diff options
| author | user <user@clank> | 2026-08-27 15:17:48 +0200 |
|---|---|---|
| committer | user <user@clank> | 2026-08-27 15:17:48 +0200 |
| commit | 7ceb4e1492922fc33d0f9d13d01eec72bfda2363 (patch) | |
| tree | f45fe0d7c953883fcc00f43c774207845850ea50 /tests/test_corner.py | |
| parent | selection: drag past an edge scrolls the viewport under it (diff) | |
| download | slosh-7ceb4e1492922fc33d0f9d13d01eec72bfda2363.tar.gz slosh-7ceb4e1492922fc33d0f9d13d01eec72bfda2363.tar.xz slosh-7ceb4e1492922fc33d0f9d13d01eec72bfda2363.zip | |
test_splash: 'landed' is the row in its final form
The h-centered check raced the assembly it was measuring. landed_row()
answered as soon as one occurrence of the opening run had a block glyph
at its start -- but the logo's first line contains that run three times,
assembly is staggered, and under a parallel suite's load the middle
section can finish whole checks before the leftmost. The test then
measured a middle occurrence's x (24) and called the logo off-centre.
The row is 'landed' when all three occurrences are present -- its final
form -- and the first block glyph is still required to sit where the
match begins, so a particle passing to the left keeps us waiting rather
than shifting the answer.
Diffstat (limited to 'tests/test_corner.py')
0 files changed, 0 insertions, 0 deletions