[PR-11] Output formats, dev infra, final polish

Final polish PR for the neoipcr upstreaming stack:
- `write_json()` for plain-JSON neoipcr interchange (minimum subset); promotes `jsonlite` to Imports.
- Deprecated-API sweep: `dplyr::case_match()` → `dplyr::recode_values()`; fix tidyselect `.data\$` usage.
- Real-data fixes: NA-safe `drop_quartiles` gate; VRE genus filter; split `unknownPathogenNames` before `finalize_to_schema` strips `name`.
- Adopt `data/` as the protected local-only directory convention (was `data/local/`); mechanical safeguards via `.gitignore` + `.Rbuildignore`.
- Universal-tier guardrails: doc-comments in-band (not as a sweep), namespace-qualification, plus 8 others.
- WIP `API-DESIGN.md` checked in as a living draft.

Author: Brar

.Rbuildignore

@@ -5,10 +5,12 @@
 ^\.Rproj\.user$
 ^LICENSE\.md$
 ^README\.Rmd$
+^data$
 ^data-raw$
 ^tools$
 ^\.covrignore$
 ^CLAUDE\.md$
+^API-DESIGN\.md$
 ^coverage\.html$
 ^lib$
 ^scripts$

.github/copilot-instructions.md

@@ -57,3 +57,6 @@ rsconnect/
 # Local coverage report
 coverage.html
 lib/
+
+# Local-only sensitive data (the protected-dir convention)
+data/

.gitignore

@@ -17,11 +17,12 @@ LazyData: true
 Roxygen: list(markdown = TRUE)
 Depends:
     R (>= 4.4)
-Imports: 
+Imports:
     askpass (>= 1.2.1),
     dplyr (>= 1.1.4),
     httr2 (>= 1.1.1),
     ids (>= 1.0.1),
+    jsonlite (>= 1.9.1),
     lubridate (>= 1.9.4),
     purrr (>= 1.0.4),
     readr (>= 2.1.5),
@@ -33,7 +34,6 @@ Imports:
     uuid (>= 1.2-1)
 Suggests:
     covr (>= 3.6.0),
-    jsonlite (>= 1.9.1),
     roxygen2 (>= 8.0.0),
     testthat (>= 3.0.0),
     withr (>= 2.5.0),
@@ -67,6 +67,7 @@ Collate:
     'import-dhis2.R'
     'import-standalone-obj-type.R'
     'import-standalone-types-check.R'
+    'json.R'
     'neoipcr-package.R'
     'pathogens.R'
     'scales.R'

API-DESIGN.md

@@ -27,5 +27,6 @@ export(is_valid_ichi_code)
 export(neoipc_poisson_ci)
 export(neoipc_wilson_ci)
 export(pretty_names)
+export(write_json)
 import(rlang)
 importFrom(rlang,.data)

CLAUDE.md

@@ -854,15 +854,15 @@ pretty_names.neoipcr_tbl_sr_ref <- function(x, ...) {
     dplyr::inner_join(pairs, dplyr::join_by("pro_cat")) |>
     dplyr::mutate(pro_cat = .data$pretty_name, .keep = "unused") |>
     dplyr::rename_with(
-      ~ dplyr::case_match(
+      ~ dplyr::recode_values(
         .x,
         "pro_cat"~col_names[["pro_cat"]],
         "n"~col_names[["n"]],
         "pooled"~col_names[["pooled"]],
         "q1"~col_names[["q1"]],
         "q2"~col_names[["q2"]],
         "q3"~col_names[["q3"]],
-        .default = .x))
+        default = .x))
 }
 
 quartile_probs <- c(0.25,0.5,0.75)

DESCRIPTION

@@ -210,7 +210,7 @@ get_procedure_category <- function(x, not_surgery_na = FALSE) {
 }
 
 get_procedure_category_pretty <- function(x) {
-  dplyr::case_match(
+  dplyr::recode_values(
     as.character(x),
     "overall" ~ gettext("Overall"),
     "abdominal_surgery" ~ gettext("Abdominal surgery"),
@@ -222,6 +222,6 @@ get_procedure_category_pretty <- function(x) {
     "other" ~ gettext("Other"),
     "not_surgery" ~ gettext("Not a surgical procedure"),
     "to_be_categorised" ~ gettext("Not yet categorised"),
-    .default = x
+    default = x
   )
 }

NAMESPACE

@@ -17,7 +17,7 @@ get_dev_ass_incidence_density_rates <- function(
       dplyr::select("event_key", "dev_ass") |>
       dplyr::filter(.data$dev_ass != 0) |>
       dplyr::mutate(
-        dev = dplyr::case_match(
+        dev = dplyr::recode_values(
           as.integer(as.character(.data$dev_ass)),
           !!!dev_map),
         .keep = "unused")
@@ -144,7 +144,9 @@ get_infectious_agent_detection_rates_with_department_quartiles <- function(
       use_cache = use_cache) |>
     dplyr::select(tidyselect::all_of(c(group_cols, "n", "inf_with_pathogen")), rate = "n_per_iwp") |>
     dplyr::mutate(
-      drop_quartiles = n_deps < 5 | round(100 / .data$rate) >= median_inf_with_pathogen)
+      drop_quartiles = tidyr::replace_na(
+        n_deps < 5 | round(100 / .data$rate) >= median_inf_with_pathogen,
+        TRUE))
 
   if(nrow(r1) < 1)
   {
@@ -427,7 +429,7 @@ get_resistance_test_rate_with_department_quartiles <- function(
         na.rm = TRUE)) |>
     dplyr::mutate(
       name=names(.data$value),
-      name=dplyr::case_match(
+      name=dplyr::recode_values(
         .data$name,
         "25%"~"q1",
         "50%"~"q2",
@@ -445,7 +447,9 @@ get_resistance_test_rate_with_department_quartiles <- function(
 
   rate <- rate |>
     dplyr::mutate(
-      drop_quartiles = .data$n_deps < 5 | round(100 / .data$rate) >= .data$median,
+      drop_quartiles = tidyr::replace_na(
+        .data$n_deps < 5 | round(100 / .data$rate) >= .data$median,
+        TRUE),
       q1 = dplyr::if_else(
         .data$drop_quartiles,
         NA,
@@ -518,7 +522,7 @@ get_resistance_test_rate <- function(
       dplyr::across(
         tidyselect::all_of(resistance),
         ~ factor(
-          dplyr::case_match(
+          dplyr::recode_values(
             as.character(.x),
             "yes" ~ "tested",
             "no" ~ "tested",
@@ -609,7 +613,7 @@ get_resistance_rate_with_department_quartiles <- function(
           na.rm = TRUE)) |>
       dplyr::mutate(
         name=names(.data$value),
-        name=dplyr::case_match(
+        name=dplyr::recode_values(
           .data$name,
           "25%"~"q1",
           "50%"~"q2",
@@ -628,7 +632,9 @@ get_resistance_rate_with_department_quartiles <- function(
 
   rate <- rate |>
     dplyr::mutate(
-      drop_quartiles = .data$n_deps < 5 | round(100 / .data$inf_rs_rate) >= .data$median,
+      drop_quartiles = tidyr::replace_na(
+        .data$n_deps < 5 | round(100 / .data$inf_rs_rate) >= .data$median,
+        TRUE),
       q1 = dplyr::if_else(
         .data$drop_quartiles,
         NA,
@@ -721,7 +727,7 @@ get_organism_resistance_rate_with_department_quartiles <- function(
           na.rm = TRUE)) |>
       dplyr::mutate(
         name=names(.data$value),
-        name=dplyr::case_match(
+        name=dplyr::recode_values(
           .data$name,
           "25%"~"q1",
           "50%"~"q2",
@@ -740,7 +746,9 @@ get_organism_resistance_rate_with_department_quartiles <- function(
 
   rate <- rate |>
     dplyr::mutate(
-      drop_quartiles = .data$n_deps < 5 | round(100 / .data$ia_rs_rate) >= .data$median,
+      drop_quartiles = tidyr::replace_na(
+        .data$n_deps < 5 | round(100 / .data$ia_rs_rate) >= .data$median,
+        TRUE),
       q1 = dplyr::if_else(
         .data$drop_quartiles,
         NA,