lh | 9ed821d | 2023-04-07 01:36:19 -0700 | [diff] [blame] | 1 | #include <stdio.h> |
| 2 | #include <ctype.h> |
| 3 | #include <errno.h> |
| 4 | #include <locale.h> |
| 5 | #include <regex.h> |
| 6 | #include <stdlib.h> |
| 7 | #include <unistd.h> |
| 8 | #include <string.h> |
| 9 | |
| 10 | /* lowercase chars mapped to uppercase */ |
| 11 | static const char casetable[] = { |
| 12 | '\000', '\001', '\002', '\003', '\004', '\005', '\006', '\007', |
| 13 | '\010', '\011', '\012', '\013', '\014', '\015', '\016', '\017', |
| 14 | '\020', '\021', '\022', '\023', '\024', '\025', '\026', '\027', |
| 15 | '\030', '\031', '\032', '\033', '\034', '\035', '\036', '\037', |
| 16 | /* ' ' '!' '"' '#' '$' '%' '&' ''' */ |
| 17 | '\040', '\041', '\042', '\043', '\044', '\045', '\046', '\047', |
| 18 | /* '(' ')' '*' '+' ',' '-' '.' '/' */ |
| 19 | '\050', '\051', '\052', '\053', '\054', '\055', '\056', '\057', |
| 20 | /* '0' '1' '2' '3' '4' '5' '6' '7' */ |
| 21 | '\060', '\061', '\062', '\063', '\064', '\065', '\066', '\067', |
| 22 | /* '8' '9' ':' ';' '<' '=' '>' '?' */ |
| 23 | '\070', '\071', '\072', '\073', '\074', '\075', '\076', '\077', |
| 24 | /* '@' 'A' 'B' 'C' 'D' 'E' 'F' 'G' */ |
| 25 | '\100', '\141', '\142', '\143', '\144', '\145', '\146', '\147', |
| 26 | /* 'H' 'I' 'J' 'K' 'L' 'M' 'N' 'O' */ |
| 27 | '\150', '\151', '\152', '\153', '\154', '\155', '\156', '\157', |
| 28 | /* 'P' 'Q' 'R' 'S' 'T' 'U' 'V' 'W' */ |
| 29 | '\160', '\161', '\162', '\163', '\164', '\165', '\166', '\167', |
| 30 | /* 'X' 'Y' 'Z' '[' '\' ']' '^' '_' */ |
| 31 | '\170', '\171', '\172', '\133', '\134', '\135', '\136', '\137', |
| 32 | /* '`' 'a' 'b' 'c' 'd' 'e' 'f' 'g' */ |
| 33 | '\140', '\141', '\142', '\143', '\144', '\145', '\146', '\147', |
| 34 | /* 'h' 'i' 'j' 'k' 'l' 'm' 'n' 'o' */ |
| 35 | '\150', '\151', '\152', '\153', '\154', '\155', '\156', '\157', |
| 36 | /* 'p' 'q' 'r' 's' 't' 'u' 'v' 'w' */ |
| 37 | '\160', '\161', '\162', '\163', '\164', '\165', '\166', '\167', |
| 38 | /* 'x' 'y' 'z' '{' '|' '}' '~' */ |
| 39 | '\170', '\171', '\172', '\173', '\174', '\175', '\176', '\177', |
| 40 | |
| 41 | /* Latin 1: */ |
| 42 | '\200', '\201', '\202', '\203', '\204', '\205', '\206', '\207', |
| 43 | '\210', '\211', '\212', '\213', '\214', '\215', '\216', '\217', |
| 44 | '\220', '\221', '\222', '\223', '\224', '\225', '\226', '\227', |
| 45 | '\230', '\231', '\232', '\233', '\234', '\235', '\236', '\237', |
| 46 | '\240', '\241', '\242', '\243', '\244', '\245', '\246', '\247', |
| 47 | '\250', '\251', '\252', '\253', '\254', '\255', '\256', '\257', |
| 48 | '\260', '\261', '\262', '\263', '\264', '\265', '\266', '\267', |
| 49 | '\270', '\271', '\272', '\273', '\274', '\275', '\276', '\277', |
| 50 | '\340', '\341', '\342', '\343', '\344', '\345', '\346', '\347', |
| 51 | '\350', '\351', '\352', '\353', '\354', '\355', '\356', '\357', |
| 52 | '\360', '\361', '\362', '\363', '\364', '\365', '\366', '\327', |
| 53 | '\370', '\371', '\372', '\373', '\374', '\375', '\376', '\337', |
| 54 | '\340', '\341', '\342', '\343', '\344', '\345', '\346', '\347', |
| 55 | '\350', '\351', '\352', '\353', '\354', '\355', '\356', '\357', |
| 56 | '\360', '\361', '\362', '\363', '\364', '\365', '\366', '\367', |
| 57 | '\370', '\371', '\372', '\373', '\374', '\375', '\376', '\377', |
| 58 | }; |
| 59 | |
| 60 | |
| 61 | static int |
| 62 | run_test (const char *pattern, struct re_registers *regs) |
| 63 | { |
| 64 | static char text[] = "1111AAAA2222bbbb"; |
| 65 | |
| 66 | struct re_pattern_buffer pat; |
| 67 | |
| 68 | const char *err; |
| 69 | int res; |
| 70 | int start2; |
| 71 | |
| 72 | memset (&pat, '\0', sizeof (pat)); |
| 73 | memset (regs, '\0', 2 * sizeof (regs[0])); |
| 74 | pat.allocated = 0; /* regex will allocate the buffer */ |
| 75 | pat.fastmap = (char *) malloc (256); |
| 76 | if (pat.fastmap == NULL) |
| 77 | { |
| 78 | puts ("out of memory"); |
| 79 | exit (1); |
| 80 | } |
| 81 | |
| 82 | pat.translate = (unsigned char *) casetable; |
| 83 | |
| 84 | err = re_compile_pattern (pattern, strlen (pattern), &pat); |
| 85 | if (err != NULL) |
| 86 | { |
| 87 | fprintf (stderr, "/%s/: %s\n", pattern, err); |
| 88 | exit (1); |
| 89 | } |
| 90 | res = re_search (&pat, text, strlen (text), 0, strlen (text), ®s[0]); |
| 91 | if (res < 0) |
| 92 | printf ("search 1: res = %d\n", res); |
| 93 | else |
| 94 | printf ("search 1: res = %d, start = %d, end = %d\n", |
| 95 | res, regs[0].start[0], regs[0].end[0]); |
| 96 | |
| 97 | if (regs[0].end == NULL) |
| 98 | start2 = 8; |
| 99 | else |
| 100 | start2 = regs[0].end[0] + 1; |
| 101 | regs[1] = regs[0]; |
| 102 | res = re_search (&pat, text, strlen (text), start2, strlen (text), ®s[1]); |
| 103 | if (res < 0) |
| 104 | printf ("search 2: res = %d\n", res); |
| 105 | else |
| 106 | printf ("search 2: res = %d, start = %d, end = %d\n", |
| 107 | res, regs[1].start[0], regs[1].end[0]); |
| 108 | |
| 109 | return res < 0 ? 1 : 0; |
| 110 | } |
| 111 | |
| 112 | |
| 113 | static int |
| 114 | do_test (void) |
| 115 | { |
| 116 | static const char lower[] = "[[:lower:]]+"; |
| 117 | static const char upper[] = "[[:upper:]]+"; |
| 118 | struct re_registers regs[4]; |
| 119 | |
| 120 | setlocale (LC_ALL, "C"); |
| 121 | |
| 122 | (void) re_set_syntax (RE_SYNTAX_GNU_AWK); |
| 123 | |
| 124 | int result; |
| 125 | #define CHECK(exp) \ |
| 126 | if (exp) { puts (#exp); result = 1; } |
| 127 | |
| 128 | result = run_test (lower, regs); |
| 129 | result |= run_test (upper, ®s[2]); |
| 130 | if (! result) |
| 131 | { |
| 132 | CHECK (regs[0].start[0] != regs[2].start[0]); |
| 133 | CHECK (regs[0].end[0] != regs[2].end[0]); |
| 134 | CHECK (regs[1].start[0] != regs[3].start[0]); |
| 135 | CHECK (regs[1].end[0] != regs[3].end[0]); |
| 136 | } |
| 137 | |
| 138 | return result; |
| 139 | } |
| 140 | |
| 141 | #define TEST_FUNCTION do_test () |
| 142 | #include "../test-skeleton.c" |