an example that calls out to the GitHub API and displays the user making the call,
$ cat gh.req
Stderr = open "/dev/stderr";
Endpoint = "https://api.github.com";
Token = env "GH_TOKEN";
if $Token == "" {
writeln $Stderr "GH_TOKEN not set";
exit 1;
}
Headers = (
Authorization: "Bearer $(Token)",
);
Resp = GET "$(Endpoint)/user" $Headers -> send;
match $Resp.StatusCode {
200 -> {
User = decode json $Resp.Body;
writeln _ "Hello $(User["login"])";
}
_ -> {
writeln $Stderr "Unexpected response: $(Resp.Status)";
exit 1;
}
}
$ GH_TOKEN=1a2b3c4d5ef req gh.req
This language hopes to fill in a gap when it comes to writing scripts for working with an HTTP service. Typically, you have a choice between a shell script that utilizes cURL, or a programmaing language and any HTTP library that may come with it.
The cURL approach can work, for simple one off requests, but when you want to do something more with the response you're left with having to munge that data with jq, grep, sed, or awk (or all of the above). Using a programming language gives you more control, but can be more cumbersome as you have far many more knobs to turn.
req provides a middleground between the two. A limited syntax, with builtin commands for working with any data you want to send/received. For more details on how to start working with req, then refer to the documentation, or you can dive right in by looking over the examples.