|
|
| version 1.129, 2004/02/11 15:33:16 | version 1.138, 2007/08/27 19:04:03 |
|---|---|
| Line 1 | Line 1 |
| /** @file | /** @file |
| Parser: String class part: untaint mechanizm. | Parser: String class part: untaint mechanizm. |
| Copyright(c) 2001-2004 ArtLebedev Group (http://www.artlebedev.com) | Copyright(c) 2001-2005 ArtLebedev Group (http://www.artlebedev.com) |
| Author: Alexandr Petrosian <paf@design.ru> (http://paf.design.ru) | Author: Alexandr Petrosian <paf@design.ru> (http://paf.design.ru) |
| */ | */ |
| Line 34 extern "C" { // author forgot to do that | Line 34 extern "C" { // author forgot to do that |
| { \ | { \ |
| bool skip=false; \ | bool skip=false; \ |
| if(optimize) switch(c) { \ | if(optimize) switch(c) { \ |
| case ' ': case '\r': case '\n': case '\t': \ | case ' ': case '\n': case '\t': \ |
| if(whitespace) \ | if(whitespace) \ |
| skip=true; /*skipping subsequent*/ \ | skip=true; /*skipping subsequent*/ \ |
| else \ | else \ |
| Line 58 extern "C" { // author forgot to do that | Line 58 extern "C" { // author forgot to do that |
| char c=CORD_pos_fetch(info->pos); \ | char c=CORD_pos_fetch(info->pos); \ |
| action \ | action \ |
| } | } |
| #define _default default: CORD_ec_append(info->result, c); break | #define _default CORD_ec_append(info->result, c) |
| #define encode(need_encode_func, prefix, otherwise) \ | #define encode(need_encode_func, prefix, otherwise) \ |
| if(need_encode_func(c)) { \ | if(need_encode_func(c)) { \ |
| static const char* hex="0123456789ABCDEF"; \ | static const char* hex="0123456789ABCDEF"; \ |
| Line 92 inline bool need_http_header_encode(unsi | Line 92 inline bool need_http_header_encode(unsi |
| return need_uri_encode(c); | return need_uri_encode(c); |
| } | } |
| inline bool need_regex_escape(unsigned char c){ | |
| return strchr("\\^$.[]|()?*+{}", c)!=0; | |
| } | |
| // String | // String |
| Line 292 struct Cstr_to_string_body_block_info { | Line 295 struct Cstr_to_string_body_block_info { |
| CORD_pos pos; | CORD_pos pos; |
| size_t fragment_begin; | size_t fragment_begin; |
| bool whitespace; | bool whitespace; |
| const char* exception; | |
| }; | }; |
| #endif | #endif |
| int cstr_to_string_body_block(char alang, size_t fragment_length, Cstr_to_string_body_block_info* info) { | int cstr_to_string_body_block(char alang, size_t fragment_length, Cstr_to_string_body_block_info* info) { |
| const String::Language fragment_lang=(String::Language)(unsigned char)alang; | const String::Language fragment_lang=(String::Language)(unsigned char)alang; |
| bool& whitespace=info->whitespace; | bool& whitespace=info->whitespace; |
| size_t fragment_end=info->fragment_begin+fragment_length; | size_t fragment_end=info->fragment_begin+fragment_length; |
| //fprintf(stderr, "%d, %d\n", fragment.lang, fragment.length); | //fprintf(stderr, "%d, %d =%s=\n", fragment_lang, fragment_length, info->body->cstr()); |
| String::Language to_lang=info->lang==String::L_UNSPECIFIED?fragment_lang:info->lang; | String::Language to_lang=info->lang==String::L_UNSPECIFIED?fragment_lang:info->lang; |
| Line 371 int cstr_to_string_body_block(char alang | Line 375 int cstr_to_string_body_block(char alang |
| bool email=false; | bool email=false; |
| uchar c; | uchar c; |
| for(const char* src=mail_ptr; (c=(uchar)*src++); ) { | for(const char* src=mail_ptr; (c=(uchar)*src++); ) { |
| //RFC + An 'encoded-word' MUST NOT appear in any portion of an 'addr-spec'. | if(c=='\r' || c=='\n') |
| if(to_quoted_printable && (c==',' || addr_spec_soon(src) || c == '"')) { | c=' '; |
| if(to_quoted_printable && (c==',' || c == '"' || addr_spec_soon(src-1/*position to 'c'*/))) { | |
| email=c=='<'; | email=c=='<'; |
| to_string("?="); | to_string("?="); |
| to_quoted_printable=false; | to_quoted_printable=false; |
| } | } |
| //RFC + An 'encoded-word' MUST NOT appear in any portion of an 'addr-spec'. | |
| if(!email && ( | if(!email && ( |
| !to_quoted_printable && (c & 0x80) // starting quote-printable-encoding on first 8bit char | !to_quoted_printable && (c & 0x80) // starting quote-printable-encoding on first 8bit char |
| || to_quoted_printable && !mail_header_char_valid_within_Qencoded(c) | || to_quoted_printable && !mail_header_char_valid_within_Qencoded(c) |
| Line 399 int cstr_to_string_body_block(char alang | Line 405 int cstr_to_string_body_block(char alang |
| } else | } else |
| ec_append(info->result, optimize, whitespace, info->pos, fragment_length); | ec_append(info->result, optimize, whitespace, info->pos, fragment_length); |
| break; | break; |
| case String::L_TABLE: | |
| // tainted, untaint language: table | |
| escape(switch(c) { | |
| case '\t': to_char(' '); break; | |
| case '\n': to_char(' '); break; | |
| _default; | |
| }); | |
| break; | |
| case String::L_SQL: | case String::L_SQL: |
| // tainted, untaint language: sql | // tainted, untaint language: sql |
| if(info->connection) { | if(info->connection) { |
| Line 415 int cstr_to_string_body_block(char alang | Line 413 int cstr_to_string_body_block(char alang |
| pa_CORD_pos_advance(info->pos, fragment_length); | pa_CORD_pos_advance(info->pos, fragment_length); |
| to_string(info->connection->quote(fragment_str, fragment_length)); | to_string(info->connection->quote(fragment_str, fragment_length)); |
| } else | } else { |
| throw Exception(0, | info->exception="untaint in SQL language failed - no connection specified"; |
| 0, | info->fragment_begin=fragment_end; |
| "untaint in SQL language failed - no connection specified"); | return 1; // stop processing. can't throw exception here |
| } | |
| break; | break; |
| case String::L_JS: | case String::L_JS: |
| escape(switch(c) { | escape(switch(c) { |
| Line 427 int cstr_to_string_body_block(char alang | Line 426 int cstr_to_string_body_block(char alang |
| case '\n': to_string("\\n"); break; | case '\n': to_string("\\n"); break; |
| case '\\': to_string("\\\\"); break; | case '\\': to_string("\\\\"); break; |
| case '\xFF': to_string("\\\xFF"); break; | case '\xFF': to_string("\\\xFF"); break; |
| _default; | default: _default; break; |
| }); | }); |
| break; | break; |
| case String::L_XML: | case String::L_XML: |
| // [2] Char ::= #x9 | #xA | #xD | [#x20-#xD7FF] | [#xE000-#xFFFD] | [#x10000-#x10FFFF] | |
| escape(switch(c) { | escape(switch(c) { |
| case '\x9': | |
| case '\xA': | |
| case '\xD': // this is usually removed on input | |
| _default; | |
| break; | |
| case '&': to_string("&"); break; | case '&': to_string("&"); break; |
| case '>': to_string(">"); break; | case '>': to_string(">"); break; |
| case '<': to_string("<"); break; | case '<': to_string("<"); break; |
| case '"': to_string("""); break; | case '"': to_string("""); break; |
| case '\'': to_string("'"); break; | case '\'': to_string("'"); break; |
| _default; | default: |
| if(((unsigned char)c)<0x20) { | |
| // fixing it, so that libxml would not result | |
| // in fatal error parsing text | |
| // though it really violates standard. | |
| // to indicate there were an error | |
| // replace bad char not to it's code, | |
| // which we can do, | |
| // but rather to '!' to show that input were actually | |
| // invalid. | |
| // life: shows that MSIE can somehow garble form values | |
| // so that they contain these chars. | |
| to_char('!'); | |
| } else { | |
| _default; | |
| } | |
| break; | |
| }); | }); |
| break; | break; |
| case String::L_HTML: | case String::L_HTML: |
| Line 446 int cstr_to_string_body_block(char alang | Line 467 int cstr_to_string_body_block(char alang |
| case '>': to_string(">"); break; | case '>': to_string(">"); break; |
| case '<': to_string("<"); break; | case '<': to_string("<"); break; |
| case '"': to_string("""); break; | case '"': to_string("""); break; |
| _default; | default: _default; break; |
| }); | }); |
| break; | break; |
| case String::L_REGEX: | |
| // tainted, untaint language: regex | |
| escape( | |
| if(need_regex_escape(c)) | |
| to_char('\\') | |
| _default; | |
| ); | |
| break; | |
| default: | default: |
| SAPI::abort("unknown untaint language #%d", | SAPI::abort("unknown untaint language #%d", |
| static_cast<int>(to_lang)); // should never | static_cast<int>(to_lang)); // should never |
| Line 464 int cstr_to_string_body_block(char alang | Line 493 int cstr_to_string_body_block(char alang |
| String::Body String::cstr_to_string_body(Language lang, | String::Body String::cstr_to_string_body(Language lang, |
| SQL_Connection* connection, | SQL_Connection* connection, |
| const Request_charsets *charsets) const { | const Request_charsets *charsets) const { |
| if(is_empty()) | |
| return String::Body(); | |
| Cstr_to_string_body_block_info info; | Cstr_to_string_body_block_info info; |
| // input | // input |
| Line 476 String::Body String::cstr_to_string_body | Line 507 String::Body String::cstr_to_string_body |
| // private | // private |
| body.set_pos(info.pos, 0); | body.set_pos(info.pos, 0); |
| info.fragment_begin=0; | info.fragment_begin=0; |
| info.exception=0; | |
| info.whitespace=true; | info.whitespace=true; |
| if(!is_empty()) | langs.for_each(body, cstr_to_string_body_block, &info); |
| langs.for_each(body, cstr_to_string_body_block, &info); | if(info.exception){ |
| throw Exception(0, | |
| 0, | |
| info.exception); | |
| } | |
| return String::Body(CORD_ec_to_cord(info.result)); | return String::Body(CORD_ec_to_cord(info.result)); |
| } | } |