shell commands - lexer rules, pipe to shell cmd redirection, tests

[fogine]

As title says... lexer rules for shell commands has been fully implemented... I've added some more tests for this as well...

I've a question, though.
Now, it's possible to define a shell command in full single/double-quoted request mutation. For example, you can do this:

httpie post "`echo name`=value"

Is that the expected behavior, @eliangcs ?
If so, I'll add tests for this...
ps: I'm asking because I couldn't think of a reason why the full single/double-quoted request mutation feature was implemented in first place...

[eliangcs]

Thanks, @fogine ! This was something I wanted to do in PR #61.

For your question, yes, it's possible to write a shell command in a full quoted mutation, see the test cases here: https://github.com/eliangcs/http-prompt/blob/e0308a106f02b20e3e62b17de2d1e344d1c4d624/tests/test_execution.py#L647

Full quoted mutation was implemented because it works on httpie. For instance, http post http://httpbin.org/post "my message=hello world".

[fogine]

I've added the "pipe to shell command redirection" to this pull request as discussed in #61 PR.
Things which works now:

localhost> httpie get | tee /tmp/test # output of `httpie get` is redirected to shell subprocess cmd: `tee /tmp/test`
localhost> get some==data| tee /tmp/test # supports "action" commands

localhost> httpie | `echo "sed 's/localhost/127.0.0.1/' "` # this will be "transformed" to httpie | sed 's/localhost/127.0.0.1/' .... Also this is the case where I'm not sure if it's implemented in the best possible way

Lexer rules covers all the possible options now.
Tests updated/added...

I think this PR implements all the remaining functionality to the #61 PR now...?

[eliangcs]

@fogine Sorry for my late reply. I'll refactor the code first (in branch eliangcs/refactor-source-and-exec) before I can review this PR. It'll take some time. Thank you for your patience in advance :)

[eliangcs]

Overall looks good. Merged. We can keep fixing details after I resolve the conflicts between PR #61 and the master branch. Thanks!