Index dictionary by first grapheme instead of first char

sillsdev · Jun 20, 2024 · 614dbd5 · 614dbd5
1 parent 2a7692b
commit 614dbd5
Show file tree

Hide file tree

Showing 3 changed files with 14 additions and 11 deletions.
diff --git a/backend/FwDataMiniLcmBridge/Api/FwDataMiniLcmApi.cs b/backend/FwDataMiniLcmBridge/Api/FwDataMiniLcmApi.cs
@@ -138,7 +138,7 @@ internal void CompleteExemplars(WritingSystems writingSystems)
     {
         var wsExemplars = writingSystems.Vernacular.Concat(writingSystems.Analysis)
             .Distinct()
-            .ToDictionary(ws => ws, ws => ws.Exemplars.Select(s => s[0]).ToHashSet());
+            .ToDictionary(ws => ws, ws => ws.Exemplars.ToHashSet());
         var wsExemplarsByHandle = wsExemplars.ToFrozenDictionary(kv => GetWritingSystemHandle(kv.Key.Id), kv => kv.Value);
 
         foreach (var entry in _entriesRepository.AllInstances())
@@ -149,7 +149,7 @@ internal void CompleteExemplars(WritingSystems writingSystems)
 
         foreach (var ws in wsExemplars.Keys)
         {
-            ws.Exemplars = [.. wsExemplars[ws].Order().Select(s => s.ToString())];
+            ws.Exemplars = [.. wsExemplars[ws].Order()];
         }
     }
 

diff --git a/backend/FwDataMiniLcmBridge/Api/LcmHelpers.cs b/backend/FwDataMiniLcmBridge/Api/LcmHelpers.cs
@@ -1,3 +1,4 @@
+using System.Globalization;
 using SIL.LCModel.Core.KernelInterfaces;
 
 namespace FwDataMiniLcmBridge.Api;
@@ -38,32 +39,33 @@ internal static bool SearchValue(this ITsMultiString multiString, string value)
         '\u200C', // Zero Width Non-Joiner
         '\u200D', // Zero Width Joiner
         '\u200E', // Left-to-Right Mark
-        '\u200F',  // Right-to-Left Mark
+        '\u200F', // Right-to-Left Mark
         '\u2028', // Line Separator
         '\u2029', // Paragraph Separator
         '\u202F', // Narrow No-Break Space
         '\u205F', // Medium Mathematical Space
-        '\u3000',  // Ideographic Space
+        '\u3000', // Ideographic Space
         '\uFEFF', // Zero Width No-Break Space / BOM
     ];
 
     internal static readonly char[] WhitespaceAndFormattingChars =
     [
         .. WhitespaceChars,
-        '\u0640', // Arabic Tatweel
+        '\u0640'.ToString().Normalize(System.Text.NormalizationForm.FormD)[0], // Arabic Tatweel
     ];
 
-    internal static void ContributeExemplars(ITsMultiString multiString, IReadOnlyDictionary<int, HashSet<char>> wsExemplars)
+    internal static void ContributeExemplars(ITsMultiString multiString, IReadOnlyDictionary<int, HashSet<string>> wsExemplars)
     {
         for (var i = 0; i < multiString.StringCount; i++)
         {
             var tsString = multiString.GetStringFromIndex(i, out var ws);
             if (string.IsNullOrEmpty(tsString.Text)) continue;
-            var value = tsString.Text.AsSpan().Trim(WhitespaceAndFormattingChars);
-            if (!value.IsEmpty && wsExemplars.TryGetValue(ws, out var exemplars))
+            var value = new StringInfo(tsString.Text).SubstringByTextElements(0, 1)
+                // in some cases we need to trim things both before and after the grapheme
+                .Trim(WhitespaceAndFormattingChars);
+            if (value.Length > 0 && wsExemplars.TryGetValue(ws, out var exemplars))
             {
-                //todo should we upper or lowercase the value?
-                exemplars.Add(value[0]);
+                exemplars.Add(value.ToUpper());
             }
         }
     }

diff --git a/frontend/viewer/src/ProjectView.svelte b/frontend/viewer/src/ProjectView.svelte
@@ -148,7 +148,8 @@
 
   function selectEntry(entry: IEntry) {
     $selectedEntry = entry;
-    $selectedIndexExemplar = headword(entry).charAt(0).toLocaleUpperCase() || undefined;
+    const indexChar: string | undefined = new Intl.Segmenter().segment(headword(entry))[Symbol.iterator]().next()?.value?.segment;
+    $selectedIndexExemplar = indexChar?.toLocaleUpperCase() ?? undefined;
     refreshEntries();
     pickedEntry = true;
   }