/* Documented in save_complete.h */
void save_complete_init(void)
{
/* Match an @import rule - see CSS 2.1 G.1. */
regcomp_wrapper(&save_complete_import_re,
"@import" /* IMPORT_SYM */
"[ \t\r\n\f]*" /* S* */
/* 1 */
"(" /* [ */
/* 2 3 */
"\"(([^\"]|[\\]\")*)\"" /* STRING (approximated) */
"|"
/* 4 5 */
"'(([^']|[\\]')*)'"
"|" /* | */
"url\\([ \t\r\n\f]*" /* URI (approximated) */
/* 6 7 */
"\"(([^\"]|[\\]\")*)\""
"[ \t\r\n\f]*\\)"
"|"
"url\\([ \t\r\n\f]*"
/* 8 9 */
"'(([^']|[\\]')*)'"
"[ \t\r\n\f]*\\)"
"|"
"url\\([ \t\r\n\f]*"
/* 10 */
"([^) \t\r\n\f]*)"
"[ \t\r\n\f]*\\)"
")", /* ] */
REG_EXTENDED | REG_ICASE);
}
void url_init(void)
{
/* regex from RFC 2396 */
regcomp_wrapper(&url_re, "^[[:space:]]*"
#define URL_RE_SCHEME 2
"(([a-zA-Z][-a-zA-Z0-9+.]*):)?"
#define URL_RE_AUTHORITY 4
"(//([^/?#[:space:]]*))?"
#define URL_RE_PATH 5
"([^?#[:space:]]*)"
#define URL_RE_QUERY 7
"(\\?([^#[:space:]]*))?"
#define URL_RE_FRAGMENT 9
"(#([^[:space:]]*))?"
"[[:space:]]*$", REG_EXTENDED);
regcomp_wrapper(&url_up_re,
"/([^/]?|[.][^./]|[^./][.]|[^./][^./]|[^/][^/][^/]+)"
"/[.][.](/|$)",
REG_EXTENDED);
}
/**
* Initialise the curl handle and compile regular expressions.
*/
void init(void)
{
struct curl_slist *request_headers = 0;
if (curl_global_init(CURL_GLOBAL_ALL))
die("Failed to initialise libcurl");
curl = curl_easy_init();
if (!curl)
die("Failed to create curl handle");
if (curl_easy_setopt(curl, CURLOPT_HEADERFUNCTION, header_callback))
die("Failed to set curl options");
if (curl_easy_setopt(curl, CURLOPT_WRITEFUNCTION, data_callback))
die("Failed to set curl options");
if (curl_easy_setopt(curl, CURLOPT_USERAGENT, "httplint"))
die("Failed to set curl options");
if (curl_easy_setopt(curl, CURLOPT_ERRORBUFFER, error_buffer))
die("Failed to set curl options");
/* remove libcurl default headers */
request_headers = curl_slist_append(request_headers, "Accept:");
request_headers = curl_slist_append(request_headers, "Pragma:");
if (curl_easy_setopt(curl, CURLOPT_HTTPHEADER, request_headers))
die("Failed to set curl options");
/* compile regular expressions */
regcomp_wrapper(&re_status_line,
"^HTTP/([0-9]+)[.]([0-9]+) ([0-9][0-9][0-9]) ([\t -~€-ÿ]*)$",
REG_EXTENDED);
regcomp_wrapper(&re_token,
"^([-0-9a-zA-Z_.!]+)",
REG_EXTENDED);
regcomp_wrapper(&re_token_value,
"^([-0-9a-zA-Z_.!]+)(=([-0-9a-zA-Z_.!]+|\"([^\"]|[\\].)*\"))?",
REG_EXTENDED);
regcomp_wrapper(&re_content_type,
"^([-0-9a-zA-Z_.]+)/([-0-9a-zA-Z_.]+)[ \t]*"
"(;[ \t]*([-0-9a-zA-Z_.]+)="
"([-0-9a-zA-Z_.]+|\"([^\"]|[\\].)*\")[ \t]*)*$",
REG_EXTENDED);
regcomp_wrapper(&re_absolute_uri,
"^[a-zA-Z0-9]+://[^ ]+$",
REG_EXTENDED);
regcomp_wrapper(&re_etag,
"^(W/[ \t]*)?\"([^\"]|[\\].)*\"$",
REG_EXTENDED);
regcomp_wrapper(&re_server,
"^((([-0-9a-zA-Z_.!]+(/[-0-9a-zA-Z_.]+)?)|(\\(.*\\)))[ \t]*)+$",
REG_EXTENDED);
regcomp_wrapper(&re_transfer_coding,
"^([-0-9a-zA-Z_.]+)[ \t]*"
"(;[ \t]*([-0-9a-zA-Z_.]+)="
"([-0-9a-zA-Z_.]+|\"([^\"]|[\\].)*\")[ \t]*)*$",
REG_EXTENDED);
regcomp_wrapper(&re_upgrade,
"^([-0-9a-zA-Z_.](/[-0-9a-zA-Z_.])?)+$",
REG_EXTENDED);
regcomp_wrapper(&re_ugly,
"^[a-zA-Z0-9]+://[^/]+[-/a-zA-Z0-9_]*$",
REG_EXTENDED);
regcomp_wrapper(&re_rfc1123,
"^(Mon|Tue|Wed|Thu|Fri|Sat|Sun), ([0123][0-9]) "
"(Jan|Feb|Mar|Apr|May|Jun|Jul|Aug|Sep|Oct|Nov|Dec) ([0-9]{4}) "
"([012][0-9]):([0-5][0-9]):([0-5][0-9]) GMT$",
REG_EXTENDED);
regcomp_wrapper(&re_rfc1036,
"^(Monday|Tuesday|Wednesday|Thursday|Friday|Saturday|Sunday), "
"([0123][0-9])-(Jan|Feb|Mar|Apr|May|Jun|Jul|Aug|Sep|Oct|Nov|Dec)-"
"([0-9][0-9]) ([012][0-9]):([0-5][0-9]):([0-5][0-9]) GMT$",
REG_EXTENDED);
regcomp_wrapper(&re_asctime,
"^(Mon|Tue|Wed|Thu|Fri|Sat|Sun) "
"(Jan|Feb|Mar|Apr|May|Jun|Jul|Aug|Sep|Oct|Nov|Dec) ([ 12][0-9]) "
"([012][0-9]):([0-5][0-9]):([0-5][0-9]) ([0-9]{4})$",
REG_EXTENDED);
regcomp_wrapper(&re_cookie_nameval,
"^[^;, ]+=[^;, ]*$",
REG_EXTENDED);
regcomp_wrapper(&re_cookie_expires,
"^(Mon|Tue|Wed|Thu|Fri|Sat|Sun), ([0123][0-9])-"
"(Jan|Feb|Mar|Apr|May|Jun|Jul|Aug|Sep|Oct|Nov|Dec)-([0-9]{4}) "
"([012][0-9]):([0-5][0-9]):([0-5][0-9]) GMT$",
REG_EXTENDED);
}