Minor Changes
-
#48
e58f17aThanks @alerizzo! - Show the repository's coverage status, not just its percentage.Codacy now reports whether a repository's coverage is up to date, still waiting
on a report, has stopped receiving them, or was never set up — and the CLI can
tell those apart:codacy reposmarks a repository whose latest commit has no report yet with a
dim⋯after its last known value, and shows a dim⊘instead of a number
for one that has stopped receiving reports. A legend under the table explains
only the states actually present in the listing.codacy repo's Metrics section spells the same states out, with the date and
commit of the last report, and notes when a stopped repository's coverage gate
is no longer being enforced. A repository that never had coverage now reads
Not set uprather than a bareN/A.codacy repo's Analysis row reads coverage state from the API's own status
field instead of inferring it from a separate request. This fixes repositories
that were reported as healthy while showing a stale percentage, drops one
request per run, and makes the coverage state available under a repository
token for the first time.
--output jsongainscoverage.status,coverage.lastCommitWithCoverage,
coverage.statusUpdatedAtandcoverage.valueUpdatedAton both commands. Under
a repository token,codacy repo'sunavailablearray is now["pullRequests"]
only. -
#45
e21f321Thanks @alerizzo! - New-k, --matches-stack [value]filter oncodacy patterns, which narrows a tool's code patterns to those that do (or don't) match the repository's detected stack.It's a tri-state flag, the same shape as
issues --false-positives:codacy patterns eslint9 --matches-stack # only patterns matching the repo stack codacy patterns eslint9 --matches-stack true # same codacy patterns eslint9 --matches-stack false # only patterns that don't match codacy patterns eslint9 # unfiltered
The filter applies in bulk mode too, so
--enable-all/--disable-allcan be scoped to the stack:codacy patterns eslint9 --disable-all --matches-stack falseThe summary printed after a bulk update still reports counts for the whole tool, not just the updated subset.
Only
trueandfalseare accepted as values. Because Commander's optional-value syntax consumes the next token, a lax parser would letcodacy patterns gh org repo --matches-stack eslintsilently swallow the tool name and then fail with a confusing positional-count error; the flag now rejects non-boolean values with a message that says what to do instead.