|
|
| version 1.60, 2001/08/28 10:30:28 | version 1.116, 2003/04/18 10:12:05 |
|---|---|
| Line 1 | Line 1 |
| /** @file | /** @file |
| Parser: String class part: untaint mechanizm. | Parser: String class part: untaint mechanizm. |
| Copyright(c) 2001 ArtLebedev Group(http://www.artlebedev.com) | Copyright(c) 2001, 2003 ArtLebedev Group (http://www.artlebedev.com) |
| Author: Alexandr Petrosian <paf@design.ru> (http://paf.design.ru) | |
| Author: Alexander Petrosyan <paf@design.ru>(http://design.ru/paf) | |
| */ | */ |
| static const char *RCSId="$Id$"; | |
| static const char* IDENT_UNTAINT_C="$Date$"; | |
| #include "pa_pool.h" | #include "pa_pool.h" |
| #include "pa_string.h" | #include "pa_string.h" |
| Line 15 static const char *RCSId="$Id$"; | Line 15 static const char *RCSId="$Id$"; |
| #include "pa_globals.h" | #include "pa_globals.h" |
| #include "pa_sql_connection.h" | #include "pa_sql_connection.h" |
| #include "pa_dictionary.h" | #include "pa_dictionary.h" |
| #include "pa_common.h" | |
| #include "pa_charset.h" | |
| //#define DEBUG_STRING_APPENDS_VS_EXPANDS | |
| #ifdef DEBUG_STRING_APPENDS_VS_EXPANDS | |
| ulong string_string_shortcut_economy=0; | |
| #endif | |
| #define escape(action) \ | #define escape(action) \ |
| { \ | { \ |
| Line 24 static const char *RCSId="$Id$"; | Line 32 static const char *RCSId="$Id$"; |
| } | } |
| #define _default default: *dest++=*src; break | #define _default default: *dest++=*src; break |
| #define encode(need_encode_func, prefix) \ | #define encode(need_encode_func, prefix) \ |
| default: \ | |
| if(need_encode_func(*src)) { \ | if(need_encode_func(*src)) { \ |
| static const char *hex="0123456789ABCDEF"; \ | static const char *hex="0123456789ABCDEF"; \ |
| char chunk[3]={prefix}; \ | char chunk[3]={prefix}; \ |
| Line 40 static const char *RCSId="$Id$"; | Line 47 static const char *RCSId="$Id$"; |
| dest+=bsize; \ | dest+=bsize; \ |
| inline bool need_file_encode(unsigned char c){ | inline bool need_file_encode(unsigned char c){ |
| if((c>='0') &&(c<='9') ||(c>='A') &&(c<='Z') ||(c>='a') &&(c<='z')) | // russian letters and space ENABLED |
| return false; | // encoding only these... |
| return strchr( | |
| return !strchr( | "*?'\"<>|" |
| #ifdef WIN32 | #ifndef WIN32 |
| ":\\~" | ":\\" |
| #endif | #endif |
| "./()_-", c); | , c)!=0; |
| } | } |
| inline bool need_uri_encode(unsigned char c){ | inline bool need_uri_encode(unsigned char c){ |
| if((c>='0') &&(c<='9') ||(c>='A') &&(c<='Z') ||(c>='a') &&(c<='z')) | if((c>='0') &&(c<='9') ||(c>='A') &&(c<='Z') ||(c>='a') &&(c<='z')) |
| Line 75 static const char * String_Untaint_lang_ | Line 82 static const char * String_Untaint_lang_ |
| just a flag, that value not stored | just a flag, that value not stored |
| */ | */ |
| "A", ///< leave all characters intact | "A", ///< leave all characters intact |
| "F", ///< filename | "F", ///< file specification |
| "H", ///< text in HTTP response header | "H", ///< ext in HTTP response header |
| "M", ///< text in mail header | "M", ///< text in mail header |
| "URI", ///< text in uri | "URI", ///< text in uri |
| "T", ///< ^table:set body | "T", ///< ^table:set body |
| "SQL", ///< ^table:sql body | "SQL", ///< ^table:sql body |
| "JS", ///< JavaScript code | "JS", ///< JavaScript code |
| "HTML", ///< HTML code (for editing) | "XML", ///< ^dom:set xml |
| "UHTML", ///< HTML code with USER chars | "HTML" ///< HTML code (for editing) |
| }; | }; |
| // String | // String |
| static bool typo_present(Array::Item *value, const void *info) { | |
| Array *row=static_cast<Array *>(value); | |
| const char *src=static_cast<const char *>(info); | |
| int partial; | |
| row->get_string(0)->cmp(partial, src); | |
| return | |
| partial==0 || // full match | |
| partial==1; // typo left column starts 'src' | |
| } | |
| /* | /* |
| HTTP-header = field-name ":" [ field-value ] CRLF | HTTP-header = field-name ":" [ field-value ] CRLF |
| Line 142 inline bool need_quote_http_header(const | Line 138 inline bool need_quote_http_header(const |
| return false; | return false; |
| } | } |
| /** @todo maybe additional check "are all pieces are clean?" would be profitable? | //#include "pa_sapi.h" |
| @todo fix potential forigins_mode buf overrun | /** |
| appends other String, | |
| marking all tainted pieces of it with @a lang. | |
| or marking ALL pieces of it with a @a lang when @a forced to, | |
| and propagating OPTIMIZE language bit. | |
| using architecture advantage: after string-to-string-append string never modified. | |
| algorithm: | |
| if no language-change specified and src not yet appended to some other string[last_chunk!=0] | |
| shrinking dest last_chunk[preparing it for linking], | |
| ///shrinking src last_chunk[preparing it to be linked, consequent dest.appends would go there], | |
| linking[dest.last_chunk = src.head.chunk] | |
| if some language-change specified or src already appended to some other string[last_chunk==0] | |
| cloning pieces. | |
| */ | */ |
| size_t String::cstr_bufsize(Untaint_lang lang) const { | String& String::append(const String& src, uchar lang, bool forced) { |
| return (lang==UL_AS_IS?size():size()*UNTAINT_TIMES_BIGGER*(forigins_mode?10:1)) +1; | if(!last_chunk) // growth stopped [we're appended as string to somebody] |
| throw Exception(0, | |
| this, | |
| "string growth stopped (append string)"); | |
| if(src.is_empty()) | |
| return *this; | |
| // without language-chage, not-appended-before, big[not fitting our tail] string? | |
| if(lang==UL_PASS_APPENDED | |
| && src.last_chunk | |
| && (uint(&last_chunk->rows[last_chunk->count]-append_here) < src.used_rows())) { | |
| #ifdef DEBUG_STRING_APPENDS_VS_EXPANDS | |
| string_string_shortcut_economy+=src.used_rows()*sizeof(String::Chunk::Row); | |
| #endif | |
| // using fact: | |
| // src.head.chunk.count initally equeals this.head.chunk.count and shrinks-only, | |
| // so can't be more than this.head.chunk.count, | |
| // which means that we know that | |
| // src.head.chunk would fit into this.head.chunk | |
| if(is_empty()) { // our head.chunk is empty | |
| // they have more than head.chunk? we need all head.chunk : we need only filled-part of head.chunk | |
| Chunk *src_head_link=src.head.chunk.rows[src.head.chunk.count].link; | |
| size_t head_count=src_head_link?src.head.chunk.count:(src.append_here-src.head.chunk.rows); | |
| // "your head.chunk is my head.chunk" | |
| memcpy(head.chunk.rows, src.head.chunk.rows, sizeof(Chunk::Row)*(head_count)); | |
| if(src_head_link) { | |
| // "your body is my body" | |
| head.chunk.rows[head.chunk.count=head_count].link=src_head_link; | |
| // "your last_chunk is mine now" | |
| last_chunk=src.last_chunk; | |
| // "your append_here is mine now" | |
| append_here=src.append_here; | |
| } else { | |
| // "your last_chunk is mine now" | |
| last_chunk=&head.chunk; | |
| // "your append_here is recalc-mine now" | |
| append_here=head.chunk.rows+head_count; | |
| } | |
| } else { // our head.chunk contains something | |
| // "chopping off my tail-reserve" | |
| last_chunk->count=append_here-last_chunk->rows; | |
| // "you is my tail" | |
| append_here->link=&src.head.chunk; | |
| // "your last_chunk is mine now" | |
| last_chunk=src.last_chunk; | |
| // "your append_here is mine now" | |
| append_here=src.append_here; | |
| } | |
| // stop-growing mark | |
| src.last_chunk=0; | |
| return *this; | |
| } | |
| // manually unrolled code to avoid do{if(const)} constructs | |
| if(forced) | |
| STRING_SRC_FOREACH_ROW( | |
| APPEND(row->item.ptr, row->item.size, | |
| lang, //forcing passed lang | |
| row->item.origin.file, row->item.origin.line); | |
| ) | |
| else if(lang==UL_PASS_APPENDED) | |
| STRING_SRC_FOREACH_ROW( | |
| APPEND(row->item.ptr, row->item.size, | |
| row->item.lang, // passing item's lang | |
| row->item.origin.file, row->item.origin.line); | |
| ) | |
| else if(lang&UL_OPTIMIZE_BIT) // main idea here | |
| // tainted piece would get OPTIMIZED bit from 'lang' | |
| // clean piece would be marked OPTIMIZED manually | |
| // pieces with determined languages [not tainted|clean] would retain theirs langs | |
| STRING_SRC_FOREACH_ROW( | |
| APPEND(row->item.ptr, row->item.size, | |
| row->item.lang==UL_TAINTED?lang:( | |
| row->item.lang==UL_CLEAN?UL_CLEAN|UL_OPTIMIZE_BIT: // ORing with OPTIMIZED flag | |
| row->item.lang | |
| ), | |
| row->item.origin.file, row->item.origin.line); | |
| ) | |
| else | |
| STRING_SRC_FOREACH_ROW( | |
| APPEND(row->item.ptr, row->item.size, | |
| row->item.lang==UL_TAINTED?lang:row->item.lang, | |
| row->item.origin.file, row->item.origin.line); | |
| ); | |
| /* | |
| for(Chunk::Row *row=last_chunk->rows; row<append_here; row++) | |
| if(row->link==(void*)0xcdcdcdcd) | |
| _asm int 3;*/ | |
| return *this; | |
| } | } |
| /** @todo fix theoretical \n mem overrun in TYPO replacements | size_t String::cstr_bufsize(Untaint_lang lang, |
| SQL_Connection *connection, | |
| Charset *buf_charset) const { | |
| size_t dest=1; // for terminating 0 | |
| STRING_FOREACH_ROW( | |
| uchar to_lang=lang==UL_UNSPECIFIED?row->item.lang:lang; | |
| switch(to_lang & ~UL_OPTIMIZE_BIT) { | |
| case UL_CLEAN: | |
| case UL_TAINTED: | |
| case UL_AS_IS: | |
| // clean piece | |
| // tainted piece, but undefined untaint language | |
| // for VString.as_double of tainted values | |
| // for ^process{body} evaluation | |
| // tainted, untaint language: as-is | |
| dest+=row->item.size; | |
| break; | |
| case UL_FILE_SPEC: | |
| // tainted, untaint language: file [name] | |
| dest+=row->item.size*3/* worst: Z->%XX */; | |
| break; | |
| case UL_URI: | |
| // tainted, untaint language: uri | |
| dest+=row->item.size*6*3/* worst utf8 x worst Z->%XX */; | |
| break; | |
| case UL_HTTP_HEADER: | |
| // tainted, untaint language: http-field-content-text | |
| dest+=row->item.size*3/* worst: Z->%XX */; | |
| break; | |
| case UL_MAIL_HEADER: | |
| // tainted, untaint language: mail-header | |
| if(buf_charset) { | |
| int parts_count=1; | |
| const char *prev=row->item.ptr; | |
| size_t size=row->item.size; | |
| while(const char *comma_at=(const char*)memchr(prev, ',', size)) { | |
| parts_count++; | |
| size-=comma_at-prev; | |
| prev=comma_at+1; | |
| } | |
| // Subject: Re: parser3: =?koi8-r?Q?=D3=C5=CD=C9=CE=C1=D2?= | |
| dest+= | |
| row->item.size*3+ | |
| parts_count*(buf_charset->name().size()+MAX_STRING/* worst: =?charset?Q?=%XX?= */); | |
| } else | |
| dest+=row->item.size; | |
| break; | |
| case UL_TABLE: | |
| // tainted, untaint language: table | |
| dest+=row->item.size; | |
| break; | |
| case UL_SQL: | |
| // tainted, untaint language: sql | |
| if(connection) | |
| dest+=connection->quote(0, row->item.ptr, row->item.size); | |
| break; | |
| case UL_JS: | |
| escape(switch(*src) { | |
| case '"': case '\'': case '\n': case '\\': case '\xFF': | |
| dest+=2; break; | |
| default: | |
| dest++; break; | |
| }); | |
| break; | |
| case UL_XML: | |
| escape(switch(*src) { | |
| case '&': case '>': case '<': case '"': case '\'': | |
| dest+= 6; break; | |
| default: | |
| dest++; break; | |
| }); | |
| break; | |
| case UL_HTML: | |
| escape(switch(*src) { | |
| case '&': | |
| case '>': | |
| case '<': | |
| case '"': | |
| dest+=6; break; | |
| default: | |
| dest++; break; | |
| }); | |
| break; | |
| } | |
| ); | |
| return dest; | |
| } | |
| /** http://www.ietf.org/rfc/rfc2047.txt | |
| RFC | |
| (3) As a replacement for a 'word' entity within a 'phrase', for example, | |
| one that precedes an address in a From, To, or Cc header. The ABNF | |
| definition for 'phrase' from RFC 822 thus becomes: | |
| phrase = 1*( encoded-word / word ) | |
| In this case the set of characters that may be used in a "Q"-encoded | |
| 'encoded-word' is restricted to: <upper and lower case ASCII | |
| letters, decimal digits, "!", "*", "+", "-", "/", "=", and "_" | |
| (underscore, ASCII 95.)>. An 'encoded-word' that appears within a | |
| 'phrase' MUST be separated from any adjacent 'word', 'text' or | |
| 'special' by 'linear-white-space'. | |
| ... | |
| (2) The 8-bit hexadecimal value 20 (e.g., ISO-8859-1 SPACE) may be | |
| represented as "_" (underscore, ASCII 95.). (This character may | |
| not pass through some internetwork mail gateways, but its use | |
| will greatly enhance readability of "Q" encoded data with mail | |
| readers that do not support this encoding.) Note that the "_" | |
| always represents hexadecimal 20, even if the SPACE character | |
| occupies a different code position in the character set in use. | |
| paf: obviously, | |
| without "=", or one could not differ "=E0" and "russian letter a" | |
| and without "_", or in would mean 0x20 | |
| */ | */ |
| static bool mail_header_char_valid_within_Qencoded(char c) { | |
| return c>='A' && c<='Z' | |
| || c>='a' && c<='Z' | |
| || c>='0' && c<='9' | |
| || strchr("!*+-/", c); | |
| } | |
| char *String::store_to(char *dest, Untaint_lang lang, | char *String::store_to(char *dest, Untaint_lang lang, |
| SQL_Connection *connection, | SQL_Connection *connection, |
| const char *charset) const { | Charset *store_to_charset, |
| // $MAIN:html-typo table | const char *store_to_charset_name) const { |
| Dictionary *user_typo_dict=static_cast<Dictionary *>(pool().tag()); | // WARNING: |
| Dictionary *typo_dict=user_typo_dict?user_typo_dict:default_typo_dict; | // before any changes check cstr_bufsize first!!! |
| bool whitespace=true; | bool whitespace=true; |
| const Chunk *chunk=&head; | //debug |
| do { | const Chunk *chunk=&(*this).head.chunk; \ |
| const Chunk::Row *row=chunk->rows; | const Chunk::Row *row=chunk->rows; \ |
| for(size_t i=0; i<chunk->count; i++, row++) { | uint countdown=chunk->count; \ |
| if(row==append_here) | while(row!=(*this).append_here) { \ |
| goto break2; | if(countdown==0) { \ |
| chunk=row->link; \ | |
| Untaint_lang to_lang=lang==UL_UNSPECIFIED?row->item.lang:lang; | row=chunk->rows; \ |
| countdown=chunk->count; \ | |
| char *dest_before_origins=dest; | }; |
| //*/ | |
| if(forigins_mode) { | // STRING_FOREACH_ROW( |
| #ifndef NO_STRING_ORIGIN | |
| if(row->item.origin.file) | uchar to_lang=lang==UL_UNSPECIFIED?row->item.lang:lang; |
| dest+=sprintf(dest, "%s(%d)", | |
| row->item.origin.file, | char *start=dest; |
| 1+row->item.origin.line); | |
| else | switch(to_lang & ~UL_OPTIMIZE_BIT) { |
| dest+=sprintf(dest, "unknown"); | case UL_CLEAN: |
| #endif | case UL_TAINTED: |
| dest+=sprintf(dest, "#%s: ", | case UL_AS_IS: |
| String_Untaint_lang_name[to_lang]); | // clean piece |
| } | |
| char *dest_after_origins=dest; | // tainted piece, but undefined untaint language |
| // for VString.as_double of tainted values | |
| // WARNING: | // for ^process{body} evaluation |
| // string can grow only UNTAINT_TIMES_BIGGER | |
| switch(to_lang) { | // tainted, untaint language: as-is |
| case UL_CLEAN: | memcpy(dest, row->item.ptr, row->item.size); |
| // clean piece | dest+=row->item.size; |
| { // optimizing whitespace | break; |
| const char *src=row->item.ptr; | case UL_FILE_SPEC: |
| for(int size=row->item.size; size--; src++) | // tainted, untaint language: file [name] |
| switch(*src) { | escape( |
| case ' ': case '\n': case '\r': case '\t': | // Macintosh has problems with small Russian letter 'r' |
| if(!whitespace) { | if( *src=='\xF0' && pool().get_source_charset().name()=="windows-1251" ) { |
| *dest++=*src; | // fixing that letter for most common charset |
| whitespace=true; | to_char('p'); |
| } | } else // fallback to default |
| break; | encode(need_file_encode, '_'); |
| default: | ); |
| whitespace=false; | break; |
| *dest++=*src; | case UL_URI: |
| break; | // tainted, untaint language: uri |
| } | const void *client_ptr; |
| } | size_t client_size; |
| break; | Charset::transcode(pool(), |
| case UL_TAINTED: | pool().get_source_charset(), row->item.ptr, row->item.size, |
| // tainted piece, but undefined untaint language | pool().get_client_charset(), client_ptr, client_size); |
| // for VString.as_double of tainted values | { |
| // for ^process{body} evaluation | const char *src=(const char *)client_ptr; |
| case UL_AS_IS: | for(int size=client_size; size--; src++) |
| // tainted, untaint language: as-is | |
| memcpy(dest, row->item.ptr, row->item.size); | |
| dest+=row->item.size; | |
| break; | |
| case UL_FILE_NAME: | |
| // tainted, untaint language: file [name] | |
| escape(switch(*src) { | |
| case ' ': to_char('_'); break; | |
| encode(need_file_encode, '+'); | |
| }); | |
| break; | |
| case UL_URI: | |
| // tainted, untaint language: uri | |
| escape(switch(*src) { | |
| case ' ': to_char('+'); break; | |
| encode(need_uri_encode, '%'); | encode(need_uri_encode, '%'); |
| }); | } |
| break; | break; |
| case UL_HTTP_HEADER: | case UL_HTTP_HEADER: |
| // tainted, untaint language: http-header | // tainted, untaint language: http-field-content-text |
| if(need_quote_http_header(row->item.ptr, row->item.size)) { | escape( |
| *dest++='\"'; | encode(need_uri_encode, '%'); |
| escape(switch(*src) { | ); |
| case '\"': to_string("\\\"", 2); break; | break; |
| _default; | case UL_MAIL_HEADER: |
| }); | // tainted, untaint language: mail-header |
| *dest++='\"'; | // http://www.ietf.org/rfc/rfc2047.txt |
| } else { | if(store_to_charset && store_to_charset_name) { |
| memcpy(dest, row->item.ptr, row->item.size); | const void *mail_ptr; |
| dest+=row->item.size; | size_t mail_size; |
| } | Charset::transcode(pool(), |
| break; | pool().get_source_charset(), row->item.ptr, row->item.size, |
| case UL_MAIL_HEADER: | *store_to_charset, mail_ptr, mail_size); |
| // tainted, untaint language: mail-header | |
| if(charset) { | // Subject: Re: parser3: =?koi8-r?Q?=D3=C5=CD=C9=CE=C1=D2?= |
| // Subject: Re: parser3: =?koi8-r?Q?=D3=C5=CD=C9=CE=C1=D2?= | const char *src=(const char *)mail_ptr; |
| const char *src=row->item.ptr; | bool to_quoted_printable=false; |
| bool to_quoted_printable=false; | |
| for(int size=row->item.size; size--; src++) { | bool email=false; |
| if(*src & 0x80) { | for(const char *end=src+mail_size; src<end; src++) { |
| if(!to_quoted_printable) { | //RFC + An 'encoded-word' MUST NOT appear in any portion of an 'addr-spec'. |
| dest+=sprintf(dest, "=?%.15s?Q?", charset); | if(to_quoted_printable && (*src==',' || *src=='<')) { |
| to_quoted_printable=true; | email=*src=='<'; |
| } | |
| dest+=sprintf(dest, "=%02X", *src & 0xFF); | |
| } else { | |
| *dest++=*src; | |
| } | |
| } | |
| if(to_quoted_printable) // close | |
| dest+=sprintf(dest, "?="); | dest+=sprintf(dest, "?="); |
| } else { | to_quoted_printable=false; |
| memcpy(dest, row->item.ptr, row->item.size); | } |
| dest+=row->item.size; | if(!email && ( |
| } | !to_quoted_printable && (*src & 0x80) // starting quote-printable-encoding on first 8bit char |
| break; | || to_quoted_printable && !mail_header_char_valid_within_Qencoded(*src) |
| case UL_TABLE: | )) { |
| // tainted, untaint language: table | if(!to_quoted_printable) { |
| escape(switch(*src) { | dest+=sprintf(dest, "=?%s?Q?", store_to_charset_name); |
| case '\t': to_char(' '); break; | to_quoted_printable=true; |
| case '\n': to_char(' '); break; | |
| _default; | |
| }); | |
| break; | |
| case UL_SQL: | |
| // tainted, untaint language: sql | |
| if(connection) | |
| dest+=connection->quote(dest, row->item.ptr, row->item.size); | |
| else | |
| THROW(0, 0, | |
| this, | |
| "untaint in SQL language failed - no connection specified"); | |
| break; | |
| case UL_JS: | |
| escape(switch(*src) { | |
| case '"': to_string("\\\"", 2); break; | |
| case '\'': to_string("\\'", 2); break; | |
| case '\n': to_string("\\n", 2); break; | |
| case '\\': to_string("\\\\", 2); break; | |
| case '\xFF': to_string("\\\xFF", 2); break; | |
| _default; | |
| }); | |
| break; | |
| case UL_HTML: | |
| escape(switch(*src) { | |
| case '&': to_string("&", 5); break; | |
| case '>': to_string(">", 4); break; | |
| case '<': to_string("<", 4); break; | |
| case '"': to_string(""", 6); break; | |
| //TODO: XSLT case '\'': to_string("'", 6); break; | |
| _default; | |
| }); | |
| break; | |
| case UL_USER_HTML: { | |
| // tainted, untaint language: html-typo | |
| if(!typo_dict) // never, always has default | |
| THROW(0, 0, | |
| this, | |
| "untaint to user-html lang failed, no typo table"); | |
| char *html_for_typo= | |
| (char *)malloc(row->item.size*2/* '\n' -> '\' 'n' */+1); | |
| // note: | |
| // there still is a possibility that user | |
| // would not replace \n as she supposed to | |
| // and rather replace \ and n into huge strings | |
| // thus causing memory overrun | |
| // this can be dealed by allocating *2 memory, but that's too expensive | |
| size_t html_for_typo_size; | |
| { // local dest | |
| char *dest=html_for_typo; | |
| escape(switch(*src) { | |
| // convinient name for typo match "\n" | |
| case '\r': | |
| to_string("\\n", 2); // \r -> "\n" | |
| if(size && src[1]=='\n') { // \r\n -> remove \n | |
| size--; src++; | |
| } | |
| break; | |
| case '\n': | |
| to_string("\\n", 2); | |
| break; | |
| //TODO: XSLT case '\'': to_string("'", 6); break; | |
| _default; | |
| }); | |
| *dest=0; | |
| html_for_typo_size=dest-html_for_typo; | |
| } | |
| // typo table replacements | |
| const char *src=html_for_typo; | |
| do { | |
| // there is a row where first column starts 'src' | |
| if(Table::Item *item=typo_dict->first_that_starts(src)) { | |
| // get a=>b values | |
| const String& a=*static_cast<Array *>(item)->get_string(0); | |
| const String& b=*static_cast<Array *>(item)->get_string(1); | |
| // overflow check: | |
| // b allowed to be max UNTAINT_TIMES_BIGGER then a | |
| if(b.size()>UNTAINT_TIMES_BIGGER*a.size()) { | |
| pool().set_tag(0); // avoid recursion | |
| THROW(0, 0, | |
| &b, | |
| "is %g times longer then '%s', " | |
| "while maximum, handled by Parser, is %d", | |
| ((double)b.size())/a.size(), | |
| a.cstr(), | |
| UNTAINT_TIMES_BIGGER); | |
| } | } |
| //RFC Upper case should be used for hexadecimal digits "A" through "F" | |
| // skip 'a' in 'src' | if(*src == 0x20) // RFC The 8-bit hexadecimal value 20 (e.g., ISO-8859-1 SPACE) |
| src+=a.size(); | *dest++='_'; // RFC may be represented as "_" |
| // write 'b' to 'dest' | else |
| b.store_to(dest); | dest+=sprintf(dest, "=%02X", *src & 0xFF); |
| // skip 'b' in 'dest' | |
| dest+=b.size(); | |
| } else | } else |
| *dest++=*src++; | *dest++=*src; |
| } while(*src); | if(*src=='>') |
| break; | email=false; |
| } | } |
| default: | if(to_quoted_printable) // close |
| THROW(0, 0, | dest+=sprintf(dest, "?="); |
| this, | |
| "unknown untaint language #%d of %d piece", | } else { |
| static_cast<int>(row->item.lang), | memcpy(dest, row->item.ptr, row->item.size); |
| i); // never | dest+=row->item.size; |
| break; // never | |
| } | } |
| break; | |
| case UL_TABLE: | |
| // tainted, untaint language: table | |
| escape(switch(*src) { | |
| case '\t': to_char(' '); break; | |
| case '\n': to_char(' '); break; | |
| _default; | |
| }); | |
| break; | |
| case UL_SQL: | |
| // tainted, untaint language: sql | |
| if(connection) | |
| dest+=connection->quote(dest, row->item.ptr, row->item.size); | |
| else | |
| throw Exception(0, | |
| this, | |
| "untaint in SQL language failed - no connection specified"); | |
| break; | |
| case UL_JS: | |
| escape(switch(*src) { | |
| case '"': to_string("\\\"", 2); break; | |
| case '\'': to_string("\\'", 2); break; | |
| case '\n': to_string("\\n", 2); break; | |
| case '\\': to_string("\\\\", 2); break; | |
| case '\xFF': to_string("\\\xFF", 2); break; | |
| _default; | |
| }); | |
| break; | |
| case UL_XML: | |
| escape(switch(*src) { | |
| case '&': to_string("&", 5); break; | |
| case '>': to_string(">", 4); break; | |
| case '<': to_string("<", 4); break; | |
| case '"': to_string(""", 6); break; | |
| case '\'': to_string("'", 6); break; | |
| _default; | |
| }); | |
| break; | |
| case UL_HTML: | |
| escape(switch(*src) { | |
| case '&': to_string("&", 5); break; | |
| case '>': to_string(">", 4); break; | |
| case '<': to_string("<", 4); break; | |
| case '"': to_string(""", 6); break; | |
| _default; | |
| }); | |
| break; | |
| default: | |
| throw Exception(0, | |
| this, | |
| "unknown untaint language #%d", | |
| static_cast<int>(to_lang)); // should never | |
| break; // never | |
| } | |
| if((lang==UL_UNSPECIFIED?row->item.lang:lang)!=UL_CLEAN) | if(to_lang & UL_OPTIMIZE_BIT) { |
| whitespace=false; | // optimizing whitespace |
| char *stop=dest; dest=start; | |
| for(char *src=start; src<stop; src++) | |
| switch(*src) { | |
| // of all consequent white space chars leaving only first one | |
| case ' ': case '\r': case '\n': case '\t': | |
| if(!whitespace) { | |
| *dest++=*src; | |
| whitespace=true; | |
| } | |
| break; | |
| default: | |
| whitespace=false; | |
| *dest++=*src; | |
| break; | |
| }; | |
| } else // piece without optimization | |
| //debug | |
| whitespace=false; | |
| row++; countdown--; \ | |
| } | |
| /* | |
| ); | |
| */ | |
| return dest; | |
| } | |
| if(forigins_mode) | char *String::cstr_debug_origins() const { |
| if(dest==dest_after_origins) // never moved==optimized space | //_asm int 3; |
| dest=dest_before_origins; | char *result=(char *)malloc(size()+used_rows()*MAX_STRING*2); |
| else { | char *dest=result; |
| for(char *p=dest_after_origins; p<dest; p++) | |
| if(*p=='\n') | STRING_FOREACH_ROW( |
| *p='|'; | IFNDEF_NO_STRING_ORIGIN( |
| if(row->item.origin.file) | |
| dest+=sprintf(dest, ORIGIN_FILE_LINE_FORMAT, | |
| row->item.origin.file, | |
| 1+row->item.origin.line); | |
| else | |
| dest+=sprintf(dest, "<unknown>"); | |
| ); | |
| uchar show_lang=row->item.lang & ~UL_OPTIMIZE_BIT; | |
| if(show_lang>=sizeof(String_Untaint_lang_name)/sizeof(String_Untaint_lang_name[0])) | |
| throw Exception(0, | |
| this, | |
| "unknown untaint language #%d", | |
| static_cast<int>(show_lang)); // sould never | |
| dest+=sprintf(dest, "#%s%s: ", | |
| String_Untaint_lang_name[show_lang], | |
| row->item.lang & UL_OPTIMIZE_BIT?".O":""); | |
| char *dest_after_origins=dest; | |
| memcpy(dest, row->item.ptr, row->item.size); | |
| dest+=row->item.size; | |
| remove_crlf(dest_after_origins, dest); | |
| to_char('\n'); | |
| ); | |
| to_char('\n'); | *dest=0; |
| } | return result; |
| } | |
| chunk=row->link; | |
| } while(chunk); | |
| break2: | |
| return dest; | |
| } | } |