Same project config for multiple items

[thibaultduponchelle]

The idea is :

• Define one project dir with a sparrowfile and a sparky.yaml
• Use it for multiple items (repos or modules) through a list (implementation to be defined) (*)

(*) An extra feature would be github/bitbucket/gitlab org/project/whatever scanning

[thibaultduponchelle]

Is the answer "downstream projects" ?

[melezhik]

if you please give more practical example, how this could be used ... ?

[thibaultduponchelle]

If you have a github org with 100 projects that compiles with ./configure && make && make install
You have to create 100 directories with same files ?

I want one project to scan them all or one project to apply to a list of names (here repo names).

[melezhik]

probably not exactly what we're talking about but this where FTP protocol could be useful.

You can spin up an supplemental application which accepts requests (say through a web hook) and dispatch all of them to the same Sparky build bypassing repo specific parameters.

For example:

.sparky/projects/make/sparky.yml

allow_manual_run: false # this project only built though an supplemental web app, using sparky  FTP proto
sparrowdo:
  docker: debian 
  no_sudo: true  

.sparky/projects/make/sparrowfile

say conf()<repo>;
say conf()<commit>;
# do something with repo and commit, you got the idea

Application's cro http route:

post -> 'queue', :%params {
 
  my $sparky-dir = "{%*ENV<HOME>}./sparky/projects/make";

  my $rnd = ('a' .. 'z').pick(20).join('');

  my $id = "{$rnd}{$*PID}";

  request-body -> (:$repo, :$commit ) {

     # this directory is used as to keep build unique configuration  
     my $build-dir = "/some/local/dir/${$id}"; 
     
    # write up build configuration
    spurt "{$build-dir}/config.pl6", %(
        repo => $repo,
        commit => $commit
     ).perl;

    # write up a trigger data with sparrowdo configuration and build CWD
    spurt "{$sparky-dir}/.triggers/$id", %(
       sparrowdo => %( 
           conf => "{$build-dir}/config.pl6", 
         ),
         description => "build for {$repo and $commit}",
        ).perl
    );
  }
}

You could also see how this is implemented in RakuPlay - https://github.com/melezhik/RakuPlay/blob/main/lib/RakuPlay.rakumod#L40-L61.

[melezhik]

Of course this supplemental application should have a write access to Sparky file system, because it's File Triggering Protocol 😄