// This file is part of OpenCV project. // It is subject to the license terms in the LICENSE file found in the top-level directory // of this distribution and at http://opencv.org/license.html #include "precomp.hpp" #include "persistence.hpp" /****************************************************************************************\ * JSON Parser * \****************************************************************************************/ static char* icvJSONSkipSpaces( CvFileStorage* fs, char* ptr ) { bool is_eof = false; bool is_completed = false; while ( is_eof == false && is_completed == false ) { switch ( *ptr ) { /* comment */ case '/' : { ptr++; if ( *ptr == '\0' ) { ptr = icvGets( fs, fs->buffer_start, static_cast(fs->buffer_end - fs->buffer_start) ); if ( !ptr ) { is_eof = true; break; } } if ( *ptr == '/' ) { while ( *ptr != '\n' && *ptr != '\r' ) { if ( *ptr == '\0' ) { ptr = icvGets( fs, fs->buffer_start, static_cast(fs->buffer_end - fs->buffer_start) ); if ( !ptr ) { is_eof = true; break; } } else { ptr++; } } } else if ( *ptr == '*' ) { ptr++; for (;;) { if ( *ptr == '\0' ) { ptr = icvGets( fs, fs->buffer_start, static_cast(fs->buffer_end - fs->buffer_start) ); if ( !ptr ) { is_eof = true; break; } } else if ( *ptr == '*' ) { ptr++; if ( *ptr == '\0' ) { ptr = icvGets( fs, fs->buffer_start, static_cast(fs->buffer_end - fs->buffer_start) ); if ( !ptr ) { is_eof = true; break; } } if ( *ptr == '/' ) { ptr++; break; } } else { ptr++; } } } else { CV_PARSE_ERROR( "Not supported escape character" ); } } break; /* whitespace */ case '\t': case ' ' : { ptr++; } break; /* newline || end mark */ case '\0': case '\n': case '\r': { ptr = icvGets( fs, fs->buffer_start, static_cast(fs->buffer_end - fs->buffer_start) ); if ( !ptr ) { is_eof = true; break; } } break; /* other character */ default: { if ( !cv_isprint(*ptr) ) CV_PARSE_ERROR( "Invalid character in the stream" ); is_completed = true; } break; } } if ( is_eof ) { ptr = fs->buffer_start; *ptr = '\0'; fs->dummy_eof = 1; } else if ( !is_completed ) { /* should not be executed */ ptr = 0; fs->dummy_eof = 1; CV_PARSE_ERROR( "Abort at parse time" ); } return ptr; } static char* icvJSONParseKey( CvFileStorage* fs, char* ptr, CvFileNode* map, CvFileNode** value_placeholder ) { if( *ptr != '"' ) CV_PARSE_ERROR( "Key must start with \'\"\'" ); char * beg = ptr + 1; char * end = beg; do { ++ptr; CV_PERSISTENCE_CHECK_END_OF_BUFFER_BUG(); } while( cv_isprint(*ptr) && *ptr != '"' ); if( *ptr != '"' ) CV_PARSE_ERROR( "Key must end with \'\"\'" ); end = ptr; ptr++; ptr = icvJSONSkipSpaces( fs, ptr ); if ( ptr == 0 || fs->dummy_eof ) return 0; if( *ptr != ':' ) CV_PARSE_ERROR( "Missing \':\' between key and value" ); /* [beg, end) */ if( end <= beg ) CV_PARSE_ERROR( "Key is empty" ); if ( end - beg == 7u && memcmp(beg, "type_id", 7u) == 0 ) { *value_placeholder = 0; } else { CvStringHashNode* str_hash_node = cvGetHashedKey( fs, beg, static_cast(end - beg), 1 ); *value_placeholder = cvGetFileNode( fs, map, str_hash_node, 1 ); } ptr++; return ptr; } static char* icvJSONParseValue( CvFileStorage* fs, char* ptr, CvFileNode* node ) { ptr = icvJSONSkipSpaces( fs, ptr ); if ( ptr == 0 || fs->dummy_eof ) CV_PARSE_ERROR( "Unexpected End-Of-File" ); memset( node, 0, sizeof(*node) ); if ( *ptr == '"' ) { /* must be string or Base64 string */ ptr++; char * beg = ptr; size_t len = 0u; for ( ; (cv_isalnum(*ptr) || *ptr == '$' ) && len <= 9u; ptr++ ) len++; if ( len >= 8u && memcmp( beg, "$base64$", 8u ) == 0 ) { /**************** Base64 string ****************/ ptr = beg += 8; std::string base64_buffer; base64_buffer.reserve( PARSER_BASE64_BUFFER_SIZE ); bool is_matching = false; while ( !is_matching ) { switch ( *ptr ) { case '\0': { base64_buffer.append( beg, ptr ); ptr = icvGets( fs, fs->buffer_start, static_cast(fs->buffer_end - fs->buffer_start) ); if ( !ptr ) CV_PARSE_ERROR( "'\"' - right-quote of string is missing" ); beg = ptr; break; } case '\"': { base64_buffer.append( beg, ptr ); beg = ptr; is_matching = true; break; } case '\n': case '\r': { CV_PARSE_ERROR( "'\"' - right-quote of string is missing" ); break; } default: { ptr++; break; } } } if ( *ptr != '\"' ) CV_PARSE_ERROR( "'\"' - right-quote of string is missing" ); else ptr++; if ( base64_buffer.size() >= base64::ENCODED_HEADER_SIZE ) { const char * base64_beg = base64_buffer.data(); const char * base64_end = base64_beg + base64_buffer.size(); /* get dt from header */ std::string dt; { std::vector header(base64::HEADER_SIZE + 1, ' '); base64::base64_decode(base64_beg, header.data(), 0U, base64::ENCODED_HEADER_SIZE); if ( !base64::read_base64_header(header, dt) || dt.empty() ) CV_PARSE_ERROR("Invalid `dt` in Base64 header"); } /* set base64_beg to beginning of base64 data */ base64_beg = &base64_buffer.at( base64::ENCODED_HEADER_SIZE ); if ( base64_buffer.size() > base64::ENCODED_HEADER_SIZE ) { if ( !base64::base64_valid( base64_beg, 0U, base64_end - base64_beg ) ) CV_PARSE_ERROR( "Invalid Base64 data." ); /* buffer for decoded data(exclude header) */ std::vector binary_buffer( base64::base64_decode_buffer_size(base64_end - base64_beg) ); int total_byte_size = static_cast( base64::base64_decode_buffer_size( base64_end - base64_beg, base64_beg, false ) ); { base64::Base64ContextParser parser(binary_buffer.data(), binary_buffer.size() ); const uchar * binary_beg = reinterpret_cast( base64_beg ); const uchar * binary_end = binary_beg + (base64_end - base64_beg); parser.read( binary_beg, binary_end ); parser.flush(); } /* save as CvSeq */ int elem_size = ::icvCalcStructSize(dt.c_str(), 0); if (total_byte_size % elem_size != 0) CV_PARSE_ERROR("Byte size not match elememt size"); int elem_cnt = total_byte_size / elem_size; /* after icvFSCreateCollection, node->tag == struct_flags */ icvFSCreateCollection(fs, CV_NODE_FLOW | CV_NODE_SEQ, node); base64::make_seq(binary_buffer.data(), elem_cnt, dt.c_str(), *node->data.seq); } else { /* empty */ icvFSCreateCollection(fs, CV_NODE_FLOW | CV_NODE_SEQ, node); } } else if ( base64_buffer.empty() ) { /* empty */ icvFSCreateCollection(fs, CV_NODE_FLOW | CV_NODE_SEQ, node); } else { CV_PARSE_ERROR("Unrecognized Base64 header"); } } else { /**************** normal string ****************/ std::string string_buffer; string_buffer.reserve( PARSER_BASE64_BUFFER_SIZE ); ptr = beg; bool is_matching = false; while ( !is_matching ) { switch ( *ptr ) { case '\\': { string_buffer.append( beg, ptr ); ptr++; switch ( *ptr ) { case '\\': case '\"': case '\'': { string_buffer.append( 1u, *ptr ); break; } case 'n' : { string_buffer.append( 1u, '\n' ); break; } case 'r' : { string_buffer.append( 1u, '\r' ); break; } case 't' : { string_buffer.append( 1u, '\t' ); break; } case 'b' : { string_buffer.append( 1u, '\b' ); break; } case 'f' : { string_buffer.append( 1u, '\f' ); break; } case 'u' : { CV_PARSE_ERROR( "'\\uXXXX' currently not supported" ); } default : { CV_PARSE_ERROR( "Invalid escape character" ); } break; } ptr++; beg = ptr; break; } case '\0': { string_buffer.append( beg, ptr ); ptr = icvGets( fs, fs->buffer_start, static_cast(fs->buffer_end - fs->buffer_start) ); if ( !ptr ) CV_PARSE_ERROR( "'\"' - right-quote of string is missing" ); beg = ptr; break; } case '\"': { string_buffer.append( beg, ptr ); beg = ptr; is_matching = true; break; } case '\n': case '\r': { CV_PARSE_ERROR( "'\"' - right-quote of string is missing" ); break; } default: { ptr++; break; } } } if ( *ptr != '\"' ) CV_PARSE_ERROR( "'\"' - right-quote of string is missing" ); else ptr++; node->data.str = cvMemStorageAllocString ( fs->memstorage, string_buffer.c_str(), static_cast(string_buffer.size()) ); node->tag = CV_NODE_STRING; } } else if ( cv_isdigit(*ptr) || *ptr == '-' || *ptr == '+' || *ptr == '.' ) { /**************** number ****************/ char * beg = ptr; if ( *ptr == '+' || *ptr == '-' ) { ptr++; CV_PERSISTENCE_CHECK_END_OF_BUFFER_BUG(); } while( cv_isdigit(*ptr) ) { ptr++; CV_PERSISTENCE_CHECK_END_OF_BUFFER_BUG(); } if (*ptr == '.' || *ptr == 'e') { node->data.f = icv_strtod( fs, beg, &ptr ); CV_PERSISTENCE_CHECK_END_OF_BUFFER_BUG(); node->tag = CV_NODE_REAL; } else { node->data.i = static_cast(strtol( beg, &ptr, 0 )); CV_PERSISTENCE_CHECK_END_OF_BUFFER_BUG(); node->tag = CV_NODE_INT; } if ( beg >= ptr ) CV_PARSE_ERROR( "Invalid numeric value (inconsistent explicit type specification?)" ); } else { /**************** other data ****************/ const char * beg = ptr; size_t len = 0u; for ( ; cv_isalpha(*ptr) && len <= 6u; ) { len++; ptr++; CV_PERSISTENCE_CHECK_END_OF_BUFFER_BUG(); } if ( len >= 4u && memcmp( beg, "null", 4u ) == 0 ) { CV_PARSE_ERROR( "Value 'null' is not supported by this parser" ); } else if ( len >= 4u && memcmp( beg, "true", 4u ) == 0 ) { node->data.i = 1; node->tag = CV_NODE_INT; } else if ( len >= 5u && memcmp( beg, "false", 5u ) == 0 ) { node->data.i = 0; node->tag = CV_NODE_INT; } else { CV_PARSE_ERROR( "Unrecognized value" ); } } return ptr; } static char* icvJSONParseSeq( CvFileStorage* fs, char* ptr, CvFileNode* node ); static char* icvJSONParseMap( CvFileStorage* fs, char* ptr, CvFileNode* node ); static char* icvJSONParseSeq( CvFileStorage* fs, char* ptr, CvFileNode* node ) { if (!ptr) CV_PARSE_ERROR( "ptr is NULL" ); if ( *ptr != '[' ) CV_PARSE_ERROR( "'[' - left-brace of seq is missing" ); else ptr++; memset( node, 0, sizeof(*node) ); icvFSCreateCollection( fs, CV_NODE_SEQ, node ); for (;;) { ptr = icvJSONSkipSpaces( fs, ptr ); if ( ptr == 0 || fs->dummy_eof ) break; if ( *ptr != ']' ) { CvFileNode* child = (CvFileNode*)cvSeqPush( node->data.seq, 0 ); if ( *ptr == '[' ) ptr = icvJSONParseSeq( fs, ptr, child ); else if ( *ptr == '{' ) ptr = icvJSONParseMap( fs, ptr, child ); else ptr = icvJSONParseValue( fs, ptr, child ); } ptr = icvJSONSkipSpaces( fs, ptr ); if ( ptr == 0 || fs->dummy_eof ) break; if ( *ptr == ',' ) ptr++; else if ( *ptr == ']' ) break; else CV_PARSE_ERROR( "Unexpected character" ); } if (!ptr) CV_PARSE_ERROR("ptr is NULL"); if ( *ptr != ']' ) CV_PARSE_ERROR( "']' - right-brace of seq is missing" ); else ptr++; return ptr; } static char* icvJSONParseMap( CvFileStorage* fs, char* ptr, CvFileNode* node ) { if (!ptr) CV_PARSE_ERROR("ptr is NULL"); if ( *ptr != '{' ) CV_PARSE_ERROR( "'{' - left-brace of map is missing" ); else ptr++; memset( node, 0, sizeof(*node) ); icvFSCreateCollection( fs, CV_NODE_MAP, node ); for ( ;; ) { ptr = icvJSONSkipSpaces( fs, ptr ); if ( ptr == 0 || fs->dummy_eof ) break; if ( *ptr == '"' ) { CvFileNode* child = 0; ptr = icvJSONParseKey( fs, ptr, node, &child ); if ( ptr == 0 || fs->dummy_eof ) break; ptr = icvJSONSkipSpaces( fs, ptr ); if ( ptr == 0 || fs->dummy_eof ) break; if ( child == 0 ) { /* type_id */ CvFileNode tmp; ptr = icvJSONParseValue( fs, ptr, &tmp ); if ( CV_NODE_IS_STRING(tmp.tag) ) { node->info = cvFindType( tmp.data.str.ptr ); if ( node->info ) node->tag |= CV_NODE_USER; // delete tmp.data.str } else { CV_PARSE_ERROR( "\"type_id\" should be of type string" ); } } else { /* normal */ if ( *ptr == '[' ) ptr = icvJSONParseSeq( fs, ptr, child ); else if ( *ptr == '{' ) ptr = icvJSONParseMap( fs, ptr, child ); else ptr = icvJSONParseValue( fs, ptr, child ); child->tag |= CV_NODE_NAMED; } } ptr = icvJSONSkipSpaces( fs, ptr ); if ( ptr == 0 || fs->dummy_eof ) break; if ( *ptr == ',' ) ptr++; else if ( *ptr == '}' ) break; else CV_PARSE_ERROR( "Unexpected character" ); } if (!ptr) CV_PARSE_ERROR("ptr is NULL"); if ( *ptr != '}' ) CV_PARSE_ERROR( "'}' - right-brace of map is missing" ); else ptr++; return ptr; } void icvJSONParse( CvFileStorage* fs ) { char* ptr = fs->buffer_start; ptr = icvJSONSkipSpaces( fs, ptr ); if ( ptr == 0 || fs->dummy_eof ) return; if ( *ptr == '{' ) { CvFileNode* root_node = (CvFileNode*)cvSeqPush( fs->roots, 0 ); ptr = icvJSONParseMap( fs, ptr, root_node ); } else if ( *ptr == '[' ) { CvFileNode* root_node = (CvFileNode*)cvSeqPush( fs->roots, 0 ); ptr = icvJSONParseSeq( fs, ptr, root_node ); } else { CV_PARSE_ERROR( "left-brace of top level is missing" ); } if ( fs->dummy_eof != 0 ) CV_PARSE_ERROR( "Unexpected End-Of-File" ); } /****************************************************************************************\ * JSON Emitter * \****************************************************************************************/ void icvJSONWrite( CvFileStorage* fs, const char* key, const char* data ) { /* check write_struct */ check_if_write_struct_is_delayed( fs ); if ( fs->state_of_writing_base64 == base64::fs::Uncertain ) { switch_to_Base64_state( fs, base64::fs::NotUse ); } else if ( fs->state_of_writing_base64 == base64::fs::InUse ) { CV_Error( CV_StsError, "At present, output Base64 data only." ); } /* check parameters */ size_t key_len = 0u; if( key && *key == '\0' ) key = 0; if ( key ) { key_len = strlen(key); if ( key_len == 0u ) CV_Error( CV_StsBadArg, "The key is an empty" ); else if ( static_cast(key_len) > CV_FS_MAX_LEN ) CV_Error( CV_StsBadArg, "The key is too long" ); } size_t data_len = 0u; if ( data ) data_len = strlen(data); int struct_flags = fs->struct_flags; if( CV_NODE_IS_COLLECTION(struct_flags) ) { if ( (CV_NODE_IS_MAP(struct_flags) ^ (key != 0)) ) CV_Error( CV_StsBadArg, "An attempt to add element without a key to a map, " "or add element with key to sequence" ); } else { fs->is_first = 0; struct_flags = CV_NODE_EMPTY | (key ? CV_NODE_MAP : CV_NODE_SEQ); } /* start to write */ char* ptr = 0; if( CV_NODE_IS_FLOW(struct_flags) ) { int new_offset; ptr = fs->buffer; if( !CV_NODE_IS_EMPTY(struct_flags) ) *ptr++ = ','; new_offset = static_cast(ptr - fs->buffer_start + key_len + data_len); if( new_offset > fs->wrap_margin && new_offset - fs->struct_indent > 10 ) { fs->buffer = ptr; ptr = icvFSFlush(fs); } else *ptr++ = ' '; } else { if ( !CV_NODE_IS_EMPTY(struct_flags) ) { ptr = fs->buffer; *ptr++ = ','; *ptr++ = '\n'; *ptr++ = '\0'; ::icvPuts( fs, fs->buffer_start ); ptr = fs->buffer = fs->buffer_start; } ptr = icvFSFlush(fs); } if( key ) { if( !cv_isalpha(key[0]) && key[0] != '_' ) CV_Error( CV_StsBadArg, "Key must start with a letter or _" ); ptr = icvFSResizeWriteBuffer( fs, ptr, static_cast(key_len) ); *ptr++ = '\"'; for( size_t i = 0u; i < key_len; i++ ) { char c = key[i]; ptr[i] = c; if( !cv_isalnum(c) && c != '-' && c != '_' && c != ' ' ) CV_Error( CV_StsBadArg, "Key names may only contain alphanumeric characters [a-zA-Z0-9], '-', '_' and ' '" ); } ptr += key_len; *ptr++ = '\"'; *ptr++ = ':'; *ptr++ = ' '; } if( data ) { ptr = icvFSResizeWriteBuffer( fs, ptr, static_cast(data_len) ); memcpy( ptr, data, data_len ); ptr += data_len; } fs->buffer = ptr; fs->struct_flags = struct_flags & ~CV_NODE_EMPTY; } void icvJSONStartWriteStruct( CvFileStorage* fs, const char* key, int struct_flags, const char* type_name) { int parent_flags; char data[CV_FS_MAX_LEN + 1024]; struct_flags = (struct_flags & (CV_NODE_TYPE_MASK|CV_NODE_FLOW)) | CV_NODE_EMPTY; if( !CV_NODE_IS_COLLECTION(struct_flags)) CV_Error( CV_StsBadArg, "Some collection type - CV_NODE_SEQ or CV_NODE_MAP, must be specified" ); if ( type_name && *type_name == '\0' ) type_name = 0; bool has_type_id = false; bool is_real_collection = true; if (type_name && memcmp(type_name, "binary", 6) == 0) { struct_flags = CV_NODE_STR; data[0] = '\0'; is_real_collection = false; } else if( type_name ) { has_type_id = true; } if ( is_real_collection ) { char c = CV_NODE_IS_MAP(struct_flags) ? '{' : '['; data[0] = c; data[1] = '\0'; } icvJSONWrite( fs, key, data ); parent_flags = fs->struct_flags; cvSeqPush( fs->write_stack, &parent_flags ); fs->struct_flags = struct_flags; fs->struct_indent += 4; if ( has_type_id ) fs->write_string( fs, "type_id", type_name, 1 ); } void icvJSONEndWriteStruct( CvFileStorage* fs ) { if( fs->write_stack->total == 0 ) CV_Error( CV_StsError, "EndWriteStruct w/o matching StartWriteStruct" ); int parent_flags = 0; int struct_flags = fs->struct_flags; cvSeqPop( fs->write_stack, &parent_flags ); fs->struct_indent -= 4; fs->struct_flags = parent_flags & ~CV_NODE_EMPTY; assert( fs->struct_indent >= 0 ); if ( CV_NODE_IS_COLLECTION(struct_flags) ) { if ( !CV_NODE_IS_FLOW(struct_flags) ) { if ( fs->buffer <= fs->buffer_start + fs->space ) { /* some bad code for base64_writer... */ *fs->buffer++ = '\n'; *fs->buffer++ = '\0'; icvPuts( fs, fs->buffer_start ); fs->buffer = fs->buffer_start; } icvFSFlush(fs); } char* ptr = fs->buffer; if( ptr > fs->buffer_start + fs->struct_indent && !CV_NODE_IS_EMPTY(struct_flags) ) *ptr++ = ' '; *ptr++ = CV_NODE_IS_MAP(struct_flags) ? '}' : ']'; fs->buffer = ptr; } } void icvJSONStartNextStream( CvFileStorage* fs ) { if( !fs->is_first ) { while( fs->write_stack->total > 0 ) icvJSONEndWriteStruct(fs); fs->struct_indent = 4; icvFSFlush(fs); fs->buffer = fs->buffer_start; } } void icvJSONWriteInt( CvFileStorage* fs, const char* key, int value ) { char buf[128]; icvJSONWrite( fs, key, icv_itoa( value, buf, 10 )); } void icvJSONWriteReal( CvFileStorage* fs, const char* key, double value ) { char buf[128]; size_t len = strlen( icvDoubleToString( buf, value ) ); if( len > 0 && buf[len-1] == '.' ) { // append zero if string ends with decimal place to match JSON standard buf[len] = '0'; buf[len+1] = '\0'; } icvJSONWrite( fs, key, buf ); } void icvJSONWriteString( CvFileStorage* fs, const char* key, const char* str, int quote) { char buf[CV_FS_MAX_LEN*4+16]; char* data = (char*)str; int i, len; if( !str ) CV_Error( CV_StsNullPtr, "Null string pointer" ); len = (int)strlen(str); if( len > CV_FS_MAX_LEN ) CV_Error( CV_StsBadArg, "The written string is too long" ); if( quote || len == 0 || str[0] != str[len-1] || (str[0] != '\"' && str[0] != '\'') ) { int need_quote = 1; data = buf; *data++ = '\"'; for( i = 0; i < len; i++ ) { char c = str[i]; switch ( c ) { case '\\': case '\"': case '\'': { *data++ = '\\'; *data++ = c; break; } case '\n': { *data++ = '\\'; *data++ = 'n'; break; } case '\r': { *data++ = '\\'; *data++ = 'r'; break; } case '\t': { *data++ = '\\'; *data++ = 't'; break; } case '\b': { *data++ = '\\'; *data++ = 'b'; break; } case '\f': { *data++ = '\\'; *data++ = 'f'; break; } default : { *data++ = c; } break; } } *data++ = '\"'; *data++ = '\0'; data = buf + !need_quote; } icvJSONWrite( fs, key, data ); } void icvJSONWriteComment( CvFileStorage* fs, const char* comment, int eol_comment ) { if( !comment ) CV_Error( CV_StsNullPtr, "Null comment" ); int len = static_cast(strlen(comment)); char* ptr = fs->buffer; const char* eol = strchr(comment, '\n'); bool multiline = eol != 0; if( !eol_comment || multiline || fs->buffer_end - ptr < len || ptr == fs->buffer_start ) ptr = icvFSFlush( fs ); else *ptr++ = ' '; while( comment ) { *ptr++ = '/'; *ptr++ = '/'; *ptr++ = ' '; if( eol ) { ptr = icvFSResizeWriteBuffer( fs, ptr, (int)(eol - comment) + 1 ); memcpy( ptr, comment, eol - comment + 1 ); fs->buffer = ptr + (eol - comment); comment = eol + 1; eol = strchr( comment, '\n' ); } else { len = (int)strlen(comment); ptr = icvFSResizeWriteBuffer( fs, ptr, len ); memcpy( ptr, comment, len ); fs->buffer = ptr + len; comment = 0; } ptr = icvFSFlush( fs ); } }