Fix solana-install init making unnecessary API requests

[acheroncrypto]

Problem

Using solana-install init makes API requests to GitHub even when the given version is already installed locally.

Summary of Changes

Only fetch releases from GitHub if the given version is not installed.

Fixes #33948

[CriesofCarrots]

Thanks for taking a crack at this!
This isn't quite right yet, as it breaks some of the other commands (solana-install info, solana-install update) that ultimately call into fn init_or_update().

Before:

$ solana-install init v1.16.13
  ✨ 1.16.13 initialized
$ solana-install info
Configuration: /.config/solana/install/config.yml
Active release directory: /.local/share/solana/install/active_release
• Release version: 1.16.13
• Release URL: https://github.com/solana-labs/solana/releases/download/v1.16.13/solana-release-aarch64-apple-darwin.tar.bz2
• Release commit: b2a38f6
  🎁 Update available: 1.16.18

With this patch:

$ cargo run --bin solana-install init v1.16.13
    Finished dev [unoptimized + debuginfo] target(s) in 0.22s
     Running `solana/target/debug/solana-install init v1.16.13`
  ✨ 1.16.13 initialized
$ cargo run --bin solana-install info
    Finished dev [unoptimized + debuginfo] target(s) in 0.22s
     Running `solana/target/debug/solana-install info`
Configuration: /.config/solana/install/config.yml
Active release directory: /.local/share/solana/install/active_release
• Release version: 1.16.13
• Release URL: https://github.com/solana-labs/solana/releases/download/v1.16.13/solana-release-aarch64-apple-darwin.tar.bz2
• Release commit: b2a38f6
  🎁 Update available: 1.16.13

I think that we need some flag to ensure your github-check bypass is only executed when the call comes from fn init(). There are probably alternatives that involve refactoring fn init_or_update(), which is kind of a kitchen sink.

[acheroncrypto]

I think that we need some flag to ensure your github-check bypass is only executed when the call comes from fn init(). There are probably alternatives that involve refactoring fn init_or_update(), which is kind of a kitchen sink.

I added is_init check(already existed in the arguments) so this change should only impact the init command now. I wanted to avoid refactoring for this small change since it would be harder to review and I also didn't want to potentially break something(failed 😅).

I didn't test the other commands, sorry for that, and thanks for pointing out some of them broke.

[CriesofCarrots]

already existed in the arguments

Oh cool, I forgot we had that already 😅

I just kicked off CI on this, so keep an eye out in case there are any complaints. Looking good, though.

[codecov]

Codecov Report

Merging #33949 (e611e06) into master (662ac8b) will decrease coverage by 0.1%.
Report is 11 commits behind head on master.
The diff coverage is 0.0%.

@@            Coverage Diff            @@
##           master   #33949     +/-   ##
=========================================
- Coverage    81.9%    81.9%   -0.1%     
=========================================
  Files         809      809             
  Lines      219070   219074      +4     
=========================================
- Hits       179476   179466     -10     
- Misses      39594    39608     +14     

[CriesofCarrots]

Thanks for this contribution, @acheroncrypto !