spring cleaning 1

Wip/unicode/1.expected

@@ -0,0 +1,2 @@
+0000000 41 e2 89 a2 ce 91 2e
+0000007

Wip/unicode/1.test

@@ -0,0 +1,3 @@
+#!/bin/sh
+
+printf '%s\n%s\n%s\n%s\n' U+0041 U+0391 U+002E | ./utf 8 | od -t x1

Wip/unicode/2.expected

@@ -0,0 +1,2 @@
+0000000 ed 95 9c ea b5 ad ec 96 b4
+0000009

Wip/unicode/2.test

@@ -0,0 +1,3 @@
+#!/bin/sh
+
+printf '%s\n%s\n%s\n' U+D55C U+AD6D U+C5B4 | ./utf 8 | od -t x1

Wip/unicode/3.expected

@@ -0,0 +1,2 @@
+0000000 e6 97 a5 e6 9c ac e8 aa 9e
+0000009

Wip/unicode/3.test

@@ -0,0 +1,3 @@
+#!/bin/sh
+
+printf '%s\n%s\n%s\n' U+65E5 U+672C U+8A9E | ./utf 8 | od -t x1

Wip/unicode/HACKING

@@ -0,0 +1 @@
+Tests are from RFC3629, section 7.

Wip/unicode/Makefile

@@ -0,0 +1,15 @@
+test: utf
+	sh ../testing/test.sh
+
+utf: utf.o libunicode.o
+	$(CC) -g -o utf libunicode.o utf.o
+
+libunicode.o:
+
+utf.o: libunicode.h utf.c
+	$(CC) -I../ascii -c -g -o utf.o utf.c
+
+clean:
+	rm -f *.o utf
+
+.PHONY: clean test

Wip/unicode/libunicode.c

@@ -0,0 +1,23 @@
+#include "libunicode.h"
+
+utf32_t utf8(utf32_t c){
+	unsigned char n; /* 4 - number of bytes - 1 */
+	utf8_t r;
+
+	if(c > UTF8_MAX) /* return 0 if c exceeds max */
+		c = 0;
+
+	switch(n = (c >= 0x010000) + (c >= 0x0800) + (c >= 0x0080)){
+	case 3: r = 0xF0 + ((c >> 18) & 0x07); /* 11110xxx */
+	case 2:	r = (r << 8) + (n == 2
+		? 0xE0 + ((c >> 12) & 0x0F) /* 1110xxxx */
+		: 0x50 + ((c >> 12) & 0x3F)); /* 10xxxxxx */
+	case 1:	r = (r << 8) + (n == 1
+		? 0xC0 + ((c >> 6) & 0x1F) /* 110xxxxx */
+		: 0x50 + ((c >> 6) & 0x3F)); /* 10xxxxxx */
+	case 0:	r = (r << 8) + (n == 0
+		? (c & 0x7F)  /* 0xxxxxxx */
+		: 0x50 + ((c >> 6) & 0x3F)); /* 10xxxxxx */
+	}
+	return r;
+}

Wip/unicode/libunicode.h

@@ -0,0 +1,16 @@
+#if defined UINT32_MAX /* indicator <stdint.h> is included */
+#	define unicode_codepoint_t uint32_t
+#else
+ /* C99 draft 5.2.4.2.1 Sizes of integer types says unsigned long must be able
+  * to hold 32b */
+#	define unicode_codepoint_t unsigned long
+#endif
+#define utf32_t unicode_codepoint_t
+/* for holding the literal numeric value of a utf8 rune, not for assembling
+ * strings; >=32b value */
+#define utf8_t unicode_codepoint_t
+
+#define UTF8_MAX 0x10FFFF
+
+/* encode UTF-32 value into UTF-8 */
+utf32_t utf8(utf32_t c);

Wip/unicode/utf.c

@@ -0,0 +1,58 @@
+#include <stdio.h> /* fprintf(3), getc(3), putc(3) stderr, stdin, EOF */
+#include <string.h> /* strchr(3) */
+#include "ascii.h" /* ASCII_HEXADECIMAL_DIGITS_LOWER,
+                    * ASCII_HEXADECIMAL_DIGITS_UPPER */
+#include "libunicode.h" /* utf8(3) */
+
+static struct {
+	char *name;
+	utf32_t (*f)(utf32_t);
+}
+
+static char *hex = ASCII_HEXADECIMAL_DIGITS_UPPER
+	ASCII_HEXADECIMAL_DIGITS_LOWER;
+
+int main(int argc, char *argv[]){
+	int c;
+	int i;
+	int l; /* line counter */
+	char *n;
+	utf32_t codepoint;
+	utf8_t encoded;
+
+	l = 0;
+
+init:	codepoint = 0;
+	i = 0;
+	++l;
+	n = NULL;
+	while((c = getc(stdin)) != EOF){
+		if(c == '\n'){
+			if(i < 2 && i > 0) /* empty lines are fine */
+				fprintf(stderr, "%s: %s: Syntax error.\n",
+					argv[0], l);
+			else if(i >= 2){
+				encoded = utf8(codepoint);
+				for(i = 3; i >= 0; --i)
+					if((encoded >> 8 * i) > 0 || i == 0)
+						putc(encoded >> 8 * i, stdout);
+			}
+			goto init;
+		}
+		if(
+				(i == 0 && c != 'U')
+				|| (i == 1 && c != '+')
+				|| i > 8 /* strlen("U+10FFFF") */
+				|| (i > 1 && ((n = strchr(hex, c)) == NULL))
+		){
+				fprintf(stderr, "%s: %s: Syntax error.\n",
+					argv[0], l);
+				while((c = getc(stdin)) != '\n' && c != EOF);
+				++l;
+				continue;
+		}
+		if(n != NULL)
+			codepoint = (codepoint << 4) + (n - hex) % 16;
+		++i;
+	}
+}