Add a helper function gh_rate_limit to extract rate limit information.

[rundel]

Adds the function gh_rate_limit which returns basic rate limit information from either a gh_response object or a call to GET /rate_limit.

Currently returns a list with:

• limit - overall limit (integer)
• remaining - remaining requests (integer)
• reset - reset time for limit (datetime)

Fixes #93 and may be helpful for #67.

[gaborcsardi]

Thanks, this is great! I made some small comments below.

[codecov-io]

Codecov Report

Merging #105 into master will decrease coverage by 4.85%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master     #105      +/-   ##
==========================================
- Coverage   50.22%   45.37%   -4.86%     
==========================================
  Files          11       12       +1     
  Lines         442      454      +12     
==========================================
- Hits          222      206      -16     
- Misses        220      248      +28

Impacted Files	Coverage Δ
R/utils.R	58.33% <ø> (ø)	⬆️
R/gh_rate_limit.R	100% <100%> (ø)
R/gh_whoami.R	41.17% <0%> (-58.83%)	⬇️
R/gh_response.R	42.37% <0%> (-25.43%)	⬇️
R/package.R	58.82% <0%> (-8.83%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update a36cccb...360fb5c. Read the comment docs.

[rundel]

One other wrinkle is that depending on the end point you hit you get slightly different details (since things like search have different rate limits).

> r  = gh("/user")
> gh_rate_limit(r)
$limit
[1] 5000
$remaining
[1] 4985
$reset
[1] "2020-01-31 14:59:35 PST"

> r2 = gh("/search/repositories", q="tidyverse")
> gh_rate_limit(r2)
$limit
[1] 30
$remaining
[1] 29
$reset
[1] "2020-01-31 14:15:26 PST"

All of this is available from "/rate_limit" and looks like the possible different options are: core, search, graphql, integration_manifest, source_import. Search is the only one that "looks" different. I'm not seeing anything obvious in the response headers about what rate limit you used.

Dealing with search separately via a gh_search (#33) may make this less of an issue.

[gaborcsardi]

One other wrinkle is that depending on the end point you hit you get slightly different details (since things like search have different rate limits).

I think we can handle these separately/later, this PR is a good step already. :)

[gaborcsardi]

Thanks much!