Mercurial > minori

annotate dep/utf8proc/test/charwidth.c @ 347:a0aa8c8c4307

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
dep/anitomy: port to use UCS-4 rather than wide strings rationale: wide strings are not the same on every platform, and might not even be Unicode. (while they usually are, its possible that they are not) I was *going* to change StringToInt to use a string stream, but outputting to an integer doesn't seem to work at all with UCS-4, even though it ought to, so I just rolled my own that uses the arabic digits only.
author Paper <paper@paper.us.eu.org>
date Sun, 23 Jun 2024 10:32:09 -0400
parents 1faa72660932
children
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
rev   line source
343
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
1 #include "tests.h"
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
2 #include <ctype.h>
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
3 #include <wchar.h>
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
4
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
5 static int my_unassigned(int c) {
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
6 int cat = utf8proc_get_property(c)->category;
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
7 return (cat == UTF8PROC_CATEGORY_CN) || (cat == UTF8PROC_CATEGORY_CO);
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
8 }
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
9
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
10 static int my_isprint(int c) {
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
11 int cat = utf8proc_get_property(c)->category;
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
12 return (UTF8PROC_CATEGORY_LU <= cat && cat <= UTF8PROC_CATEGORY_ZS) ||
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
13 (c == 0x0601 || c == 0x0602 || c == 0x0603 || c == 0x06dd || c == 0x00ad) ||
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
14 (cat == UTF8PROC_CATEGORY_CN) || (cat == UTF8PROC_CATEGORY_CO);
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
15 }
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
16
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
17 int main(int argc, char **argv)
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
18 {
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
19 int c, error = 0, updates = 0;
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
20
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
21 (void) argc; /* unused */
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
22 (void) argv; /* unused */
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
23
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
24 /* some simple sanity tests of the character widths */
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
25 for (c = 0; c <= 0x110000; ++c) {
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
26 int cat = utf8proc_get_property(c)->category;
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
27 int w = utf8proc_charwidth(c);
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
28 if ((cat == UTF8PROC_CATEGORY_MN || cat == UTF8PROC_CATEGORY_ME) && w > 0) {
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
29 fprintf(stderr, "nonzero width %d for combining char %x\n", w, c);
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
30 error += 1;
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
31 }
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
32 if (w == 0 &&
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
33 ((cat >= UTF8PROC_CATEGORY_LU && cat <= UTF8PROC_CATEGORY_LO) ||
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
34 (cat >= UTF8PROC_CATEGORY_ND && cat <= UTF8PROC_CATEGORY_SC) ||
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
35 (cat >= UTF8PROC_CATEGORY_SO && cat <= UTF8PROC_CATEGORY_ZS))) {
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
36 fprintf(stderr, "zero width for symbol-like char %x\n", c);
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
37 error += 1;
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
38 }
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
39 if (c <= 127 && ((!isprint(c) && w > 0) || (isprint(c) && wcwidth(c) != w))) {
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
40 fprintf(stderr, "wcwidth %d mismatch %d for %s ASCII %x\n",
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
41 wcwidth(c), w,
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
42 isprint(c) ? "printable" : "non-printable", c);
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
43 error += 1;
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
44 }
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
45 if (!my_isprint(c) && w > 0) {
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
46 fprintf(stderr, "non-printing %x had width %d\n", c, w);
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
47 error += 1;
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
48 }
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
49 if (my_unassigned(c) && w != 1) {
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
50 fprintf(stderr, "unexpected width %d for unassigned char %x\n", w, c);
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
51 error += 1;
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
52 }
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
53 }
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
54 check(!error, "utf8proc_charwidth FAILED %d tests.", error);
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
55
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
56 check(utf8proc_charwidth(0x00ad) == 1, "incorrect width for U+00AD (soft hyphen)");
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
57 check(utf8proc_charwidth(0xe000) == 1, "incorrect width for U+e000 (PUA)");
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
58
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
59 /* print some other information by compariing with system wcwidth */
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
60 printf("Mismatches with system wcwidth (not necessarily errors):\n");
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
61 for (c = 0; c <= 0x110000; ++c) {
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
62 int w = utf8proc_charwidth(c);
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
63 int wc = wcwidth(c);
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
64 if (sizeof(wchar_t) == 2 && c >= (1<<16)) continue;
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
65 /* lots of these errors for out-of-date system unicode tables */
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
66 if (wc == -1 && my_isprint(c) && !my_unassigned(c) && w > 0)
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
67 updates += 1;
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
68 if (wc == -1 && !my_isprint(c) && w > 0)
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
69 printf(" wcwidth(%x) = -1 for non-printable width-%d char\n", c, w);
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
70 if (wc >= 0 && wc != w)
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
71 printf(" wcwidth(%x) = %d != charwidth %d\n", c, wc, w);
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
72 }
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
73 printf(" ... (positive widths for %d chars unknown to wcwidth) ...\n", updates);
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
74 printf("Character-width tests SUCCEEDED.\n");
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
75
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
76 return 0;
1faa72660932 *: transfer back to cmake from autotools
Paper <paper@paper.us.eu.org>
parents:
diff changeset
77 }