myvtm/libtoolkit/url.c

#include "precompile.h"
#include "Url.h"

#include <stdlib.h>
#include <string.h>
#include <ctype.h>

/* from libasynchttp */

TOOLKIT_API char* urlencode(unsigned char *string) {
	int escapecount = 0;
	unsigned char *src, *dest;
	unsigned char *newstr;
	
	char hextable[] = { '0', '1', '2', '3', '4', '5', '6', '7',
					'8', '9', 'A', 'B', 'C', 'D', 'E', 'F' };
	
	if (string == NULL) return NULL;

	for (src = string; *src != 0; src++) 
		if (!isalnum(*src)) escapecount++;
	
	newstr = (unsigned char*)malloc(strlen((char*)string) - escapecount + (escapecount * 3) + 1);
	
	src = string;
	dest = newstr;
	while (*src != 0) {
		if (!isalnum(*src)) {
			*dest++ = '%';
			*dest++ = hextable[*src >> 4];
			*dest++ = hextable[*src & 0x0F];
			src++;
		} else {
			*dest++ = *src++;
		}
	}
	*dest = 0;

	return (char*)newstr;
}

TOOLKIT_API unsigned char* urldecode(char *string) {
	int destlen = 0;
	unsigned char *src, *dest;
	unsigned char *newstr;
	
	if (string == NULL) return NULL;

	for (src = (unsigned char*)string; *src != 0; src++) {
		if (*src == '%') { src+=2; } /* FIXME: this isn't robust. should check
										the next two chars for 0 */
		destlen++;
	}
	
	newstr = (unsigned char*)malloc(destlen + 1);
	src = (unsigned char*)string;
	dest = newstr;
	while (*src != 0) {
		if (*src == '%') {
			char h = toupper(src[1]);
			char l = toupper(src[2]);
			int vh, vl;
			vh = isalpha(h) ? (10+(h-'A')) : (h-'0');
			vl = isalpha(l) ? (10+(l-'A')) : (l-'0');
			*dest++ = ((vh<<4)+vl);
			src += 3;
		} else if (*src == '+') {
			*dest++ = ' ';
			src++;
		} else {
			*dest++ = *src++;
		}
	}
	*dest = 0;

	return newstr;
}

TOOLKIT_API char* util_skipline(char *text) {
	while (*text != 0) {
		if (*text == '\n') {
			text++;
			break;
		}
		text++;
	}
	return text;
}

TOOLKIT_API char* util_getline(char *text) {
	int len;
	char *str;

	for (len = 0; text[len] != 0; len++)
		if (text[len] == '\r' || text[len] == '\n') break;

	str = (char*)malloc(len+1);
	memcpy(str, text, len+1);
	str[len] = 0;
	return str;
}

static char*strgrab(char *src, int len) {
	char *ret = (char*)malloc(len+1);
	memcpy(ret, src, len);
	ret[len] = 0;
	return ret;
}

static char* strscan(char *src, char *origsearch) {
	char *search;
	while (*src) {
		search = origsearch;
		while (*search) {
			if (*src == *search)
				return src;
			search++;
		}
		src++;
	}
	return src;
}

TOOLKIT_API int url_parse(char *url, url_fields *uf) {
	char *s, *e;

	memset(uf, 0, sizeof(url_fields));

	/* search for end of scheme */
	s = url;
	e = strscan(s, ":");
	uf->scheme = strgrab(s, e-s);
	s = e+1;
	if (*s == 0) goto err;

	if (*s == '/' && *(s+1) == '/') {
		/* it's a net_path */
	} else goto err;
	s = s+2;

	/* search for / or : trailing hostname */
	e = strscan(s, "/:");
	if (e == s) goto err;

	uf->host = strgrab(s, e-s);
	s = e;

	if (*s == ':') {
		char *port;

		s++;
		/* grab the port */
		e = strscan(s, "/");
		if (e == s) goto err;
		port = strgrab(s, e-s);
		uf->port = atoi(port);
		free(port);

		s = e;
	}

	/* grab remainder into path */
	e = strscan(s, "");
	uf->path = strgrab(s, e-s);

	/* if they've not specified a path, default to / */
	if (uf->path[0] == 0) {
		free(uf->path);
		uf->path = _strdup("/");
	}
	return 0;
err:
	if (uf->path) free(uf->path);
	if (uf->host) free(uf->host);
	if (uf->scheme) free(uf->scheme);
	return -(e-url);
}

TOOLKIT_API void url_free_fields(url_fields *uf) {
	if (uf->scheme) free(uf->scheme);
	if (uf->host) free(uf->host);
	if (uf->path) free(uf->path);
}