|
|
| version 1.104, 2002/08/01 11:41:19 | version 1.115.2.12.2.6, 2003/03/20 14:19:10 |
|---|---|
| Line 1 | Line 1 |
| /** @file | /** @file |
| Parser: String class part: untaint mechanizm. | Parser: String class part: untaint mechanizm. |
| Copyright(c) 2001, 2002 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: Alexandr Petrosian <paf@design.ru> (http://paf.design.ru) |
| */ | */ |
| static const char* IDENT_UNTAINT_C="$Date$"; | static const char* IDENT_UNTAINT_C="$Date$"; |
| #include "pa_pool.h" | |
| #include "pa_string.h" | #include "pa_string.h" |
| #include "pa_hash.h" | #include "pa_hash.h" |
| #include "pa_exception.h" | #include "pa_exception.h" |
| #include "pa_table.h" | #include "pa_table.h" |
| #include "pa_globals.h" | #include "pa_globals.h" |
| #include "pa_sql_connection.h" | |
| #include "pa_dictionary.h" | #include "pa_dictionary.h" |
| #include "pa_common.h" | #include "pa_common.h" |
| #include "pa_charset.h" | #include "pa_charset.h" |
| #include "pa_request_charsets.h" | |
| extern "C" { // author forgot to do that | |
| #include "ec.h" | |
| } | |
| //#define DEBUG_STRING_APPENDS_VS_EXPANDS | //#define PA_SQL |
| #ifdef DEBUG_STRING_APPENDS_VS_EXPANDS | #ifdef PA_SQL |
| ulong string_string_shortcut_economy=0; | #include "pa_sql_connection.h" |
| #endif | #endif |
| #define escape(action) \ | #define escape(action) \ |
| { \ | for(; fragment_length--; CORD_next(pos)) { \ |
| const char *src=row->item.ptr; \ | char c=CORD_pos_fetch(pos); \ |
| for(int size=row->item.size; size--; src++) \ | action \ |
| action \ | |
| } | } |
| #define _default default: *dest++=*src; break | #define _default default: CORD_ec_append(result, c); break |
| #define encode(need_encode_func, prefix) \ | #define encode(need_encode_func, prefix) \ |
| if(need_encode_func(*src)) { \ | if(need_encode_func(c)) { \ |
| static const char *hex="0123456789ABCDEF"; \ | static const char* hex="0123456789ABCDEF"; \ |
| char chunk[3]={prefix}; \ | CORD_ec_append(result, prefix); \ |
| chunk[1]=hex[((unsigned char)*src)/0x10]; \ | CORD_ec_append(result, hex[((unsigned char)c)/0x10]); \ |
| chunk[2]=hex[((unsigned char)*src)%0x10]; \ | CORD_ec_append(result, hex[((unsigned char)c)%0x10]); \ |
| memcpy(dest, chunk, 3); dest+=3; \ | } else \ |
| } else \ | CORD_ec_append(result, c); |
| *dest++=*src; \ | #define to_char(c) CORD_ec_append(result, c) |
| break | #define to_string(s) CORD_ec_append_cord(result, s) |
| #define to_char(c) *dest++=c | |
| #define to_string(b, bsize) \ | |
| memcpy(dest, b, bsize); \ | |
| dest+=bsize; \ | |
| inline bool need_file_encode(unsigned char c){ | inline bool need_file_encode(unsigned char c){ |
| // theoretical problem with, for instance, "_2B" and "." fragments, | // russian letters and space ENABLED |
| // they would yield the same | // encoding only these... |
| // because need_file_encode('_')=false | return strchr( |
| // but we need to delete such files somehow, getting names from ^index | "*?'\"<>|" |
| #ifndef WIN32 | |
| if((c>='0') &&(c<='9') ||(c>='A') &&(c<='Z') ||(c>='a') &&(c<='z')) | ":\\~" |
| return false; | |
| return !strchr( | |
| " _./()-" | |
| #ifdef 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 77 inline bool need_http_header_encode(unsi | Line 70 inline bool need_http_header_encode(unsi |
| // | // |
| static const char * String_Untaint_lang_name[]={ | static const char* String_Untaint_lang_name[]={ |
| "U", ///< zero value handy for hash lookup @see untaint_lang_name2enum | "U", ///< zero value handy for hash lookup @see untaint_lang_name2enum |
| "C", ///< clean | "C", ///< clean |
| "T", ///< tainted, untaint language as assigned later | "T", ///< tainted, untaint language as assigned later |
| Line 137 quoted-pair = "\" CHAR | Line 130 quoted-pair = "\" CHAR |
| if(strchr("()<>@,;:\\\"/[]?={} \t", *ptr)) | if(strchr("()<>@,;:\\\"/[]?={} \t", *ptr)) |
| */ | */ |
| inline bool need_quote_http_header(const char *ptr, size_t size) { | inline bool need_quote_http_header(const char* ptr, size_t size) { |
| for(; size--; ptr++) | for(; size--; ptr++) |
| if(strchr(";\\\"= \t" /* excluded ()<>@, :/ ? []{} */, *ptr)) | if(strchr(";\\\"= \t" /* excluded ()<>@, :/ ? []{} */, *ptr)) |
| return true; | return true; |
| return false; | return false; |
| } | } |
| //#include "pa_sapi.h" | |
| /** | /** |
| appends other String, | appends to other String, |
| marking all tainted pieces of it with @a lang. | marking all tainted pieces of it with @a lang. |
| or marking ALL pieces of it with a @a lang when @a forced to, | or marking ALL pieces of it with a @a lang when @a forced to, |
| and propagating OPTIMIZE language bit. | 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. | |
| */ | */ |
| String& String::append(const String& src, uchar lang, bool forced) { | String& String::append_to(String& dest, Untaint_lang lang, bool forced) const { |
| if(!last_chunk) // growth stopped [we're appended as string to somebody] | if(!*this) |
| throw Exception(0, | return dest; |
| 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: | // first: letters |
| // src.head.chunk.count initally equeals this.head.chunk.count and shrinks-only, | dest.body=CORD_cat(dest.body, body); |
| // 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 | // next: fragment infos |
| src.last_chunk=0; | |
| return *this; | if(lang==UL_PASS_APPENDED) // without language-chage string? |
| dest.fragments.append(fragments, 0, fragments.count()); | |
| else if(forced) //forcing passed lang? | |
| dest.fragments+=Fragment(lang, length()); | |
| else if(lang&UL_OPTIMIZE_BIT) { | |
| Array_iterator<ArrayFragment::element_type> i(fragments); | |
| while(i.has_next()) { | |
| const Fragment& fragment=i.next(); | |
| // 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 | |
| dest.fragments+=Fragment(static_cast<Untaint_lang>( | |
| fragment.lang==UL_TAINTED?lang:( | |
| fragment.lang==UL_CLEAN?UL_CLEAN|UL_OPTIMIZE_BIT: // ORing with OPTIMIZED flag | |
| fragment.lang)), | |
| fragment.length); | |
| } | |
| } else { // The core idea: tainted pieces got marked with context's lang | |
| Array_iterator<ArrayFragment::element_type> i(fragments); | |
| while(i.has_next()) { | |
| const Fragment& fragment=i.next(); | |
| dest.fragments+=Fragment( | |
| fragment.lang==UL_TAINTED?lang:fragment.lang, | |
| fragment.length); | |
| } | |
| } | } |
| // manually unrolled code to avoid do{if(const)} constructs | return dest; |
| 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; | |
| } | } |
| size_t String::cstr_bufsize(Untaint_lang lang, | /** http://www.ietf.org/rfc/rfc2047.txt |
| SQL_Connection *connection, | RFC |
| Charset *buf_charset) const { | (3) As a replacement for a 'word' entity within a 'phrase', for example, |
| size_t dest=1; // for terminating 0 | one that precedes an address in a From, To, or Cc header. The ABNF |
| STRING_FOREACH_ROW( | definition for 'phrase' from RFC 822 thus becomes: |
| uchar to_lang=lang==UL_UNSPECIFIED?row->item.lang:lang; | |
| phrase = 1*( encoded-word / word ) | |
| switch(to_lang & ~UL_OPTIMIZE_BIT) { | |
| case UL_CLEAN: | In this case the set of characters that may be used in a "Q"-encoded |
| case UL_TAINTED: | 'encoded-word' is restricted to: <upper and lower case ASCII |
| case UL_AS_IS: | letters, decimal digits, "!", "*", "+", "-", "/", "=", and "_" |
| // clean piece | (underscore, ASCII 95.)>. An 'encoded-word' that appears within a |
| 'phrase' MUST be separated from any adjacent 'word', 'text' or | |
| // tainted piece, but undefined untaint language | 'special' by 'linear-white-space'. |
| // for VString.as_double of tainted values | ... |
| // for ^process{body} evaluation | (2) The 8-bit hexadecimal value 20 (e.g., ISO-8859-1 SPACE) may be |
| represented as "_" (underscore, ASCII 95.). (This character may | |
| // tainted, untaint language: as-is | not pass through some internetwork mail gateways, but its use |
| dest+=row->item.size; | will greatly enhance readability of "Q" encoded data with mail |
| break; | readers that do not support this encoding.) Note that the "_" |
| case UL_FILE_SPEC: | always represents hexadecimal 20, even if the SPACE character |
| // tainted, untaint language: file [name] | occupies a different code position in the character set in use. |
| dest+=row->item.size*3/* worst: Z->%XX */; | |
| break; | paf: obviously, |
| case UL_URI: | without "=", or one could not differ "=E0" and "russian letter a" |
| // tainted, untaint language: uri | and without "_", or in would mean 0x20 |
| dest+=row->item.size*6*3/* worst utf8 x worst Z->%XX */; | */ |
| break; | inline bool mail_header_char_valid_within_Qencoded(char c) { |
| case UL_HTTP_HEADER: | return c>='A' && c<='Z' |
| // tainted, untaint language: http-field-content-text | || c>='a' && c<='Z' |
| dest+=row->item.size*3/* worst: Z->%XX */; | || c>='0' && c<='9' |
| break; | || strchr("!*+-/", c); |
| case UL_MAIL_HEADER: | } |
| // tainted, untaint language: mail-header | inline void ec_append(CORD_ec result, CORD_pos& pos, size_t size) { |
| if(buf_charset) { | while(size--) { |
| // Subject: Re: parser3: =?koi8-r?Q?=D3=C5=CD=C9=CE=C1=D2?= | CORD_ec_append(result, CORD_pos_fetch(pos)); |
| dest+= | CORD_next(pos); |
| row->item.size*3+ | } |
| 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; | |
| } | } |
| CORD String::cstr_to_cord(Untaint_lang lang, | |
| SQL_Connection *connection, | |
| const Request_charsets *charsets) const { | |
| CORD_ec result; CORD_ec_init(result); | |
| CORD_pos pos; CORD_set_pos(pos, body, 0); | |
| char *String::store_to(char *dest, Untaint_lang lang, | |
| SQL_Connection *connection, | |
| Charset *store_to_charset, | |
| const char *store_to_charset_name) const { | |
| // WARNING: | |
| // before any changes check cstr_bufsize first!!! | |
| bool whitespace=true; | bool whitespace=true; |
| STRING_FOREACH_ROW( | |
| uchar to_lang=lang==UL_UNSPECIFIED?row->item.lang:lang; | |
| char *start=dest; | size_t fragment_start=0; |
| size_t fragment_end; | |
| for(Array_iterator<Fragment> i(fragments); i.has_next(); fragment_start=fragment_end) { | |
| const Fragment& fragment=i.next(); | |
| size_t fragment_length=fragment.length; | |
| fragment_end=fragment_start+fragment_length; | |
| Untaint_lang to_lang=lang==UL_UNSPECIFIED?fragment.lang:lang; | |
| switch(to_lang & ~UL_OPTIMIZE_BIT) { | switch(to_lang & ~UL_OPTIMIZE_BIT) { |
| case UL_CLEAN: | case UL_CLEAN: |
| Line 361 char *String::store_to(char *dest, Untai | Line 251 char *String::store_to(char *dest, Untai |
| // for ^process{body} evaluation | // for ^process{body} evaluation |
| // tainted, untaint language: as-is | // tainted, untaint language: as-is |
| memcpy(dest, row->item.ptr, row->item.size); | ec_append(result, pos, fragment_length); |
| dest+=row->item.size; | |
| break; | break; |
| case UL_FILE_SPEC: | case UL_FILE_SPEC: |
| // tainted, untaint language: file [name] | // tainted, untaint language: file [name] |
| escape( | escape( |
| encode(need_file_encode, '_'); | // Macintosh has problems with small Russian letter 'r' |
| if( c=='\xF0' && charsets && charsets->source().name()=="windows-1251" ) { | |
| // fixing that letter for most common charset | |
| to_char('p'); | |
| } else // fallback to default | |
| encode(need_file_encode, '_'); | |
| ); | ); |
| break; | break; |
| #ifdef LATER | |
| case UL_URI: | case UL_URI: |
| // tainted, untaint language: uri | // tainted, untaint language: uri |
| const void *client_ptr; | |
| size_t client_size; | |
| Charset::transcode(pool(), | |
| pool().get_source_charset(), row->item.ptr, row->item.size, | |
| pool().get_client_charset(), client_ptr, client_size); | |
| { | { |
| const char *src=(const char *)client_ptr; | const char *substr= |
| for(int size=client_size; size--; src++) | CORD_to_const_char_star( |
| switch(*src) { | CORD_substr(body, fragment_start, fragment_length)); |
| case ' ': to_char('+'); break; | const void *client_ptr; |
| default: encode(need_uri_encode, '%'); | size_t client_size; |
| }; | if(charsets) |
| Charset::transcode( | |
| charsets->source(), substr, fragment_length, | |
| charsets->client(), client_ptr, client_size); | |
| else { | |
| client_ptr=substr; | |
| client_size=fragment_length; | |
| } | |
| char c; | |
| for(const char* src=(const char* )client_ptr; c=*src++; ) | |
| encode(need_uri_encode, '%'); | |
| } | } |
| break; | break; |
| #endif | |
| case UL_HTTP_HEADER: | case UL_HTTP_HEADER: |
| // tainted, untaint language: http-field-content-text | // tainted, untaint language: http-field-content-text |
| escape(switch(*src) { | escape( |
| case ' ': to_char('+'); break; | encode(need_uri_encode, '%'); |
| default: encode(need_uri_encode, '%'); | ); |
| }); | |
| break; | break; |
| #ifdef LATER | |
| case UL_MAIL_HEADER: | case UL_MAIL_HEADER: |
| // tainted, untaint language: mail-header | // tainted, untaint language: mail-header |
| if(store_to_charset && store_to_charset_name) { | // http://www.ietf.org/rfc/rfc2047.txt |
| const void *mail_ptr; | if(charsets) { |
| size_t mail_size; | const void *mail_ptr=fragment.ptr; |
| Charset::transcode(pool(), | size_t mail_size=fragment_length; |
| pool().get_source_charset(), row->item.ptr, row->item.size, | const char* charset_name=charsets->mail().name()->cstr(); |
| *store_to_charset, mail_ptr, mail_size); | |
| // 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=(const char* )mail_ptr; |
| bool to_quoted_printable=false; | bool to_quoted_printable=false; |
| for(int size=mail_size; size--; src++) { | |
| if((*src & 0x80) // starting quote-printable-encoding on first 8bit char | bool email=false; |
| || (to_quoted_printable && (*src=='?' || *src=='=')) // additionally encoding '?' and '|' | for(const char* end=src+mail_size; src<end; src++) { |
| ) { | //RFC + An 'encoded-word' MUST NOT appear in any portion of an 'addr-spec'. |
| if(to_quoted_printable && (c==',' || c=='<')) { | |
| email=c=='<'; | |
| dest+=sprintf(dest, "?="); | |
| to_quoted_printable=false; | |
| } | |
| if(!email && ( | |
| !to_quoted_printable && (c & 0x80) // starting quote-printable-encoding on first 8bit char | |
| || to_quoted_printable && !mail_header_char_valid_within_Qencoded(c) | |
| )) { | |
| if(!to_quoted_printable) { | if(!to_quoted_printable) { |
| dest+=sprintf(dest, "=?%s?Q?", store_to_charset_name); | dest+=sprintf(dest, "=?%s?Q?", charset_name.get()); |
| to_quoted_printable=true; | to_quoted_printable=true; |
| } | } |
| dest+=sprintf(dest, "=%02X", *src & 0xFF); | //RFC Upper case should be used for hexadecimal digits "A" through "F" |
| } else { | if(c == 0x20) // RFC The 8-bit hexadecimal value 20 (e.g., ISO-8859-1 SPACE) |
| *dest++=*src; | *dest++='_'; // RFC may be represented as "_" |
| } | else |
| dest+=sprintf(dest, "=%02X", c & 0xFF); | |
| } else | |
| *dest++=c; | |
| if(c=='>') | |
| email=false; | |
| } | } |
| if(to_quoted_printable) // close | if(to_quoted_printable) // close |
| dest+=sprintf(dest, "?="); | dest+=sprintf(dest, "?="); |
| } else { | } else { |
| memcpy(dest, row->item.ptr, row->item.size); | memcpy(dest, fragment.ptr, fragment_length); |
| dest+=row->item.size; | dest+=fragment_length; |
| } | } |
| break; | break; |
| #endif | |
| case UL_TABLE: | case UL_TABLE: |
| // tainted, untaint language: table | // tainted, untaint language: table |
| escape(switch(*src) { | escape(switch(c) { |
| case '\t': to_char(' '); break; | case '\t': to_char(' '); break; |
| case '\n': to_char(' '); break; | case '\n': to_char(' '); break; |
| _default; | _default; |
| }); | }); |
| break; | break; |
| #ifdef PA_SQL | |
| case UL_SQL: | case UL_SQL: |
| // tainted, untaint language: sql | // tainted, untaint language: sql |
| if(connection) | if(connection) { |
| dest+=connection->quote(dest, row->item.ptr, row->item.size); | dest+=connection->quote(0, |
| CORD_substr(body, fragment_start, fragment_length), | |
| fragment_length); | |
| ...^^^ | |
| dest+=connection->quote(dest, fragment.ptr, fragment_length); | |
| else | else |
| throw Exception(0, | throw Exception(0, |
| this, | 0, |
| "untaint in SQL language failed - no connection specified"); | "untaint in SQL language failed - no connection specified"); |
| break; | break; |
| #endif | |
| case UL_JS: | case UL_JS: |
| escape(switch(*src) { | escape(switch(c) { |
| case '"': to_string("\\\"", 2); break; | case '"': to_string("\\\""); break; |
| case '\'': to_string("\\'", 2); break; | case '\'': to_string("\\'"); break; |
| case '\n': to_string("\\n", 2); break; | case '\n': to_string("\\n"); break; |
| case '\\': to_string("\\\\", 2); break; | case '\\': to_string("\\\\"); break; |
| case '\xFF': to_string("\\\xFF", 2); break; | case '\xFF': to_string("\\\xFF"); break; |
| _default; | _default; |
| }); | }); |
| break; | break; |
| case UL_XML: | case UL_XML: |
| escape(switch(*src) { | escape(switch(c) { |
| case '&': to_string("&", 5); break; | case '&': to_string("&"); break; |
| case '>': to_string(">", 4); break; | case '>': to_string(">"); break; |
| case '<': to_string("<", 4); break; | case '<': to_string("<"); break; |
| case '"': to_string(""", 6); break; | case '"': to_string("""); break; |
| case '\'': to_string("'", 6); break; | case '\'': to_string("'"); break; |
| _default; | _default; |
| }); | }); |
| break; | break; |
| case UL_HTML: | case UL_HTML: |
| escape(switch(*src) { | escape(switch(c) { |
| case '&': to_string("&", 5); break; | case '&': to_string("&"); break; |
| case '>': to_string(">", 4); break; | case '>': to_string(">"); break; |
| case '<': to_string("<", 4); break; | case '<': to_string("<"); break; |
| case '"': to_string(""", 6); break; | case '"': to_string("""); break; |
| _default; | _default; |
| }); | }); |
| break; | break; |
| default: | default: |
| throw Exception(0, | throw Exception(0, |
| this, | 0, |
| "unknown untaint language #%d", | "unknown untaint language #%d", |
| static_cast<int>(row->item.lang)); // sould never | static_cast<int>(to_lang)); // should never |
| break; // never | break; // never |
| } | } |
| #ifdef LATER | |
| if(to_lang & UL_OPTIMIZE_BIT) { | if(to_lang & UL_OPTIMIZE_BIT) { |
| // optimizing whitespace | // optimizing whitespace |
| char *stop=dest; dest=start; | char *stop=dest; dest=start; |
| for(char *src=start; src<stop; src++) | for(char c=start; src<stop; src++) |
| switch(*src) { | switch(c) { |
| // of all consequent white space chars leaving only first one | // of all consequent white space chars leaving only first one |
| case ' ': case '\r': case '\n': case '\t': | case ' ': case '\r': case '\n': case '\t': |
| if(!whitespace) { | if(!whitespace) { |
| *dest++=*src; | *dest++=c; |
| whitespace=true; | whitespace=true; |
| } | } |
| break; | break; |
| default: | default: |
| whitespace=false; | whitespace=false; |
| *dest++=*src; | *dest++=c; |
| break; | break; |
| }; | }; |
| } else // piece without optimization | } else // piece without optimization |
| whitespace=false; | whitespace=false; |
| ); | #endif |
| } | |
| return dest; | |
| } | |
| char *String::cstr_debug_origins() const { | |
| //_asm int 3; | |
| char *result=(char *)malloc(size()+used_rows()*MAX_STRING*2); | |
| char *dest=result; | |
| STRING_FOREACH_ROW( | |
| 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'); | |
| ); | |
| *dest=0; | return CORD_ec_to_cord(result); |
| return result; | |
| } | } |