- 论坛徽章:
- 0
|
本帖最后由 bittertea 于 2011-03-02 18:31 编辑
字符串分割的, 来自google code search, 后来根据这个改了一版带split后执行回调的给自己用:- #include <stdio.h>
- #include <string.h>
- /*
- - split - divide a string into fields, like awk split()
- = int split(char *string, char *fields[], int nfields, char *sep);
- */
- int /* number of fields, including overflow */
- split(string, fields, nfields, sep)
- char *string;
- char *fields[]; /* list is not NULL-terminated */
- int nfields; /* number of entries available in fields[] */
- char *sep; /* "" white, "c" single char, "ab" [ab]+ */
- {
- register char *p = string;
- register char c; /* latest character */
- register char sepc = sep[0];
- register char sepc2;
- register int fn;
- register char **fp = fields;
- register char *sepp;
- register int trimtrail;
- /* white space */
- if (sepc == '\0') {
- while ((c = *p++) == ' ' || c == '\t')
- continue;
- p--;
- trimtrail = 1;
- sep = " \t"; /* note, code below knows this is 2 long */
- sepc = ' ';
- } else
- trimtrail = 0;
- sepc2 = sep[1]; /* now we can safely pick this up */
- /* catch empties */
- if (*p == '\0')
- return(0);
- /* single separator */
- if (sepc2 == '\0') {
- fn = nfields;
- for (;;) {
- *fp++ = p;
- fn--;
- if (fn == 0)
- break;
- while ((c = *p++) != sepc)
- if (c == '\0')
- return(nfields - fn);
- *(p-1) = '\0';
- }
- /* we have overflowed the fields vector -- just count them */
- fn = nfields;
- for (;;) {
- while ((c = *p++) != sepc)
- if (c == '\0')
- return(fn);
- fn++;
- }
- /* not reached */
- }
- /* two separators */
- if (sep[2] == '\0') {
- fn = nfields;
- for (;;) {
- *fp++ = p;
- fn--;
- while ((c = *p++) != sepc && c != sepc2)
- if (c == '\0') {
- if (trimtrail && **(fp-1) == '\0')
- fn++;
- return(nfields - fn);
- }
- if (fn == 0)
- break;
- *(p-1) = '\0';
- while ((c = *p++) == sepc || c == sepc2)
- continue;
- p--;
- }
- /* we have overflowed the fields vector -- just count them */
- fn = nfields;
- while (c != '\0') {
- while ((c = *p++) == sepc || c == sepc2)
- continue;
- p--;
- fn++;
- while ((c = *p++) != '\0' && c != sepc && c != sepc2)
- continue;
- }
- /* might have to trim trailing white space */
- if (trimtrail) {
- p--;
- while ((c = *--p) == sepc || c == sepc2)
- continue;
- p++;
- if (*p != '\0') {
- if (fn == nfields+1)
- *p = '\0';
- fn--;
- }
- }
- return(fn);
- }
- /* n separators */
- fn = 0;
- for (;;) {
- if (fn < nfields)
- *fp++ = p;
- fn++;
- for (;;) {
- c = *p++;
- if (c == '\0')
- return(fn);
- sepp = sep;
- while ((sepc = *sepp++) != '\0' && sepc != c)
- continue;
- if (sepc != '\0') /* it was a separator */
- break;
- }
- if (fn < nfields)
- *(p-1) = '\0';
- for (;;) {
- c = *p++;
- sepp = sep;
- while ((sepc = *sepp++) != '\0' && sepc != c)
- continue;
- if (sepc == '\0') /* it wasn't a separator */
- break;
- }
- p--;
- }
- /* not reached */
- }
- #ifdef TEST_SPLIT
- /*
- * test program
- * pgm runs regression
- * pgm sep splits stdin lines by sep
- * pgm str sep splits str by sep
- * pgm str sep n splits str by sep n times
- */
- int
- main(argc, argv)
- int argc;
- char *argv[];
- {
- char buf[512];
- register int n;
- # define MNF 10
- char *fields[MNF];
- if (argc > 4)
- for (n = atoi(argv[3]); n > 0; n--) {
- (void) strcpy(buf, argv[1]);
- }
- else if (argc > 3)
- for (n = atoi(argv[3]); n > 0; n--) {
- (void) strcpy(buf, argv[1]);
- (void) split(buf, fields, MNF, argv[2]);
- }
- else if (argc > 2)
- dosplit(argv[1], argv[2]);
- else if (argc > 1)
- while (fgets(buf, sizeof(buf), stdin) != NULL) {
- buf[strlen(buf)-1] = '\0'; /* stomp newline */
- dosplit(buf, argv[1]);
- }
- else
- regress();
- exit(0);
- }
- dosplit(string, seps)
- char *string;
- char *seps;
- {
- # define NF 5
- char *fields[NF];
- register int nf;
- nf = split(string, fields, NF, seps);
- print(nf, NF, fields);
- }
- print(nf, nfp, fields)
- int nf;
- int nfp;
- char *fields[];
- {
- register int fn;
- register int bound;
- bound = (nf > nfp) ? nfp : nf;
- printf("%d:\t", nf);
- for (fn = 0; fn < bound; fn++)
- printf("\"%s\"%s", fields[fn], (fn+1 < nf) ? ", " : "\n");
- }
- #define RNF 5 /* some table entries know this */
- struct {
- char *str;
- char *seps;
- int nf;
- char *fi[RNF];
- } tests[] = {
- "", " ", 0, { "" },
- " ", " ", 2, { "", "" },
- "x", " ", 1, { "x" },
- "xy", " ", 1, { "xy" },
- "x y", " ", 2, { "x", "y" },
- "abc def g ", " ", 5, { "abc", "def", "", "g", "" },
- " a bcd", " ", 4, { "", "", "a", "bcd" },
- "a b c d e f", " ", 6, { "a", "b", "c", "d", "e f" },
- " a b c d ", " ", 6, { "", "a", "b", "c", "d " },
- "", " _", 0, { "" },
- " ", " _", 2, { "", "" },
- "x", " _", 1, { "x" },
- "x y", " _", 2, { "x", "y" },
- "ab _ cd", " _", 2, { "ab", "cd" },
- " a_b c ", " _", 5, { "", "a", "b", "c", "" },
- "a b c_d e f", " _", 6, { "a", "b", "c", "d", "e f" },
- " a b c d ", " _", 6, { "", "a", "b", "c", "d " },
- "", " _~", 0, { "" },
- " ", " _~", 2, { "", "" },
- "x", " _~", 1, { "x" },
- "x y", " _~", 2, { "x", "y" },
- "ab _~ cd", " _~", 2, { "ab", "cd" },
- " a_b c~", " _~", 5, { "", "a", "b", "c", "" },
- "a b_c d~e f", " _~", 6, { "a", "b", "c", "d", "e f" },
- "~a b c d ", " _~", 6, { "", "a", "b", "c", "d " },
- "", " _~-", 0, { "" },
- " ", " _~-", 2, { "", "" },
- "x", " _~-", 1, { "x" },
- "x y", " _~-", 2, { "x", "y" },
- "ab _~- cd", " _~-", 2, { "ab", "cd" },
- " a_b c~", " _~-", 5, { "", "a", "b", "c", "" },
- "a b_c-d~e f", " _~-", 6, { "a", "b", "c", "d", "e f" },
- "~a-b c d ", " _~-", 6, { "", "a", "b", "c", "d " },
- "", " ", 0, { "" },
- " ", " ", 2, { "", "" },
- "x", " ", 1, { "x" },
- "xy", " ", 1, { "xy" },
- "x y", " ", 2, { "x", "y" },
- "abc def g ", " ", 4, { "abc", "def", "g", "" },
- " a bcd", " ", 3, { "", "a", "bcd" },
- "a b c d e f", " ", 6, { "a", "b", "c", "d", "e f" },
- " a b c d ", " ", 6, { "", "a", "b", "c", "d " },
- "", "", 0, { "" },
- " ", "", 0, { "" },
- "x", "", 1, { "x" },
- "xy", "", 1, { "xy" },
- "x y", "", 2, { "x", "y" },
- "abc def g ", "", 3, { "abc", "def", "g" },
- "\t a bcd", "", 2, { "a", "bcd" },
- " a \tb\t c ", "", 3, { "a", "b", "c" },
- "a b c d e ", "", 5, { "a", "b", "c", "d", "e" },
- "a b\tc d e f", "", 6, { "a", "b", "c", "d", "e f" },
- " a b c d e f ", "", 6, { "a", "b", "c", "d", "e f " },
- NULL, NULL, 0, { NULL },
- };
- regress()
- {
- char buf[512];
- register int n;
- char *fields[RNF+1];
- register int nf;
- register int i;
- register int printit;
- register char *f;
- for (n = 0; tests[n].str != NULL; n++) {
- (void) strcpy(buf, tests[n].str);
- fields[RNF] = NULL;
- nf = split(buf, fields, RNF, tests[n].seps);
- printit = 0;
- if (nf != tests[n].nf) {
- printf("split `%s' by `%s' gave %d fields, not %d\n",
- tests[n].str, tests[n].seps, nf, tests[n].nf);
- printit = 1;
- } else if (fields[RNF] != NULL) {
- printf("split() went beyond array end\n");
- printit = 1;
- } else {
- for (i = 0; i < nf && i < RNF; i++) {
- f = fields[i];
- if (f == NULL)
- f = "(NULL)";
- if (strcmp(f, tests[n].fi[i]) != 0) {
- printf("split `%s' by `%s', field %d is `%s', not `%s'\n",
- tests[n].str, tests[n].seps,
- i, fields[i], tests[n].fi[i]);
- printit = 1;
- }
- }
- }
- if (printit)
- print(nf, RNF, fields);
- }
- }
- #endif
复制代码 根据那个改的带split执行回调的,:- //定义拆分后单元结构
- typedef struct field
- {
- char *value;
- unsigned int len;
- } field_t;
- #define SPLIT_CB_OK (0)
- #define SPLIT_CB_ERR (-1)
- //拆分成功后回调
- typedef int (*split_cb_t) (field_t *fields, unsigned int nfields, void *arg);
- #define my_split(string, sep, fields, nfields) \
- my_split2(string, sep, fields, nfields, NULL, NULL)
- /*函数名称: my_split2
- *说 明: 拆分字符串
- *参 数: string 母串
- * sep 子串
- * fields 保存拆分后的单元结构数组
- * nfields 该数组个数
- * split_cb拆分成功回调
- * arg 回调函数用户数据
- *返 回: 拆分后的单元个数
- */
- unsigned int my_split2(char *string, char *sep, field_t fields[],
- unsigned int nfields, split_cb_t split_cb, void *arg)
- {
- char *p = string;
- char c;
- char sepc;
- char sepc2;
- unsigned int fn;
- field_t *fp = fields;
- char *sepp;
- int trimtrail;
- if (!p || *p == '\0')
- return 0;
- if (!sep || sep[0] == '\0')
- {
- //未指定分隔字符, 则去掉开始的white space
- while ((c = *p) == ' ' || c == '\t')
- {
- p++;
- continue;
- }
- trimtrail = 1;
- sep = " \t"; //分隔字符为" \t"
- }
- else
- trimtrail = 0;
- sepc = sep[0];
- sepc2 = sep[1];
- if (sepc2 == '\0')
- {
- //分隔字符为单字符
- fn = nfields;
- for (;;)
- {
- fp -> value = p;
- fn--;
- while ((c = *p) != sepc)
- {
- if (c == '\0')
- {
- fp -> len = p - fp -> value;
- do_split_cb(split_cb, fp, nfields - fn, arg);
- return (nfields - fn);
- }
- p++;
- }
- fp -> len = p - fp -> value; //可能为0
- do_split_cb(split_cb, fp, nfields - fn, arg);
- p++;
- if (fn == 0)
- break;
- fp++;
- }
- //若仍有未处理的字符串, 则返回总token个数
- fn = nfields;
- for (;;)
- {
- while ((c = *p) != sepc)
- {
- if (c == '\0') return fn;
- p++;
- }
- fn++;
- }
- }
- if (sep[2] == '\0')
- {
- //分隔字符串含有两个字符
- fn = nfields;
- for (; ; )
- {
- fp -> value = p;
- fn--;
- while ((c = *p) != sepc && c != sepc2)
- {
- if (c == '\0')
- {
- if (trimtrail && (fp -> value) == '\0')
- {
- fn++;
- }
- else
- {
- fp -> len = p - fp -> value;
- do_split_cb(split_cb, fp, nfields - fn, arg);
- }
- return (nfields - fn);
- }
- p++;
- }
- fp -> len = p - fp -> value;
- do_split_cb(split_cb, fp, nfields - fn, arg);
- p++;
- if (fn == 0)
- break;
- fp++;
- while ((c = *p) == sepc || c == sepc2)
- {
- p++;
- continue;
- }
- }
- //若仍有未处理的字符串, 则返回总token个数
- fn = nfields;
- while (c != '\0')
- {
- while ((c = *p) == sepc || c == sepc2)
- {
- p++;
- continue;
- }
- fn++;
- while ((c = *p) != '\0' && c != sepc && c != sepc2)
- {
- p++;
- continue;
- }
- p++;
- }
- p--; //p指向'\0'
- if (trimtrail)
- {
- p--; //p指向'\0'前一个字节
- while ((c = *p) == sepc || c == sepc2)
- {
- //去除尾部white space
- p--;
- continue;
- }
- p++;
- if (*p != '\0')
- {
- //尾部有white space
- fn--;
- }
- }
- return fn;
- }
- //分隔字符串含有多个字符
- fn = 0;
- for (; ; )
- {
- if (fn < nfields)
- {
- fp -> value = p;
- //*fp++ = p;
- }
- fn++;
- for (; ; )
- {
- c = *p; //获取母串当前字符
- if (c == '\0')
- {
- if (fn <= nfields)
- {
- fp -> len = p - fp -> value;
- do_split_cb(split_cb, fp, fn, arg);
- }
- return fn;
- }
- //匹配当前字符是否在分隔符中
- sepp = sep;
- while ((sepc = *sepp) != '\0' && sepc != c)
- {
- sepp++;
- continue;
- }
- //母串中遇到分隔符则跳出循环, 否则母串字符指针递增
- if (sepc != '\0')
- break;
- p++;
- }
- if (fn <= nfields)
- {
- fp -> len = p - fp -> value;
- do_split_cb(split_cb, fp, nfields - fn, arg);
- fp++;
- }
- p++;
- for (; ; )
- {
- //略去母串中连续得分隔符
- c = *p;
- sepp = sep;
- while ((sepc = *sepp) != '\0' && sepc != c)
- {
- sepp++;
- continue;
- }
- if (sepc == '\0')
- break;
- p++;
- }
- }
- /* not reached */
- }
复制代码 |
|