diff options
author | Steven G. Johnson <stevenj@alum.mit.edu> | 2014-12-07 22:25:31 -0500 |
---|---|---|
committer | Steven G. Johnson <stevenj@mit.edu> | 2014-12-12 16:29:29 -0500 |
commit | 539d2cc2024f494b1e3292d4730bdc96390e1361 (patch) | |
tree | 84876217e3476899f6ea2d2e5646c2e845c6843a /tests.h | |
parent | 1b3992ebe5c587446aaa962a314ef9244d86fb0d (diff) | |
download | libutf8proc-539d2cc2024f494b1e3292d4730bdc96390e1361.tar.gz libutf8proc-539d2cc2024f494b1e3292d4730bdc96390e1361.tar.bz2 |
grapheme test for UAX#29
Diffstat (limited to 'tests.h')
-rw-r--r-- | tests.h | 53 |
1 files changed, 53 insertions, 0 deletions
@@ -0,0 +1,53 @@ +/* Common functions and includes for our test programs. */ + +#include <stdio.h> +#include <stdlib.h> +#include <ctype.h> +#include <string.h> +#include <stdarg.h> + +#include "mojibake.h" + +size_t lineno = 0; + +void check(int cond, const char *format, ...) +{ + if (!cond) { + va_list args; + fprintf(stderr, "line %zd: ", lineno); + va_start(args, format); + vfprintf(stderr, format, args); + va_end(args); + fprintf(stderr, "\n"); + exit(1); + } +} + +size_t skipspaces(const char *buf, size_t i) +{ + while (isspace(buf[i])) ++i; + return i; +} + +/* if buf points to a sequence of codepoints encoded as hexadecimal strings, + separated by whitespace, and terminated by any character not in + [0-9a-fA-F] or whitespace, then stores the corresponding utf8 string + in dest, returning the number of bytes read from buf */ +size_t encode(char *dest, const char *buf) +{ + size_t i = 0, j, d = 0; + do { + int c; + i = skipspaces(buf, i); + for (j=i; buf[j] && strchr("0123456789abcdef", tolower(buf[j])); ++j) + ; /* find end of hex input */ + if (j == i) { /* no codepoint found */ + dest[d] = 0; /* NUL-terminate destination string */ + return i + 1; + } + check(sscanf(buf + i, "%x", &c) == 1, "invalid hex input %s", buf+i); + i = j; /* skip to char after hex input */ + d += utf8proc_encode_char(c, (uint8_t *) (dest + d)); + } while (1); +} + |