kicad/include/dsnlexer.h

531 lines
18 KiB
C
Raw Normal View History

/*
* This program source code file is part of KICAD, a free EDA CAD application.
*
* Copyright (C) 2007-2010 SoftPLC Corporation, Dick Hollenbeck <dick@softplc.com>
* Copyright (C) 2007-2021 Kicad Developers, see change_log.txt for contributors.
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License
* as published by the Free Software Foundation; either version 2
* of the License, or (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, you may find one here:
* http://www.gnu.org/licenses/old-licenses/gpl-2.0.html
* or you may search the http://www.gnu.org website for the version 2 license,
* or you may write to the Free Software Foundation, Inc.,
* 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA
*/
#ifndef DSNLEXER_H_
#define DSNLEXER_H_
#include <cstdio>
#include <hashtables.h>
#include <string>
#include <vector>
#include <richio.h>
#ifndef SWIG
/**
2020-12-18 14:04:03 +00:00
* Hold a keyword string and its unique integer token.
*/
struct KEYWORD
{
const char* name; ///< unique keyword.
int token; ///< a zero based index into an array of KEYWORDs
};
2020-11-18 19:50:36 +00:00
#endif // SWIG
// something like this macro can be used to help initialize a KEYWORD table.
// see SPECCTRA_DB::keywords[] as an example.
//#define TOKDEF(x) { #x, T_##x }
/**
2020-12-18 14:04:03 +00:00
* List all the DSN lexer's tokens that are supported in lexing.
*
* It is up to the parser if it wants also to support them.
*/
2020-11-18 19:50:36 +00:00
enum DSN_SYNTAX_T
{
DSN_NONE = -11,
DSN_COMMENT = -10,
DSN_STRING_QUOTE = -9,
2020-11-18 19:50:36 +00:00
DSN_QUOTE_DEF = -8,
DSN_DASH = -7,
DSN_SYMBOL = -6,
DSN_NUMBER = -5,
DSN_RIGHT = -4, // right bracket, ')'
DSN_LEFT = -3, // left bracket, '('
DSN_STRING = -2, // a quoted string, stripped of the quotes
DSN_EOF = -1 // special case for end of file
};
/**
2020-12-18 14:04:03 +00:00
* Implement a lexical analyzer for the SPECCTRA DSN file format.
*
* It reads lexical tokens from the current #LINE_READER through the #NextTok() function.
*/
class DSNLEXER
{
public:
/**
2020-12-18 14:04:03 +00:00
* Initialize a DSN lexer and prepares to read from aFile which is already open and has
* \a aFilename.
*
* @param aKeywordTable is an array of KEYWORDS holding \a aKeywordCount. This
* token table need not contain the lexer separators such as '(' ')', etc.
* @param aKeywordCount is the count of tokens in aKeywordTable.
* @param aFile is an open file, which will be closed when this is destructed.
* @param aFileName is the name of the file
*/
DSNLEXER( const KEYWORD* aKeywordTable, unsigned aKeywordCount,
FILE* aFile, const wxString& aFileName );
/**
2020-12-18 14:04:03 +00:00
* Initialize a DSN lexer and prepares to read from @a aSExpression.
*
* @param aKeywordTable is an array of KEYWORDS holding \a aKeywordCount. This
* token table need not contain the lexer separators such as '(' ')', etc.
* @param aKeywordCount is the count of tokens in aKeywordTable.
* @param aSExpression is text to feed through a STRING_LINE_READER
* @param aSource is a description of aSExpression, used for error reporting.
*/
DSNLEXER( const KEYWORD* aKeywordTable, unsigned aKeywordCount,
const std::string& aSExpression, const wxString& aSource = wxEmptyString );
/**
2020-12-18 14:04:03 +00:00
* Initialize a DSN lexer and prepares to read from @a aSExpression.
*
2020-12-18 14:04:03 +00:00
* Use this one without a keyword table with the DOM parser in ptree.h.
*
* @param aSExpression is text to feed through a #STRING_LINE_READER
* @param aSource is a description of aSExpression, used for error reporting.
*/
DSNLEXER( const std::string& aSExpression, const wxString& aSource = wxEmptyString );
/**
2020-12-18 14:04:03 +00:00
* Initialize a DSN lexer and prepares to read from @a aLineReader which is already
* open, and may be in use by other DSNLEXERs also.
*
2020-12-18 14:04:03 +00:00
* No ownership is taken of @a aLineReader. This enables it to be used by other DSNLEXERs.
*
2020-12-18 14:04:03 +00:00
* @param aKeywordTable is an array of #KEYWORDS holding \a aKeywordCount. This
* token table need not contain the lexer separators such as '(' ')', etc.
* @param aKeywordCount is the count of tokens in aKeywordTable.
* @param aLineReader is any subclassed instance of LINE_READER, such as
2020-12-18 14:04:03 +00:00
* #STRING_LINE_READER or #FILE_LINE_READER. No ownership is taken.
*/
DSNLEXER( const KEYWORD* aKeywordTable, unsigned aKeywordCount,
LINE_READER* aLineReader = nullptr );
virtual ~DSNLEXER();
/**
2020-12-18 14:04:03 +00:00
* Usable only for DSN lexers which share the same #LINE_READER.
*
* Synchronizes the pointers handling the data read by the #LINE_READER. Allows 2
* #DNSLEXER objects to share the same current line, when switching from a #DNSLEXER
* to another #DNSLEXER
* @param aLexer the model.
* @return true if the sync can be made ( at least the same line reader ).
*/
bool SyncLineReaderWith( DSNLEXER& aLexer );
/**
2020-12-18 14:04:03 +00:00
* Change the behavior of this lexer into or out of "specctra mode".
*
* If specctra mode, then:
* -#) stringDelimiter can be changed.
* -#) KiCad quoting protocol is not in effect.
* -#) space_in_quoted_tokens is functional else none of the above are true.
*
* The default mode is non-specctra mode, meaning:
* -#) stringDelimiter cannot be changed.
* -#) KiCad quoting protocol is in effect.
* -#) space_in_quoted_tokens is not functional.
*/
void SetSpecctraMode( bool aMode );
/**
2020-12-18 14:04:03 +00:00
* Manage a stack of LINE_READERs in order to handle nested file inclusion.
*
* This function pushes aLineReader onto the top of a stack of LINE_READERs and makes
2020-12-18 14:04:03 +00:00
* it the current #LINE_READER with its own #GetSource(), line number and line text.
* A grammar must be designed such that the "include" token (whatever its various names),
* and any of its parameters are not followed by anything on that same line,
* because PopReader always starts reading from a new line upon returning to
2020-12-18 14:04:03 +00:00
* the original #LINE_READER.
*/
void PushReader( LINE_READER* aLineReader );
/**
2020-12-18 14:04:03 +00:00
* Delete the top most #LINE_READER from an internal stack of LINE_READERs and
* in the case of #FILE_LINE_READER this means the associated FILE is closed.
*
* The most recently used former #LINE_READER on the stack becomes the
* current #LINE_READER and its previous position in its input stream and the
* its latest line number should pertain. PopReader always starts reading
2020-12-18 14:04:03 +00:00
* from a new line upon returning to the previous #LINE_READER. A pop is only
* possible if there are at least 2 #LINE_READERs on the stack, since popping
* the last one is not supported.
*
2020-12-18 14:04:03 +00:00
* @return the LINE_READER that was in use before the pop, or NULL
* if there was not at least two readers on the stack and therefore the
* pop failed.
*/
LINE_READER* PopReader();
/**
2020-12-18 14:04:03 +00:00
* Return the next token found in the input file or DSN_EOF when reaching the end of
* file.
*
* Users should wrap this function to return an enum to aid in grammar debugging while
* running under a debugger, but leave this lower level function returning an int (so
* the enum does not collide with another usage).
*
* @return the type of token found next.
* @throw IO_ERROR only if the #LINE_READER throws it.
*/
int NextTok();
/**
2020-12-18 14:04:03 +00:00
* Call #NextTok() and then verifies that the token read in satisfies #IsSymbol().
*
* @return the actual token read in.
* @throw IO_ERROR if the next token does not satisfy IsSymbol().
*/
int NeedSYMBOL();
/**
2020-12-18 14:04:03 +00:00
* Call #NextTok() and then verifies that the token read in satisfies bool IsSymbol() or
* the next token is #DSN_NUMBER.
*
* @return the actual token read in.
* @throw IO_ERROR if the next token does not satisfy the above test.
*/
int NeedSYMBOLorNUMBER();
2011-02-14 06:39:51 +00:00
/**
2020-12-18 14:04:03 +00:00
* Call #NextTok() and then verifies that the token read is type #DSN_NUMBER.
*
* @return the actual token read in.
* @throw IO_ERROR if the next token does not satisfy the above test.
2011-02-14 06:39:51 +00:00
*/
int NeedNUMBER( const char* aExpectation );
2011-02-14 06:39:51 +00:00
/**
2020-12-18 14:04:03 +00:00
* Return whatever #NextTok() returned the last time it was called.
*/
2020-11-18 19:50:36 +00:00
int CurTok() const
{
return curTok;
}
/**
2020-12-18 14:04:03 +00:00
* Return whatever NextTok() returned the 2nd to last time it was called.
*/
2020-11-18 19:50:36 +00:00
int PrevTok() const
{
return prevTok;
}
/**
2020-12-18 14:04:03 +00:00
* Used to support "loose" matches (quoted tokens).
*/
2020-11-18 19:50:36 +00:00
int GetCurStrAsToken() const
{
return findToken( curText );
}
/**
2020-12-18 14:04:03 +00:00
* Change the string delimiter from the default " to some other character and return
* the old value.
*
* @param aStringDelimiter The character in lowest 8 bits.
2020-12-18 14:04:03 +00:00
* @return The old delimiter in the lowest 8 bits.
*/
char SetStringDelimiter( char aStringDelimiter )
{
int old = stringDelimiter;
2020-12-18 14:04:03 +00:00
if( specctraMode )
stringDelimiter = aStringDelimiter;
2020-12-18 14:04:03 +00:00
return old;
}
/**
2020-12-18 14:04:03 +00:00
* Change the setting controlling whether a space in a quoted string isa terminator.
*
* @param val If true, means
*/
bool SetSpaceInQuotedTokens( bool val )
{
bool old = space_in_quoted_tokens;
2020-12-18 14:04:03 +00:00
if( specctraMode )
space_in_quoted_tokens = val;
2020-12-18 14:04:03 +00:00
return old;
}
2010-01-21 07:41:30 +00:00
/**
2020-12-18 14:04:03 +00:00
* Change the handling of comments.
*
* If set true, comments are returned as single line strings with a terminating newline.
* Otherwise they are consumed by the lexer and not returned.
2010-01-21 07:41:30 +00:00
*/
bool SetCommentsAreTokens( bool val )
{
bool old = commentsAreTokens;
commentsAreTokens = val;
2010-01-21 07:41:30 +00:00
return old;
}
/**
2020-12-18 14:04:03 +00:00
* Check the next sequence of tokens and reads them into a wxArrayString if they are
* comments.
*
* Reading continues until a non-comment token is encountered, and such last read token
* remains as #CurTok() and as #CurText(). No push back or "un get" mechanism is used
* for this support. Upon return you simply avoid calling NextTok() for the next token,
* but rather #CurTok().
*
2020-12-18 14:04:03 +00:00
* @return Heap allocated block of comments or NULL if none. The caller owns the
* allocation and must delete if not NULL.
*/
wxArrayString* ReadCommentLines();
/**
2020-12-18 14:04:03 +00:00
* Test a token to see if it is a symbol.
*
* This means it cannot be a special delimiter character such as #DSN_LEFT, #DSN_RIGHT,
* #DSN_QUOTE, etc. It may however, coincidentally match a keyword and still be a symbol.
*/
static bool IsSymbol( int aTok );
/**
2020-12-18 14:04:03 +00:00
* Throw an #IO_ERROR exception with an input file specific error message.
*
* @param aTok is the token/keyword type which was expected at the current input location.
* @throw IO_ERROR with the location within the input file of the problem.
*/
2020-11-18 19:50:36 +00:00
void Expecting( int aTok ) const;
/**
2020-12-18 14:04:03 +00:00
* Throw an #IO_ERROR exception with an input file specific error message.
*
* @param aTokenList is the token/keyword type which was expected at the
2020-12-18 14:04:03 +00:00
* current input location, e.g. "pin|graphic|property".
* @throw IO_ERROR with the location within the input file of the problem.
*/
2020-11-18 19:50:36 +00:00
void Expecting( const char* aTokenList ) const;
/**
2020-12-18 14:04:03 +00:00
* Throw an #IO_ERROR exception with an input file specific error message.
*
* @param aTok is the token/keyword type which was not expected at the
2020-12-18 14:04:03 +00:00
* current input location.
* @throw IO_ERROR with the location within the input file of the problem.
*/
2020-11-18 19:50:36 +00:00
void Unexpected( int aTok ) const;
/**
2020-12-18 14:04:03 +00:00
* Throw an #IO_ERROR exception with an input file specific error message.
*
* @param aToken is the token which was not expected at the current input location.
* @throw IO_ERROR with the location within the input file of the problem.
*/
2020-11-18 19:50:36 +00:00
void Unexpected( const char* aToken ) const;
/**
2020-12-18 14:04:03 +00:00
* Throw an #IO_ERROR exception with a message saying specifically that \a aTok
2011-03-20 01:59:17 +00:00
* is a duplicate of one already seen in current context.
2020-12-18 14:04:03 +00:00
*
* @param aTok is the token/keyword type which was not expected at the current input
* location.
* @throw IO_ERROR with the location within the input file of the problem.
*/
void Duplicate( int aTok );
/**
2020-12-18 14:04:03 +00:00
* Call #NextTok() and then verifies that the token read in is a #DSN_LEFT.
*
* @throw IO_ERROR if the next token is not a #DSN_LEFT
*/
void NeedLEFT();
/**
2020-12-18 14:04:03 +00:00
* Call #NextTok() and then verifies that the token read in is a #DSN_RIGHT.
*
* @throw IO_ERROR if the next token is not a #DSN_RIGHT
*/
void NeedRIGHT();
/**
2020-12-18 14:04:03 +00:00
* Return the C string representation of a #DSN_T value.
*/
2020-11-18 19:50:36 +00:00
const char* GetTokenText( int aTok ) const;
/**
2020-12-18 14:04:03 +00:00
* Return a quote wrapped wxString representation of a token value.
*/
2020-11-18 19:50:36 +00:00
wxString GetTokenString( int aTok ) const;
static const char* Syntax( int aTok );
/**
2020-12-18 14:04:03 +00:00
* Return a pointer to the current token's text.
*/
2020-11-18 19:50:36 +00:00
const char* CurText() const
{
return curText.c_str();
}
/**
2020-12-18 14:04:03 +00:00
* Return a reference to current token in std::string form.
*/
2020-11-18 19:50:36 +00:00
const std::string& CurStr() const
{
return curText;
}
/**
2020-12-18 14:04:03 +00:00
* Return the current token text as a wxString, assuming that the input byte stream
* is UTF8 encoded.
*/
2020-11-18 19:50:36 +00:00
wxString FromUTF8() const
{
return wxString::FromUTF8( curText.c_str() );
}
/**
2020-12-18 14:04:03 +00:00
* Return the current line number within my #LINE_READER.
*/
2020-11-18 19:50:36 +00:00
int CurLineNumber() const
{
return reader->LineNumber();
}
/**
2020-12-18 14:04:03 +00:00
* Return the current line of text from which the #CurText() would return its token.
*/
2020-11-18 19:50:36 +00:00
const char* CurLine() const
{
return (const char*)(*reader);
}
/**
2020-12-18 14:04:03 +00:00
* Return the current #LINE_READER source.
*
* @return source of the lines of text, e.g. a filename or "clipboard".
*/
2020-11-18 19:50:36 +00:00
const wxString& CurSource() const
{
return reader->GetSource();
}
/**
2020-12-18 14:04:03 +00:00
* Return the byte offset within the current line, using a 1 based index.
*
* @return a one based index into the current line.
*/
2020-11-18 19:50:36 +00:00
int CurOffset() const
{
return curOffset + 1;
}
2020-12-18 14:04:03 +00:00
#ifndef SWIG
protected:
void init();
int readLine()
{
if( reader )
{
reader->ReadLine();
unsigned len = reader->Length();
// start may have changed in ReadLine(), which can resize and
// relocate reader's line buffer.
start = reader->Line();
next = start;
limit = next + len;
return len;
}
return 0;
}
/**
* Take @a aToken string and looks up the string in the keywords table.
*
* @param aToken is a string to lookup in the keywords table.
* @return with a value from the enum #DSN_T matching the keyword text,
* or #DSN_SYMBOL if @a aToken is not in the keywords table.
*/
int findToken( const std::string& aToken ) const;
bool isStringTerminator( char cc ) const
{
if( !space_in_quoted_tokens && cc == ' ' )
return true;
if( cc == stringDelimiter )
return true;
return false;
}
bool iOwnReaders; ///< on readerStack, should I delete them?
const char* start;
const char* next;
const char* limit;
char dummy[1]; ///< when there is no reader.
typedef std::vector<LINE_READER*> READER_STACK;
READER_STACK readerStack; ///< all the LINE_READERs by pointer.
///< no ownership. ownership is via readerStack, maybe, if iOwnReaders
LINE_READER* reader;
bool specctraMode; ///< if true, then:
///< 1) stringDelimiter can be changed
///< 2) Kicad quoting protocol is not in effect
///< 3) space_in_quoted_tokens is functional
///< else not.
char stringDelimiter;
bool space_in_quoted_tokens; ///< blank spaces within quoted strings
bool commentsAreTokens; ///< true if should return comments as tokens
int prevTok; ///< curTok from previous NextTok() call.
int curOffset; ///< offset within current line of the current token
int curTok; ///< the current token obtained on last NextTok()
std::string curText; ///< the text of the current token
const KEYWORD* keywords; ///< table sorted by CMake for bsearch()
unsigned keywordCount; ///< count of keywords table
KEYWORD_MAP keyword_hash; ///< fast, specialized "C string" hashtable
#endif // SWIG
};
#endif // DSNLEXER_H_