Split JSON tidying into core parser and tibblify wrapper

[Copilot]

resp_tidy_json() previously always tibblified output, which made it unsuitable for simple non-tabular JSON payloads. This PR separates concerns by introducing a plain JSON tidier and moving tibblify behavior into an explicit companion API.

• API split: base JSON vs tibblify JSON

  • Added new base functions:
    • resp_tidy_json(resp, subset_path = NULL, simplifyVector = FALSE)
    • tidy_policy_json(subset_path = NULL, simplifyVector = FALSE)
  • Renamed former tibblifying functions to:
    • resp_tidy_json_tibblify()
    • tidy_policy_json_tibblify()

• Behavior changes

  • resp_tidy_json() now:
    • parses via httr2::resp_body_json()
    • supports and forwards simplifyVector
    • applies subset_path
    • returns parsed result directly (or NULL if empty)
    • no longer accepts spec / unspecified
  • resp_tidy_json_tibblify() now delegates JSON extraction to resp_tidy_json(..., simplifyVector = FALSE) and only performs tibblify-specific steps (check_installed, tibblify()).

• Docs and usage updates

  • Split source/test files to match function families:
    • R/resp_tidy_json.R
    • R/resp_tidy_json_tibblify.R
    • tests/testthat/test-resp_tidy_json.R
    • tests/testthat/test-resp_tidy_json_tibblify.R
  • Regenerated Rd/NAMESPACE exports for new names.
  • Updated examples that expect tibble output to use tidy_policy_json_tibblify().

• Example

  # raw parsed JSON (no tibblify)
  resp_tidy_json(resp, subset_path = c("message", "items"), simplifyVector = FALSE)
  
  # tibble-oriented parsing via explicit wrapper
  resp_tidy_json_tibblify(resp, subset_path = c("message", "items"))

[github-actions]

{qcthat} Report: User Acceptance

No issues are awaiting UAT.

This report was generated by this GitHub Actions job.

Last updated: 2026-06-07 16:13:30 UTC

[github-actions]

{qcthat} Reports

PR-Associated Issues

✅ A qcthat issue test matrix with 1 milestone, 1 issue, and 13 tests

└─█─Milestone: 0.1 (1 issue, 13 tests)
  └─📥─Feature 95: `resp_tidy_json()` vs `resp_tidy_json_tibblify()`
    ├─✅─resp_tidy_json_tibblify fails gracefully with a bad subset_path (#40, #95)
    ├─✅─resp_tidy_json_tibblify returns NULL for an empty body (#40, #95)
    ├─✅─resp_tidy_json_tibblify tidies a response (#40, #95)
    ├─✅─resp_tidy_json_tibblify subsets a response (#40, #95)
    ├─✅─resp_tidy_json_tibblify tidies a response with a spec (#40, #95)
    ├─✅─resp_tidy_json_tibblify calls resp_tidy_json with simplifyVector = FALSE (#40, #95)
    ├─✅─tidy_policy_json_tibblify() prepares parser for resp_tidy() (#40, #95)
    ├─✅─resp_tidy_json fails gracefully with a bad subset_path (#95)
    ├─✅─resp_tidy_json returns NULL for an empty body (#95)
    ├─✅─resp_tidy_json returns parsed JSON directly (#95)
    ├─✅─resp_tidy_json subsets a response (#95)
    ├─✅─resp_tidy_json passes simplifyVector to httr2::resp_body_json (#95)
    └─✅─tidy_policy_json() prepares parser for resp_tidy() (#95)
# Issue state: 📥 = open, ☑️ = closed (completed), ⛔ = closed (won't fix)
# Test disposition: ✅ = passed, ❌ = failed, 🚫 = skipped

✅ All tests passed

🟢 All issues have at least one test

---

Completed Issues

✅ A qcthat issue test matrix with 1 milestone, 18 issues, and 91 tests

├─█─Milestone: 0.1 (17 issues, 88 tests)
│ ├─☑️─Feature 92: `header` and `cookie` params in `req_modify()` and `req_prepare()`
│ │ ├─✅─req_modify() applies headers (#92)
│ │ ├─✅─req_modify() uses NULL headers to remove previously-set headers (#92)
│ │ ├─✅─req_modify() applies cookies (#92)
│ │ ├─✅─req_modify() removes NULL cookies (#92)
│ │ ├─✅─req_prepare() applies headers (#92)
│ │ ├─✅─req_prepare() uses NULL headers to remove previously-set headers (#92)
│ │ ├─✅─req_prepare() applies cookies (#92)
│ │ └─✅─req_prepare() removes NULL cookies (#92)
│ ├─☑️─Feature 30: Simpler cookies in req_auth_api_key()
│ │ └─✅─req_auth_api_key works for cookies (#30)
│ ├─☑️─Feature 29: User Agent Modifier
│ │ ├─✅─req_pkg_user_agent adds a package agent (#29)
│ │ ├─✅─req_pkg_user_agent adds a package agent and url (#29)
│ │ ├─✅─Corner case of no supplied name works (#29)
│ │ ├─✅─Corner case of no supplied agent works (#29)
│ │ └─✅─req_prepare() applies user agent (#10, #29)
│ ├─☑️─Feature 19: Export more autoreq pieces
│ │ ├─✅─req_prepare() uses query parameters (#5, #10, #11, #19)
│ │ ├─✅─req_prepare() uses the .multi arg (#5, #11, #19)
│ │ ├─✅─req_prepare() removes empty query parameters (#5, #11, #19)
│ │ └─✅─req_prepare() applies methods (#5, #10, #11, #19)
│ ├─☑️─Feature 11: Reconcile body parser
│ │ ├─✅─req_prepare() uses query parameters (#5, #10, #11, #19)
│ │ ├─✅─req_prepare() uses the .multi arg (#5, #11, #19)
│ │ ├─✅─req_prepare() removes empty query parameters (#5, #11, #19)
│ │ ├─✅─req_prepare() applies methods (#5, #10, #11, #19)
│ │ └─✅─resp_parse works for raw results (#11)
│ ├─☑️─Feature 10: Split call_api()
│ │ ├─✅─req_prepare() applies user agent (#10, #29)
│ │ ├─✅─req_prepare() deals with paths (#10)
│ │ ├─✅─req_prepare() uses query parameters (#5, #10, #11, #19)
│ │ ├─✅─req_prepare() uses body parameters (#10)
│ │ ├─✅─req_prepare() applies methods (#5, #10, #11, #19)
│ │ ├─✅─resp_parse parses json-containing httr2_response objects (#10)
│ │ ├─✅─resp_parse parses httr2_response objects with specified parser (#10)
│ │ ├─✅─resp_parse returns raw resp if NULL parser specified (#10)
│ │ ├─✅─resp_parse accepts parser args (#10)
│ │ └─✅─resp_parse parses lists of httr2_responses (#10)
│ ├─☑️─Feature 8: Authenticate with API Keys
│ │ ├─✅─req_auth_api_key works for header (#8)
│ │ └─✅─req_auth_api_key works for query (#8)
│ ├─☑️─Feature 6: Standardize endpoint argument
│ │ └─✅─bodies with paths are handled properly (#6)
│ ├─☑️─Feature 5: req_perform() args
│ │ ├─✅─req_prepare() uses query parameters (#5, #10, #11, #19)
│ │ ├─✅─req_prepare() uses the .multi arg (#5, #11, #19)
│ │ ├─✅─req_prepare() removes empty query parameters (#5, #11, #19)
│ │ └─✅─req_prepare() applies methods (#5, #10, #11, #19)
│ ├─☑️─Bug 90: Clean up `resp_parse()`
│ │ ├─✅─resp_parse drops NULL parsed pages before combining (#90)
│ │ └─✅─resp_parse returns NULL when all parsed pages are NULL (#90)
│ ├─☑️─Feature 86: `tidy_policy_prepare()` function
│ │ ├─✅─req_prepare() applies prepared tidying (#86)
│ │ ├─✅─req_prepare() errors for unsupported tidy policy objects (#86)
│ │ ├─✅─req_tidy_policy applies resp_body_auto by default (#44, #86)
│ │ ├─✅─req_tidy_policy applies the specified policy (#44, #86)
│ │ ├─✅─tidy_policy_body_auto() prepares resp_body_auto for resp_tidy() (#86)
│ │ ├─✅─tidy_policy_unknown() prepares resp_tidy_unknown for resp_tidy() (#86)
│ │ ├─✅─tidy_policy_prepare() constructs nectar_tidy_policy objects (#86)
│ │ ├─✅─.as_nectar_tidy_policy() returns nectar_tidy_policy objects unchanged (#86)
│ │ ├─✅─.as_nectar_tidy_policy() returns NULL for NULL input (#86)
│ │ ├─✅─.as_nectar_tidy_policy() errors for list without tidy_fn (#86)
│ │ ├─✅─.as_nectar_tidy_policy() converts function input to nectar_tidy_policy (#86)
│ │ ├─✅─.as_nectar_tidy_policy() merges tidy_args with additional tidy policy fields (#86)
│ │ └─✅─.as_nectar_tidy_policy() errors for non-listable tidy_args (#86)
│ ├─☑️─Feature 81: `auth_prepare()` function
│ │ ├─✅─auth_prepare() constructs nectar_auth objects (#81)
│ │ ├─✅─.as_nectar_auth() returns nectar_auth objects unchanged (#81)
│ │ ├─✅─.as_nectar_auth() returns NULL auth_fn and empty auth_args for NULL input (#81)
│ │ ├─✅─.as_nectar_auth() errors for list without auth_fn (#81)
│ │ ├─✅─.as_nectar_auth() converts function input to nectar_auth (#81)
│ │ ├─✅─.as_nectar_auth() merges auth_args with additional auth fields (#81)
│ │ ├─✅─.as_nectar_auth() errors for non-listable auth_args (#81)
│ │ ├─✅─auth_api_key() prepares req_auth_api_key auth (#81)
│ │ ├─✅─req_prepare() applies prepared auth (#81)
│ │ └─✅─req_prepare() errors for unsupported auth objects (#81)
│ ├─☑️─Bug 76: Remove api key on `NULL`
│ │ ├─✅─req_auth_api_key returns req unchanged if api_key is NA or empty (#76)
│ │ └─✅─req_auth_api_key removes the key when api_key is NULL (#76)
│ ├─☑️─Feature 44: resp_tidy policies
│ │ ├─✅─req_tidy_policy applies resp_body_auto by default (#44, #86)
│ │ └─✅─req_tidy_policy applies the specified policy (#44, #86)
│ ├─☑️─Feature 40: resp_tidy()
│ │ ├─✅─resp_body_auto works for json (#40)
│ │ ├─✅─resp_body_auto works for xml (#40)
│ │ ├─✅─resp_body_auto works for html (#40)
│ │ ├─✅─resp_body_auto works for svg (#40)
│ │ ├─✅─resp_body_auto works for csv (#40)
│ │ ├─✅─resp_body_auto works for tsv (#40)
│ │ ├─✅─resp_body_auto works for json subtypes (#40)
│ │ ├─✅─resp_body_auto works for other strings (#40)
│ │ ├─✅─resp_body_auto works for other things (#40)
│ │ ├─✅─resp_body_csv extracts csv data (#40)
│ │ ├─✅─resp_body_csv fails gracefully for bad data (#40)
│ │ ├─✅─resp_body_tsv extracts tsv data (#40)
│ │ ├─✅─resp_body_tsv fails gracefully for bad data (#40)
│ │ ├─✅─resp_body_separate listifies responses (#40)
│ │ ├─✅─resp_parse fails gracefully for unsupported classes (#40)
│ │ ├─✅─resp_tidy_json_tibblify fails gracefully with a bad subset_path (#40, #95)
│ │ ├─✅─resp_tidy_json_tibblify returns NULL for an empty body (#40, #95)
│ │ ├─✅─resp_tidy_json_tibblify tidies a response (#40, #95)
│ │ ├─✅─resp_tidy_json_tibblify subsets a response (#40, #95)
│ │ ├─✅─resp_tidy_json_tibblify tidies a response with a spec (#40, #95)
│ │ ├─✅─resp_tidy_json_tibblify calls resp_tidy_json with simplifyVector = FALSE (#40, #95)
│ │ ├─✅─tidy_policy_json_tibblify() prepares parser for resp_tidy() (#40, #95)
│ │ ├─✅─resp_tidy parses json-containing httr2_response objects (#40, #88)
│ │ ├─✅─resp_tidy parses httr2_response objects with resp_tidy policy (#40, #88)
│ │ └─✅─resp_tidy uses policies$resp_tidy$tidy_args (#40, #88)
│ ├─☑️─Feature 35: req_pagination()
│ │ ├─✅─req_pagination_policy errors informatively for bad fn (#35)
│ │ ├─✅─req_pagination_policy applies the specified policy (#35)
│ │ ├─✅─choose_pagination_fn returns NULL for no policy (#35)
│ │ └─✅─choose_pagination_fn extracts the pagination_fn (#35)
│ └─☑️─Feature 32: Feature: smarter cursor pagination
│   ├─✅─iterate_with_json_cursor repeats with non-empty cursor (#32)
│   ├─✅─iterate_with_json_cursor stops with NULL cursor (#32)
│   └─✅─iterate_with_json_cursor stops with blank cursor (#32)
└─█─Milestone: <none> (1 issue, 6 tests)
  └─☑️─Issue 88: Clarify `resp_parse()` vs `resp_tidy()`
    ├─✅─resp_parse fails gracefully for list of non-responses (#88)
    ├─✅─resp_tidy returns NULL for NULL input (#88)
    ├─✅─resp_tidy errors for non-response input (#88)
    ├─✅─resp_tidy parses json-containing httr2_response objects (#40, #88)
    ├─✅─resp_tidy parses httr2_response objects with resp_tidy policy (#40, #88)
    └─✅─resp_tidy uses policies$resp_tidy$tidy_args (#40, #88)
# Issue state: 📥 = open, ☑️ = closed (completed), ⛔ = closed (won't fix)
# Test disposition: ✅ = passed, ❌ = failed, 🚫 = skipped

✅ All tests passed

🟢 All issues have at least one test

🙈 7 issues with label "qcthat-nocov" were ignored

---

This report was generated by this GitHub Actions job.

Last updated: 2026-06-07 16:13:31 UTC

[github-actions]

🌐 PR pkgdown deployed

This report was generated by this GitHub Actions job.

Last updated: 2026-06-07 16:13:54 UTC

[jonthegeek]

@copilot address these change requests.

[Copilot]

@copilot address these change requests.

Addressed in 80ef569.

Updated the roxygen inheritance in R/resp_tidy_json.R, moved rlang::check_installed() inside the non-empty branch in resp_tidy_json_tibblify(), retagged tibblify tests as (#40, #95), and removed nectar_0.0.0.9007.tar.gz from the repo.