X-Git-Url: http://www.chiark.greenend.org.uk/ucgi/~mdw/git/disorder/blobdiff_plain/bcf9ed7f5b44c177d927d147f87c5c08e377adfa..ad2f8275a5fc0dba29ca97bc02342d1f1627e590:/lib/test.c diff --git a/lib/test.c b/lib/test.c index 6249444..f4043b9 100644 --- a/lib/test.c +++ b/lib/test.c @@ -38,7 +38,6 @@ #include "charset.h" #include "mime.h" #include "hex.h" -#include "words.h" #include "heap.h" #include "unicode.h" #include "inputline.h" @@ -88,12 +87,8 @@ static const char *format_utf32(const uint32_t *s) { 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; @@ -139,15 +134,16 @@ static void test_utf8(void) { 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 */ @@ -392,8 +388,10 @@ static void test_casefold(void) { 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, @@ -401,7 +399,8 @@ static void test_casefold(void) { 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, @@ -411,7 +410,70 @@ static void test_casefold(void) { ++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] : ""; + const char *g = i < ngot ? got[i] : ""; + fprintf(stderr, " %2zu | %-30s | %-30s\n", i, e, g); + } + count_error(); + } + ++tests; + } } /** @brief Less-than comparison function for integer heap */ @@ -467,7 +529,7 @@ static FILE *open_unicode_test(const char *path) { return fp; } -/** @brief Run breaking tests for utf32_is_gcb() etc */ +/** @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); @@ -510,8 +572,10 @@ static void breaktest(const char *path, for(n = 0; n <= bn; ++n) { if(breakfn(buffer, bn, n) != break_allowed[n]) { fprintf(stderr, - "%s:%d: offset %zu: mismatch\n", - path, lineno, n); + "%s:%d: offset %zu: mismatch\n" + "%s\n" + "\n", + path, lineno, n, l); count_error(); } ++tests; @@ -527,7 +591,7 @@ static void test_unicode(void) { 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; fprintf(stderr, "test_unicode\n"); @@ -560,6 +624,8 @@ static void test_unicode(void) { 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; \ @@ -567,9 +633,11 @@ static void test_unicode(void) { 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])); \ count_error(); \ } \ @@ -584,6 +652,16 @@ static void test_unicode(void) { 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]); @@ -591,7 +669,7 @@ static void test_unicode(void) { xfree(l); } fclose(fp); - breaktest("auxiliary/GraphemeBreakTest.txt", utf32_is_gcb); + breaktest("auxiliary/GraphemeBreakTest.txt", utf32_is_grapheme_boundary); breaktest("auxiliary/WordBreakTest.txt", utf32_is_word_boundary); } @@ -641,6 +719,7 @@ int main(void) { /* vector.c */ /* words.c */ test_casefold(); + test_words(); /* XXX words() */ /* wstat.c */ fprintf(stderr, "%d errors out of %d tests\n", errors, tests);