Part of #81. Gated on the next aozora pin bump that crosses feat/devex-inner-loop. Lightweight / optional.
Background
The 12 new authoring diagnostics (aozora f810ee0 8 "clean" + 25dc003
4 model-dependent) already surface through afm with no code change:
render_to_string returns them in Rendered.diagnostics, and afm check
prints each generically as diagnostic [code]: …
(crates/afm-cli/src/main.rs). So this issue is polish, not a blocker.
Optional work
- Parity with aozora's
aozora check (6ccbc22, graphical + json/short
output): add --format {human,json,short} to afm check in
crates/afm-cli/src/main.rs; human stays the default
(back-compatible).
- Document the 12 diagnostic codes (coordinate with the docs issue).
- Extend
crates/afm-cli/tests/cli_integration.rs — assert at least one
new code round-trips, and (if added) the --format json shape.
Decision to make
Whether afm-cli should own a --format flag at all, or defer diagnostic
formatting entirely to downstream consumers of the library
diagnostics. Resolve before implementing.
Re-confirm the diagnostic code strings at the pinned SHA.
Part of #81. Gated on the next aozora pin bump that crosses
feat/devex-inner-loop. Lightweight / optional.Background
The 12 new authoring diagnostics (aozora
f810ee08 "clean" +25dc0034 model-dependent) already surface through afm with no code change:
render_to_stringreturns them inRendered.diagnostics, andafm checkprints each generically as
diagnostic [code]: …(
crates/afm-cli/src/main.rs). So this issue is polish, not a blocker.Optional work
aozora check(6ccbc22, graphical + json/shortoutput): add
--format {human,json,short}toafm checkincrates/afm-cli/src/main.rs;humanstays the default(back-compatible).
crates/afm-cli/tests/cli_integration.rs— assert at least onenew code round-trips, and (if added) the
--format jsonshape.Decision to make
Whether afm-cli should own a
--formatflag at all, or defer diagnosticformatting entirely to downstream consumers of the library
diagnostics. Resolve before implementing.