Retain discovery prefixes in usage diagnostics
This commit is contained in:
@@ -98,7 +98,7 @@ invocation metadata were added.
|
||||
|
||||
## R-2: Retain discovery-prefix context in usage diagnostics
|
||||
|
||||
Status: open. Priority: P3. Category: presentation consistency.
|
||||
Status: complete. Priority: P3. Category: presentation consistency.
|
||||
|
||||
**Evidence:** Discovery parsing uses the backend root with only trailing flag
|
||||
tokens, so parser failures have an empty canonical path.
|
||||
@@ -130,6 +130,14 @@ identify `demo list` in usage. Repeat through a composed group shortcut after
|
||||
R-1. Check identifying fragments rather than exact formatting, and confirm that
|
||||
ordinary root and canonical command diagnostics retain their existing behavior.
|
||||
|
||||
**Completed:** Usage presentation now uses the selected discovery prefix rather
|
||||
than asking the canonical backend for its usage. Regressions first failed for
|
||||
all three error forms at both direct and chained prefixes, and now retain the
|
||||
entered prefix and flag spelling without inventing a canonical identity.
|
||||
Usage status 2, completed diagnostic delivery, metadata-only failures, and
|
||||
ordinary root/canonical usage remain covered. The complete offline root suite
|
||||
passed on Go 1.26.8.
|
||||
|
||||
## R-3: Preserve empty and dash operands before child-name tokens
|
||||
|
||||
Status: open. Priority: P2. Category: inherited parser/routing inconsistency.
|
||||
|
||||
Reference in New Issue
Block a user