• Home
  • Features
  • Pricing
  • Docs
  • Announcements
  • Sign In

MapColonies / shigola / 08f1e3958
44%

Build:
DEFAULT BRANCH: master
Ran 30 Aug 2026 01:27PM UTC
Jobs 1
Files 197
Run time 1min
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

30 Aug 2026 01:22PM UTC coverage: 47.414% (+0.3%) from 47.123%
08f1e3958

push

github

web-flow
refactor(atlas): make the tiling scheme explicit on the seed and encode seams (MAPCO-11486) (#10)

* refactor(atlas): pass the tiling scheme to encode, seed and purge

Encode, SeedMapTile and PurgeMapTile each resolved the grid themselves,
by reading the map's first TileMatrixSet. Callers that meant a different
scheme said so by handing them a copy of the map with its list replaced
— Map.InGrid — so the argument travelled as hidden state on a struct
that has nothing else to do with it. Two seams then had to agree about
where to look: the bytes came from one read of the map and the cache key
from another, and a divergence between them files a tile under a name
nothing serves, with no error anywhere.

Every one of those calls now takes the grid as a parameter. InGrid and
TileGrid are gone, and all three seams reject a nil grid with the same
atlas.ErrNilGrid rather than defaulting to WebMercatorQuad: a caller
that has not resolved a scheme has not decided what it is asking for.

TileGrids no longer falls back to DefaultTileGrid either. That fallback
is what put a panicking resolver on the request path — HandleTileSets
and SupportsTileGrid both reach it — to cover a map with no schemes,
which registration cannot produce: it fills the list with every
available grid when the config omits the key. Such a map now offers no
tileset and 404s, which is also the honest answer, since a tile served
under a grid nothing declared is one no tileset document describes. The
same map fails a seed run instead of being seeded on WebMercatorQuad;
both are pinned by test rather than left to be discovered.

The CLI keeps the default it had for every map it can actually be given.
It just says so itself now, rather than borrowing an accessor that made
the choice for everyone.

Cache keys are unchanged, pinned as literals in TestSeedAndPurgeKeys —
which passed against the old signatures before the refactor and passes
against the new ones after.

Refs MAPCO-11486

Co-Authored-By:... (continued)

54 of 69 new or added lines in 5 files covered. (78.26%)

2 existing lines in 1 file now uncovered.

9342 of 19703 relevant lines covered (47.41%)

210.1 hits per line

Uncovered Changes

Lines Coverage ∆ File
8
38.93
13.54% atlas/atlas.go
3
62.47
0.73% atlas/map.go
3
12.63
0.63% cmd/shigola/cmd/cache/worker.go
1
56.35
4.38% cmd/shigola/cmd/cache/seed_purge.go

Coverage Regressions

Lines Coverage ∆ File
2
38.93
13.54% atlas/atlas.go
Jobs
ID Job ID Ran Files Coverage
1 08f1e3958.1 30 Aug 2026 01:27PM UTC 197
47.41
GitHub Action Run
Source Files on build 08f1e3958
  • Tree
  • List 197
  • Changed 8
  • Source Changed 0
  • Coverage Changed 8
Coverage ∆ File Lines Relevant Covered Missed Hits/Line
  • Back to Repo
  • 08f1e395 on github
  • Prev Build on master (#C5BAC289C)
  • Next Build on master (#CBBCD55E6)
  • Delete
STATUS · Troubleshooting · Open an Issue · Sales · Support · CAREERS · ENTERPRISE · START FREE TRIAL · SCHEDULE DEMO
ANNOUNCEMENTS · TWITTER · TOS & SLA · Supported CI Services · What's a CI service? · Automated Testing

© 2026 Coveralls, Inc