From ee844db84596e8ba8e0b84f770d66b6e6fabb028 Mon Sep 17 00:00:00 2001
From: david <lacasadepepito7@gmail.com>
Date: Tue, 6 Oct 2026 17:50:28 -0300
Subject: fixed utf-8 characters returning the wrong width with strlen. added a
 bunch of stuff

---
 src/helper.c | 78 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 78 insertions(+)
 create mode 100644 src/helper.c

(limited to 'src/helper.c')

diff --git a/src/helper.c b/src/helper.c
new file mode 100644
index 0000000..2749daa
--- /dev/null
+++ b/src/helper.c
@@ -0,0 +1,78 @@
+#include "../defs/all.h"
+#include "../defs/widechar.h"
+
+/* this reads one utf81 character from the start of bytes
+   it puts the character's number in *codepoint and returns how many bytes it used (its always 1 or more)
+   invalid input it stores U+FFFD (�) and returns 1, so the caller always moves forward. */
+static int utf8_decode(const unsigned char *bytes, unsigned long *codepoint){
+    unsigned long value = bytes[0];
+    unsigned long minallowed = 0;   //rejects overlong encodings 
+    int sequence_length;
+    int byte_index;
+
+    if(value < 0x80){
+        *codepoint = value;
+        return 1;
+    } else if(value >= 0xC2 && value <= 0xDF){
+        sequence_length = 2;
+        value &= 0x1F;
+        minallowed = 0x80;
+    } else if(value >= 0xE0 && value <= 0xEF){
+        sequence_length = 3;
+        value &= 0x0F;
+        minallowed = 0x800;
+    } else if(value >= 0xF0 && value <= 0xF4){
+        sequence_length = 4;
+        value &= 0x07;
+        minallowed = 0x10000;
+    } else {
+        *codepoint = 0xFFFD;
+        return 1;
+    }
+
+    for(byte_index = 1; byte_index < sequence_length; byte_index++){
+        if((bytes[byte_index] & 0xC0) != 0x80){
+            *codepoint = 0xFFFD;
+            return 1;
+        }
+        value = (value << 6) | (bytes[byte_index] & 0x3F);
+    }
+
+    if(value < minallowed || value > 0x10FFFF || (value >= 0xD800 && value <= 0xDFFF)){
+        *codepoint = 0xFFFD;
+        return 1;
+    }
+
+    *codepoint = value;
+    return sequence_length;
+}
+
+/* this turns the library's return value into a column count so 0 1 or 2 */
+static int codepoint_width(unsigned long codepoint){
+    int width = widechar_wcwidth(codepoint);
+    if(width > 0)
+        return width;
+
+    switch(width){
+        case widechar_ambiguous:     
+        case widechar_private_use:
+            return 1;
+        case widechar_widened_in_9:
+            return 2;
+        default:                     
+            return 0;
+    }
+}
+
+/* how many terminal columns a string takes up */
+int displaywidth(const char *text){
+    const unsigned char *current_byte = (const unsigned char *)text;
+    int total_width = 0;
+
+    while(*current_byte){
+        unsigned long codepoint;
+        current_byte += utf8_decode(current_byte, &codepoint);
+        total_width += codepoint_width(codepoint);
+    }
+    return total_width;
+}
-- 
cgit v1.3-2-g0d8e