|
|
| version 1.122.2.2, 2004/01/22 08:28:20 | version 1.133, 2004/07/01 07:20:10 |
|---|---|
| Line 1 | Line 1 |
| /** @file | /** @file |
| Parser: String class part: untaint mechanizm. | Parser: String class part: untaint mechanizm. |
| Copyright(c) 2001-2003 ArtLebedev Group (http://www.artlebedev.com) | Copyright(c) 2001-2004 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) |
| */ | */ |
| static const char* IDENT_UNTAINT_C="$Date$"; | static const char * const IDENT_UNTAINT_C="$Date$"; |
| #include "pa_string.h" | #include "pa_string.h" |
| Line 23 extern "C" { // author forgot to do that | Line 23 extern "C" { // author forgot to do that |
| #include "ec.h" | #include "ec.h" |
| } | } |
| #define PA_SQL | |
| #ifdef PA_SQL | |
| #include "pa_sql_connection.h" | #include "pa_sql_connection.h" |
| #endif | |
| // defines | // defines |
| #undef CORD_ec_append | #undef CORD_ec_append |
| // redefining to intercept flushes and implement whitespace optimization | // redefining to intercept flushes and implement whitespace optimization |
| // of all consequent white space chars leaving only first one | // of all consequent white space chars leaving only first one |
| Line 39 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 63 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 98 inline bool need_http_header_encode(unsi | Line 93 inline bool need_http_header_encode(unsi |
| return need_uri_encode(c); | return need_uri_encode(c); |
| } | } |
| // | |
| static const char* String_Untaint_lang_name[]={ | |
| "U", ///< zero value handy for hash lookup @see untaint_lang_name2enum | |
| "C", ///< clean | |
| "T", ///< tainted, untaint language as assigned later | |
| // untaint languages. assigned by ^untaint[lang]{...} | |
| "P", | |
| /**< | |
| leave language built into string being appended. | |
| just a flag, that value not stored | |
| */ | |
| "A", ///< leave all characters intact | |
| "F", ///< file specification | |
| "H", ///< ext in HTTP response header | |
| "M", ///< text in mail header | |
| "URI", ///< text in uri | |
| "T", ///< ^table:set body | |
| "SQL", ///< ^table:sql body | |
| "JS", ///< JavaScript code | |
| "XML", ///< ^dom:set xml | |
| "HTML" ///< HTML code (for editing) | |
| }; | |
| // String | // String |
| /* | /* |
| HTTP-header = field-name ":" [ field-value ] CRLF | HTTP-header = field-name ":" [ field-value ] CRLF |
| field-name = token | field-name = token |
| Line 268 inline bool mail_header_char_valid_withi | Line 237 inline bool mail_header_char_valid_withi |
| || strchr("!*+-/", c); | || strchr("!*+-/", c); |
| } | } |
| inline bool addr_spec_soon(const char *src) { | inline bool addr_spec_soon(const char *src) { |
| for(char c; c=*src; src++) | for(char c; (c=*src); src++) |
| if(c=='<') | if(c=='<') |
| return true; | return true; |
| else if(!(c==' ' || c=='\t')) | else if(!(c==' ' || c=='\t')) |
| Line 284 inline bool addr_spec_soon(const char *s | Line 253 inline bool addr_spec_soon(const char *s |
| inline bool mail_header_nonspace_char(char c) { | inline bool mail_header_nonspace_char(char c) { |
| return c != 0x20; | return c != 0x20; |
| } | } |
| inline void ec_append(CORD_ec& result, bool& optimize, bool& whitespace, CORD_pos pos, size_t size) { | inline void ec_append(CORD_ec& result, bool& optimize, bool& whitespace, CORD_pos pos, size_t size) { |
| while(size--) { | while(size--) { |
| CORD_ec_append(result, CORD_pos_fetch(pos)); | CORD_ec_append(result, CORD_pos_fetch(pos)); |
| Line 292 inline void ec_append(CORD_ec& result, b | Line 262 inline void ec_append(CORD_ec& result, b |
| } | } |
| inline void pa_CORD_pos_advance(CORD_pos pos, size_t n) { | inline void pa_CORD_pos_advance(CORD_pos pos, size_t n) { |
| while(true) { | while(true) { |
| size_t avail=CORD_pos_chars_left(pos); | long avail=CORD_pos_chars_left(pos); |
| if(avail==0) { | if(avail<=0) { |
| CORD_next(pos); | CORD_next(pos); |
| if(!--n) | if(!--n) |
| break; | break; |
| } else if(avail<n) { | } else if((size_t)avail<n) { |
| CORD_pos_advance(pos, avail); | CORD_pos_advance(pos, avail); |
| n-=avail; | n-=avail; |
| } else { // avail>=n | } else { // avail>=n |
| Line 373 int cstr_to_string_body_block(char alang | Line 343 int cstr_to_string_body_block(char alang |
| info->charsets->client()); | info->charsets->client()); |
| char c; | char c; |
| for(const char* src=output.str; c=*src++; ) | for(const char* src=output.str; (c=*src++); ) |
| encode(need_uri_encode, '%', c); | encode(need_uri_encode, '%', c); |
| } | } |
| break; | break; |
| Line 400 int cstr_to_string_body_block(char alang | Line 370 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(to_quoted_printable && (c==',' || c == '"' || addr_spec_soon(src-1/*position to 'c'*/))) { |
| if(to_quoted_printable && (c==',' || addr_spec_soon(src) || 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 429 int cstr_to_string_body_block(char alang | Line 399 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; | |
| #ifdef PA_SQL | |
| case String::L_SQL: | case String::L_SQL: |
| // tainted, untaint language: sql | // tainted, untaint language: sql |
| if(info->connection) { | if(info->connection) { |
| Line 451 int cstr_to_string_body_block(char alang | Line 412 int cstr_to_string_body_block(char alang |
| 0, | 0, |
| "untaint in SQL language failed - no connection specified"); | "untaint in SQL language failed - no connection specified"); |
| break; | break; |
| #endif | |
| case String::L_JS: | case String::L_JS: |
| escape(switch(c) { | escape(switch(c) { |
| case '"': to_string("\\\""); break; | case '"': to_string("\\\""); break; |
| Line 459 int cstr_to_string_body_block(char alang | Line 419 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 478 int cstr_to_string_body_block(char alang | Line 460 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; |
| default: | default: |
| assert(!"should never"); | SAPI::abort("unknown untaint language #%d", |
| SAPI::die("unknown untaint language #%d", | |
| static_cast<int>(to_lang)); // should never | static_cast<int>(to_lang)); // should never |
| break; // never | break; // never |
| } | } |