Merge branch 'bugfix/LF-3015/enable-unicode-aware-mode-for-regular-ex…

…pressions' into 'master'

Enabled unicode-aware mode for regular expressions in fhirpath

See merge request lfor/fhirpath.js!13

CHANGELOG.md

@@ -3,6 +3,12 @@
 This log documents significant changes for each release.  This project follows
 [Semantic Versioning](http://semver.org/).
 
+## [3.13.1] - 2024-04-25
+### Fixed
+- Added flag 'u' for regular expressions in the specification's `matches` and
+  `replaceMatches` functions to support the use of unicode character class
+  escapes.
+
 ## [3.13.0] - 2024-04-10
 ### Added
 - Function `defineVariable(name: String [, expr: expression])`.

package.json

@@ -1,6 +1,6 @@
 {
   "name": "fhirpath",
-  "version": "3.13.0",
+  "version": "3.13.1",
   "description": "A FHIRPath engine",
   "main": "src/fhirpath.js",
   "dependencies": {

src/strings.js

@@ -162,7 +162,7 @@ if (dotAllIsSupported) {
     if (util.isEmpty(regex) || util.isEmpty(str)) {
       return [];
     }
-    const reg = new RegExp(regex, 's');
+    const reg = new RegExp(regex, 'su');
     return reg.test(str);
   };
 } else {
@@ -171,7 +171,7 @@ if (dotAllIsSupported) {
     if (util.isEmpty(regex) || util.isEmpty(str)) {
       return [];
     }
-    const reg = new RegExp(rewritePatternForDotAll(regex));
+    const reg = new RegExp(rewritePatternForDotAll(regex), 'u');
     return reg.test(str);
   };
 }
@@ -190,7 +190,7 @@ engine.replaceMatches = function (coll, regex, repl) {
   if (util.isEmpty(regex) || util.isEmpty(repl) || util.isEmpty(str)) {
     return [];
   }
-  const reg = new RegExp(regex, 'g');
+  const reg = new RegExp(regex, 'gu');
   return str.replace(reg, repl);
 };
 

test/cases/5.6_string_manipulation.yaml

@@ -258,6 +258,10 @@ tests:
     expression: Patient.name.given[0].matches('.*')
     result: []
 
+  - desc: '** matches with unicode character class escapes'
+    expression: "'Taupō'.matches('^\\\\p{Lu}\\\\p{Ll}*$')"
+    result: [true]
+
   - desc: '5.6.10. replaceMatches(regex : string, substitution: string) : string'
 # If the input collection contains a single item of type string, the function will match the input using the regular expression in regex and replace each match with the substitution string. The substitution may refer to identified match groups in the regular expression.
 # This example of replaceMatches() will convert a string with a date formatted as MM/dd/yy to dd-MM-yy:
@@ -299,6 +303,10 @@ tests:
     expression: Patient.name.given[0].replaceMatches('(.*)', '$1')
     result: []
 
+  - desc: '** replaceMatches with a unicode character class escape'
+    expression: "'Taupō'.replaceMatches('(\\\\p{Lu})', 'city: $1')"
+    result: ['city: Taupō']
+
   - desc: '5.6.11. length() : integer'
 # If the input collection contains a single item of type string, the function will return the length of the string. If the input collection is empty ({ }), the result is empty.
   - desc: '** length'