Repository navigation
fix: reject trailing input in beacon, gnomAD and SPDI translators - #670
Merged
Merged
Conversation
The translator regexes were applied with re.match, which only anchors at the start. Input with trailing content was silently truncated, so a multi-allelic gnomAD expression like 1-55516888-G-GA,GT was translated as G>GA. Use fullmatch so such input is rejected as unparseable.
korikuzma
requested changes
Oct 19, 2026
korikuzma
left a comment
korikuzma
left a comment
Contributor
There was a problem hiding this comment.
Thanks @syahra712 for the great catch (surprised it's gone this long unnoticed)
Changes related to this PR look good. Requesting that the generated test files be removed from this PR (I have a fix in #671 so that these files are written to temp dirs)
Contributor
Author
|
Thanks! Both generated files are removed in 47ef310, so the PR now only touches |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Bug
AlleleTranslatorsilently accepts beacon, gnomAD and SPDI expressions that have extra content after the variant, and translates only the leading part. For example:returns the allele for
G>GAand quietly drops the second ALT. A caller passing a multi-allelic VCF-style record therefore gets a wrong but plausible-looking variant back, with no error. The same thing happens with19-44908822-C-T;foo(gnomAD),NC_000019.10:44908821:1:T:extra(SPDI) and19 : 44908822 C > T,G(beacon).Root cause
_from_beacon,_from_gnomadand_from_spdiapplybeacon_re,gnomad_reandspdi_rewithre.match(). That anchors only at the start of the string, and none of the patterns end with$, so any matching prefix is enough.Fix
Use
fullmatch()in all three parsers. Input with trailing content now fails to parse, and the parser returnsNonebefore it makes any data proxy call.translate_from(..., fmt=...)then raises the usualValueError: Unable to parse data as <fmt> variation, and format auto-detection falls through as it does for any other unparseable string. Well-formed input is unaffected.Test
Adds
test_from_trailing_input_rejectedintests/extras/test_allele_translator.py, parametrized over the four inputs above. It needs no cassette because rejection happens before any data proxy lookup. All four cases fail onmainand pass with this change.Locally, with
SEQREPO_ROOT_DIR=./tests/data/seqrepo/latest pytest --vcr-record=none, the full suite gives 153 passed.ruff checkandruff format --check(ruff 0.9.4) are clean.