From caf5fc8d8d7899a6beec1fd52b5a6d6c98021f9e Mon Sep 17 00:00:00 2001 From: Dick Hollenbeck Date: Mon, 1 Jul 2013 01:47:36 -0500 Subject: [PATCH] DSNLEXER::NextTok() organizes the specctraMode code better, into one if block mostly. This keeps it out of the KiCad mode path, making that leaner and less confusing about what is supported in KiCad mode. Within KiCad mode, treat quoted vs. non-quoted tokens as the two general categories, with non-quoted having sub-categories. Eliminate an unimplemented, unused function declaration in DSNLEXER. Improve the output formatting of THROW_PARSE_ERROR(). --- common/dsnlexer.cpp | 446 ++++++++++++++++++++++---------------------- include/dsnlexer.h | 18 +- include/richio.h | 2 +- pcbnew/files.cpp | 7 +- 4 files changed, 229 insertions(+), 244 deletions(-) diff --git a/common/dsnlexer.cpp b/common/dsnlexer.cpp index 27ee92a6c5..c44ca0ca09 100644 --- a/common/dsnlexer.cpp +++ b/common/dsnlexer.cpp @@ -196,19 +196,18 @@ int DSNLEXER::findToken( const std::string& tok ) if( findings ) return findings->token; else - return -1; + return DSN_SYMBOL; // not a keyword, some arbitrary symbol. } #else -int DSNLEXER::findToken( const std::string& tok ) +inline int DSNLEXER::findToken( const std::string& tok ) { - KEYWORD_MAP::const_iterator it = keyword_hash.find( tok.c_str() ); - if( it == keyword_hash.end() ) - return -1; + if( it != keyword_hash.end() ) + return it->second; - return it->second; + return DSN_SYMBOL; // not a keyword, some arbitrary symbol. } #endif @@ -300,25 +299,24 @@ bool DSNLEXER::IsSymbol( int aTok ) void DSNLEXER::Expecting( int aTok ) throw( IO_ERROR ) { - wxString errText; - errText.Printf( _("Expecting '%s'"), GetChars( GetTokenString( aTok ) ) ); + wxString errText = wxString::Format( + _("Expecting '%s'"), GetChars( GetTokenString( aTok ) ) ); THROW_PARSE_ERROR( errText, CurSource(), CurLine(), CurLineNumber(), CurOffset() ); } void DSNLEXER::Expecting( const char* text ) throw( IO_ERROR ) { - wxString errText; - errText.Printf( _("Expecting '%s'"), - GetChars( wxString::FromUTF8( text ) ) ); + wxString errText = wxString::Format( + _("Expecting '%s'"), GetChars( wxString::FromUTF8( text ) ) ); THROW_PARSE_ERROR( errText, CurSource(), CurLine(), CurLineNumber(), CurOffset() ); } void DSNLEXER::Unexpected( int aTok ) throw( IO_ERROR ) { - wxString errText; - errText.Printf( _("Unexpected '%s'"), GetChars( GetTokenString( aTok ) ) ); + wxString errText = wxString::Format( + _("Unexpected '%s'"), GetChars( GetTokenString( aTok ) ) ); THROW_PARSE_ERROR( errText, CurSource(), CurLine(), CurLineNumber(), CurOffset() ); } @@ -333,9 +331,8 @@ void DSNLEXER::Duplicate( int aTok ) throw( IO_ERROR ) void DSNLEXER::Unexpected( const char* text ) throw( IO_ERROR ) { - wxString errText; - errText.Printf( _("Unexpected '%s'"), - GetChars( wxString::FromUTF8( text ) ) ); + wxString errText = wxString::Format( + _("Unexpected '%s'"), GetChars( wxString::FromUTF8( text ) ) ); THROW_PARSE_ERROR( errText, CurSource(), CurLine(), CurLineNumber(), CurOffset() ); } @@ -433,11 +430,11 @@ inline bool isSep( char cc ) * at the first non-number character, even if it is not whitespace. * * @param cp is the start of the current token. - * @param limit is the end of the current line of text. - * @return const char* - if initial text was a number, then pointer to the text - * after this number, else NULL if initial text was not a number. + * @param limit is the end of the current token. + * + * @return bool - true if input token is a number, else false. */ -static const char* isNumber( const char* cp, const char* limit ) +static bool isNumber( const char* cp, const char* limit ) { // regex for a float: "^[-+]?[0-9]*\\.?[0-9]+([eE][-+]?[0-9]+)?" i.e. any number, // code traversal manually here: @@ -483,7 +480,7 @@ static const char* isNumber( const char* cp, const char* limit ) } } - return sawNumber ? cp : NULL; + return sawNumber; } @@ -494,73 +491,179 @@ int DSNLEXER::NextTok() throw( IO_ERROR ) prevTok = curTok; - if( curTok != DSN_EOF ) + if( curTok == DSN_EOF ) + goto exit; + + if( cur >= limit ) { - if( cur >= limit ) - { L_read: - // blank lines are returned as "\n" and will have a len of 1. - // EOF will have a len of 0 and so is detectable. - int len = readLine(); - if( len == 0 ) - { - cur = start; // after readLine(), since start can change, set cur offset to start - curTok = DSN_EOF; - goto exit; - } - - cur = start; // after readLine() since start can change. - - // skip leading whitespace - while( cur= limit ) - goto L_read; - - if( *cur == '(' ) - { - curText = *cur; - curTok = DSN_LEFT; - head = cur+1; + cur = start; // after readLine(), since start can change, set cur offset to start + curTok = DSN_EOF; goto exit; } - if( *cur == ')' ) + cur = start; // after readLine() since start can change. + + // skip leading whitespace + while( cur= limit ) + goto L_read; + + if( *cur == '(' ) + { + curText = *cur; + curTok = DSN_LEFT; + head = cur+1; + goto exit; + } + + if( *cur == ')' ) + { + curText = *cur; + curTok = DSN_RIGHT; + head = cur+1; + goto exit; + } + + // Non-specctraMode, understands and deciphers escaped \, \r, \n, and \". + // Strips off leading and trailing double quotes + if( !specctraMode ) + { + // a quoted string, will return DSN_STRING + if( *cur == stringDelimiter ) + { + // copy the token, character by character so we can remove doubled up quotes. + curText.clear(); + + ++cur; // skip over the leading delimiter, which is always " in non-specctraMode + + head = cur; + + while( head= limit ) + break; // throw exception at L_unterminated + + switch( *head++ ) + { + case '"': + case '\\': c = head[-1]; break; + case 'a': c = '\x07'; break; + case 'b': c = '\x08'; break; + case 'f': c = '\x0c'; break; + case 'n': c = '\n'; break; + case 'r': c = '\r'; break; + case 't': c = '\x09'; break; + case 'v': c = '\x0b'; break; + + case 'x': // 1 or 2 byte hex escape sequence + for( i=0; i<2; ++i ) + { + if( !isxdigit( head[i] ) ) + break; + tbuf[i] = head[i]; + } + tbuf[i] = '\0'; + if( i > 0 ) + c = (char) strtoul( tbuf, NULL, 16 ); + else + c = 'x'; // a goofed hex escape sequence, interpret as 'x' + head += i; + break; + + default: // 1-3 byte octal escape sequence + --head; + for( i=0; i<3; ++i ) + { + if( head[i] < '0' || head[i] > '7' ) + break; + tbuf[i] = head[i]; + } + tbuf[i] = '\0'; + if( i > 0 ) + c = (char) strtoul( tbuf, NULL, 8 ); + else + c = '\\'; // a goofed octal escape sequence, interpret as '\' + head += i; + break; + } + + curText += c; + } + + else if( *head == '"' ) // end of the non-specctraMode DSN_STRING + { + curTok = DSN_STRING; + ++head; // omit this trailing double quote + goto exit; + } + + else + curText += *head++; + + } // while + + // L_unterminated: + wxString errtxt( _( "Un-terminated delimited string" ) ); + THROW_PARSE_ERROR( errtxt, CurSource(), CurLine(), CurLineNumber(), CurOffset() ); + } + } + + else // is specctraMode, tests in this block should not occur in KiCad mode. + { + /* get the dash out of a which is embedded for example + like: U2-14 or "U2"-"14" + This is detectable by a non-space immediately preceeding the dash. + */ + if( *cur == '-' && cur>start && !isSpace( cur[-1] ) ) + { + curText = '-'; + curTok = DSN_DASH; head = cur+1; goto exit; } @@ -594,166 +697,53 @@ L_read: goto exit; } - /* get the dash out of a which is embedded for example - like: U2-14 or "U2"-"14" - This is detectable by a non-space immediately preceeding the dash. - */ - if( *cur == '-' && cur>start && !isSpace( cur[-1] ) ) - { - curText = '-'; - curTok = DSN_DASH; - head = cur+1; - goto exit; - } - - if( ( head = isNumber( cur, limit ) ) != NULL && - // token can only be a number if trailing text is a separator or line end - ( head==limit || isSep( *head ) ) ) - { - curText.clear(); - curText.append( cur, head ); - curTok = DSN_NUMBER; - goto exit; - } - - // a quoted string, will return DSN_STRING + // specctraMode DSN_STRING if( *cur == stringDelimiter ) { - // Non-specctraMode, understands and deciphers escaped \, \r, \n, and \". - // Strips off leading and trailing double quotes - if( !specctraMode ) + ++cur; // skip over the leading delimiter: ",', or $ + + head = cur; + + while( head= limit ) { - // copy the token, character by character so we can remove doubled up quotes. - curText.clear(); - - ++cur; // skip over the leading delimiter, which is always " in non-specctraMode - - head = cur; - - while( head= limit ) - break; // throw exception at L_unterminated - - switch( *head++ ) - { - case '"': - case '\\': c = head[-1]; break; - case 'a': c = '\x07'; break; - case 'b': c = '\x08'; break; - case 'f': c = '\x0c'; break; - case 'n': c = '\n'; break; - case 'r': c = '\r'; break; - case 't': c = '\x09'; break; - case 'v': c = '\x0b'; break; - - case 'x': // 1 or 2 byte hex escape sequence - for( i=0; i<2; ++i ) - { - if( !isxdigit( head[i] ) ) - break; - tbuf[i] = head[i]; - } - tbuf[i] = '\0'; - if( i > 0 ) - c = (char) strtoul( tbuf, NULL, 16 ); - else - c = 'x'; // a goofed hex escape sequence, interpret as 'x' - head += i; - break; - - default: // 1-3 byte octal escape sequence - --head; - for( i=0; i<3; ++i ) - { - if( head[i] < '0' || head[i] > '7' ) - break; - tbuf[i] = head[i]; - } - tbuf[i] = '\0'; - if( i > 0 ) - c = (char) strtoul( tbuf, NULL, 8 ); - else - c = '\\'; // a goofed octal escape sequence, interpret as '\' - head += i; - break; - } - - curText += c; - } - - else if( *head == '"' ) // end of the non-specctraMode DSN_STRING - { - curTok = DSN_STRING; - ++head; // omit this trailing double quote - goto exit; - } - - else - curText += *head++; - - } // while - - // L_unterminated: wxString errtxt( _( "Un-terminated delimited string" ) ); THROW_PARSE_ERROR( errtxt, CurSource(), CurLine(), CurLineNumber(), CurOffset() ); } - else // specctraMode DSN_STRING - { - ++cur; // skip over the leading delimiter: ",', or $ - - head = cur; - - while( head= limit ) - { - wxString errtxt( _( "Un-terminated delimited string" ) ); - THROW_PARSE_ERROR( errtxt, CurSource(), CurLine(), CurLineNumber(), CurOffset() ); - } - - curText.clear(); - curText.append( cur, head ); - - ++head; // skip over the trailing delimiter - - curTok = DSN_STRING; - goto exit; - } - } - - // Maybe it is a token we will find in the token table. - // If not, then call it a DSN_SYMBOL. - { - head = cur+1; - while( headAdd( FromUTF8() ); + } while( ( tok = NextTok() ) == DSN_COMMENT ); } diff --git a/include/dsnlexer.h b/include/dsnlexer.h index 4c01ed03df..896a8bf2f4 100644 --- a/include/dsnlexer.h +++ b/include/dsnlexer.h @@ -134,19 +134,6 @@ protected: return 0; } - - /** - * Function readLineOrCmt - * reads a line from the LINE_READER and returns either: - *
    - *
  1. a positive line length (a +1 if empty line) - *
  2. zero of end of file. - *
  3. DSN_COMMENT if the line is a comment - *
- */ - int readLineOrCmt(); - - /** * Function findToken * takes a string and looks up the string in the list of expected @@ -154,7 +141,8 @@ protected: * * @param tok A string holding the token text to lookup, in an * unpredictable case: uppercase or lowercase - * @return int - DSN_T or -1 if argument string is not a recognized token. + * @return int - DSN_T matching the keyword text, or DSN_SYMBOL if argument + * string is not a recognized token. */ int findToken( const std::string& tok ); @@ -168,6 +156,8 @@ protected: return false; } + + #endif public: diff --git a/include/richio.h b/include/richio.h index 7fd0f6a265..72bd733298 100644 --- a/include/richio.h +++ b/include/richio.h @@ -45,7 +45,7 @@ #define IO_FORMAT _( "IO_ERROR: %s\nfrom %s : %s" ) -#define PARSE_FORMAT _( "PARSE_ERROR: %s in input/source \"%s\", line %d, offset %d\nfrom %s : %s" ) +#define PARSE_FORMAT _( "PARSE_ERROR: %s in input/source\n'%s'\nline %d\noffset %d\nfrom %s : %s" ) // references: // http://stackoverflow.com/questions/2670816/how-can-i-use-the-compile-time-constant-line-in-a-string diff --git a/pcbnew/files.cpp b/pcbnew/files.cpp index 16bf9a40ef..bdd507fae0 100644 --- a/pcbnew/files.cpp +++ b/pcbnew/files.cpp @@ -50,6 +50,9 @@ #include #include // LEGACY_BOARD_FILE_VERSION +//#define USE_INSTRUMENTATION true +#define USE_INSTRUMENTATION false + static const wxString backupFileExtensionSuffix( wxT( "-bak" ) ); static const wxString autosaveFilePrefix( wxT( "_autosave-" ) ); @@ -283,7 +286,7 @@ bool PCB_EDIT_FRAME::LoadOnePcbFile( const wxString& aFileName, bool aAppend, props["page_width"] = wxString::Format( wxT( "%d" ), GetPageSizeIU().x ); props["page_height"] = wxString::Format( wxT( "%d" ), GetPageSizeIU().y ); -#if 0 +#if USE_INSTRUMENTATION // measure the time to load a BOARD. unsigned startTime = GetRunningMicroSecs(); #endif @@ -291,7 +294,7 @@ bool PCB_EDIT_FRAME::LoadOnePcbFile( const wxString& aFileName, bool aAppend, // load or append either: loadedBoard = pi->Load( GetBoard()->GetFileName(), aAppend ? GetBoard() : NULL, &props ); -#if 0 +#if USE_INSTRUMENTATION unsigned stopTime = GetRunningMicroSecs(); printf( "PLUGIN::Load(): %u usecs\n", stopTime - startTime ); #endif