Add rudimentary DSL for Rye

[randywallace]

• bin/rye loads dsl config(s) specified on command line
• lib/rye/dsl must be loaded since functions are global
• example of working dsl in tst/dsl_example

Still much work to be done, but it works pretty well in the example.

Inspiration somewhat due to Fabric (Python) and the features implemented here:

http://tav.espians.com/fabric-python-with-cleaner-api-and-parallel-deployment-support.html

-- Let me know what you think, and whether or not I should continue, thoughts, whatnot

[delano]

Looks good. I need to think more about how rye should work (there was an executable in previous versions) but I like it so far.

[randywallace]

I see that you use http://stackoverflow.com/questions/3386233/how-to-get-exit-status-with-rubys-netssh-library to get the exit status. Likewise, its mentioned in the code that Net::SSH based Rye::Rap objects return the exit code of SSH, not the actual command. Is there any way to get the 'correct' exit code?

[delano]

I might be misunderstanding the question, but Rye does get the exit status from the executed command (via channel.on_request("exit-status"). However, the exit codes returned by standard utils differ between implementations (some versions of ls return 1 if a file path is not found, others return 0).

On Sat, Dec 29, 2012 at 11:42 AM, Randy D. Wallace Jr. < notifications@github.com> wrote:

I see that you use http://stackoverflow.com/questions/3386233/how-to-get-exit-status-with-rubys-netssh-libraryto get the exit status. Likewise, its mentioned in the code that Net::SSH based Rye::Rap objects return the exit code of SSH, not the actual command. Is there any way to get the 'correct' exit code?

— Reply to this email directly or view it on GitHubhttps://github.com/delano/rye/pull/27#issuecomment-11756849.

[randywallace]

Ah, I know what I did. Been doing this long enough that I should have remembered that grep will grep itself in the ps list. At two in the morning I should have caught that.

[delano]

No prob.

On Sun, Dec 30, 2012 at 1:32 AM, Randy D. Wallace Jr. < notifications@github.com> wrote:

Ah, I know what I did. Been doing this long enough that I should have remembered that grep will grep itself in the ps list. At two in the morning I should have caught that.

— Reply to this email directly or view it on GitHubhttps://github.com/delano/rye/pull/27#issuecomment-11762945.

[delano]

If you can remove the colorize dependency, the '#puts data.read_long' line in rye/box.rb, and (for now) bin/rye, I'll pull in the dsl for the next release.

This is definitely the direction to go. If you like I'd be happy to chat/email about next steps.

[randywallace]

I'll remove bin/rye and colorize from the Gemfile, refactor the output sections to quasi-hard-code the colorizing (which are just terminal codes) and leave it commented for future use, and fix that debug line I put in rye/box.rb (I forgot to remove that!).

I'll PM you my email.

[randywallace]

I rebased out the changes you requested. For colorizing, it just puts in the ANSI codes directly now, which should, at most, cause minor annoyance.

[delano]

Thanks. I pulled in the patch. I also removed bin/rye (as mentioned above) and the color codes. The colors can improve readability in a lot of cases but can be a pain in others (I would look for a to do it via include, a la require "rye/rgb").