summaryrefslogtreecommitdiffabout
authorMichael Krelin <hacker@klever.net>2007-12-29 21:10:48 (UTC)
committer Michael Krelin <hacker@klever.net>2008-01-04 18:21:38 (UTC)
commitff04188567b117c28d54d6f81a9dca40ff0b1730 (patch) (unidiff)
tree80fe40af5c38a35f7cdfec710a6eb937fd14be88
parente429f672f681953cd3f9390482090e8f03cf2b45 (diff)
downloadlibopkele-ff04188567b117c28d54d6f81a9dca40ff0b1730.zip
libopkele-ff04188567b117c28d54d6f81a9dca40ff0b1730.tar.gz
libopkele-ff04188567b117c28d54d6f81a9dca40ff0b1730.tar.bz2
adjust w3c_to_time output by timezone, so that it returns local time
Signed-off-by: Michael Krelin <hacker@klever.net>
Diffstat (more/less context) (ignore whitespace changes)
-rw-r--r--lib/util.cc20
1 files changed, 15 insertions, 5 deletions
diff --git a/lib/util.cc b/lib/util.cc
index e5ca62d..ee75d29 100644
--- a/lib/util.cc
+++ b/lib/util.cc
@@ -1,305 +1,315 @@
1#include <errno.h> 1#include <errno.h>
2#include <cassert> 2#include <cassert>
3#include <cctype> 3#include <cctype>
4#include <cstring> 4#include <cstring>
5#include <vector> 5#include <vector>
6#include <string> 6#include <string>
7#include <stack> 7#include <stack>
8#include <openssl/bio.h> 8#include <openssl/bio.h>
9#include <openssl/evp.h> 9#include <openssl/evp.h>
10#include <curl/curl.h> 10#include <curl/curl.h>
11#include "opkele/util.h" 11#include "opkele/util.h"
12#include "opkele/exception.h" 12#include "opkele/exception.h"
13 13
14namespace opkele { 14namespace opkele {
15 using namespace std; 15 using namespace std;
16 16
17 namespace util { 17 namespace util {
18 18
19 /* 19 /*
20 * base64 20 * base64
21 */ 21 */
22 string encode_base64(const void *data,size_t length) { 22 string encode_base64(const void *data,size_t length) {
23 BIO *b64 = 0, *bmem = 0; 23 BIO *b64 = 0, *bmem = 0;
24 try { 24 try {
25 b64 = BIO_new(BIO_f_base64()); 25 b64 = BIO_new(BIO_f_base64());
26 if(!b64) 26 if(!b64)
27 throw exception_openssl(OPKELE_CP_ "failed to BIO_new() base64 encoder"); 27 throw exception_openssl(OPKELE_CP_ "failed to BIO_new() base64 encoder");
28 BIO_set_flags(b64,BIO_FLAGS_BASE64_NO_NL); 28 BIO_set_flags(b64,BIO_FLAGS_BASE64_NO_NL);
29 bmem = BIO_new(BIO_s_mem()); 29 bmem = BIO_new(BIO_s_mem());
30 BIO_set_flags(b64,BIO_CLOSE); 30 BIO_set_flags(b64,BIO_CLOSE);
31 if(!bmem) 31 if(!bmem)
32 throw exception_openssl(OPKELE_CP_ "failed to BIO_new() memory buffer"); 32 throw exception_openssl(OPKELE_CP_ "failed to BIO_new() memory buffer");
33 BIO_push(b64,bmem); 33 BIO_push(b64,bmem);
34 if(((size_t)BIO_write(b64,data,length))!=length) 34 if(((size_t)BIO_write(b64,data,length))!=length)
35 throw exception_openssl(OPKELE_CP_ "failed to BIO_write()"); 35 throw exception_openssl(OPKELE_CP_ "failed to BIO_write()");
36 if(BIO_flush(b64)!=1) 36 if(BIO_flush(b64)!=1)
37 throw exception_openssl(OPKELE_CP_ "failed to BIO_flush()"); 37 throw exception_openssl(OPKELE_CP_ "failed to BIO_flush()");
38 char *rvd; 38 char *rvd;
39 long rvl = BIO_get_mem_data(bmem,&rvd); 39 long rvl = BIO_get_mem_data(bmem,&rvd);
40 string rv(rvd,rvl); 40 string rv(rvd,rvl);
41 BIO_free_all(b64); 41 BIO_free_all(b64);
42 return rv; 42 return rv;
43 }catch(...) { 43 }catch(...) {
44 if(b64) BIO_free_all(b64); 44 if(b64) BIO_free_all(b64);
45 throw; 45 throw;
46 } 46 }
47 } 47 }
48 48
49 void decode_base64(const string& data,vector<unsigned char>& rv) { 49 void decode_base64(const string& data,vector<unsigned char>& rv) {
50 BIO *b64 = 0, *bmem = 0; 50 BIO *b64 = 0, *bmem = 0;
51 rv.clear(); 51 rv.clear();
52 try { 52 try {
53 bmem = BIO_new_mem_buf((void*)data.data(),data.size()); 53 bmem = BIO_new_mem_buf((void*)data.data(),data.size());
54 if(!bmem) 54 if(!bmem)
55 throw exception_openssl(OPKELE_CP_ "failed to BIO_new_mem_buf()"); 55 throw exception_openssl(OPKELE_CP_ "failed to BIO_new_mem_buf()");
56 b64 = BIO_new(BIO_f_base64()); 56 b64 = BIO_new(BIO_f_base64());
57 if(!b64) 57 if(!b64)
58 throw exception_openssl(OPKELE_CP_ "failed to BIO_new() base64 decoder"); 58 throw exception_openssl(OPKELE_CP_ "failed to BIO_new() base64 decoder");
59 BIO_set_flags(b64,BIO_FLAGS_BASE64_NO_NL); 59 BIO_set_flags(b64,BIO_FLAGS_BASE64_NO_NL);
60 BIO_push(b64,bmem); 60 BIO_push(b64,bmem);
61 unsigned char tmp[512]; 61 unsigned char tmp[512];
62 size_t rb = 0; 62 size_t rb = 0;
63 while((rb=BIO_read(b64,tmp,sizeof(tmp)))>0) 63 while((rb=BIO_read(b64,tmp,sizeof(tmp)))>0)
64 rv.insert(rv.end(),tmp,&tmp[rb]); 64 rv.insert(rv.end(),tmp,&tmp[rb]);
65 BIO_free_all(b64); 65 BIO_free_all(b64);
66 }catch(...) { 66 }catch(...) {
67 if(b64) BIO_free_all(b64); 67 if(b64) BIO_free_all(b64);
68 throw; 68 throw;
69 } 69 }
70 } 70 }
71 71
72 /* 72 /*
73 * big numerics 73 * big numerics
74 */ 74 */
75 75
76 BIGNUM *base64_to_bignum(const string& b64) { 76 BIGNUM *base64_to_bignum(const string& b64) {
77 vector<unsigned char> bin; 77 vector<unsigned char> bin;
78 decode_base64(b64,bin); 78 decode_base64(b64,bin);
79 BIGNUM *rv = BN_bin2bn(&(bin.front()),bin.size(),0); 79 BIGNUM *rv = BN_bin2bn(&(bin.front()),bin.size(),0);
80 if(!rv) 80 if(!rv)
81 throw failed_conversion(OPKELE_CP_ "failed to BN_bin2bn()"); 81 throw failed_conversion(OPKELE_CP_ "failed to BN_bin2bn()");
82 return rv; 82 return rv;
83 } 83 }
84 84
85 BIGNUM *dec_to_bignum(const string& dec) { 85 BIGNUM *dec_to_bignum(const string& dec) {
86 BIGNUM *rv = 0; 86 BIGNUM *rv = 0;
87 if(!BN_dec2bn(&rv,dec.c_str())) 87 if(!BN_dec2bn(&rv,dec.c_str()))
88 throw failed_conversion(OPKELE_CP_ "failed to BN_dec2bn()"); 88 throw failed_conversion(OPKELE_CP_ "failed to BN_dec2bn()");
89 return rv; 89 return rv;
90 } 90 }
91 91
92 string bignum_to_base64(const BIGNUM *bn) { 92 string bignum_to_base64(const BIGNUM *bn) {
93 vector<unsigned char> bin(BN_num_bytes(bn)+1); 93 vector<unsigned char> bin(BN_num_bytes(bn)+1);
94 unsigned char *binptr = &(bin.front())+1; 94 unsigned char *binptr = &(bin.front())+1;
95 int l = BN_bn2bin(bn,binptr); 95 int l = BN_bn2bin(bn,binptr);
96 if(l && (*binptr)&0x80){ 96 if(l && (*binptr)&0x80){
97 (*(--binptr)) = 0; ++l; 97 (*(--binptr)) = 0; ++l;
98 } 98 }
99 return encode_base64(binptr,l); 99 return encode_base64(binptr,l);
100 } 100 }
101 101
102 /* 102 /*
103 * w3c times 103 * w3c times
104 */ 104 */
105 105
106 string time_to_w3c(time_t t) { 106 string time_to_w3c(time_t t) {
107 struct tm tm_t; 107 struct tm tm_t;
108 if(!gmtime_r(&t,&tm_t)) 108 if(!gmtime_r(&t,&tm_t))
109 throw failed_conversion(OPKELE_CP_ "failed to BN_dec2bn()"); 109 throw failed_conversion(OPKELE_CP_ "failed to BN_dec2bn()");
110 char rv[25]; 110 char rv[25];
111 if(!strftime(rv,sizeof(rv)-1,"%Y-%m-%dT%H:%M:%SZ",&tm_t)) 111 if(!strftime(rv,sizeof(rv)-1,"%Y-%m-%dT%H:%M:%SZ",&tm_t))
112 throw failed_conversion(OPKELE_CP_ "failed to strftime()"); 112 throw failed_conversion(OPKELE_CP_ "failed to strftime()");
113 return rv; 113 return rv;
114 } 114 }
115 115
116 time_t w3c_to_time(const string& w) { 116 time_t w3c_to_time(const string& w) {
117 int fraction;
117 struct tm tm_t; 118 struct tm tm_t;
118 memset(&tm_t,0,sizeof(tm_t)); 119 memset(&tm_t,0,sizeof(tm_t));
119 if( 120 if( (
121 sscanf(
122 w.c_str(),
123 "%04d-%02d-%02dT%02d:%02d:%02dZ",
124 &tm_t.tm_year,&tm_t.tm_mon,&tm_t.tm_mday,
125 &tm_t.tm_hour,&tm_t.tm_min,&tm_t.tm_sec
126 ) != 6
127 ) && (
120 sscanf( 128 sscanf(
121 w.c_str(), 129 w.c_str(),
122 "%04d-%02d-%02dT%02d:%02d:%02dZ", 130 "%04d-%02d-%02dT%02d:%02d:%02d.%03dZ",
123 &tm_t.tm_year,&tm_t.tm_mon,&tm_t.tm_mday, 131 &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 132 &tm_t.tm_hour,&tm_t.tm_min,&tm_t.tm_sec,
125 ) != 6 ) 133 &fraction
134 ) != 7
135 ) )
126 throw failed_conversion(OPKELE_CP_ "failed to sscanf()"); 136 throw failed_conversion(OPKELE_CP_ "failed to sscanf()");
127 tm_t.tm_mon--; 137 tm_t.tm_mon--;
128 tm_t.tm_year-=1900; 138 tm_t.tm_year-=1900;
129 time_t rv = mktime(&tm_t); 139 time_t rv = mktime(&tm_t);
130 if(rv==(time_t)-1) 140 if(rv==(time_t)-1)
131 throw failed_conversion(OPKELE_CP_ "failed to mktime()"); 141 throw failed_conversion(OPKELE_CP_ "failed to mktime()");
132 return rv; 142 return rv-timezone;
133 } 143 }
134 144
135 /* 145 /*
136 * 146 *
137 */ 147 */
138 148
139 string url_encode(const string& str) { 149 string url_encode(const string& str) {
140 char * t = curl_escape(str.c_str(),str.length()); 150 char * t = curl_escape(str.c_str(),str.length());
141 if(!t) 151 if(!t)
142 throw failed_conversion(OPKELE_CP_ "failed to curl_escape()"); 152 throw failed_conversion(OPKELE_CP_ "failed to curl_escape()");
143 string rv(t); 153 string rv(t);
144 curl_free(t); 154 curl_free(t);
145 return rv; 155 return rv;
146 } 156 }
147 157
148 string long_to_string(long l) { 158 string long_to_string(long l) {
149 char rv[32]; 159 char rv[32];
150 int r=snprintf(rv,sizeof(rv),"%ld",l); 160 int r=snprintf(rv,sizeof(rv),"%ld",l);
151 if(r<0 || r>=(int)sizeof(rv)) 161 if(r<0 || r>=(int)sizeof(rv))
152 throw failed_conversion(OPKELE_CP_ "failed to snprintf()"); 162 throw failed_conversion(OPKELE_CP_ "failed to snprintf()");
153 return rv; 163 return rv;
154 } 164 }
155 165
156 long string_to_long(const string& s) { 166 long string_to_long(const string& s) {
157 char *endptr = 0; 167 char *endptr = 0;
158 long rv = strtol(s.c_str(),&endptr,10); 168 long rv = strtol(s.c_str(),&endptr,10);
159 if((!endptr) || endptr==s.c_str()) 169 if((!endptr) || endptr==s.c_str())
160 throw failed_conversion(OPKELE_CP_ "failed to strtol()"); 170 throw failed_conversion(OPKELE_CP_ "failed to strtol()");
161 return rv; 171 return rv;
162 } 172 }
163 173
164 /* 174 /*
165 * Normalize URL according to the rules, described in rfc 3986, section 6 175 * Normalize URL according to the rules, described in rfc 3986, section 6
166 * 176 *
167 * - uppercase hex triplets (e.g. %ab -> %AB) 177 * - uppercase hex triplets (e.g. %ab -> %AB)
168 * - lowercase scheme and host 178 * - lowercase scheme and host
169 * - decode %-encoded characters, specified as unreserved in rfc 3986, section 2.3, 179 * - decode %-encoded characters, specified as unreserved in rfc 3986, section 2.3,
170 * that is - [:alpha:][:digit:]._~- 180 * that is - [:alpha:][:digit:]._~-
171 * - remove dot segments 181 * - remove dot segments
172 * - remove empty and default ports 182 * - remove empty and default ports
173 * - if there's no path component, add '/' 183 * - if there's no path component, add '/'
174 */ 184 */
175 string rfc_3986_normalize_uri(const string& uri) { 185 string rfc_3986_normalize_uri(const string& uri) {
176 static const char *whitespace = " \t\r\n"; 186 static const char *whitespace = " \t\r\n";
177 string rv; 187 string rv;
178 string::size_type ns = uri.find_first_not_of(whitespace); 188 string::size_type ns = uri.find_first_not_of(whitespace);
179 if(ns==string::npos) 189 if(ns==string::npos)
180 throw bad_input(OPKELE_CP_ "Can't normalize empty URI"); 190 throw bad_input(OPKELE_CP_ "Can't normalize empty URI");
181 string::size_type colon = uri.find(':',ns); 191 string::size_type colon = uri.find(':',ns);
182 if(colon==string::npos) 192 if(colon==string::npos)
183 throw bad_input(OPKELE_CP_ "No scheme specified in URI"); 193 throw bad_input(OPKELE_CP_ "No scheme specified in URI");
184 transform( 194 transform(
185 uri.begin()+ns, uri.begin()+colon+1, 195 uri.begin()+ns, uri.begin()+colon+1,
186 back_inserter(rv), ::tolower ); 196 back_inserter(rv), ::tolower );
187 bool s; 197 bool s;
188 string::size_type ul = uri.find_last_not_of(whitespace)+1; 198 string::size_type ul = uri.find_last_not_of(whitespace)+1;
189 if(ul <= (colon+3)) 199 if(ul <= (colon+3))
190 throw bad_input(OPKELE_CP_ "Unexpected end of URI being normalized encountered"); 200 throw bad_input(OPKELE_CP_ "Unexpected end of URI being normalized encountered");
191 if(uri[colon+1]!='/' || uri[colon+2]!='/') 201 if(uri[colon+1]!='/' || uri[colon+2]!='/')
192 throw bad_input(OPKELE_CP_ "Unexpected input in URI being normalized after scheme component"); 202 throw bad_input(OPKELE_CP_ "Unexpected input in URI being normalized after scheme component");
193 if(rv=="http:") 203 if(rv=="http:")
194 s = false; 204 s = false;
195 else if(rv=="https:") 205 else if(rv=="https:")
196 s = true; 206 s = true;
197 else{ 207 else{
198 /* TODO: support more schemes. 208 /* TODO: support more schemes.
199 * e.g. xri. How do we normalize 209 * e.g. xri. How do we normalize
200 * xri? 210 * xri?
201 */ 211 */
202 rv.append(uri,colon+1,ul-colon-1); 212 rv.append(uri,colon+1,ul-colon-1);
203 return rv; 213 return rv;
204 } 214 }
205 rv += "//"; 215 rv += "//";
206 string::size_type interesting = uri.find_first_of(":/#?",colon+3); 216 string::size_type interesting = uri.find_first_of(":/#?",colon+3);
207 if(interesting==string::npos) { 217 if(interesting==string::npos) {
208 transform( 218 transform(
209 uri.begin()+colon+3,uri.begin()+ul, 219 uri.begin()+colon+3,uri.begin()+ul,
210 back_inserter(rv), ::tolower ); 220 back_inserter(rv), ::tolower );
211 rv += '/'; return rv; 221 rv += '/'; return rv;
212 } 222 }
213 transform( 223 transform(
214 uri.begin()+colon+3,uri.begin()+interesting, 224 uri.begin()+colon+3,uri.begin()+interesting,
215 back_inserter(rv), ::tolower ); 225 back_inserter(rv), ::tolower );
216 bool qf = false; 226 bool qf = false;
217 char ic = uri[interesting]; 227 char ic = uri[interesting];
218 if(ic==':') { 228 if(ic==':') {
219 string::size_type ni = uri.find_first_of("/#?%",interesting+1); 229 string::size_type ni = uri.find_first_of("/#?%",interesting+1);
220 const char *nptr = uri.data()+interesting+1; 230 const char *nptr = uri.data()+interesting+1;
221 char *eptr = 0; 231 char *eptr = 0;
222 long port = strtol(nptr,&eptr,10); 232 long port = strtol(nptr,&eptr,10);
223 if( (port>0) && (port<65535) && port!=(s?443:80) ) { 233 if( (port>0) && (port<65535) && port!=(s?443:80) ) {
224 char tmp[8]; 234 char tmp[8];
225 snprintf(tmp,sizeof(tmp),":%ld",port); 235 snprintf(tmp,sizeof(tmp),":%ld",port);
226 rv += tmp; 236 rv += tmp;
227 } 237 }
228 if(ni==string::npos) { 238 if(ni==string::npos) {
229 rv += '/'; return rv; 239 rv += '/'; return rv;
230 } 240 }
231 interesting = ni; 241 interesting = ni;
232 }else if(ic!='/') { 242 }else if(ic!='/') {
233 rv += '/'; rv += ic; 243 rv += '/'; rv += ic;
234 qf = true; 244 qf = true;
235 ++interesting; 245 ++interesting;
236 } 246 }
237 string::size_type n = interesting; 247 string::size_type n = interesting;
238 char tmp[3] = { 0,0,0 }; 248 char tmp[3] = { 0,0,0 };
239 stack<string::size_type> psegs; psegs.push(rv.length()); 249 stack<string::size_type> psegs; psegs.push(rv.length());
240 string pseg; 250 string pseg;
241 for(;n<ul;) { 251 for(;n<ul;) {
242 string::size_type unsafe = uri.find_first_of(qf?"%":"%/?#",n); 252 string::size_type unsafe = uri.find_first_of(qf?"%":"%/?#",n);
243 if(unsafe==string::npos) { 253 if(unsafe==string::npos) {
244 pseg.append(uri,n,ul-n-1); n = ul-1; 254 pseg.append(uri,n,ul-n-1); n = ul-1;
245 }else{ 255 }else{
246 pseg.append(uri,n,unsafe-n); 256 pseg.append(uri,n,unsafe-n);
247 n = unsafe; 257 n = unsafe;
248 } 258 }
249 char c = uri[n++]; 259 char c = uri[n++];
250 if(c=='%') { 260 if(c=='%') {
251 if((n+1)>=ul) 261 if((n+1)>=ul)
252 throw bad_input(OPKELE_CP_ "Unexpected end of URI encountered while parsing percent-encoded character"); 262 throw bad_input(OPKELE_CP_ "Unexpected end of URI encountered while parsing percent-encoded character");
253 tmp[0] = uri[n++]; 263 tmp[0] = uri[n++];
254 tmp[1] = uri[n++]; 264 tmp[1] = uri[n++];
255 if(!( isxdigit(tmp[0]) && isxdigit(tmp[1]) )) 265 if(!( isxdigit(tmp[0]) && isxdigit(tmp[1]) ))
256 throw bad_input(OPKELE_CP_ "Invalid percent-encoded character in URI being normalized"); 266 throw bad_input(OPKELE_CP_ "Invalid percent-encoded character in URI being normalized");
257 int cc = strtol(tmp,0,16); 267 int cc = strtol(tmp,0,16);
258 if( isalpha(cc) || isdigit(cc) || strchr("._~-",cc) ) 268 if( isalpha(cc) || isdigit(cc) || strchr("._~-",cc) )
259 pseg += cc; 269 pseg += cc;
260 else{ 270 else{
261 pseg += '%'; 271 pseg += '%';
262 pseg += toupper(tmp[0]); pseg += toupper(tmp[1]); 272 pseg += toupper(tmp[0]); pseg += toupper(tmp[1]);
263 } 273 }
264 }else if(qf) { 274 }else if(qf) {
265 rv += pseg; rv += c; 275 rv += pseg; rv += c;
266 pseg.clear(); 276 pseg.clear();
267 }else if(n>=ul || strchr("?/#",c)) { 277 }else if(n>=ul || strchr("?/#",c)) {
268 if(pseg.empty() || pseg==".") { 278 if(pseg.empty() || pseg==".") {
269 }else if(pseg=="..") { 279 }else if(pseg=="..") {
270 if(psegs.size()>1) { 280 if(psegs.size()>1) {
271 rv.resize(psegs.top()); psegs.pop(); 281 rv.resize(psegs.top()); psegs.pop();
272 } 282 }
273 }else{ 283 }else{
274 psegs.push(rv.length()); 284 psegs.push(rv.length());
275 if(c!='/') { 285 if(c!='/') {
276 pseg += c; 286 pseg += c;
277 qf = true; 287 qf = true;
278 } 288 }
279 rv += '/'; rv += pseg; 289 rv += '/'; rv += pseg;
280 } 290 }
281 if(c=='/' && (n>=ul || strchr("?#",uri[n])) ) { 291 if(c=='/' && (n>=ul || strchr("?#",uri[n])) ) {
282 rv += '/'; 292 rv += '/';
283 if(n<ul) 293 if(n<ul)
284 qf = true; 294 qf = true;
285 }else if(strchr("?#",c)) { 295 }else if(strchr("?#",c)) {
286 if(psegs.size()==1 && psegs.top()==rv.length()) 296 if(psegs.size()==1 && psegs.top()==rv.length())
287 rv += '/'; 297 rv += '/';
288 if(pseg.empty()) 298 if(pseg.empty())
289 rv += c; 299 rv += c;
290 qf = true; 300 qf = true;
291 } 301 }
292 pseg.clear(); 302 pseg.clear();
293 }else{ 303 }else{
294 pseg += c; 304 pseg += c;
295 } 305 }
296 } 306 }
297 if(!pseg.empty()) { 307 if(!pseg.empty()) {
298 rv += '/'; rv += pseg; 308 rv += '/'; rv += pseg;
299 } 309 }
300 return rv; 310 return rv;
301 } 311 }
302 312
303 } 313 }
304 314
305} 315}