aboutsummaryrefslogtreecommitdiffstats
path: root/src
diff options
context:
space:
mode:
authorEugene Triguba <hello@eugenetriguba.com>2025-10-11 14:19:39 -0400
committerEugene Triguba <hello@eugenetriguba.com>2025-10-11 14:31:37 -0400
commitd3ec525b7d6ecd8de064d3f3a54d112769d1d3b3 (patch)
tree935b51f95afb513604c82c9e1efca223af60c109 /src
parentFix git clone link in readme (diff)
downloadcli-d3ec525b7d6ecd8de064d3f3a54d112769d1d3b3.zip
Add unicode support to justify
Diffstat (limited to 'src')
-rw-r--r--src/linebuffer.c27
-rw-r--r--src/utf8.c74
2 files changed, 92 insertions, 9 deletions
diff --git a/src/linebuffer.c b/src/linebuffer.c
index 6773ca1..3fa08b0 100644
--- a/src/linebuffer.c
+++ b/src/linebuffer.c
@@ -1,4 +1,5 @@
#include "linebuffer.h"
+#include "utf8.h"
#include <ctype.h>
#include <stdbool.h>
#include <stdio.h>
@@ -6,19 +7,23 @@
#include <string.h>
#define NULL_TERMINATOR_LEN 1
+#define UTF8_MAX_BYTES_PER_CHAR 4
-LineBuffer *linebuffer_create(size_t buffer_capacity) {
+LineBuffer *linebuffer_create(size_t line_width) {
LineBuffer *lb = malloc(sizeof(LineBuffer));
if (lb == NULL) {
return NULL;
}
+ size_t buffer_capacity = line_width * UTF8_MAX_BYTES_PER_CHAR;
lb->_buffer = malloc(buffer_capacity + NULL_TERMINATOR_LEN);
if (lb->_buffer == NULL) {
free(lb);
return NULL;
}
lb->_num_words = 0;
+ lb->_line_width = line_width;
lb->_buffer_length = 0;
+ lb->_display_width = 0;
lb->_buffer_capacity = buffer_capacity;
return lb;
}
@@ -31,6 +36,7 @@ void linebuffer_destroy(LineBuffer *lb) {
void linebuffer_clear(LineBuffer *lb) {
lb->_num_words = 0;
lb->_buffer_length = 0;
+ lb->_display_width = 0;
}
bool linebuffer_append_word(LineBuffer *lb, const char *word) {
@@ -40,25 +46,28 @@ bool linebuffer_append_word(LineBuffer *lb, const char *word) {
if (lb->_num_words > 0) {
lb->_buffer[lb->_buffer_length] = ' ';
lb->_buffer_length++;
+ lb->_display_width++;
}
- size_t word_length = strlen(word);
- memcpy(lb->_buffer + lb->_buffer_length, word, word_length);
- lb->_buffer_length += word_length;
+ size_t word_byte_length = strlen(word);
+ size_t word_display_width = utf8_display_width(word);
+ memcpy(lb->_buffer + lb->_buffer_length, word, word_byte_length);
+ lb->_buffer_length += word_byte_length;
+ lb->_display_width += word_display_width;
lb->_num_words++;
return true;
}
bool linebuffer_can_fit(LineBuffer *lb, const char *word) {
- size_t append_len = strlen(word);
+ size_t word_display_width = utf8_display_width(word);
+ size_t required_width = word_display_width;
if (lb->_num_words > 0) {
- // Account for space before the word
- append_len++;
+ required_width++;
}
- return append_len <= linebuffer_space_remaining(lb);
+ return required_width <= linebuffer_space_remaining(lb);
}
size_t linebuffer_space_remaining(LineBuffer *lb) {
- return lb->_buffer_capacity - lb->_buffer_length;
+ return lb->_line_width - lb->_display_width;
}
int linebuffer_write(LineBuffer *lb, FILE *fp) {
diff --git a/src/utf8.c b/src/utf8.c
new file mode 100644
index 0000000..07b751e
--- /dev/null
+++ b/src/utf8.c
@@ -0,0 +1,74 @@
+#include "utf8.h"
+#include <unicode/ustring.h>
+#include <unicode/uchar.h>
+#include <unicode/ubrk.h>
+#include <string.h>
+#include <stdlib.h>
+
+size_t utf8_display_width(const char *str) {
+ if (str == NULL) {
+ return 0;
+ }
+
+ UErrorCode status = U_ZERO_ERROR;
+ int32_t utf8_len = (int32_t)strlen(str);
+
+ // Convert UTF-8 to UTF-16 for ICU processing
+ UChar *utf16 = NULL;
+ int32_t utf16_len = 0;
+
+ // First call to get required length
+ u_strFromUTF8(NULL, 0, &utf16_len, str, utf8_len, &status);
+ if (status != U_BUFFER_OVERFLOW_ERROR && U_FAILURE(status)) {
+ return 0;
+ }
+
+ status = U_ZERO_ERROR;
+ utf16 = (UChar *)malloc((utf16_len + 1) * sizeof(UChar));
+ if (utf16 == NULL) {
+ return 0;
+ }
+
+ u_strFromUTF8(utf16, utf16_len + 1, NULL, str, utf8_len, &status);
+ if (U_FAILURE(status)) {
+ free(utf16);
+ return 0;
+ }
+
+ // Use break iterator to count grapheme clusters
+ UBreakIterator *bi = ubrk_open(UBRK_CHARACTER, NULL, utf16, utf16_len, &status);
+ if (U_FAILURE(status)) {
+ free(utf16);
+ return 0;
+ }
+
+ size_t width = 0;
+ int32_t start = ubrk_first(bi);
+ int32_t end = ubrk_next(bi);
+
+ while (end != UBRK_DONE) {
+ // Get the first code point of this grapheme cluster
+ UChar32 ch;
+ int32_t idx = start;
+ U16_NEXT(utf16, idx, utf16_len, ch);
+
+ // Determine display width based on East Asian Width property
+ int32_t ea_width = u_getIntPropertyValue(ch, UCHAR_EAST_ASIAN_WIDTH);
+
+ // Full-width and Wide characters take 2 columns
+ if (ea_width == U_EA_FULLWIDTH || ea_width == U_EA_WIDE) {
+ width += 2;
+ } else {
+ // All others (Narrow, Halfwidth, Neutral, Ambiguous) take 1 column
+ width += 1;
+ }
+
+ start = end;
+ end = ubrk_next(bi);
+ }
+
+ ubrk_close(bi);
+ free(utf16);
+
+ return width;
+}