|
|
| version 1.75, 2001/11/16 12:38:44 | version 1.104, 2002/08/01 11:41:19 |
|---|---|
| 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, 2002 ArtLebedev Group (http://www.artlebedev.com) |
| Author: Alexander Petrosyan <paf@design.ru>(http://paf.design.ru) | Author: Alexandr Petrosian <paf@design.ru> (http://paf.design.ru) |
| $Id$ | |
| */ | */ |
| static const char* IDENT_UNTAINT_C="$Date$"; | |
| #include "pa_pool.h" | #include "pa_pool.h" |
| #include "pa_string.h" | #include "pa_string.h" |
| #include "pa_hash.h" | #include "pa_hash.h" |
| Line 16 | Line 16 |
| #include "pa_sql_connection.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" | |
| //#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 25 | Line 32 |
| } | } |
| #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 41 | Line 47 |
| dest+=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, | |
| // they would yield the same | |
| // because need_file_encode('_')=false | |
| // but we need to delete such files somehow, getting names from ^index | |
| 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')) |
| return false; | return false; |
| return !strchr( | return !strchr( |
| " _./()-" | |
| #ifdef WIN32 | #ifdef WIN32 |
| ":\\~" | ":\\~" |
| #endif | #endif |
| "./()_-", c); | , c); |
| } | } |
| 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 84 static const char * String_Untaint_lang_ | Line 96 static const char * String_Untaint_lang_ |
| "SQL", ///< ^table:sql body | "SQL", ///< ^table:sql body |
| "JS", ///< JavaScript code | "JS", ///< JavaScript code |
| "XML", ///< ^dom:set xml | "XML", ///< ^dom:set xml |
| "HTML", ///< HTML code (for editing) | "HTML" ///< HTML code (for editing) |
| "UHTML", ///< HTML code with USER chars | |
| }; | }; |
| Line 133 inline bool need_quote_http_header(const | Line 144 inline bool need_quote_http_header(const |
| return false; | return false; |
| } | } |
| /// @test UL_OPTIMIZED_HTML optimize | //#include "pa_sapi.h" |
| size_t String::cstr_bufsize(Untaint_lang lang, | /** |
| SQL_Connection *connection, | appends other String, |
| const char *charset) const { | marking all tainted pieces of it with @a lang. |
| size_t dest=1; | or marking ALL pieces of it with a @a lang when @a forced to, |
| bool whitespace=true; | and propagating OPTIMIZE language bit. |
| const Chunk *chunk=&head; | |
| do { | using architecture advantage: after string-to-string-append string never modified. |
| const Chunk::Row *row=chunk->rows; | algorithm: |
| for(uint i=0; i<chunk->count; i++, row++) { | if no language-change specified and src not yet appended to some other string[last_chunk!=0] |
| if(row==append_here) | shrinking dest last_chunk[preparing it for linking], |
| goto break2; | ///shrinking src last_chunk[preparing it to be linked, consequent dest.appends would go there], |
| linking[dest.last_chunk = src.head.chunk] | |
| Untaint_lang to_lang=lang==UL_UNSPECIFIED?(Untaint_lang)row->item.lang:lang; | if some language-change specified or src already appended to some other string[last_chunk==0] |
| cloning pieces. | |
| if(forigins_mode) { | */ |
| #ifndef NO_STRING_ORIGIN | String& String::append(const String& src, uchar lang, bool forced) { |
| dest+=MAX_STRING; | 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 | #endif |
| dest+=MAX_STRING; | |
| } | |
| switch(to_lang) { | // using fact: |
| case UL_CLEAN: | // src.head.chunk.count initally equeals this.head.chunk.count and shrinks-only, |
| // clean piece | // so can't be more than this.head.chunk.count, |
| { // optimizing whitespace | // which means that we know that |
| escape(switch(*src) { | // src.head.chunk would fit into this.head.chunk |
| case ' ': case '\n': case '\t': | if(is_empty()) { // our head.chunk is empty |
| if(!whitespace) { | // they have more than head.chunk? we need all head.chunk : we need only filled-part of head.chunk |
| dest++; | Chunk *src_head_link=src.head.chunk.rows[src.head.chunk.count].link; |
| whitespace=true; | 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" |
| break; | memcpy(head.chunk.rows, src.head.chunk.rows, sizeof(Chunk::Row)*(head_count)); |
| default: | if(src_head_link) { |
| whitespace=false; | // "your body is my body" |
| dest++; | head.chunk.rows[head.chunk.count=head_count].link=src_head_link; |
| break; | // "your last_chunk is mine now" |
| }); | last_chunk=src.last_chunk; |
| } | // "your append_here is mine now" |
| break; | append_here=src.append_here; |
| case UL_TAINTED: | } else { |
| // tainted piece, but undefined untaint language | // "your last_chunk is mine now" |
| // for VString.as_double of tainted values | last_chunk=&head.chunk; |
| // for ^process{body} evaluation | // "your append_here is recalc-mine now" |
| case UL_AS_IS: | append_here=head.chunk.rows+head_count; |
| // 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*3/* 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(charset) { | |
| // Subject: Re: parser3: =?koi8-r?Q?=D3=C5=CD=C9=CE=C1=D2?= | |
| dest+=row->item.size*3+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: | |
| case UL_OPTIMIZED_HTML: | |
| escape(switch(*src) { | |
| case '&': | |
| case '>': | |
| case '<': | |
| case '"': | |
| dest+=6; break; | |
| default: | |
| dest++; break; | |
| }); | |
| break; | |
| } | } |
| } else { // our head.chunk contains something | |
| if((lang==UL_UNSPECIFIED?row->item.lang:lang)!=UL_CLEAN) | // "chopping off my tail-reserve" |
| whitespace=false; | 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; | |
| } | } |
| chunk=row->link; | |
| } while(chunk); | |
| break2: | // 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; | |
| } | |
| 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) { | |
| // Subject: Re: parser3: =?koi8-r?Q?=D3=C5=CD=C9=CE=C1=D2?= | |
| dest+= | |
| 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; | return dest; |
| } | } |
| /// @test UL_OPTIMIZED_HTML optimize | |
| 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, |
| const char *store_to_charset_name) const { | |
| // WARNING: | // WARNING: |
| // before any changes check cstr_bufsize first!!! | // before any changes check cstr_bufsize first!!! |
| bool whitespace=true; | bool whitespace=true; |
| const Chunk *chunk=&head; | STRING_FOREACH_ROW( |
| do { | uchar to_lang=lang==UL_UNSPECIFIED?row->item.lang:lang; |
| const Chunk::Row *row=chunk->rows; | |
| for(uint i=0; i<chunk->count; i++, row++) { | |
| if(row==append_here) | |
| goto break2; | |
| Untaint_lang to_lang=lang==UL_UNSPECIFIED?(Untaint_lang)row->item.lang:lang; | |
| char *dest_before_origins=dest; | |
| if(forigins_mode) { | |
| #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>"); | |
| #endif | |
| dest+=sprintf(dest, "#%s: ", | |
| String_Untaint_lang_name[to_lang]); | |
| } | |
| char *dest_after_origins=dest; | |
| switch(to_lang) { | char *start=dest; |
| case UL_CLEAN: | |
| // clean piece | switch(to_lang & ~UL_OPTIMIZE_BIT) { |
| { // optimizing whitespace | case UL_CLEAN: |
| escape(switch(*src) { | case UL_TAINTED: |
| case ' ': case '\n': case '\t': | case UL_AS_IS: |
| if(!whitespace) { | // clean piece |
| *dest++=*src; | |
| whitespace=true; | // tainted piece, but undefined untaint language |
| } | // for VString.as_double of tainted values |
| break; | // for ^process{body} evaluation |
| default: | |
| whitespace=false; | // tainted, untaint language: as-is |
| *dest++=*src; | memcpy(dest, row->item.ptr, row->item.size); |
| break; | dest+=row->item.size; |
| }); | break; |
| } | case UL_FILE_SPEC: |
| break; | // tainted, untaint language: file [name] |
| case UL_TAINTED: | escape( |
| // tainted piece, but undefined untaint language | encode(need_file_encode, '_'); |
| // for VString.as_double of tainted values | ); |
| // for ^process{body} evaluation | break; |
| case UL_AS_IS: | case UL_URI: |
| // tainted, untaint language: as-is | // tainted, untaint language: uri |
| memcpy(dest, row->item.ptr, row->item.size); | const void *client_ptr; |
| dest+=row->item.size; | size_t client_size; |
| break; | Charset::transcode(pool(), |
| case UL_FILE_SPEC: | pool().get_source_charset(), row->item.ptr, row->item.size, |
| // tainted, untaint language: file [name] | pool().get_client_charset(), client_ptr, client_size); |
| escape(switch(*src) { | { |
| case ' ': to_char('_'); break; | const char *src=(const char *)client_ptr; |
| encode(need_file_encode, '+'); | for(int size=client_size; size--; src++) |
| }); | switch(*src) { |
| break; | case ' ': to_char('+'); break; |
| case UL_URI: | default: encode(need_uri_encode, '%'); |
| // tainted, untaint language: uri | }; |
| escape(switch(*src) { | } |
| case ' ': to_char('+'); break; | break; |
| encode(need_uri_encode, '%'); | case UL_HTTP_HEADER: |
| }); | // tainted, untaint language: http-field-content-text |
| break; | escape(switch(*src) { |
| case UL_HTTP_HEADER: | case ' ': to_char('+'); break; |
| // tainted, untaint language: http-field-content-text | default: encode(need_uri_encode, '%'); |
| escape(switch(*src) { | }); |
| case ' ': to_char('+'); break; | break; |
| encode(need_uri_encode, '%'); | case UL_MAIL_HEADER: |
| }); | // tainted, untaint language: mail-header |
| break; | if(store_to_charset && store_to_charset_name) { |
| case UL_MAIL_HEADER: | const void *mail_ptr; |
| // tainted, untaint language: mail-header | size_t mail_size; |
| if(charset) { | Charset::transcode(pool(), |
| // Subject: Re: parser3: =?koi8-r?Q?=D3=C5=CD=C9=CE=C1=D2?= | pool().get_source_charset(), row->item.ptr, row->item.size, |
| const char *src=row->item.ptr; | *store_to_charset, mail_ptr, mail_size); |
| bool to_quoted_printable=false; | |
| for(int size=row->item.size; size--; src++) { | // Subject: Re: parser3: =?koi8-r?Q?=D3=C5=CD=C9=CE=C1=D2?= |
| if(*src & 0x80) { | const char *src=(const char *)mail_ptr; |
| if(!to_quoted_printable) { | bool to_quoted_printable=false; |
| dest+=sprintf(dest, "=?%.15s?Q?", charset); | for(int size=mail_size; size--; src++) { |
| to_quoted_printable=true; | if((*src & 0x80) // starting quote-printable-encoding on first 8bit char |
| } | || (to_quoted_printable && (*src=='?' || *src=='=')) // additionally encoding '?' and '|' |
| dest+=sprintf(dest, "=%02X", *src & 0xFF); | ) { |
| } else { | if(!to_quoted_printable) { |
| *dest++=*src; | dest+=sprintf(dest, "=?%s?Q?", store_to_charset_name); |
| to_quoted_printable=true; | |
| } | } |
| dest+=sprintf(dest, "=%02X", *src & 0xFF); | |
| } else { | |
| *dest++=*src; | |
| } | } |
| if(to_quoted_printable) // close | |
| dest+=sprintf(dest, "?="); | |
| } else { | |
| memcpy(dest, row->item.ptr, row->item.size); | |
| dest+=row->item.size; | |
| } | } |
| break; | if(to_quoted_printable) // close |
| case UL_TABLE: | dest+=sprintf(dest, "?="); |
| // tainted, untaint language: table | |
| escape(switch(*src) { | } else { |
| case '\t': to_char(' '); break; | memcpy(dest, row->item.ptr, row->item.size); |
| case '\n': to_char(' '); break; | dest+=row->item.size; |
| _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, 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: | |
| case UL_OPTIMIZED_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, 0, | |
| this, | |
| "unknown untaint language #%d of %d piece", | |
| static_cast<int>(row->item.lang), | |
| i); // never | |
| 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>(row->item.lang)); // sould 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 | |
| whitespace=false; | |
| ); | |
| if(forigins_mode) | return dest; |
| if(dest==dest_after_origins) // never moved==optimized space | } |
| dest=dest_before_origins; | |
| else { | |
| remove_crlf(dest_after_origins, dest); | |
| to_char('\n'); | char *String::cstr_debug_origins() const { |
| } | //_asm int 3; |
| } | char *result=(char *)malloc(size()+used_rows()*MAX_STRING*2); |
| chunk=row->link; | char *dest=result; |
| } while(chunk); | |
| 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'); | |
| ); | |
| break2: | *dest=0; |
| return dest; | return result; |
| } | } |