author | Michael Krelin <hacker@klever.net> | 2008-03-02 23:52:14 (UTC) |
---|---|---|
committer | Michael Krelin <hacker@klever.net> | 2008-03-02 23:52:14 (UTC) |
commit | 5fd5ecad8c2bd1e8846c11fa9b281f0f4ab8a4a7 (patch) (side-by-side diff) | |
tree | 76ba24c589d0ce7de6cab113787cf7a70572f994 /lib | |
parent | f6482fb9003e4953f838ba4ef2c110190355622f (diff) | |
download | libopkele-5fd5ecad8c2bd1e8846c11fa9b281f0f4ab8a4a7.zip libopkele-5fd5ecad8c2bd1e8846c11fa9b281f0f4ab8a4a7.tar.gz libopkele-5fd5ecad8c2bd1e8846c11fa9b281f0f4ab8a4a7.tar.bz2 |
aded util::url_decode()
Signed-off-by: Michael Krelin <hacker@klever.net>
-rw-r--r-- | lib/util.cc | 28 |
1 files changed, 28 insertions, 0 deletions
diff --git a/lib/util.cc b/lib/util.cc index a6e08e2..3e7f3aa 100644 --- a/lib/util.cc +++ b/lib/util.cc @@ -131,128 +131,156 @@ namespace opkele { w.c_str(), "%04d-%02d-%02dT%02d:%02d:%02dZ", &tm_t.tm_year,&tm_t.tm_mon,&tm_t.tm_mday, &tm_t.tm_hour,&tm_t.tm_min,&tm_t.tm_sec ) != 6 ) && ( sscanf( w.c_str(), "%04d-%02d-%02dT%02d:%02d:%02d.%03dZ", &tm_t.tm_year,&tm_t.tm_mon,&tm_t.tm_mday, &tm_t.tm_hour,&tm_t.tm_min,&tm_t.tm_sec, &fraction ) != 7 ) ) throw failed_conversion(OPKELE_CP_ "failed to sscanf()"); tm_t.tm_mon--; tm_t.tm_year-=1900; time_t rv = mktime(&tm_t); if(rv==(time_t)-1) throw failed_conversion(OPKELE_CP_ "failed to mktime()"); return rv-timezone; } /* * */ static inline bool isrfc3986unreserved(int c) { if(c<'-') return false; if(c<='.') return true; if(c<'0') return false; if(c<='9') return true; if(c<'A') return false; if(c<='Z') return true; if(c<'_') return false; if(c=='_') return true; if(c<'a') return false; if(c<='z') return true; if(c=='~') return true; return false; } struct __url_encoder : public unary_function<char,void> { public: string& rv; __url_encoder(string& r) : rv(r) { } result_type operator()(argument_type c) { if(isrfc3986unreserved(c)) rv += c; else{ char tmp[4]; snprintf(tmp,sizeof(tmp),"%%%02X", (c&0xff)); rv += tmp; } } }; string url_encode(const string& str) { string rv; for_each(str.begin(),str.end(), __url_encoder(rv)); return rv; } + string url_decode(const string& str) { + string rv; + back_insert_iterator<string> ii(rv); + for(string::const_iterator i=str.begin(),ie=str.end(); + i!=ie;++i) { + switch(*i) { + case '+': + *(ii++) = ' '; break; + case '%': + ++i; + static char tmp[3] = {0,0,0}; + if(i==ie) + throw failed_conversion(OPKELE_CP_ "trailing percent in the url-encoded string"); + tmp[0] == *(i++); + if(i==ie) + throw failed_conversion(OPKELE_CP_ "not enough hexadecimals after the percent sign in url-encoded string"); + tmp[1] == *i; + if(!(isxdigit(tmp[0]) && isxdigit(tmp[1]))) + throw failed_conversion(OPKELE_CP_ "non-hex follows percent in url-encoded string"); + *(ii++) = strtol(tmp,0,16); + break; + default: + *(ii++) = *i; break; + } + } + return rv; + } + string attr_escape(const string& str) { static const char *unsafechars = "<>&\n\"'"; string rv; string::size_type p=0; while(true) { string::size_type us = str.find_first_of(unsafechars,p); if(us==string::npos) { if(p!=str.length()) rv.append(str,p,str.length()-p); return rv; } rv.append(str,p,us-p); rv += "&#"; rv += long_to_string((long)str[us]); rv += ';'; p = us+1; } } string long_to_string(long l) { char rv[32]; int r=snprintf(rv,sizeof(rv),"%ld",l); if(r<0 || r>=(int)sizeof(rv)) throw failed_conversion(OPKELE_CP_ "failed to snprintf()"); return rv; } long string_to_long(const string& s) { char *endptr = 0; long rv = strtol(s.c_str(),&endptr,10); if((!endptr) || endptr==s.c_str()) throw failed_conversion(OPKELE_CP_ "failed to strtol()"); return rv; } /* * Normalize URL according to the rules, described in rfc 3986, section 6 * * - uppercase hex triplets (e.g. %ab -> %AB) * - lowercase scheme and host * - decode %-encoded characters, specified as unreserved in rfc 3986, section 2.3, * that is - [:alpha:][:digit:]._~- * - remove dot segments * - remove empty and default ports * - if there's no path component, add '/' */ string rfc_3986_normalize_uri(const string& uri) { string rv; string::size_type ns = uri.find_first_not_of(data::_whitespace_chars); if(ns==string::npos) throw bad_input(OPKELE_CP_ "Can't normalize empty URI"); string::size_type colon = uri.find(':',ns); if(colon==string::npos) throw bad_input(OPKELE_CP_ "No scheme specified in URI"); transform( uri.begin()+ns, uri.begin()+colon+1, back_inserter(rv), ::tolower ); bool s; string::size_type ul = uri.find_last_not_of(data::_whitespace_chars)+1; if(ul <= (colon+3)) throw bad_input(OPKELE_CP_ "Unexpected end of URI being normalized encountered"); if(uri[colon+1]!='/' || uri[colon+2]!='/') throw bad_input(OPKELE_CP_ "Unexpected input in URI being normalized after scheme component"); if(rv=="http:") |