Though I'm sure our C_STRING.pattern values are not going to be complex regexes, different implementations could choose to use a different regex language, generating incompatibilities at the archetype level.
To avoid this we might:
1. Suggest a preferred language 2. Specify which subset of the regex language should be used 3. If 2. is really small, maybe we can add a syntax/grammar definition to the spec
Activity
Show:
Thomas Beale
November 4, 2021 at 8:41 AM
Usually PERL regex is what we assume in openEHR. I didn’t realise that was not stated in that part of the spec, so we’ll need to update that. Anyway, PREs allow \d, \w etc.
The description of C_STRING.pattern is very vague "Regular expression pattern for proposed instances of String to match." (https://specifications.openehr.org/releases/AM/Release-2.2.0/AOM1.4.html#_c_string_class).
Considering there are many regex languages with different features and levels of compatibility:
https://en.wikipedia.org/wiki/Comparison_of_regular-expression_engines#Language_features
https://docs.oracle.com/javase/7/docs/api/java/util/regex/Pattern.html#jcc
Though I'm sure our C_STRING.pattern values are not going to be complex regexes, different implementations could choose to use a different regex language, generating incompatibilities at the archetype level.
To avoid this we might:
1. Suggest a preferred language
2. Specify which subset of the regex language should be used
3. If 2. is really small, maybe we can add a syntax/grammar definition to the spec