Details
-
Change Request
-
Resolution: Persuasive
-
Medium
-
FHIR Core (FHIR)
-
R5
-
Terminology Infrastructure
-
CodeSystem
-
4.8.19
-
-
Reuben Daniels/Marc Duteau: 4-0-0
-
Clarification
-
Non-substantive
Description
I think we should explicitly state in the Comments that character case definitions and conversions are out of scope for the FHIR standard, and thus the results of doing a case-insensitive code comparison are implementation dependent. This isn't an issue in the US realm with the ASCII character set, but character case handling varies widely in other locales. If the client and server have different locales configured, or the concept codes are in a different language, then results of case-insensitive code comparisons might not be what the user expects.
In practice many FHIR servers are implemented using Java. The String.equalsIgnoreCase documentation explicitly warns about this issue.
Attachments
Issue Links
- relates to
-
FHIR-38688 Warn about locale inconsistencies with case-insensitive searches
- Published