archetech / archon
92%
main: 94%

LAST BUILD BRANCH: ci/gatekeeper-parity
DEFAULT BRANCH: main
Repo Added
Token VKRdMqMwqPTOZprp4p5Cb3v3uLt6LWUtk regen
Build 1946 Last
Files 88
Badge
Embed ▾
README BADGES
x

If you need to use a raster PNG badge, change the '.svg' to '.png' in the link

Markdown

Textile

RDoc

HTML

Rst

LAST BUILD ON BRANCH test/ipfs-kubo-client
branch: test/ipfs-kubo-client
CHANGE BRANCH
x
Sync Branches

coverage: 91.949% (+1.3%) from 90.652%
30720144293

Pull #824

github

macterra 
test(ipfs): cover the kubo client

packages/ipfs/src/kubo-client.ts had zero coverage — the second-largest
gap surfaced by collectCoverageFrom in #822, and the largest remaining
after #823. 31 tests take it to 99.1% (115/116).

The class is a thin wrapper over kubo-rpc-client, so the tests mock that
one module and exercise everything else for real: connect and the
readiness wait, content add/get for text, binary and streams, JSON block
round-trips, identity, peer dialling, and peering configuration.

Notable coverage:

- addJSON derives the CID locally from the encoded block rather than
  trusting the daemon, so the tests assert equal payloads yield equal
  CIDs, different payloads differ, and the stored block decodes back to
  the original.
- getAddresses filters private addresses; covered across IPv4 loopback
  and RFC1918, IPv6 loopback, public v4 and v6, and a /dns4/ multiaddr
  that matches no ip pattern and is dropped.
- addPeer rejects a malformed address or a private IP *without dialling*,
  which is the SSRF-shaped guard, plus the unrecognized-response and
  dial-failure paths.
- addPeeringPeer skips self, keeps only dialable addresses, writes
  nothing when none are dialable, and merges into an existing entry
  without duplicating.

waitUntilReady is driven with intervalSeconds: 0 and a maxRetries bound
so the retry loop neither sleeps nor leaves a timer behind — CI runs
without --forceExit.

No findings this time: unlike the sqlite backend, the error handling here
matches what the types promise. Repo total 93.84% -> 95.19%.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Pull Request #824: test(ipfs): cover the kubo client

3458 of 4060 branches covered (85.17%)

Branch coverage included in aggregate %.

8100 of 8510 relevant lines covered (95.18%)

668.99 hits per line

Relevant lines Covered
8510 RELEVANT LINES 8100 COVERED LINES
668.99 HITS PER LINE
Source Files on test/ipfs-kubo-client

Recent builds

Builds Branch Commit Type Ran Committer Via Coverage
30720144293 test/ipfs-kubo-client test(ipfs): cover the kubo client packages/ipfs/src/kubo-client.ts had zero coverage — the second-largest gap surfaced by collectCoverageFrom in #822, and the largest remaining after #823. 31 tests take it to 99.1% (115/116). The class is a thin... Pull #824 macterra github
91.95
See All Builds (1939)