From ccada385c4fdc2dc320b0ab468c7413ff27bcc7d Mon Sep 17 00:00:00 2001
From: Bruce Hill <bruce@bruce-hill.com>
Date: Tue, 19 Nov 2024 13:27:27 -0500
Subject: Add Text.at(i) for getting a single cluster

---
 stdlib/text.c | 52 ++++++++++++++++++++++++++++++++++++++++++++++++++++
 stdlib/text.h |  1 +
 2 files changed, 53 insertions(+)

(limited to 'stdlib')

diff --git a/stdlib/text.c b/stdlib/text.c
index 44179fa7..69e54ff6 100644
--- a/stdlib/text.c
+++ b/stdlib/text.c
@@ -563,6 +563,58 @@ public Text_t Text$slice(Text_t text, Int_t first_int, Int_t last_int)
     }
 }
 
+public Text_t Text$cluster(Text_t text, Int_t index_int)
+{
+    int64_t index = Int_to_Int64(index_int, false);
+    if (index == 0) fail("Invalid index: 0");
+
+    if (index < 0) index = text.length + index + 1;
+
+    if (index > text.length || index < 1)
+        fail("Invalid index: %ld is beyond the length of the text (length = %ld)",
+             Int_to_Int64(index_int, false), text.length);
+
+    switch (text.tag) {
+    case TEXT_SHORT_ASCII: {
+        return (Text_t) {
+            .tag=TEXT_SHORT_ASCII,
+            .length=1,
+            .short_ascii={text.short_ascii[index-1]},
+        };
+    }
+    case TEXT_ASCII: {
+        return (Text_t) {
+            .tag=TEXT_SHORT_ASCII,
+            .length=1,
+            .short_ascii={text.ascii[index-1]},
+        };
+    }
+    case TEXT_SHORT_GRAPHEMES: {
+        return (Text_t) {
+            .tag=TEXT_SHORT_GRAPHEMES,
+            .length=1,
+            .short_graphemes={text.short_graphemes[index-1]},
+        };
+    }
+    case TEXT_GRAPHEMES: {
+        return (Text_t) {
+            .tag=TEXT_SHORT_GRAPHEMES,
+            .length=1,
+            .short_graphemes={text.graphemes[index-1]},
+        };
+    }
+    case TEXT_SUBTEXT: {
+        Text_t *subtext = text.subtexts;
+        while (index > subtext[0].length) {
+            index -= subtext[0].length;
+            ++subtext;
+        }
+        return Text$cluster(*subtext, I(index));
+    }
+    default: errx(1, "Invalid tag");
+    }
+}
+
 Text_t text_from_u32(ucs4_t *codepoints, int64_t num_codepoints, bool normalize)
 {
     // Normalization is apparently guaranteed to never exceed 3x in the input length
diff --git a/stdlib/text.h b/stdlib/text.h
index 45aa00ca..4f23834f 100644
--- a/stdlib/text.h
+++ b/stdlib/text.h
@@ -28,6 +28,7 @@ Text_t Text$_concat(int n, Text_t items[n]);
 #define Text$concat(...) Text$_concat(sizeof((Text_t[]){__VA_ARGS__})/sizeof(Text_t), (Text_t[]){__VA_ARGS__})
 #define Texts(...) Text$concat(__VA_ARGS__)
 Text_t Text$slice(Text_t text, Int_t first_int, Int_t last_int);
+Text_t Text$cluster(Text_t text, Int_t index_int);
 OptionalText_t Text$from_str(const char *str);
 OptionalText_t Text$from_strn(const char *str, size_t len);
 PUREFUNC uint64_t Text$hash(Text_t *text);
-- 
cgit v1.2.3