From 5498793a56b19da99b7b6856c953933e50b8d572 Mon Sep 17 00:00:00 2001 From: Thomas Voss Date: Wed, 15 May 2024 00:43:54 +0200 Subject: Add ucsnorm_nfkd() --- test/_norm-test.h | 107 ++++++++++++++++++++++++++++++++++++++++++++++++++ test/norm-nfd-test.c | 100 +--------------------------------------------- test/norm-nfkd-test.c | 2 + test/run-tests | 2 +- 4 files changed, 112 insertions(+), 99 deletions(-) create mode 100644 test/_norm-test.h create mode 100644 test/norm-nfkd-test.c (limited to 'test') diff --git a/test/_norm-test.h b/test/_norm-test.h new file mode 100644 index 0000000..68209f1 --- /dev/null +++ b/test/_norm-test.h @@ -0,0 +1,107 @@ +#if !defined(NORMTYPE) +# error "NORMTYPE must be defined" +#endif + +#define _GNU_SOURCE +#include +#include + +#include +#include +#include +#include +#include +#include +#include + +#define TESTFILE "norm.in" +#define FUNC CONCAT(ucsnorm_, NORMTYPE) + +static bool test(struct u8view, int); + +int +main(int, char **argv) +{ + int rv; + size_t n; + ssize_t nr; + char *line; + FILE *fp; + + rv = EXIT_SUCCESS; + line = nullptr; + mlib_setprogname(argv[0]); + + if ((fp = fopen(TESTFILE, "r")) == nullptr) + err("fopen: %s:", TESTFILE); + + for (int id = 1; (nr = getline(&line, &n, fp)) > 0; id++) { + if (line[nr - 1] == '\n') + line[--nr] = '\0'; + + if (!test((struct u8view){line, (size_t)nr}, id)) { + rv = EXIT_FAILURE; + break; + } + } + if (ferror(fp)) + err("getline: %s:", TESTFILE); + + free(line); + fclose(fp); + return rv; +} + +bool +test(struct u8view sv, int id) +{ + bool rv = true; + arena a = mkarena(0); + struct arena_ctx ctx = {.a = &a}; + + dynarr(struct u8view) columns = { + .alloc = alloc_arena, + .ctx = &ctx, + }; + + struct u8view column; + while (ucscut(&column, &sv, U";", 1) != MBEND) { + dynarr(char8_t) s = { + .alloc = alloc_arena, + .ctx = &ctx, + }; + + rune _; + struct u8view cp; + do { + rune ch; + _ = ucscut(&cp, &column, U" ", 1); + sscanf(cp.p, "%" SCNxRUNE, &ch); + char8_t buf[U8_LEN_MAX]; + int w = rtoucs(buf, sizeof(buf), ch); + DAEXTEND(&s, buf, w); + } while (_ != MBEND); + + DAPUSH(&columns, ((struct u8view){s.buf, s.len})); + } + + for (size_t i = 0; i < 5; i++) { + size_t base; + if (streq(STR(NORMTYPE), "nfkd")) + base = 4; + else + base = i < 3 ? 2 : 4; + struct u8view normd = {}; + normd.p = FUNC(&normd.len, columns.buf[i], alloc_arena, &ctx); + if (!ucseq(columns.buf[base], normd)) { + warn("case %d: expected c%zu to be ‘%.*s’ but got ‘%.*s’", id, + i + 1, SV_PRI_ARGS(columns.buf[base]), SV_PRI_ARGS(normd)); + rv = false; + goto out; + } + } + +out: + arena_free(&a); + return rv; +} diff --git a/test/norm-nfd-test.c b/test/norm-nfd-test.c index 95bc8d5..6067352 100644 --- a/test/norm-nfd-test.c +++ b/test/norm-nfd-test.c @@ -1,98 +1,2 @@ -#define _GNU_SOURCE -#include -#include - -#include -#include -#include -#include -#include -#include -#include - -#define TESTFILE "norm-nfd.in" - -static bool test(struct u8view, int); - -int -main(int, char **argv) -{ - int rv; - size_t n; - ssize_t nr; - char *line; - FILE *fp; - - rv = EXIT_SUCCESS; - line = nullptr; - mlib_setprogname(argv[0]); - - if ((fp = fopen(TESTFILE, "r")) == nullptr) - err("fopen: %s:", TESTFILE); - - for (int id = 1; (nr = getline(&line, &n, fp)) > 0; id++) { - if (line[nr - 1] == '\n') - line[--nr] = '\0'; - - if (!test((struct u8view){line, (size_t)nr}, id)) { - rv = EXIT_FAILURE; - break; - } - } - if (ferror(fp)) - err("getline: %s:", TESTFILE); - - free(line); - fclose(fp); - return rv; -} - -bool -test(struct u8view sv, int id) -{ - bool rv = true; - arena a = mkarena(0); - struct arena_ctx ctx = {.a = &a}; - - dynarr(struct u8view) columns = { - .alloc = alloc_arena, - .ctx = &ctx, - }; - - struct u8view column; - while (ucscut(&column, &sv, U";", 1) != MBEND) { - dynarr(char8_t) s = { - .alloc = alloc_arena, - .ctx = &ctx, - }; - - rune _; - struct u8view cp; - do { - rune ch; - _ = ucscut(&cp, &column, U" ", 1); - sscanf(cp.p, "%" SCNxRUNE, &ch); - char8_t buf[U8_LEN_MAX]; - int w = rtoucs(buf, sizeof(buf), ch); - DAEXTEND(&s, buf, w); - } while (_ != MBEND); - - DAPUSH(&columns, ((struct u8view){s.buf, s.len})); - } - - for (size_t i = 0; i < 5; i++) { - size_t base = i < 3 ? 2 : 4; - struct u8view normd = {}; - normd.p = ucsnorm_nfd(&normd.len, columns.buf[i], alloc_arena, &ctx); - if (!ucseq(columns.buf[base], normd)) { - warn("case %d: expected c%zu to be ‘%.*s’ but got ‘%.*s’", id, - i + 1, SV_PRI_ARGS(columns.buf[base]), SV_PRI_ARGS(normd)); - rv = false; - goto out; - } - } - -out: - arena_free(&a); - return rv; -} +#define NORMTYPE nfd +#include "_norm-test.h" diff --git a/test/norm-nfkd-test.c b/test/norm-nfkd-test.c new file mode 100644 index 0000000..3fe8ff2 --- /dev/null +++ b/test/norm-nfkd-test.c @@ -0,0 +1,2 @@ +#define NORMTYPE nfkd +#include "_norm-test.h" diff --git a/test/run-tests b/test/run-tests index 860d243..ae9c96e 100755 --- a/test/run-tests +++ b/test/run-tests @@ -32,7 +32,7 @@ grep '^[^#]' data/UppercaseTest >upper.in grep '^[^#]' data/WordHumanBreakTest >wbrk-human.in sed -En 's/\s+//g; s/÷?#.*//g; /./p' data/GraphemeBreakTest >gbrk.in sed -En 's/\s+//g; s/÷?#.*//g; /./p' data/WordBreakTest >wbrk.in -sed -En 's/(#|^@).*//; s/\s+$//; /./p' data/NormalizationTest >norm-nfd.in +sed -En 's/(#|^@).*//; s/\s+$//; /./p' data/NormalizationTest >norm.in for src in *.c do -- cgit v1.2.3