Skip to content

test(integration): metadata accessor parity, one module per RS_ function - #1236

Draft
james-willis wants to merge 1 commit into
apache:mainfrom
james-willis:jw/parity-metadata-accessors
Draft

test(integration): metadata accessor parity, one module per RS_ function#1236
james-willis wants to merge 1 commit into
apache:mainfrom
james-willis:jw/parity-metadata-accessors

Conversation

@james-willis

Copy link
Copy Markdown
Contributor

Probe-first parity for the eleven metadata scalar accessors (RS_Width, RS_Height, RS_NumBands, RS_ScaleX/Y, RS_SkewX/Y, RS_UpperLeftX/Y, RS_SRID, RS_BandPixelType), one module per function, RS-only SQL. Every value probed bit-identical across engines — standard, sheared (raw-transform), bottom-up, and EPSG:3857 fixtures; pixel-type names identical for all six GeoTIFF dtypes; band addressing checked on band 2. 17 passed, all anchored, no xfails or tolerances. Locally verified with pyspark 4.0.4 + Sedona 1.9.1; CI runs the parity lane.

RS_Width, RS_Height, RS_NumBands, RS_ScaleX/Y, RS_SkewX/Y,
RS_UpperLeftX/Y, RS_SRID, and RS_BandPixelType. Probed first: every
value is bit-identical across engines on the standard, sheared,
bottom-up, and EPSG:3857 fixtures, and both engines spell the
PostGIS-style pixel-type names identically for all six GeoTIFF dtypes —
17 anchored passing tests, no xfails.
@github-actions
github-actions Bot requested a review from paleolimbot September 4, 2026 20:51
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant