Merge pull request #1167 from nyaruka/fix-matching-french-words-locat…

…ions Support hyphen in locations name as one word
nyaruka · May 3, 2023 · acc7ebb · acc7ebb
2 parents 89d0546 + c3b3a47
commit acc7ebb
Show file tree

Hide file tree

Showing 2 changed files with 9 additions and 1 deletion.
diff --git a/flows/environment.go b/flows/environment.go
@@ -58,7 +58,8 @@ func (r *assetLocationResolver) FindLocationsFuzzy(text string, level envs.Locat
 	}
 
 	// try on each tokenized word
-	words := regexp.MustCompile(`[\s\p{P}]+`).Split(text, -1)
+	re := regexp.MustCompile(`[\p{L}\d]+(-[\p{L}\d]+)*`)
+	words := re.FindAllString(text, -1)
 	for _, word := range words {
 		if locations := r.FindLocations(word, level, parent); len(locations) > 0 {
 			return locations

diff --git a/flows/routers/cases/tests_test.go b/flows/routers/cases/tests_test.go
@@ -61,6 +61,11 @@ var locationHierarchyJSON = `{
 			"aliases": ["Q.C", "Le Québec", "Quebec", "Que,", "Que", "Qc", "Québec"],
 			"children": []
 		},
+		{
+			"name": "Île-de-France",
+			"aliases": [],
+			"children": []
+		},
 		{
 			"name": "Paktika",
 			"aliases": ["Janikhel", "Terwa", "Yahyakhel", "Yusufkhel", "\u067e\u06a9\u062a\u06cc\u06a9\u0627", "\u062a\u0631\u0648\u0648", "\u06cc\u062d\u06cc\u06cc \u062e\u06cc\u0644", "\u06cc\u0648\u0633\u0641 \u062e\u06cc\u0644"],
@@ -300,6 +305,8 @@ var testTests = []struct {
 	{"has_state", []types.XValue{xs("Québec")}, result(xs("Rwanda > Québec"))},
 	{"has_state", []types.XValue{xs("Je suis dans la province du Québec")}, result(xs("Rwanda > Québec"))},
 
+	{"has_state", []types.XValue{xs("Je suis dans la province de l'Île-de-France")}, result(xs("Rwanda > Île-de-France"))},
+
 	{"has_state", []types.XValue{xs("kigali city")}, result(xs("Rwanda > Kigali City"))},
 	{"has_state", []types.XValue{xs("¡Kigali!")}, result(xs("Rwanda > Kigali City"))},