Handle zipped jsons in `read_azure_json()`

[matt-dray]

Might be handy if read_azure_json() could handle zipped jsons, given that the old results format is .json.gz.

For example, I'd like something like this to work:

container <- azkit::get_container(Sys.getenv("AZ_RESULTS_CONTAINER"))

x <- azkit::read_azure_json(
  container,
  "prod/v4.2/RCX/test-md-v42-20251001_105805.json.gz"
)

Sidenote: curiously that errors with:

Error in `check_blob_exists()`:
! `check_scalar_type()`: multiple matching json files found

This is a uniquely named file, so I don't know why that error would be generated.

[matt-dray]

In nhp_tagged_runs_params we do it with jsonlite::parse_gzjson_raw(): https://github.com/The-Strategy-Unit/nhp_tagged_runs_params/blob/c1904295f648d32503d9804ce36f53ed708e4fa2/R/azure.R#L79-L88

[O-Mohammed]

I did a little digging to try and understand why I was getting issues with azkit::get_container(). I had the same issue as Matt when trying to get .json.gz results.

With a small amount of tweaking I was able to download .json.gz results. I didn't want to create a PR as I've just 'hacked' my way to getting the functionality I wanted.

I replaced the below with check_blob_exists(container, file, "json.gz", info, path)

azkit/R/read_azure_files.R

Line 45 in 2429e29

check_blob_exists(container, file, "json", info, path) |>

I replaced the below with jsonlite::parse_gzjson_raw(...)

azkit/R/read_azure_files.R

Line 48 in 2429e29

yyjsonr::read_json_raw(...)

[francisbarton]

This is going to be handled by read_azure_jsongz() once #57 is merged

[francisbarton]

The error you mentioned in your initial comment Matt is due to me thinking that

x <- character(0)
purrr::every(x, rlang::is_scalar_character)
#> [1] TRUE

would return FALSE, so the intended error message for "no file found" was not getting triggered.
I have replaced with a better, simpler check