void
BUrl::_ExplodeUrlString(const BString& url)
{
int16 urlIndex = 0;
_ResetFields();
_ExtractProtocol(url, &urlIndex);
_ExtractAuthority(url, &urlIndex);
_ExtractPath(url, &urlIndex);
_ExtractRequestAndFragment(url, &urlIndex);
}
BUrl::BUrl()
:
fUrlString(),
fProtocol(),
fUser(),
fPassword(),
fHost(),
fPort(0),
fPath(),
fRequest(),
fHasAuthority(false)
{
_ResetFields();
}
BUrl::BUrl()
:
fUrlString(),
fProtocol(),
fUser(),
fPassword(),
fHost(),
fPort(0),
fPath(),
fRequest(),
fHasHost(false),
fHasFragment(false)
{
_ResetFields();
}
void
BUrl::_ExplodeUrlString(const BString& url)
{
// The regexp is provided in RFC3986 (URI generic syntax), Appendix B
static RegExp urlMatcher(
"^(([^:/?#]+):)?(//([^/?#]*))?([^?#]*)(\\?([^#]*))?(#(.*))?");
_ResetFields();
RegExp::MatchResult match = urlMatcher.Match(url.String());
if (!match.HasMatched())
return; // TODO error reporting
// Scheme/Protocol
url.CopyInto(fProtocol, match.GroupStartOffsetAt(1),
match.GroupEndOffsetAt(1) - match.GroupStartOffsetAt(1));
if (!_IsProtocolValid()) {
fHasProtocol = false;
fProtocol.Truncate(0);
} else
fHasProtocol = true;
// Authority (including user credentials, host, and port
url.CopyInto(fAuthority, match.GroupStartOffsetAt(3),
match.GroupEndOffsetAt(3) - match.GroupStartOffsetAt(3));
SetAuthority(fAuthority);
// Path
url.CopyInto(fPath, match.GroupStartOffsetAt(4),
match.GroupEndOffsetAt(4) - match.GroupStartOffsetAt(4));
if (!fPath.IsEmpty())
fHasPath = true;
// Query
url.CopyInto(fRequest, match.GroupStartOffsetAt(6),
match.GroupEndOffsetAt(6) - match.GroupStartOffsetAt(6));
if (!fRequest.IsEmpty())
fHasRequest = true;
// Fragment
url.CopyInto(fFragment, match.GroupStartOffsetAt(8),
match.GroupEndOffsetAt(8) - match.GroupStartOffsetAt(8));
if (!fFragment.IsEmpty())
fHasFragment = true;
}
BUrl::BUrl(BMessage* archive)
:
fUrlString(),
fProtocol(),
fUser(),
fPassword(),
fHost(),
fPort(0),
fPath(),
fRequest(),
fHasAuthority(false)
{
BString url;
if (archive->FindString(kArchivedUrl, &url) == B_OK)
SetUrlString(url);
else
_ResetFields();
}
/*
* This function takes a URL in string-form and parses the components of the URL out.
*/
status_t
BUrl::_ExplodeUrlString(const BString& url)
{
_ResetFields();
// RFC3986, Appendix C; the URL should not contain whitespace or delimiters
// by this point.
if (_ContainsDelimiter(url))
return B_BAD_VALUE;
explode_url_parse_state state = EXPLODE_PROTOCOL;
int32 offset = 0;
int32 length = url.Length();
const char *url_c = url.String();
// The regexp is provided in RFC3986 (URI generic syntax), Appendix B
// ^(([^:/?#]+):)?(//([^/?#]*))?([^?#]*)(\\?([^#]*))?(#(.*))?
// The ensuing logic attempts to simulate the behaviour of extracting the groups
// from the string without requiring a group-capable regex engine.
while (offset < length) {
switch (state) {
case EXPLODE_PROTOCOL:
{
int32 end_protocol = char_offset_until_fn_false(url_c, length,
offset, explode_is_protocol_char);
if (end_protocol < length) {
SetProtocol(BString(&url_c[offset], end_protocol - offset));
state = EXPLODE_PROTOCOLTERMINATOR;
offset = end_protocol;
} else {
// No protocol was found, try parsing from the string
// start, beginning with authority or path
SetProtocol("");
offset = 0;
state = EXPLODE_AUTHORITYORPATH;
}
break;
}
case EXPLODE_PROTOCOLTERMINATOR:
{
if (url[offset] == ':') {
offset++;
} else {
// No protocol was found, try parsing from the string
// start, beginning with authority or path
SetProtocol("");
offset = 0;
}
state = EXPLODE_AUTHORITYORPATH;
break;
}
case EXPLODE_AUTHORITYORPATH:
{
// The authority must start with //. If it isn't there, skip
// to parsing the path.
if (strncmp(&url_c[offset], "//", 2) == 0) {
state = EXPLODE_AUTHORITY;
offset += 2;
} else {
state = EXPLODE_PATH;
}
break;
}
case EXPLODE_AUTHORITY:
{
int end_authority = char_offset_until_fn_false(url_c, length,
offset, explode_is_authority_char);
SetAuthority(BString(&url_c[offset], end_authority - offset));
state = EXPLODE_PATH;
offset = end_authority;
break;
}
case EXPLODE_PATH:
{
int end_path = char_offset_until_fn_false(url_c, length, offset,
explode_is_path_char);
SetPath(BString(&url_c[offset], end_path - offset));
state = EXPLODE_REQUEST;
offset = end_path;
break;
}
case EXPLODE_REQUEST: // query
{
if (url_c[offset] == '?') {
offset++;
int end_request = char_offset_until_fn_false(url_c, length,
offset, explode_is_request_char);
SetRequest(BString(&url_c[offset], end_request - offset));
offset = end_request;
}
state = EXPLODE_FRAGMENT;
break;
}
case EXPLODE_FRAGMENT:
{
if (url_c[offset] == '#') {
offset++;
SetFragment(BString(&url_c[offset], length - offset));
offset = length;
}
state = EXPLODE_COMPLETE;
break;
}
case EXPLODE_COMPLETE:
// should never be reached - keeps the compiler happy
break;
}
}
return B_OK;
}