Repository navigation
Conversation
Step 50 now writes "call filtered out (FLAGS)" for a catalog site whose only calls MEGAnE flagged. The note under the MEI section explained "not callable" and "not seen" only. Both renderers now add one sentence: MEGAnE called an insertion there but flagged it, so it is not a finding, and not proof of absence either. The Python and Rust notes stay the same, byte for byte. Bead: sv-34
… did not Two mei-report cases put a call MEGAnE flagged LC inside the FKTN window: alone (og: call filtered out (LC); bin/mei_report.py: seen), and before a PASS call in the same window (og names the PASS call; the oracle named the first). Each has a golden s.tsv, so the harness fails if og goes back to the oracle's reading, and fails as stale if the golden is changed back to what the oracle writes. The calls table is the same on both sides. Bead: sv-34
… note The MEI note changed in both renderers in dd9290f. The report-render oracle was render_report.py at 65095e9, which still had the old note, so 28 cases would differ from og only by that sentence. The oracle and the modules it imports now come from dd9290f, where the Python and Rust notes are the same: every report-render case still compares byte for byte. Bead: sv-34
The MEI note must name every status a catalog row can have, and both reports must show a "call filtered out (LC)" row with the sentence that explains it. Red on the note before dd9290f: both tests failed. Bead: sv-34
…e listed The overview table and the design brief listed seen, not seen and not callable only. docs/50-mei.md now says what the report note explains and names the new parity cases and test. The changelog says what changed in the reports and why not callable keeps its name. Bead: sv-34
|
Warning Review limit reachedYou've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository. Next included review available in 50 minutes. View limit details
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. |
This branch has not been deployed
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.
Problem
Step 50 now writes
call filtered out (LC)for a disease insertion site (FKTN, TAF1, BBS1) when MEGAnE's only calls there are ones it flagged. The reports printed that status with no word on what it means: the note under the MEI section explainednot callableandnot seenonly. The overview table and the design brief listed three statuses. And no parity case recorded thatog mei reportnow differs on purpose from the Python it replaced.Fix
rust/crates/og-render/src/notes.rsandbin/render_report.py, the same text) adds one sentence: "Call filtered out: MEGAnE called an insertion there but flagged it (LC is low confidence), so it is not a finding, and not proof of absence either."docs/pipeline-overview.md,docs/design/journal/BRIEF.mdanddocs/50-mei.mdlist the status and say what the note explains.CHANGELOG.mdhas a line.mei-reportparity cases put anLCSVA in the FKTN window: alone (filtered), and before aPASScall in the same window (filtered-pass). The generator istests/parity/gen/mei-filtered.sh. Each case has a goldens.tsvundertests/parity/golden/mei-report/: og writescall filtered out (LC)where the oracle wroteseen, and names thePASScall where the oracle named the first one. The calls table is the same on both sides.Output change: the text and HTML reports gain that one sentence in the MEI section. No column, file or status value changes.
Decisions
tests/parity/oracle.refandORACLE_SHAintests/parity/gen/report-corpus.sh). The old oracle still had the old note, so 48 items in 28 report-render cases would differ from og by that sentence alone. The alternative was 48 golden copies of whole reports. With the move, the Python and the Rust reports still compare byte for byte. The move also brings two later PRS wording commits and onecollect_summary.pycommit into the oracle; og already matches them, and every report-render case agrees (run below). This PR merges with a merge commit, so dd9290f stays in main's history.tests/parity/cases.tsv(two rows; without a row a golden is never read) andtests/parity/oracle.ref(one sha). Both are one-row mechanical files.not callablekeeps its name. The plan proposed renaming it tonot assessed. The rename would change an output value of step 50 and needsrust/crates/og-reports/src/mei.rs, its tests andtests/e2e/mei-1.sh, which are outside this package. It means "too few reads to tell" already, and 1.0.0 prefers keeping a value to renaming it. fix(mei): mark a disease insertion seen only when MEGAnE kept the call #518 made the same call.How it was checked
On a Mac mini, og built from this branch, the oracle in
PYTHON_IMAGEunder Docker:notes.rsas on main,cargo test -p og-render --test i411_mei_filtered_status_docsfailed 2 of 2 (the_note_explains_every_statusat line 52,both_reports_explain_a_filtered_out_callat the text report check). With the fix: 2 passed.optin_report12 passed.cargo fmt --checkandcargo clippy -p og-render --all-targets -- -D warningsare clean.scripts/ci/oracle-diff.sh mei-report/filtered mei-report/filtered-passgaveDIFF mei-report/filtered s.tsv(ogcall filtered out (LC), oracleseen) andDIFF mei-report/filtered-pass s.tsv(og names the 105639660 PASS call, oracle the 105639400 LC call); exit 1. With the goldens:all 90 items in 20 cases agreeovermei-reportandreport-render/m-optin-ok.STALE mei-report/filtered s.tsv: the golden is what the oracle writes: delete it, exit 1.48 of 763 items in 154 cases differ, all at the note. With the oracle at dd9290f:all 763 items in 154 cases agree.shellcheck tests/parity/gen/mei-filtered.shis clean.E2E: no e2e case runs these files' code beyond the reports; the batch tip runs
tests/e2e/mei-1.shas usual.Beads
sv-34 (the report note, the status lists and the parity record;
not callablekept, see Decisions).