#include "charset.h"
#include "mime.h"
#include "hex.h"
-#include "words.h"
#include "heap.h"
#include "unicode.h"
#include "inputline.h"
#include "wstat.h"
static int tests, errors;
+static int fail_first;
+
+static void count_error() {
+ ++errors;
+ if(fail_first)
+ abort();
+}
/** @brief Checks that @p expr is nonzero */
#define insist(expr) do { \
if(!(expr)) { \
- ++errors; \
+ count_error(); \
fprintf(stderr, "%s:%d: error checking %s\n", \
__FILE__, __LINE__, #expr); \
} \
dynstr_init(&d);
while((c = *s++)) {
- if(c >= 32 && c <= 127)
- dynstr_append(&d, c);
- else {
- sprintf(buf, "\\x%04lX", (unsigned long)c);
- dynstr_append_string(&d, buf);
- }
+ sprintf(buf, " %04lX", (long)c);
+ dynstr_append_string(&d, buf);
}
dynstr_terminate(&d);
return d.vec;
if(w == 0) { \
fprintf(stderr, "%s:%d: %s returned 0\n", \
__FILE__, __LINE__, #GOT); \
- ++errors; \
+ count_error(); \
} else if(strcmp(w, g)) { \
fprintf(stderr, "%s:%d: %s returned:\n%s\nexpected:\n%s\n", \
__FILE__, __LINE__, #GOT, format(g), format(w)); \
- ++errors; \
+ count_error(); \
} \
++tests; \
} while(0)
char *u8; \
\
insist(validutf8(CHARS)); \
- ucs = utf82ucs4(CHARS); \
+ ucs = utf8_to_utf32(CHARS, strlen(CHARS), 0); \
insist(ucs != 0); \
- insist(!ucs4cmp(w, ucs)); \
- u8 = ucs42utf8(ucs); \
+ insist(!utf32_cmp(w, ucs)); \
+ u8 = utf32_to_utf8(ucs, utf32_len(ucs), 0); \
insist(u8 != 0); \
insist(!strcmp(u8, CHARS)); \
} while(0)
fprintf(stderr, "test_utf8\n");
+#define validutf8(S) utf8_valid((S), strlen(S))
/* empty string */
break;
}
if(l) {
+ uint32_t *d;
/* Case-folded data is now normalized */
- canon_expected = ucs42utf8(utf32_decompose_canon(&l, 1, 0));
+ d = utf32_decompose_canon(&l, 1, 0);
+ canon_expected = utf32_to_utf8(d, utf32_len(d), 0);
if(strcmp(canon_folded, canon_expected)) {
fprintf(stderr, "%s:%d: canon-casefolding %#lx got '%s', expected '%s'\n",
__FILE__, __LINE__, (unsigned long)c,
format(canon_folded), format(canon_expected));
- ++errors;
+ count_error();
}
++tests;
- compat_expected = ucs42utf8(utf32_decompose_compat(&l, 1, 0));
+ d = utf32_decompose_compat(&l, 1, 0);
+ compat_expected = utf32_to_utf8(d, utf32_len(d), 0);
if(strcmp(compat_folded, compat_expected)) {
fprintf(stderr, "%s:%d: compat-casefolding %#lx got '%s', expected '%s'\n",
__FILE__, __LINE__, (unsigned long)c,
format(compat_folded), format(compat_expected));
- ++errors;
+ count_error();
}
++tests;
}
}
- check_string(casefold(""), "");
+ check_string(utf8_casefold_canon("", 0, 0), "");
+}
+
+struct {
+ const char *in;
+ const char *expect[10];
+} wtest[] = {
+ /* Empty string */
+ { "", { 0 } },
+ /* Only whitespace and punctuation */
+ { " ", { 0 } },
+ { " ' ", { 0 } },
+ { " ! ", { 0 } },
+ { " \"\" ", { 0 } },
+ { " @ ", { 0 } },
+ /* Basics */
+ { "wibble", { "wibble", 0 } },
+ { " wibble", { "wibble", 0 } },
+ { " wibble ", { "wibble", 0 } },
+ { "wibble ", { "wibble", 0 } },
+ { "wibble spong", { "wibble", "spong", 0 } },
+ { " wibble spong", { "wibble", "spong", 0 } },
+ { " wibble spong ", { "wibble", "spong", 0 } },
+ { "wibble spong ", { "wibble", "spong", 0 } },
+ { "wibble spong splat foo zot ", { "wibble", "spong", "splat", "foo", "zot", 0 } },
+ /* Apostrophes */
+ { "wibble 'spong", { "wibble", "spong", 0 } },
+ { " wibble's", { "wibble's", 0 } },
+ { " wibblespong' ", { "wibblespong", 0 } },
+ { "wibble sp''ong ", { "wibble", "sp", "ong", 0 } },
+};
+#define NWTEST (sizeof wtest / sizeof *wtest)
+
+static void test_words(void) {
+ size_t t, nexpect, ngot, i;
+ int right;
+
+ fprintf(stderr, "test_words\n");
+ for(t = 0; t < NWTEST; ++t) {
+ char **got = utf8_word_split(wtest[t].in, strlen(wtest[t].in), &ngot, 0);
+
+ for(nexpect = 0; wtest[t].expect[nexpect]; ++nexpect)
+ ;
+ if(nexpect == ngot) {
+ for(i = 0; i < ngot; ++i)
+ if(strcmp(wtest[t].expect[i], got[i]))
+ break;
+ right = i == ngot;
+ } else
+ right = 0;
+ if(!right) {
+ fprintf(stderr, "word split %zu failed\n", t);
+ fprintf(stderr, "input: %s\n", wtest[t].in);
+ fprintf(stderr, " | %-30s | %-30s\n",
+ "expected", "got");
+ for(i = 0; i < nexpect || i < ngot; ++i) {
+ const char *e = i < nexpect ? wtest[t].expect[i] : "<none>";
+ const char *g = i < ngot ? got[i] : "<none>";
+ fprintf(stderr, " %2zu | %-30s | %-30s\n", i, e, g);
+ }
+ count_error();
+ }
+ ++tests;
+ }
}
/** @brief Less-than comparison function for integer heap */
return fp;
}
+/** @brief Run breaking tests for utf32_grapheme_boundary() etc */
+static void breaktest(const char *path,
+ int (*breakfn)(const uint32_t *, size_t, size_t)) {
+ FILE *fp = open_unicode_test(path);
+ int lineno = 0;
+ char *l, *lp;
+ size_t bn, n;
+ char break_allowed[1024];
+ uint32_t buffer[1024];
+
+ while(!inputline(path, fp, &l, '\n')) {
+ ++lineno;
+ if(l[0] == '#') continue;
+ bn = 0;
+ lp = l;
+ while(*lp) {
+ if(*lp == ' ' || *lp == '\t') {
+ ++lp;
+ continue;
+ }
+ if(*lp == '#')
+ break;
+ if((unsigned char)*lp == 0xC3 && (unsigned char)lp[1] == 0xB7) {
+ /* 00F7 DIVISION SIGN */
+ break_allowed[bn] = 1;
+ lp += 2;
+ continue;
+ }
+ if((unsigned char)*lp == 0xC3 && (unsigned char)lp[1] == 0x97) {
+ /* 00D7 MULTIPLICATION SIGN */
+ break_allowed[bn] = 0;
+ lp += 2;
+ continue;
+ }
+ if(isxdigit((unsigned char)*lp)) {
+ buffer[bn++] = strtoul(lp, &lp, 16);
+ continue;
+ }
+ fatal(0, "%s:%d: evil line: %s", path, lineno, l);
+ }
+ for(n = 0; n <= bn; ++n) {
+ if(breakfn(buffer, bn, n) != break_allowed[n]) {
+ fprintf(stderr,
+ "%s:%d: offset %zu: mismatch\n"
+ "%s\n"
+ "\n",
+ path, lineno, n, l);
+ count_error();
+ }
+ ++tests;
+ }
+ xfree(l);
+ }
+ fclose(fp);
+}
+
/** @brief Tests for @ref lib/unicode.h */
static void test_unicode(void) {
FILE *fp;
int lineno = 0;
char *l, *lp;
uint32_t buffer[1024];
- uint32_t *c[6], *NFD_c[6], *NFKD_c[6]; /* 1-indexed */
+ uint32_t *c[6], *NFD_c[6], *NFKD_c[6], *NFC_c[6], *NFKC_c[6]; /* 1-indexed */
int cn, bn;
- char break_allowed[1024];
fprintf(stderr, "test_unicode\n");
fp = open_unicode_test("NormalizationTest.txt");
for(cn = 1; cn <= 5; ++cn) {
NFD_c[cn] = utf32_decompose_canon(c[cn], utf32_len(c[cn]), 0);
NFKD_c[cn] = utf32_decompose_compat(c[cn], utf32_len(c[cn]), 0);
+ NFC_c[cn] = utf32_compose_canon(c[cn], utf32_len(c[cn]), 0);
+ NFKC_c[cn] = utf32_compose_compat(c[cn], utf32_len(c[cn]), 0);
}
#define unt_check(T, A, B) do { \
++tests; \
fprintf(stderr, \
"NormalizationTest.txt:%d: c%d != "#T"(c%d)\n", \
lineno, A, B); \
- fprintf(stderr, " c%d: %s\n", \
+ fprintf(stderr, " c%d:%s\n", \
A, format_utf32(c[A])); \
- fprintf(stderr, "%4s(c%d): %s\n", \
+ fprintf(stderr, " c%d:%s\n", \
+ B, format_utf32(c[B])); \
+ fprintf(stderr, "%4s(c%d):%s\n", \
#T, B, format_utf32(T##_c[B])); \
- ++errors; \
+ count_error(); \
} \
} while(0)
unt_check(NFD, 3, 1);
unt_check(NFKD, 5, 3);
unt_check(NFKD, 5, 4);
unt_check(NFKD, 5, 5);
+ unt_check(NFC, 2, 1);
+ unt_check(NFC, 2, 2);
+ unt_check(NFC, 2, 3);
+ unt_check(NFC, 4, 4);
+ unt_check(NFC, 4, 5);
+ unt_check(NFKC, 4, 1);
+ unt_check(NFKC, 4, 2);
+ unt_check(NFKC, 4, 3);
+ unt_check(NFKC, 4, 4);
+ unt_check(NFKC, 4, 5);
for(cn = 1; cn <= 5; ++cn) {
xfree(NFD_c[cn]);
xfree(NFKD_c[cn]);
xfree(l);
}
fclose(fp);
- fp = open_unicode_test("auxiliary/GraphemeBreakTest.txt");
- lineno = 0;
- while(!inputline("GraphemeBreakTest.txt.txt", fp, &l, '\n')) {
- ++lineno;
- if(l[0] == '#') continue;
- bn = 0;
- lp = l;
- while(*lp) {
- if(*lp == ' ' || *lp == '\t') {
- ++lp;
- continue;
- }
- if(*lp == '#')
- break;
- if((unsigned char)*lp == 0xC3 && (unsigned char)lp[1] == 0xB7) {
- /* 00F7 DIVISION SIGN */
- break_allowed[bn] = 1;
- lp += 2;
- continue;
- }
- if((unsigned char)*lp == 0xC3 && (unsigned char)lp[1] == 0x97) {
- /* 00D7 MULTIPLICATION SIGN */
- break_allowed[bn] = 0;
- lp += 2;
- continue;
- }
- if(isxdigit((unsigned char)*lp)) {
- buffer[bn++] = strtoul(lp, &lp, 16);
- continue;
- }
- fatal(0, "GraphemeBreakTest.txt:%d: evil line: %s", lineno, l);
- }
- for(cn = 0; cn <= bn; ++cn) {
- if(utf32_is_gcb(buffer, bn, cn) != break_allowed[cn]) {
- fprintf(stderr,
- "GraphemeBreakTest.txt:%d: offset %d: utf32_is_gcb wrong\n",
- lineno, cn);
- ++errors;
- }
- ++tests;
- }
- xfree(l);
- }
+ breaktest("auxiliary/GraphemeBreakTest.txt", utf32_is_grapheme_boundary);
+ breaktest("auxiliary/WordBreakTest.txt", utf32_is_word_boundary);
}
int main(void) {
+ fail_first = !!getenv("FAIL_FIRST");
insist('\n' == 0x0A);
insist('\r' == 0x0D);
insist(' ' == 0x20);
/* vector.c */
/* words.c */
test_casefold();
+ test_words();
/* XXX words() */
/* wstat.c */
fprintf(stderr, "%d errors out of %d tests\n", errors, tests);