-rw-r--r-- | lib/util.cc | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/lib/util.cc b/lib/util.cc index eacf6d7..ac70938 100644 --- a/lib/util.cc +++ b/lib/util.cc | |||
@@ -119,177 +119,177 @@ namespace opkele { | |||
119 | if( | 119 | if( |
120 | sscanf( | 120 | sscanf( |
121 | w.c_str(), | 121 | w.c_str(), |
122 | "%04d-%02d-%02dT%02d:%02d:%02dZ", | 122 | "%04d-%02d-%02dT%02d:%02d:%02dZ", |
123 | &tm_t.tm_year,&tm_t.tm_mon,&tm_t.tm_mday, | 123 | &tm_t.tm_year,&tm_t.tm_mon,&tm_t.tm_mday, |
124 | &tm_t.tm_hour,&tm_t.tm_min,&tm_t.tm_sec | 124 | &tm_t.tm_hour,&tm_t.tm_min,&tm_t.tm_sec |
125 | ) != 6 ) | 125 | ) != 6 ) |
126 | throw failed_conversion(OPKELE_CP_ "failed to sscanf()"); | 126 | throw failed_conversion(OPKELE_CP_ "failed to sscanf()"); |
127 | tm_t.tm_mon--; | 127 | tm_t.tm_mon--; |
128 | tm_t.tm_year-=1900; | 128 | tm_t.tm_year-=1900; |
129 | time_t rv = mktime(&tm_t); | 129 | time_t rv = mktime(&tm_t); |
130 | if(rv==(time_t)-1) | 130 | if(rv==(time_t)-1) |
131 | throw failed_conversion(OPKELE_CP_ "failed to mktime()"); | 131 | throw failed_conversion(OPKELE_CP_ "failed to mktime()"); |
132 | return rv; | 132 | return rv; |
133 | } | 133 | } |
134 | 134 | ||
135 | /* | 135 | /* |
136 | * | 136 | * |
137 | */ | 137 | */ |
138 | 138 | ||
139 | string url_encode(const string& str) { | 139 | string url_encode(const string& str) { |
140 | char * t = curl_escape(str.c_str(),str.length()); | 140 | char * t = curl_escape(str.c_str(),str.length()); |
141 | if(!t) | 141 | if(!t) |
142 | throw failed_conversion(OPKELE_CP_ "failed to curl_escape()"); | 142 | throw failed_conversion(OPKELE_CP_ "failed to curl_escape()"); |
143 | string rv(t); | 143 | string rv(t); |
144 | curl_free(t); | 144 | curl_free(t); |
145 | return rv; | 145 | return rv; |
146 | } | 146 | } |
147 | 147 | ||
148 | string long_to_string(long l) { | 148 | string long_to_string(long l) { |
149 | char rv[32]; | 149 | char rv[32]; |
150 | int r=snprintf(rv,sizeof(rv),"%ld",l); | 150 | int r=snprintf(rv,sizeof(rv),"%ld",l); |
151 | if(r<0 || r>=(int)sizeof(rv)) | 151 | if(r<0 || r>=(int)sizeof(rv)) |
152 | throw failed_conversion(OPKELE_CP_ "failed to snprintf()"); | 152 | throw failed_conversion(OPKELE_CP_ "failed to snprintf()"); |
153 | return rv; | 153 | return rv; |
154 | } | 154 | } |
155 | 155 | ||
156 | long string_to_long(const string& s) { | 156 | long string_to_long(const string& s) { |
157 | char *endptr = 0; | 157 | char *endptr = 0; |
158 | long rv = strtol(s.c_str(),&endptr,10); | 158 | long rv = strtol(s.c_str(),&endptr,10); |
159 | if((!endptr) || endptr==s.c_str()) | 159 | if((!endptr) || endptr==s.c_str()) |
160 | throw failed_conversion(OPKELE_CP_ "failed to strtol()"); | 160 | throw failed_conversion(OPKELE_CP_ "failed to strtol()"); |
161 | return rv; | 161 | return rv; |
162 | } | 162 | } |
163 | 163 | ||
164 | /* | 164 | /* |
165 | * Normalize URL according to the rules, described in rfc 3986, section 6 | 165 | * Normalize URL according to the rules, described in rfc 3986, section 6 |
166 | * | 166 | * |
167 | * - uppercase hext triplets (e.g. %ab -> %AB) | 167 | * - uppercase hext triplets (e.g. %ab -> %AB) |
168 | * - lowercase scheme and host | 168 | * - lowercase scheme and host |
169 | * - decode %-encoded characters, specified as unreserved in rfc 3986, section 2.3, | 169 | * - decode %-encoded characters, specified as unreserved in rfc 3986, section 2.3, |
170 | * that is - [:alpha:][:digit:]._~- | 170 | * that is - [:alpha:][:digit:]._~- |
171 | * - remove dot segments | 171 | * - remove dot segments |
172 | * - remove empty and default ports | 172 | * - remove empty and default ports |
173 | * - if there's no path component, add '/' | 173 | * - if there's no path component, add '/' |
174 | */ | 174 | */ |
175 | string rfc_3986_normalize_uri(const string& uri) { | 175 | string rfc_3986_normalize_uri(const string& uri) { |
176 | string rv; | 176 | string rv; |
177 | string::size_type colon = uri.find(':'); | 177 | string::size_type colon = uri.find(':'); |
178 | if(colon==string::npos) | 178 | if(colon==string::npos) |
179 | throw bad_input(OPKELE_CP_ "No scheme specified in URI"); | 179 | throw bad_input(OPKELE_CP_ "No scheme specified in URI"); |
180 | transform( | 180 | transform( |
181 | uri.begin(), uri.begin()+colon+1, | 181 | uri.begin(), uri.begin()+colon+1, |
182 | back_inserter(rv), ::tolower ); | 182 | back_inserter(rv), ::tolower ); |
183 | bool s; | 183 | bool s; |
184 | if(rv=="http:") | 184 | if(rv=="http:") |
185 | s = false; | 185 | s = false; |
186 | else if(rv=="https:") | 186 | else if(rv=="https:") |
187 | s = true; | 187 | s = true; |
188 | else | 188 | else |
189 | throw not_implemented(OPKELE_CP_ "Only http(s) URIs can be normalized here"); | 189 | throw not_implemented(OPKELE_CP_ "Only http(s) URIs can be normalized here"); |
190 | string::size_type ul = uri.length(); | 190 | string::size_type ul = uri.length(); |
191 | if(ul <= (colon+3)) | 191 | if(ul <= (colon+3)) |
192 | throw bad_input(OPKELE_CP_ "Unexpected end of URI being normalized encountered"); | 192 | throw bad_input(OPKELE_CP_ "Unexpected end of URI being normalized encountered"); |
193 | if(uri[colon+1]!='/' || uri[colon+2]!='/') | 193 | if(uri[colon+1]!='/' || uri[colon+2]!='/') |
194 | throw bad_input(OPKELE_CP_ "Unexpected input in URI being normalized after scheme component"); | 194 | throw bad_input(OPKELE_CP_ "Unexpected input in URI being normalized after scheme component"); |
195 | rv += "//"; | 195 | rv += "//"; |
196 | string::size_type interesting = uri.find_first_of(":/#?",colon+3); | 196 | string::size_type interesting = uri.find_first_of(":/#?",colon+3); |
197 | if(interesting==string::npos) { | 197 | if(interesting==string::npos) { |
198 | transform( | 198 | transform( |
199 | uri.begin()+colon+3,uri.end(), | 199 | uri.begin()+colon+3,uri.end(), |
200 | back_inserter(rv), ::tolower ); | 200 | back_inserter(rv), ::tolower ); |
201 | rv += '/'; return rv; | 201 | rv += '/'; return rv; |
202 | } | 202 | } |
203 | transform( | 203 | transform( |
204 | uri.begin()+colon+3,uri.begin()+interesting, | 204 | uri.begin()+colon+3,uri.begin()+interesting, |
205 | back_inserter(rv), ::tolower ); | 205 | back_inserter(rv), ::tolower ); |
206 | bool qf = false; | 206 | bool qf = false; |
207 | char ic = uri[interesting]; | 207 | char ic = uri[interesting]; |
208 | if(ic==':') { | 208 | if(ic==':') { |
209 | string::size_type ni = uri.find_first_of("/#?%",interesting+1); | 209 | string::size_type ni = uri.find_first_of("/#?%",interesting+1); |
210 | const char *nptr = uri.data()+interesting+1; | 210 | const char *nptr = uri.data()+interesting+1; |
211 | char *eptr = 0; | 211 | char *eptr = 0; |
212 | long port = strtol(nptr,&eptr,10); | 212 | long port = strtol(nptr,&eptr,10); |
213 | if( (port>0) && (port<65535) && port!=(s?443:80) ) { | 213 | if( (port>0) && (port<65535) && port!=(s?443:80) ) { |
214 | char tmp[6]; | 214 | char tmp[6]; |
215 | snprintf(tmp,sizeof(tmp),"%d",port); | 215 | snprintf(tmp,sizeof(tmp),"%ld",port); |
216 | rv += ':'; rv += tmp; | 216 | rv += ':'; rv += tmp; |
217 | } | 217 | } |
218 | if(ni==string::npos) { | 218 | if(ni==string::npos) { |
219 | rv += '/'; return rv; | 219 | rv += '/'; return rv; |
220 | } | 220 | } |
221 | interesting = ni; | 221 | interesting = ni; |
222 | }else if(ic!='/') { | 222 | }else if(ic!='/') { |
223 | rv += '/'; rv += ic; | 223 | rv += '/'; rv += ic; |
224 | qf = true; | 224 | qf = true; |
225 | ++interesting; | 225 | ++interesting; |
226 | } | 226 | } |
227 | string::size_type n = interesting; | 227 | string::size_type n = interesting; |
228 | char tmp[3] = { 0,0,0 }; | 228 | char tmp[3] = { 0,0,0 }; |
229 | stack<string::size_type> psegs; psegs.push(rv.length()); | 229 | stack<string::size_type> psegs; psegs.push(rv.length()); |
230 | string pseg; | 230 | string pseg; |
231 | for(;n<ul;) { | 231 | for(;n<ul;) { |
232 | string::size_type unsafe = uri.find_first_of(qf?"%":"%/?#",n); | 232 | string::size_type unsafe = uri.find_first_of(qf?"%":"%/?#",n); |
233 | if(unsafe==string::npos) { | 233 | if(unsafe==string::npos) { |
234 | pseg.append(uri,n,ul-n-1); n = ul-1; | 234 | pseg.append(uri,n,ul-n-1); n = ul-1; |
235 | }else{ | 235 | }else{ |
236 | pseg.append(uri,n,unsafe-n); | 236 | pseg.append(uri,n,unsafe-n); |
237 | n = unsafe; | 237 | n = unsafe; |
238 | } | 238 | } |
239 | char c = uri[n++]; | 239 | char c = uri[n++]; |
240 | if(c=='%') { | 240 | if(c=='%') { |
241 | if((n+1)>=ul) | 241 | if((n+1)>=ul) |
242 | throw bad_input(OPKELE_CP_ "Unexpected end of URI encountered while parsing percent-encoded character"); | 242 | throw bad_input(OPKELE_CP_ "Unexpected end of URI encountered while parsing percent-encoded character"); |
243 | tmp[0] = uri[n++]; | 243 | tmp[0] = uri[n++]; |
244 | tmp[1] = uri[n++]; | 244 | tmp[1] = uri[n++]; |
245 | if(!( isxdigit(tmp[0]) && isxdigit(tmp[1]) )) | 245 | if(!( isxdigit(tmp[0]) && isxdigit(tmp[1]) )) |
246 | throw bad_input(OPKELE_CP_ "Invalid percent-encoded character in URI being normalized"); | 246 | throw bad_input(OPKELE_CP_ "Invalid percent-encoded character in URI being normalized"); |
247 | int cc = strtol(tmp,0,16); | 247 | int cc = strtol(tmp,0,16); |
248 | if( isalpha(cc) || isdigit(cc) || strchr("._~-",cc) ) | 248 | if( isalpha(cc) || isdigit(cc) || strchr("._~-",cc) ) |
249 | pseg += cc; | 249 | pseg += cc; |
250 | else{ | 250 | else{ |
251 | pseg += '%'; | 251 | pseg += '%'; |
252 | pseg += toupper(tmp[0]); pseg += toupper(tmp[1]); | 252 | pseg += toupper(tmp[0]); pseg += toupper(tmp[1]); |
253 | } | 253 | } |
254 | }else if(qf) { | 254 | }else if(qf) { |
255 | rv += pseg; rv += c; | 255 | rv += pseg; rv += c; |
256 | pseg.clear(); | 256 | pseg.clear(); |
257 | }else if(n>=ul || strchr("?/#",c)) { | 257 | }else if(n>=ul || strchr("?/#",c)) { |
258 | if(pseg.empty() || pseg==".") { | 258 | if(pseg.empty() || pseg==".") { |
259 | }else if(pseg=="..") { | 259 | }else if(pseg=="..") { |
260 | if(psegs.size()>1) { | 260 | if(psegs.size()>1) { |
261 | rv.resize(psegs.top()); psegs.pop(); | 261 | rv.resize(psegs.top()); psegs.pop(); |
262 | } | 262 | } |
263 | }else{ | 263 | }else{ |
264 | psegs.push(rv.length()); | 264 | psegs.push(rv.length()); |
265 | if(c!='/') { | 265 | if(c!='/') { |
266 | pseg += c; | 266 | pseg += c; |
267 | qf = true; | 267 | qf = true; |
268 | } | 268 | } |
269 | rv += '/'; rv += pseg; | 269 | rv += '/'; rv += pseg; |
270 | } | 270 | } |
271 | if(c=='/' && (n>=ul || strchr("?#",uri[n])) ) { | 271 | if(c=='/' && (n>=ul || strchr("?#",uri[n])) ) { |
272 | rv += '/'; | 272 | rv += '/'; |
273 | if(n<ul) | 273 | if(n<ul) |
274 | qf = true; | 274 | qf = true; |
275 | }else if(strchr("?#",c)) { | 275 | }else if(strchr("?#",c)) { |
276 | if(psegs.size()==1 && psegs.top()==rv.length()) | 276 | if(psegs.size()==1 && psegs.top()==rv.length()) |
277 | rv += '/'; | 277 | rv += '/'; |
278 | if(pseg.empty()) | 278 | if(pseg.empty()) |
279 | rv += c; | 279 | rv += c; |
280 | qf = true; | 280 | qf = true; |
281 | } | 281 | } |
282 | pseg.clear(); | 282 | pseg.clear(); |
283 | }else{ | 283 | }else{ |
284 | pseg += c; | 284 | pseg += c; |
285 | } | 285 | } |
286 | } | 286 | } |
287 | if(!pseg.empty()) { | 287 | if(!pseg.empty()) { |
288 | rv += '/'; rv += pseg; | 288 | rv += '/'; rv += pseg; |
289 | } | 289 | } |
290 | return rv; | 290 | return rv; |
291 | } | 291 | } |
292 | 292 | ||
293 | } | 293 | } |
294 | 294 | ||
295 | } | 295 | } |