Bug #13452
closed
pasta does not properly report "startTile" flag field when tilepath is only one tile long
Description
Tile path 0x0021 has only one tile in it and the emitted FastJ tile from the "pasta" tool does not correctly report that the emitted tile is a "startTile". The "endTile" flag is properly set and there is correctly no "startTag" (or a zero length "startTag").
I believe the culprit is in the pasta_fastj.go file. There's a condition when it's processing the end tile to make sure it fills in the beginning tag with the appropriate value if it's able. Instead of doing nothing in the case of not being able to fill the beginning tag, it should set the "start_tile_flag" to true so the appropriate field can be populated in the FastJ header.
It's a special case of when the tile path is only one tile long so the error is not critical but it would be nice to fix.
Updated by Abram Connelly almost 6 years ago
- Status changed from New to In Progress
Updated by Abram Connelly almost 6 years ago
Added test data and tests. Currently the FastJ tests need some files not stored in the repo, like the tile assembly and the tagset, so it might be difficult for outside parties to test. We should add what we can as small test files so the pasta program is as isolated as possible and can be tested independent of the (large-ish) data files.
#13485 is the review ticket. Since the fix was relatively small and I was in the neighborhood, I also fixed #12245 which is included in that review ticket.
Updated by Jiayong Li
Updated by Abram Connelly almost 6 years ago
- Status changed from In Progress to Closed