summaryrefslogtreecommitdiffabout
path: root/lib
authorMichael Krelin <hacker@klever.net>2007-08-09 00:01:12 (UTC)
committer Michael Krelin <hacker@klever.net>2007-08-09 00:01:12 (UTC)
commit94e566c912f97b4c9e47353e10e86649ee8060a4 (patch) (unidiff)
treea11910bf677723a7eabccad161ec0b85426293ff /lib
parent4696aae9da3500f600cedd482501a89d41fd27ec (diff)
downloadlibopkele-94e566c912f97b4c9e47353e10e86649ee8060a4.zip
libopkele-94e566c912f97b4c9e47353e10e86649ee8060a4.tar.gz
libopkele-94e566c912f97b4c9e47353e10e86649ee8060a4.tar.bz2
make canonicalize follow redirects
- renamed old consumer_t::canonicalize to consumer_t::normalize - made the real canonicalize virtual to allow caching layer
Diffstat (limited to 'lib') (more/less context) (ignore whitespace changes)
-rw-r--r--lib/consumer.cc26
1 files changed, 25 insertions, 1 deletions
diff --git a/lib/consumer.cc b/lib/consumer.cc
index 282f0cc..dd8e150 100644
--- a/lib/consumer.cc
+++ b/lib/consumer.cc
@@ -249,97 +249,121 @@ namespace opkele {
249 if(pp.has_param("invalidate_handle")) 249 if(pp.has_param("invalidate_handle"))
250 invalidate_assoc(server,pp.get_param("invalidate_handle")); 250 invalidate_assoc(server,pp.get_param("invalidate_handle"));
251 if(pp.has_param("is_valid")) { 251 if(pp.has_param("is_valid")) {
252 if(pp.get_param("is_valid")=="true") 252 if(pp.get_param("is_valid")=="true")
253 return; 253 return;
254 }else if(pp.has_param("lifetime")) { 254 }else if(pp.has_param("lifetime")) {
255 if(util::string_to_long(pp.get_param("lifetime"))) 255 if(util::string_to_long(pp.get_param("lifetime")))
256 return; 256 return;
257 } 257 }
258 throw failed_check_authentication(OPKELE_CP_ "failed to verify response"); 258 throw failed_check_authentication(OPKELE_CP_ "failed to verify response");
259 } 259 }
260 260
261 void consumer_t::retrieve_links(const string& url,string& server,string& delegate) { 261 void consumer_t::retrieve_links(const string& url,string& server,string& delegate) {
262 server.erase(); 262 server.erase();
263 delegate.erase(); 263 delegate.erase();
264 curl_t curl = curl_easy_init(); 264 curl_t curl = curl_easy_init();
265 if(!curl) 265 if(!curl)
266 throw exception_curl(OPKELE_CP_ "failed to curl_easy_init()"); 266 throw exception_curl(OPKELE_CP_ "failed to curl_easy_init()");
267 string html; 267 string html;
268 CURLcode r; 268 CURLcode r;
269 (r=curl_misc_sets(curl)) 269 (r=curl_misc_sets(curl))
270 || (r=curl_easy_setopt(curl,CURLOPT_URL,url.c_str())) 270 || (r=curl_easy_setopt(curl,CURLOPT_URL,url.c_str()))
271 || (r=curl_easy_setopt(curl,CURLOPT_WRITEFUNCTION,_curl_tostring)) 271 || (r=curl_easy_setopt(curl,CURLOPT_WRITEFUNCTION,_curl_tostring))
272 || (r=curl_easy_setopt(curl,CURLOPT_WRITEDATA,&html)) 272 || (r=curl_easy_setopt(curl,CURLOPT_WRITEDATA,&html))
273 ; 273 ;
274 if(r) 274 if(r)
275 throw exception_curl(OPKELE_CP_ "failed to curl_easy_setopt()",r); 275 throw exception_curl(OPKELE_CP_ "failed to curl_easy_setopt()",r);
276 r = curl_easy_perform(curl); 276 r = curl_easy_perform(curl);
277 if(r && r!=CURLE_WRITE_ERROR) 277 if(r && r!=CURLE_WRITE_ERROR)
278 throw exception_curl(OPKELE_CP_ "failed to curl_easy_perform()",r); 278 throw exception_curl(OPKELE_CP_ "failed to curl_easy_perform()",r);
279 pcrepp::Pcre bre("<body\\b",PCRE_CASELESS); 279 pcrepp::Pcre bre("<body\\b",PCRE_CASELESS);
280 // strip out everything past body 280 // strip out everything past body
281 if(bre.search(html)) 281 if(bre.search(html))
282 html.erase(bre.get_match_start()); 282 html.erase(bre.get_match_start());
283 pcrepp::Pcre hdre("<head[^>]*>",PCRE_CASELESS); 283 pcrepp::Pcre hdre("<head[^>]*>",PCRE_CASELESS);
284 if(!hdre.search(html)) 284 if(!hdre.search(html))
285 throw bad_input(OPKELE_CP_ "failed to find head"); 285 throw bad_input(OPKELE_CP_ "failed to find head");
286 html.erase(0,hdre.get_match_end()+1); 286 html.erase(0,hdre.get_match_end()+1);
287 pcrepp::Pcre lre("<link\\b([^>]+)>",PCRE_CASELESS), 287 pcrepp::Pcre lre("<link\\b([^>]+)>",PCRE_CASELESS),
288 rre("\\brel=['\"]([^'\"]+)['\"]",PCRE_CASELESS), 288 rre("\\brel=['\"]([^'\"]+)['\"]",PCRE_CASELESS),
289 hre("\\bhref=['\"]([^'\"]+)['\"]",PCRE_CASELESS); 289 hre("\\bhref=['\"]([^'\"]+)['\"]",PCRE_CASELESS);
290 while(lre.search(html)) { 290 while(lre.search(html)) {
291 string attrs = lre[0]; 291 string attrs = lre[0];
292 html.erase(0,lre.get_match_end()+1); 292 html.erase(0,lre.get_match_end()+1);
293 if(!(rre.search(attrs)&&hre.search(attrs))) 293 if(!(rre.search(attrs)&&hre.search(attrs)))
294 continue; 294 continue;
295 if(rre[0]=="openid.server") { 295 if(rre[0]=="openid.server") {
296 server = hre[0]; 296 server = hre[0];
297 if(!delegate.empty()) 297 if(!delegate.empty())
298 break; 298 break;
299 }else if(rre[0]=="openid.delegate") { 299 }else if(rre[0]=="openid.delegate") {
300 delegate = hre[0]; 300 delegate = hre[0];
301 if(!server.empty()) 301 if(!server.empty())
302 break; 302 break;
303 } 303 }
304 } 304 }
305 if(server.empty()) 305 if(server.empty())
306 throw failed_assertion(OPKELE_CP_ "The location has no openid.server declaration"); 306 throw failed_assertion(OPKELE_CP_ "The location has no openid.server declaration");
307 } 307 }
308 308
309 assoc_t consumer_t::find_assoc(const string& server) { 309 assoc_t consumer_t::find_assoc(const string& server) {
310 throw failed_lookup(OPKELE_CP_ "no find_assoc() provided"); 310 throw failed_lookup(OPKELE_CP_ "no find_assoc() provided");
311 } 311 }
312 312
313 string consumer_t::canonicalize(const string& url) { 313 string consumer_t::normalize(const string& url) {
314 string rv = url; 314 string rv = url;
315 // strip leading and trailing spaces 315 // strip leading and trailing spaces
316 string::size_type i = rv.find_first_not_of(" \t\r\n"); 316 string::size_type i = rv.find_first_not_of(" \t\r\n");
317 if(i==string::npos) 317 if(i==string::npos)
318 throw bad_input(OPKELE_CP_ "empty URL"); 318 throw bad_input(OPKELE_CP_ "empty URL");
319 if(i) 319 if(i)
320 rv.erase(0,i); 320 rv.erase(0,i);
321 i = rv.find_last_not_of(" \t\r\n"); 321 i = rv.find_last_not_of(" \t\r\n");
322 assert(i!=string::npos); 322 assert(i!=string::npos);
323 if(i<(rv.length()-1)) 323 if(i<(rv.length()-1))
324 rv.erase(i+1); 324 rv.erase(i+1);
325 // add missing http:// 325 // add missing http://
326 i = rv.find("://"); 326 i = rv.find("://");
327 if(i==string::npos) { // primitive. but do we need more? 327 if(i==string::npos) { // primitive. but do we need more?
328 rv.insert(0,"http://"); 328 rv.insert(0,"http://");
329 i = sizeof("http://")-1; 329 i = sizeof("http://")-1;
330 }else{ 330 }else{
331 i += sizeof("://")-1; 331 i += sizeof("://")-1;
332 } 332 }
333 string::size_type qm = rv.find('?',i); 333 string::size_type qm = rv.find('?',i);
334 string::size_type sl = rv.find('/',i); 334 string::size_type sl = rv.find('/',i);
335 if(qm!=string::npos) { 335 if(qm!=string::npos) {
336 if(sl==string::npos || sl>qm) 336 if(sl==string::npos || sl>qm)
337 rv.insert(qm,1,'/'); 337 rv.insert(qm,1,'/');
338 }else{ 338 }else{
339 if(sl==string::npos) 339 if(sl==string::npos)
340 rv += '/'; 340 rv += '/';
341 } 341 }
342 return rv; 342 return rv;
343 } 343 }
344 344
345 string consumer_t::canonicalize(const string& url) {
346 string rv = normalize(url);
347 curl_t curl = curl_easy_init();
348 if(!curl)
349 throw exception_curl(OPKELE_CP_ "failed to curl_easy_init()");
350 string html;
351 CURLcode r;
352 (r=curl_misc_sets(curl))
353 || (r=curl_easy_setopt(curl,CURLOPT_URL,rv.c_str()))
354 || (r=curl_easy_setopt(curl,CURLOPT_NOBODY,1))
355 ;
356 if(r)
357 throw exception_curl(OPKELE_CP_ "failed to curl_easy_setopt()",r);
358 r = curl_easy_perform(curl);
359 if(r)
360 throw exception_curl(OPKELE_CP_ "failed to curl_easy_perform()",r);
361 const char *eu = 0;
362 r = curl_easy_getinfo(curl,CURLINFO_EFFECTIVE_URL,&eu);
363 if(r)
364 throw exception_curl(OPKELE_CP_ "failed to curl_easy_getinfo(..CURLINFO_EFFECTIVE_URL..)",r);
365 rv = eu;
366 return normalize(rv);
367 }
368
345} 369}