/* * Copyright © 2006-2016 SplinterGU (Fenix/Bennugd) * Copyright © 2002-2006 Fenix Team (Fenix) * Copyright © 1999-2002 José Luis Cebrián Pagüe (Fenix) * * This file is part of Bennu - Game Development * * This software is provided 'as-is', without any express or implied * warranty. In no event will the authors be held liable for any damages * arising from the use of this software. * * Permission is granted to anyone to use this software for any purpose, * including commercial applications, and to alter it and redistribute it * freely, subject to the following restrictions: * * 1. The origin of this software must not be misrepresented; you must not * claim that you wrote the original software. If you use this software * in a product, an acknowledgment in the product documentation would be * appreciated but is not required. * * 2. Altered source versions must be plainly marked as such, and must not be * misrepresented as being the original software. * * 3. This notice may not be removed or altered from any source * distribution. * */ #include #ifdef TARGET_BEOS #include #else #include #endif #include #include #include "bgdc.h" #include "xctype.h" /* ---------------------------------------------------------------------- */ /* Este módulo contiene las funciones que compilan definiciones de datos: */ /* declaraciones de variables e inicialización de las mismas */ /* ---------------------------------------------------------------------- */ int compile_array_data( VARSPACE * n, segment * data, int size, int subsize, BASETYPE *t ) { int block, count = 0, base, remaining = size ; int base_offset, total_length ; expresion_result res ; for ( ;; ) { if ( !remaining && size ) { token_back(); break; // compile_error( MSG_TOO_MANY_INIT ) ; } token_next() ; /* if (*t == TYPE_UNDEFINED && token.type == STRING) { *t = typedef_base(typedef_new(TYPE_STRING)) ; } */ if ( token.type == STRING && *t == TYPE_CHAR ) { const char * str = string_get( token.code ) ; int subcount = 0 ; if ( subsize == 0 ) subsize = strlen( str ) + 1; if (( int )strlen( str ) > subsize ) compile_error( MSG_TOO_MANY_INIT ) ; while ( *str ) { segment_add_as( data, *str++, *t ) ; subcount++ ; } while ( subcount++ < subsize ) segment_add_as( data, 0, *t ) ; count += subsize ; remaining -= subsize ; } else if ( token.type == IDENTIFIER && ( token.code == identifier_rightp || token.code == identifier_semicolon ) ) { token_back() ; break ; } else if ( token.type == IDENTIFIER && token.code == identifier_leftp ) { block = compile_array_data( n, data, remaining, remaining, t ) ; remaining -= block ; count += block ; token_next() ; if ( token.type != IDENTIFIER || token.code != identifier_rightp ) compile_error( MSG_EXPECTED, ")" ) ; } else { token_back() ; res = compile_expresion( 1, 0, 0, *t ) ; if ( *t == TYPE_UNDEFINED ) { *t = typedef_base( res.type ) ; if ( *t == TYPE_UNDEFINED ) { compile_error( MSG_INCOMP_TYPE ) ; } } base = res.value ; if ( *t == TYPE_FLOAT ) base = *( int * ) & res.fvalue ; token_next() ; if ( token.type == IDENTIFIER && token.code == identifier_dup ) { token_next() ; if ( token.type != IDENTIFIER || token.code != identifier_leftp ) compile_error( MSG_EXPECTED, "(" ) ; base_offset = data->current ; block = compile_array_data( n, data, remaining, remaining, t ) ; total_length = data->current - base_offset ; if ( size && block * base > remaining ) { break; /* MSG_TOO_MANY_INIT */ } count += block * base ; if ( size ) remaining -= block * base ; while ( base-- > 1 ) { segment_copy( data, base_offset, total_length ) ; base_offset += total_length ; } token_next() ; if ( token.type != IDENTIFIER || token.code != identifier_rightp ) compile_error( MSG_EXPECTED, ")" ) ; } else { token_back() ; if ( *t == TYPE_STRING ) varspace_varstring( n, data->current ) ; segment_add_as( data, base, *t ) ; count ++ ; if ( size ) remaining -- ; } } token_next() ; if ( token.type == IDENTIFIER && token.code == identifier_comma ) { if ( !size && *t == TYPE_CHAR ) compile_error( MSG_TOO_MANY_INIT ); continue ; } token_back() ; break ; } return count ; } static BASETYPE get_basetype( VARSPACE * v ) { TYPEDEF t ; BASETYPE type = TYPE_UNDEFINED, newtype ; int n ; for ( n = 0 ; n < v->count ; n++ ) { t = v->vars[n].type ; while ( typedef_is_array( t ) ) { t = typedef_reduce( t ) ; } if ( typedef_is_struct( t ) ) { newtype = get_basetype( typedef_members( t ) ) ; } else { newtype = typedef_base( t ) ; } if ( type != TYPE_UNDEFINED && type != newtype ) { return TYPE_UNDEFINED ; } type = newtype ; } return type ; } /* * FUNCTION : compile_struct_data * * Compile the values of an struct initialization * Does not suppor parenthized sections or the DUP operator * * PARAMS: * n Current variable space * data Current data segment * size Size (total number of variables) of the array * including all dimensions (1 for single struct) * sub 1 if the struct we'are initializing is part of * another one * (the function will then ignore a trailing comma) * * RETURN VALUE: * Number of structs initialized */ int compile_struct_data( VARSPACE * n, segment * data, int size, int sub ) { int elements = 0 ; int position = 0 ; expresion_result res ; for ( ;; ) { token_next() ; /* Allow parenthized struct initialization */ if ( token.type == IDENTIFIER && token.code == identifier_leftp ) { if (( elements % n->count ) != 0 ) compile_error( MSG_NOT_ENOUGH_INIT ); /* Note - don't ignore a trailing comma! */ elements = compile_struct_data( n, data, size, 0 ) ; if ( elements >= n->count ) size -= ( elements / n->count ); token_next() ; if ( token.type != IDENTIFIER || token.code != identifier_rightp ) compile_error( MSG_EXPECTED, ")" ) ; token_next() ; if (( elements % n->count ) == 0 && size > 0 && token.type == IDENTIFIER && token.code == identifier_comma ) continue; token_back() ; return elements ; } /* Allow empty initialization */ if ( token.type == IDENTIFIER && token.code == identifier_semicolon ) return 0 ; token_back() ; for ( ;; ) { TYPEDEF next_type = n->vars[position].type ; /* Next variable is a pointer */ if ( typedef_is_pointer( next_type ) ) { res = compile_expresion( 1, 0, 0, TYPE_DWORD ) ; if ( !res.constant ) compile_error( MSG_INCORRECT_PTR_INIT ); segment_add_as( data, 0, TYPE_POINTER ) ; } else if ( typedef_is_array( next_type ) ) /* Next variable is an array */ { int elements = typedef_tcount( next_type ) ; BASETYPE base; /* Get the array base type */ while ( typedef_is_array( next_type ) ) next_type = typedef_reduce( next_type ); base = typedef_base( next_type ); /* Special case: array of structs */ if ( base == TYPE_STRUCT ) { compile_struct_data( next_type.varspace, data, elements, 1 ); } else { token_next(); /* Special case: intializing char[] strings */ if ( token.type == STRING && next_type.chunk[1].type == TYPE_CHAR ) { const char * str = string_get( token.code ) ; int subcount = 0 ; if (( int )strlen( str ) > typedef_count( next_type ) - 1 ) compile_error( MSG_TOO_MANY_INIT ) ; while ( *str ) { segment_add_as( data, *str++, TYPE_CHAR ) ; subcount++ ; } while ( subcount++ < typedef_count( next_type ) ) segment_add_as( data, 0, TYPE_CHAR ) ; } else /* Initializing normal arrays */ { int has_parents = 1; if ( token.type != IDENTIFIER || token.code != identifier_leftp ) { has_parents = 0; token_back(); } compile_array_data( n, data, elements, elements, &base ) ; if ( has_parents ) { token_next() ; if ( token.type != IDENTIFIER || token.code != identifier_rightp ) compile_error( MSG_EXPECTED, ")" ) ; } } } } else if ( typedef_is_struct( next_type ) ) /* Next variable is another struct */ { compile_struct_data( next_type.varspace, data, 1, 1 ) ; } else /* Next variable is a single type */ { res = compile_expresion( 1, 0, 0, typedef_base( next_type ) ) ; if ( !res.constant ) compile_error( MSG_CONSTANT_EXP ); segment_add_as( data, typedef_base( next_type ) == TYPE_FLOAT ? *( int* )&res.fvalue : res.value, typedef_base( next_type ) ) ; } position++ ; elements++ ; if ( position == n->count && size < 2 && sub ) break ; /* A comma should be here */ token_next() ; if ( token.type == IDENTIFIER && token.code == identifier_semicolon ) { token_back(); break ; } if ( token.type == IDENTIFIER && token.code == identifier_rightp ) { token_back() ; break ; } if ( token.type != IDENTIFIER || token.code != identifier_comma ) compile_error( MSG_EXPECTED, "," ) ; /* Wrap around for the next struct */ if ( position == n->count ) { if ( size == 1 && !sub ) compile_error( MSG_TOO_MANY_INIT ) ; size--; position = 0; } } break; } return elements ; } static void set_type( TYPEDEF * t, BASETYPE type ) { /* int n ; for ( n = 0 ; n < t->depth ; n++ ) if ( t->chunk[n].type == TYPE_UNDEFINED ) { t->chunk[n].type = type ; break ; } return ; */ t->chunk[t->depth-1].type = type ; } /* * FUNCTION : compile_varspace * * Compile a variable space (a LOCAL, PRIVATE or GLOBAL section, or part of a STRUCT data) * * PARAMS : * n Pointer to the VARSPACE object (already initialized) * data Pointer to the data segment. All variables are added at end. * additive 1 if the compiling is additive (the struct members have non-local offset; GLOBAL case) * or not (STRUCT case) * copies Number of data copies (if the STRUCT is part of an array) * collision Check this varspace array for name collisions (NULL = end of array), * if nonzero (2006/11/19 19:34 GMT-03:00, Splinter - jj_arg@yahoo.com) * alignment Byte size for member alignment. Empty data will be added * to the varspace if its size is not multiple of this value. * * In any case, nested varspaces (structs) will have an alignment of 0. * * RETURN VALUE : * None * */ int compile_varspace( VARSPACE * n, segment * data, int additive, int copies, int padding, VARSPACE ** collision, int alignment, int duplicateignore ) { int i, j, total_count, last_count = 0 ; int base_offset = data->current ; int total_length ; int size, count ; int code ; expresion_result res ; VARIABLE * var ; int unsigned_prefix = 0; int signed_prefix = 0; BASETYPE basetype = TYPE_UNDEFINED ; TYPEDEF type, typeaux; segment * segm = NULL ; PROCDEF * proc = NULL; /* Backup vars */ BASETYPE basetypeb = TYPE_UNDEFINED ; TYPEDEF typeb; segment * segmb = NULL ; /* Initialize some stuffs */ type = typedef_new( TYPE_UNDEFINED ) ; typeb = typedef_new( TYPE_UNDEFINED ) ; for ( ;; ) { if ( n->reserved == n->count ) varspace_alloc( n, 16 ) ; if ( alignment && ( n->size % alignment ) > 0 ) { int extra = alignment - ( n->size % alignment ); if ( n->reserved <= n->count + extra ) { varspace_alloc( n, extra + 16 ); } if ( data->reserved <= data->current + extra ) { segment_alloc( data, extra + 16 ); } n->size += extra; data->current += extra; } token_next() ; /* Se salta comas y puntos y coma */ if ( token.type == NOTOKEN ) break ; if ( token.type != IDENTIFIER ) compile_error( MSG_INCOMP_TYPE ) ; if ( token.code == identifier_comma ) { basetype = basetypeb ; type = typeb ; segm = segmb ; continue ; } else if ( token.code == identifier_semicolon ) { basetype = TYPE_UNDEFINED ; set_type( &type, TYPE_UNDEFINED ) ; basetypeb = basetype ; typeb = type ; segm = NULL ; proc = NULL ; continue ; } else if ( token.code == identifier_end ) { break ; } /* "Unsigned" */ if ( token.code == identifier_unsigned ) { unsigned_prefix = 1; token_next(); } else if ( token.code == identifier_signed ) { signed_prefix = 1; token_next(); } /* Tipos de datos básicos */ if ( token.code == identifier_dword ) { basetype = signed_prefix ? TYPE_INT : TYPE_DWORD ; signed_prefix = unsigned_prefix = 0; token_next() ; } else if ( token.code == identifier_word ) { basetype = signed_prefix ? TYPE_SHORT : TYPE_WORD ; signed_prefix = unsigned_prefix = 0; token_next() ; } else if ( token.code == identifier_byte ) { basetype = signed_prefix ? TYPE_SBYTE : TYPE_BYTE ; signed_prefix = unsigned_prefix = 0; token_next() ; } else if ( token.code == identifier_int ) { basetype = unsigned_prefix ? TYPE_DWORD : TYPE_INT; signed_prefix = unsigned_prefix = 0; token_next() ; } else if ( token.code == identifier_short ) { basetype = unsigned_prefix ? TYPE_WORD : TYPE_SHORT; signed_prefix = unsigned_prefix = 0; token_next() ; } else if ( token.code == identifier_char ) { basetype = TYPE_CHAR ; token_next() ; } else if ( token.code == identifier_float ) { basetype = TYPE_FLOAT ; token_next() ; } else if ( token.code == identifier_string ) { basetype = TYPE_STRING ; token_next() ; } else { if ( !proc && ( proc = procdef_search( token.code ) ) ) /* Variables tipo proceso, Splinter */ { basetype = TYPE_INT ; token_next(); } else { if ( token.type == IDENTIFIER && token.code >= reserved_words && !segment_by_name( token.code ) ) { int code = token.code; token_next(); if ( token.type == IDENTIFIER && token.code >= reserved_words ) { proc = procdef_new( procdef_getid(), code ); basetype = TYPE_INT ; } else { token_back(); } } } } if ( signed_prefix || unsigned_prefix ) compile_error( MSG_INVALID_TYPE ); if ( basetype != TYPE_STRUCT ) type = typedef_new( basetype ) ; if ( basetype == TYPE_UNDEFINED ) type = typedef_new( TYPE_INT ) ; /* Tipos de datos definidos por el usuario */ if ( basetype != TYPE_STRUCT && ( segm = segment_by_name( token.code ) ) ) { basetype = TYPE_STRUCT ; type = * typedef_by_name( token.code ) ; token_next() ; } if ( token.type == IDENTIFIER && token.code == identifier_struct ) { type.chunk[0].type = TYPE_STRUCT ; type.chunk[0].count = 1 ; type.depth = 1 ; token_next() ; segm = 0 ; } basetypeb = basetype ; typeb = type ; segmb = segm ; /* Tipos de datos derivados */ while ( token.type == IDENTIFIER && ( token.code == identifier_pointer || token.code == identifier_multiply ) ) { type = typedef_enlarge( type ) ; type.chunk[0].type = TYPE_POINTER ; basetype = TYPE_POINTER ; segm = NULL ; token_next() ; } /* Nombre del dato */ if ( token.type != IDENTIFIER ) compile_error( MSG_IDENTIFIER_EXP ) ; if ( token.code < reserved_words ) { if ( proc ) compile_error( MSG_VARIABLE_ERROR ); token_back() ; break ; } if (( var = varspace_search( n, token.code ) ) ) { if ( duplicateignore ) { int skip_all_until_semicolon = 0; int skip_equal = 0; if ( debug ) compile_warning( 0, MSG_VARIABLE_REDECLARE ) ; for ( ;; ) { token_next() ; /* Se salta todo hasta el puntos y coma o en el end o la coma si no hay asignacion */ if ( token.type == NOTOKEN ) break ; if ( token.type == IDENTIFIER ) { if ( !skip_equal && token.code == identifier_equal ) { skip_all_until_semicolon = 1; continue; } if ( !skip_all_until_semicolon && token.code == identifier_comma ) break; if ( token.code == identifier_semicolon ) break ; if ( token.code == identifier_end ) break ; } } token_back(); continue; } else compile_error( MSG_VARIABLE_REDECLARE ) ; } /* (2006/11/19 19:34 GMT-03:00, Splinter - jj_arg@yahoo.com) */ if ( collision ) for ( i = 0; collision[i];i++ ) if ( varspace_search( collision[i], token.code ) ) compile_error( MSG_VARIABLE_REDECLARE ) ; /* (2006/11/19 19:34 GMT-03:00, Splinter - jj_arg@yahoo.com) */ if ( constants_search( token.code ) ) compile_error( MSG_CONSTANT_REDECLARED_AS_VARIABLE ) ; code = token.code ; n->vars[n->count].code = token.code ; n->vars[n->count].offset = data->current; /* Non-additive STRUCT; use zero-based member offsets */ if ( !additive ) n->vars[n->count].offset -= base_offset; token_next() ; /* Compila una estructura no predefinida */ if ( !segm && typedef_is_struct( type ) ) { VARSPACE * members ; type.chunk[0].count = 1 ; count = 1 ; while ( token.type == IDENTIFIER && token.code == identifier_leftb ) { res = compile_expresion( 1, 0, 0, TYPE_INT ) ; if ( !typedef_is_integer( res.type ) ) compile_error( MSG_INTEGER_REQUIRED ) ; count *= res.value + 1 ; type = typedef_enlarge( type ) ; type.chunk[0].type = TYPE_ARRAY ; type.chunk[0].count = res.value + 1 ; token_next() ; if ( token.type != IDENTIFIER || token.code != identifier_rightb ) compile_error( MSG_EXPECTED, "]" ) ; token_next() ; } token_back() ; /* Da la vuelta a los índices [10][5] -> [5][10] */ for ( i = 0 ; i < type.depth ; i++ ) if ( type.chunk[i].type != TYPE_ARRAY ) break ; i--; for ( j = 0 ; j <= i ; j++ ) typeaux.chunk[ j ] = type.chunk[ i - j ]; for ( j = 0 ; j <= i ; j++ ) type.chunk[ j ] = typeaux.chunk[ j ]; members = ( VARSPACE * )calloc( 1, sizeof( VARSPACE ) ) ; if ( !members ) { fprintf( stdout, "compile_varspace: out of memory\n" ) ; exit( 1 ) ; } varspace_init( members ) ; size = compile_varspace( members, data, 0, count, 0, NULL, 0, duplicateignore ) ; type.varspace = members ; token_next() ; if ( token.type == IDENTIFIER && token.code == identifier_equal ) { i = data->current ; data->current = n->vars[n->count].offset ; compile_struct_data( members, data, count, 0 ); data->current = i ; } else { token_back() ; } for ( i = 0 ; i < members->stringvar_count ; i++ ) varspace_varstring( n, members->stringvars[i] ) ; n->size += typedef_size( type ) ; n->vars[n->count].type = type ; n->count++ ; continue ; /* No ; */ } else if ( token.type == IDENTIFIER && token.code == identifier_leftb ) /* Compila un array */ { total_count = 1 ; while ( token.type == IDENTIFIER && token.code == identifier_leftb ) { if ( type.depth == MAX_TYPECHUNKS ) compile_error( MSG_TOO_MANY_AL ) ; type = typedef_enlarge( type ) ; type.chunk[0].type = TYPE_ARRAY ; token_next() ; if ( token.type == IDENTIFIER && token.code == identifier_rightb ) { type.chunk[0].count = 0 ; if ( total_count != 1 ) compile_error( MSG_VTA ) ; total_count = 0 ; last_count = 0 ; } else { token_back() ; res = compile_expresion( 1, 0, 0, TYPE_DWORD ) ; if ( !total_count ) compile_error( MSG_VTA ) ; total_count *= res.value + 1 ; last_count = res.value + 1 ; type.chunk[0].count = res.value + 1 ; token_next() ; if ( token.type != IDENTIFIER || token.code != identifier_rightb ) compile_error( MSG_EXPECTED, "]" ) ; } token_next() ; } /* Da la vuelta a los índices [10][5] -> [5][10] */ for ( i = 0 ; i < type.depth ; i++ ) if ( type.chunk[i].type != TYPE_ARRAY ) break ; i--; for ( j = 0 ; j <= i ; j++ ) typeaux.chunk[ j ] = type.chunk[ i - j ]; for ( j = 0 ; j <= i ; j++ ) type.chunk[ j ] = typeaux.chunk[ j ]; if ( segm && token.type == IDENTIFIER && token.code == identifier_equal ) { for ( i = 0 ; i < total_count ; i++ ) segment_add_from( data, segm ) ; i = data->current ; data->current = n->vars[n->count].offset ; compile_struct_data( type.varspace, data, typedef_count( type ), 0 ); if ( !type.chunk[0].count ) type.chunk[0].count = ( data->current - i ) / typedef_size( typeb ); else data->current = i; /* Solo si ya habia sido alocada */ } else if ( token.type == IDENTIFIER && token.code == identifier_equal ) { /* if (basetype == TYPE_UNDEFINED) basetype = TYPE_INT; */ i = compile_array_data( n, data, total_count, last_count, &basetype ) ; assert( basetype != TYPE_UNDEFINED ) ; set_type( &type, basetype ) ; if ( total_count == 0 ) { type.chunk[0].count = i; } else if ( i < total_count ) { for ( ; i < total_count; i++ ) { if ( basetype == TYPE_STRING ) varspace_varstring( n, data->current ) ; segment_add_as( data, 0, basetype ) ; } } } else if ( segm ) { int string_offset = 0, j; if ( total_count == 0 ) compile_error( MSG_EXPECTED, "=" ) ; for ( i = 0; i < total_count; i++ ) { segment_add_from( data, segm ) ; for ( j = 0; j < type.varspace->stringvar_count; j++ ) varspace_varstring( n, type.varspace->stringvars[j] + string_offset ); string_offset += type.varspace->size; } token_back() ; } else { if ( basetype == TYPE_UNDEFINED ) { basetype = TYPE_INT; set_type( &type, basetype ) ; } if ( type.chunk[0].count == 0 ) compile_error( MSG_EXPECTED, "=" ) ; for ( i = 0; i < total_count; i++ ) { if ( basetype == TYPE_STRING ) varspace_varstring( n, data->current ) ; segment_add_as( data, 0, basetype ) ; } token_back() ; } } else if ( segm && token.type == IDENTIFIER && token.code == identifier_equal ) /* Compila una asignación de valores por defecto */ { segment_add_from( data, segm ) ; i = data->current ; data->current = n->vars[n->count].offset ; if ( !additive ) data->current += base_offset; compile_struct_data( type.varspace, data, 1, 0 ); data->current = i ; } else if ( token.type == IDENTIFIER && token.code == identifier_equal ) { res = compile_expresion( 1, 0, 0, basetype ) ; if ( basetype == TYPE_UNDEFINED ) { basetype = typedef_base( res.type ) ; if ( basetype == TYPE_UNDEFINED ) compile_error( MSG_INCOMP_TYPE ) ; set_type( &type, basetype ) ; } if ( basetype == TYPE_FLOAT ) { segment_add_as( data, *( int * )&res.fvalue, basetype ) ; } else { if ( basetype == TYPE_STRING ) varspace_varstring( n, data->current ) ; segment_add_as( data, res.value, basetype ) ; } } else if ( !segm ) /* Asigna valores por defecto (0) */ { if ( basetype == TYPE_UNDEFINED ) { basetype = TYPE_INT; set_type( &type, basetype ) ; } if ( basetype == TYPE_STRING ) varspace_varstring( n, data->current ) ; segment_add_as( data, 0, basetype ) ; token_back() ; } else { if ( typedef_is_struct( type ) ) for ( i = 0 ; i < type.varspace->stringvar_count ; i++ ) varspace_varstring( n, type.varspace->stringvars[i] + n->size ); segment_add_from( data, segm ) ; token_back() ; } n->size += typedef_size( type ) ; n->vars[n->count].type = type ; /* Variables tipo proceso, asigno varspace al tipo. Splinter */ if ( proc ) n->vars[n->count].type.varspace = proc->pubvars; n->count++ ; token_next() ; if ( token.type == IDENTIFIER && token.code == identifier_comma ) { token_back() ; continue ; } if ( token.type == IDENTIFIER && token.code == identifier_semicolon ) { token_back() ; continue ; } compile_error( MSG_EXPECTED, ";" ) ; token_back() ; break ; } if ( padding && ( data->current % padding ) > 0 ) { padding -= data->current % padding; data->current += padding; n->size += padding; if ( data->reserved <= data->current ) segment_alloc( data, data->reserved - data->current + 32 ); } n->last_offset = data->current ; total_length = data->current - base_offset ; /* n->size *= copies ; */ while ( copies-- > 1 ) { int i ; for ( i = 0 ; i < n->stringvar_count ; i++ ) { if ( n->stringvars[i] >= base_offset && n->stringvars[i] < base_offset + total_length ) { varspace_varstring( n, n->stringvars[i] - base_offset + data->current ) ; } } segment_copy( data, base_offset, total_length ) ; base_offset += total_length ; } return total_length ; }