Add tests for u8wnext_human()

3 files changed, 96 insertions, 6 deletions

diff --git a/test/data/WordHumanBreakTest b/test/data/WordHumanBreakTest
new file mode 100644
index 0000000..8eea00f
--- /dev/null
+++ b/test/data/WordHumanBreakTest
@@ -0,0 +1,11 @@
+# Empty input
+;
+
+# Latin alphabet
+C23 (ISO/IEC 9899:2024), the next C standard, replaces C17 (ISO/IEC 9899:2018).;C23|ISO|IEC|9899|2024|the|next|C|standard|replaces|C17|ISO|IEC|9899|2018
+
+# Greek alphabet
+Το ιουλιανό ημερολόγιο (365,25 ημέρες);Το|ιουλιανό|ημερολόγιο|365,25|ημέρες
+
+# Maltese with Arabic
+Il-lingwist Malti Ġużè Aquilina kien jemmen li 'Mnajdra' (bl-Għarbi: منيدرة);Il|lingwist|Malti|Ġużè|Aquilina|kien|jemmen|li|Mnajdra|bl|Għarbi|منيدرة
diff --git a/test/run-tests b/test/run-tests
index f4c53c4..2562a32 100755
--- a/test/run-tests
+++ b/test/run-tests
@@ -23,12 +23,13 @@ readonly FLAGS='
 download 'auxiliary/GraphemeBreakTest.txt'
 download 'auxiliary/WordBreakTest.txt'
 
-grep '^[^#]'                         data/CasefoldTest      >casefold.in
-grep '^[^#]'                         data/LowercaseTest     >lower.in
-grep '^[^#]'                         data/TitlecaseTest     >title.in
-grep '^[^#]'                         data/UppercaseTest     >upper.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
+grep '^[^#]'                         data/CasefoldTest       >casefold.in
+grep '^[^#]'                         data/LowercaseTest      >lower.in
+grep '^[^#]'                         data/TitlecaseTest      >title.in
+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
 
 for src in *.c
 do
diff --git a/test/wbrk-human-test.c b/test/wbrk-human-test.c
new file mode 100644
index 0000000..1fa9bd8
--- /dev/null
+++ b/test/wbrk-human-test.c
@@ -0,0 +1,78 @@
+#define _GNU_SOURCE
+#include <stdio.h>
+#include <stdlib.h>
+
+#include <dynarr.h>
+#include <errors.h>
+#include <macros.h>
+#include <mbstring.h>
+#include <unicode/string.h>
+
+#define TESTFILE "wbrk-human.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, nr}, id))
+			rv = EXIT_FAILURE;
+	}
+	if (ferror(fp))
+		err("getline: %s:", TESTFILE);
+
+	free(line);
+	fclose(fp);
+	return rv;
+}
+
+bool
+test(struct u8view sv, int id)
+{
+	struct u8view src;
+	u8cut(&src, &sv, U";", 1);
+
+	struct u8view w;
+	dynarr(struct u8view) ws = {};
+
+	while (u8cut(&w, &sv, U"|", 1) != MBEND)
+		DAPUSH(&ws, w);
+	if (w.len > 0)
+		DAPUSH(&ws, w);
+
+	/* Assert the word count is correct */
+	size_t n;
+	if ((n = u8wcnt_human(src)) != ws.len) {
+		warn("case %d: expected %zu words but got %zu", id, ws.len, n);
+		return false;
+	}
+
+	/* Assert the individual words are correct */
+	for (size_t i = 0; u8wnext_human(&w, &src) != 0; i++) {
+		if (!u8eq(w, ws.buf[i])) {
+			warn("case %d: expected word %zu to be ‘%.*s’ but got ‘%.*s’", id,
+			     i, SV_PRI_ARGS(ws.buf[i]), SV_PRI_ARGS(w));
+			return false;
+		}
+	}
+
+	free(ws.buf);
+	return true;
+}