Repository navigation
obo: the RF2 snapshot reader — one parser, three refusals - #262
Merged
Merged
Conversation
The reader every SNOMED-package consumer would otherwise duplicate wrongly, deliberately column-blind past the shared five (the file family is in the FILENAME; the consumer knows which file it opened). What it owns is the part that goes wrong silently: - the active filter, with the excluded count REPORTED so a bake census can say what it left out instead of looking short; - snapshot uniqueness — a duplicate id means Full export or corrupt concatenation, and the whole file is refused rather than keeping whichever row the file order favored; - column integrity — a ragged row refuses with its line number, because a permissive reader mis-aligns every cell after a stray tab and each still LOOKS like a value. SCTIDs read as u64 (a real 54-bit id is in the fixtures); non-numeric or absent cells are refused, never truncated. No meaning is interpreted: no concept model, no map semantics — types exist only before the bake, and this is the join stage's input surface. Generated by [Claude Code](https://claude.ai/code)
|
You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard. |
Bugbot couldn't run - usage limit reachedBugbot is counted against Cursor usage for this user or team, and this run hit a usage or spend limit. A user or team admin can review and increase usage limits in the Cursor dashboard. (requestId: serverGenReqId_05e998ee-fd8b-4d34-8516-9e0e07bfa30c) |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The
rf2-readerrow from medcare's open-points dependency map — the one parser gating three downstream bakes (GP/FP, the MedDRA maps both directions, EDQM), built here because consumers consume and never hand-parse.Deliberately column-blind
Past the five columns every
der2_*refset shares, the reader interprets nothing — the file family is encoded in the filename (sRefset,iissscc, …) and the consumer knows which file it opened. No concept model, no map semantics, no vocabulary: types exist only before the bake, and this is the join stage's input surface.What it owns: the three silent failure modes
inactiveis a field.idmeans a Full export or a corrupt concatenation; keeping either row would bake whichever the file order favored, so the whole file is refused with the id named.SCTIDs read as
u64— a real 54-bit id sits in the fixtures — and a non-numeric or absent cell is refused, never truncated.Verified: 17/17
ogar-obotests (4 new, each naming its falsifier).🤖 Generated with Claude Code
Generated by Claude Code