BASE declarations are now handled correctly
#1660
Merged
Conversation
This file contains 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
hannahbast
force-pushed
the
base-iri-fix
branch
from
7c98092 to
62f7f96
Compare
When a BASE IRI is defined, relative IRIs should be resolved against it. For example, with `BASE <http://example.org/>`, the relative IRI `<a>` should be resolved to `<http://example.org/a>`. This is a fist step in this direction. So far, most of the test fail because they use relative IRIs extensively, assuming that they are left untouched. PS: UniProt uses BASE extensively, so this is a blocker for parsing UniProt via Turtle.
hannahbast
force-pushed
the
base-iri-fix
branch
from
62f7f96 to
a2f12a4
Compare
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## master #1660 +/- ##
==========================================
+ Coverage 89.64% 89.68% +0.04%
==========================================
Files 383 383
Lines 36889 36942 +53
Branches 4167 4174 +7
==========================================
+ Hits 33068 33132 +64
+ Misses 2522 2510 -12
- Partials 1299 1300 +1 ☔ View full report in Codecov by Sentry. |
hannahbast
changed the title
BASE declarations are now handled correctly
added 2 commits
December 8, 2024 21:47
TODO: I still don't understand how the `prefixMap_` can become empty again, despite the default initialization in the `TurtleParser` class.
joka921
requested changes
Dec 10, 2024
src/parser/RdfParser.h
Outdated
Comment on lines
204
to
217
| // TODO<hannah> I would prefer to just call `prefixMap_.at(...)`, but then | ||
| // some of the tests fails because the keys are not in the map (despite the | ||
| // initialization above). | ||
| const TripleComponent::Iri& baseForRelativeIri() { | ||
| // return prefixMap_.at(baseForRelativeIriKey_); | ||
| return prefixMap_ | ||
| .try_emplace(baseForRelativeIriKey_, TripleComponent::Iri{}) | ||
| .first->second; | ||
| } | ||
| const TripleComponent::Iri& baseForAbsoluteIri() { | ||
| // return prefixMap_.at(baseForAbsoluteIriKey_); | ||
| return prefixMap_ | ||
| .try_emplace(baseForAbsoluteIriKey_, TripleComponent::Iri{}) | ||
| .first->second; |
There was a problem hiding this comment.
I agree, we should definitely fix this.
There was a problem hiding this comment.
Can you tell me, which tests those are (or I can see myself after the next iteration).
There was a problem hiding this comment.
I suspect, that these test are FRIENDS and explicitly set the prefix map. This can be reimplemented by a setter function which inserts the prefixes if necessary.
src/index/CompressedRelation.cpp
Outdated
| @@ -1332,7 +1332,7 @@ auto CompressedRelationWriter::createPermutationPair( | |||
| // TODO<joka921> Use `CALL_FIXED_SIZE`. | |||
| ad_utility::CompressedExternalIdTableSorter<decltype(compare), 0> | |||
| twinRelationSorter(basename + ".twin-twinRelationSorter", numColumns, | |||
| 4_GB, alloc); | |||
| 30_GB, alloc); | |||
There was a problem hiding this comment.
This is definitely part of a different PR.
src/parser/Iri.cpp
Outdated
| pos = iri_.find('/', pos + 3); | ||
| } | ||
| // If there is no `/` after the scheme or the first such `/` is the last | ||
| // character, the base IRI is the IRI itself. |
There was a problem hiding this comment.
- This also is the return for the case
no scheme found. Do you want to assert that we either have an absolute IRI as the base, or the empty base (for backwards compatibility) but nothing ill-formed either ?
(might be that you have this check elsewhere).
test/RdfParserTest.cpp
Outdated
| forAllParsers(testWithParser, "<missing> <object> ."); | ||
| // Make sure that the `.` is not the final character because we (need to) | ||
| // handle that as a special case in `RdfStreamParser<T>::getLineImpl`. | ||
| forAllParsers(testWithParser, "<missing> <object> . "); |
There was a problem hiding this comment.
This should be working again with the discussed end regex fix
- it should be possible to add a test that checks, that we have fixed the uniprot bug with dots in the middle of an IRIref correctly (e.g. in the StreamAndParallelparser test, where we manually fiddle with blocks anyway.
Co-authored-by: Johannes Kalmbach <[email protected]>
Conformance check passed ✅No test result changes. |
|
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.
When a
BASEIRI is defined, IRIs without a scheme (like<http://...>) should be resolved. For example, withBASE <http://purl.uniprot.org/uniprot/>, the relative IRI<UPI001AF4585D>should be resolved to<http://purl.uniprot.org/uniprot/UPI001AF4585D>, and the absolute IRI</prosite/PS51927>should be resolved to<http://purl.uniprot.org/prosite/PS51927>.NOTE: Without
BASEdeclaration, relative IRIs like<a>are accepted and left unchanged. This is used extensively in our tests. This is not 100% compliant with the SPARQL standard, but we consider being slightly more accepting than the standard harmless in this case.