source: OGRE/trunk/ogrenew/OgreMain/include/OgreCompiler2Pass.h @ 692

Revision 692, 25.9 KB checked in by mattausch, 19 years ago (diff)

adding ogre 1.2 and dependencies

RevLine 
[692]1/*
2-----------------------------------------------------------------------------
3This source file is part of OGRE
4    (Object-oriented Graphics Rendering Engine)
5For the latest info, see http://www.stevestreeting.com/ogre/
6
7Copyright (c) 2000-2005 The OGRE Team
8Also see acknowledgements in Readme.html
9
10This program is free software; you can redistribute it and/or modify it under
11the terms of the GNU General Public License as published by the Free Software
12Foundation; either version 2 of the License, or (at your option) any later
13version.
14
15This program is distributed in the hope that it will be useful, but WITHOUT
16ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
17FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
18
19You should have received a copy of the GNU General Public License along with
20this program; if not, write to the Free Software Foundation, Inc., 59 Temple
21Place - Suite 330, Boston, MA 02111-1307, USA, or go to
22http://www.gnu.org/copyleft/gpl.html.
23-----------------------------------------------------------------------------
24*/
25
26
27#ifndef __Compiler2Pass_H__
28#define __Compiler2Pass_H__
29
30#include <vector>
31#include "OgrePrerequisites.h"
32
33namespace Ogre {
34
35    /** Compiler2Pass is a generic 2 pass compiler/assembler
36    @remarks
37            provides a tokenizer in pass 1 and relies on the subclass to provide the virtual method for pass 2
38
39            PASS 1 - tokenize source: this is a simple brute force lexical scanner/analyzer that also parses
40                            the formed token for proper semantics and context in one pass
41                            it uses Look Ahead Left-Right (LALR) ruling based on Backus - Naur Form (BNF) notation for semantic
42                            checking and also performs  context checking allowing for language dialects
43
44            PASS 2 - generate application specific instructions ie native instructions based on the tokens in the instruction container.
45
46    @par
47            this class must be subclassed with the subclass providing some implementation details for Pass 2.  The subclass
48            is responsible for setting up the token libraries along with defining the language syntax and handling
49        token actions during the second pass.
50
51    @par
52        The sub class normally supplies a simplified BNF text description in its constructor prior to doing any parsing/tokenizing of source.
53        The simplified BNF text description defines the language syntax and rule structure.
54        The meta-symbols used in the BNF text description are:
55
56        ::=  meaning "is defined as". "::=" starts the definition of a rule.  The left side of ::= must contain an <identifier>
57
58        <>   angle brackets are used to surround syntax rule names. A syntax rule name is also called a non-terminal in that
59             it does not generate a terminal token in the instruction container for pass 2 processing.
60
61        |    meaning "or". if the item on the left of the | fails then the item on the right is tested.
62             Example: <true_false> ::= 'true' | 'false';
63             whitespace is used to imply AND operation between left and right items.
64             Example: <terrain_shadaws> ::= 'terrain_shadows' <true_false>
65             the 'terrain_shadows' terminal token must be found and <true_false> rule must pass in order for <terrain_shadows> rule
66             to pass.
67
68
69        []   optional rule identifier is enclosed in meta symbols [ and ].
70             Note that only one identifier or terminal token can take [] modifier.
71        {}   repetitive identifier (zero or more times) is enclosed in meta symbols { and }
72             Note that only one identifier or terminal token can take {} modifier.
73        ''   terminal tokens are surrounded by single quotes.  A terminal token is always one or more characters.
74             For example: 'Colour' defines a character sequence that must be matched in whole.  Note that matching is case
75             sensitive.
76        @    turn on single character scanning and don't skip white space.
77             Mainly used for label processing that allow white space.
78             Example: '@ ' prevents the white space between the quotes from being skipped
79        -''  no terminal token is generated when a - precedes the first single quote but the text in between the quotes is still
80             tested against the characters in the source being parsed.
81        (?! ) negative lookahead (not test) inspired by Perl 5. Scans ahead for a non-terminal or terminal expression
82             that should fail in order to make the rule production pass.
83             Does not generate a token or advance the cursur.  If the lookahead result fails ie token is found,
84             then the current rule fails and rollback occurs.  Mainly used to solve multiple contexts of a token.
85             An Example of where not test is used to solve multiple contexts:
86
87             <rule>       ::=  <identifier>  "::="  <expression>
88             <expression> ::=  <and_term> { <or_term> }
89             <or_term>    ::=  "|" <and_term>
90             <and_term>   ::=  <term> { <term> }
91             <term>       ::=  <identifier_right> | <terminal_symbol> | <repeat_expression> | <optional_expression>
92             <identifier_right> ::= <identifier> (?!"::=")
93
94             <identiefier> appears on both sides of the ::= so (?!"::=") test to make sure that ::= is not on the
95             right which would indicate that a new rule was being formed.
96
97             Works on both terminals and non-terminals.
98             Note: lookahead failure cause the whole rule to fail and rollback to occur
99
100        <#name> # indicates that a numerical value is to be parsed to form a terminal token.  Name is optional and is just a descriptor
101             to help with understanding what the value will be used for.
102             Example: <Colour> ::= <#red> <#green> <#blue>
103
104        ()   parentheses enclose a set of characters that can be used to generate a user identifier. for example:
105             (0123456789) matches a single character found in that set.
106             An example of a user identifier:
107
108             <Label> ::= <Character> {<Character>}
109             <Character> ::= (abcdefghijklmnopqrstuvwxyz)
110
111             This will generate a rule that accepts one or more lowercase letters to make up the Label.  The User identifier
112             stops collecting the characters into a string when a match cannot be found in the rule.
113    */
114    class _OgreExport Compiler2Pass
115    {
116
117    protected:
118
119            // BNF operation types
120            enum OperationType {otUNKNOWN, otRULE, otAND, otOR, otOPTIONAL, otREPEAT, otDATA, otNOT_TEST, otEND};
121
122            /** structure used to build rule paths
123
124            */
125            struct TokenRule
126        {
127                    OperationType operation;
128                    size_t tokenID;
129
130            TokenRule(void) : operation(otUNKNOWN), tokenID(0) {}
131            TokenRule(const OperationType ot, const size_t token)
132                : operation(ot), tokenID(token) {}
133            };
134
135            typedef std::vector<TokenRule> TokenRuleContainer;
136            typedef TokenRuleContainer::iterator TokenRuleIterator;
137
138        static const size_t SystemTokenBase = 1000;
139        enum SystemRuleToken {
140            _no_token_ = SystemTokenBase,
141            _character_,
142            _value_,
143            _no_space_skip_
144        };
145
146            enum BNF_ID {BNF_UNKOWN = 0,
147            BNF_SYNTAX, BNF_RULE, BNF_IDENTIFIER, BNF_IDENTIFIER_RIGHT, BNF_IDENTIFIER_CHARACTERS, BNF_ID_BEGIN, BNF_ID_END,
148            BNF_CONSTANT_BEGIN, BNF_SET_RULE, BNF_EXPRESSION,
149            BNF_AND_TERM, BNF_OR_TERM, BNF_TERM, BNF_TERM_ID, BNF_CONSTANT, BNF_OR, BNF_TERMINAL_SYMBOL, BNF_TERMINAL_START,
150            BNF_REPEAT_EXPRESSION, BNF_REPEAT_BEGIN, BNF_REPEAT_END, BNF_SET, BNF_SET_BEGIN, BNF_SET_END,
151            BNF_NOT_TEST, BNF_NOT_TEST_BEGIN, BNF_OPTIONAL_EXPRESSION, BNF_NOT_EXPRESSION, BNF_NOT_CHK,
152            BNF_OPTIONAL_BEGIN, BNF_OPTIONAL_END, BNF_NO_TOKEN_START, BNF_SINGLEQUOTE, BNF_SINGLE_QUOTE_EXC, BNF_SET_END_EXC,
153            BNF_ANY_CHARACTER, BNF_SPECIAL_CHARACTERS1,
154            BNF_SPECIAL_CHARACTERS2, BNF_WHITE_SPACE_CHK,
155
156            BNF_LETTER, BNF_LETTER_DIGIT, BNF_DIGIT, BNF_WHITE_SPACE,
157            BNF_ALPHA_SET, BNF_NUMBER_SET, BNF_SPECIAL_CHARACTER_SET1,
158                        BNF_SPECIAL_CHARACTER_SET2, BNF_SPECIAL_CHARACTER_SET3, BNF_NOT_CHARS
159        };
160
161
162            /** structure used to build lexeme Type library */
163            struct LexemeTokenDef
164        {
165                size_t ID;                                      /// Token ID which is the index into the Lexeme Token Definition Container
166            bool hasAction;            /// has an action associated with it. only applicable to terminal tokens
167            bool isNonTerminal;        /// if true then token is non-terminal
168                size_t ruleID;                          /// index into Rule database for non-terminal token rulepath and lexeme
169                bool isCaseSensitive;        /// if true use case sensitivity when comparing lexeme to source
170            String lexeme;             /// text representation of token or valid characters for label parsing
171
172            LexemeTokenDef(void) : ID(0), hasAction(false), isNonTerminal(false), ruleID(0), isCaseSensitive(false) {}
173            LexemeTokenDef( const size_t ID, const String& lexeme, const bool hasAction = false, const bool caseSensitive = false )
174                : ID(ID)
175                , hasAction(hasAction)
176                , isNonTerminal(false)
177                , ruleID(0)
178                , isCaseSensitive(caseSensitive)
179                , lexeme(lexeme)
180            {
181            }
182
183            };
184
185        typedef std::vector<LexemeTokenDef> LexemeTokenDefContainer;
186        typedef LexemeTokenDefContainer::iterator LexemeTokenDefIterator;
187
188        typedef std::map<std::string, size_t> LexemeTokenMap;
189        typedef LexemeTokenMap::iterator TokenKeyIterator;
190        /// map used to lookup client token based on previously defined lexeme
191
192
193            /** structure for Token instructions that are constructed during first pass*/
194            struct TokenInst
195        {
196            size_t NTTRuleID;                   /// Non-Terminal Token Rule ID that generated Token
197            size_t tokenID;                                     /// expected Token ID. Could be UNKNOWN if valid token was not found.
198            size_t line;                                /// line number in source code where Token was found
199            size_t pos;                         /// Character position in source where Token was found
200        bool found;                /// is true if expected token was found
201            };
202
203            typedef std::vector<TokenInst> TokenInstContainer;
204            typedef TokenInstContainer::iterator TokenInstIterator;
205
206        // token que, definitions, rules
207        struct TokenState
208        {
209            TokenInstContainer       tokenQue;
210            LexemeTokenDefContainer  lexemeTokenDefinitions;
211                TokenRuleContainer       rootRulePath;
212            LexemeTokenMap           lexemeTokenMap;
213        };
214
215        TokenState* mClientTokenState;
216
217            /// Active token que, definitions, rules currntly being used by parser
218        TokenState* mActiveTokenState;
219        /// the location within the token instruction container where pass 2 is
220        size_t mPass2TokenQuePosition;
221        /** the que position of the previous token that had an action.
222            A token's action is fired on the next token having an action.
223        */
224        size_t mPreviousActionQuePosition;
225
226            /// pointer to the source to be compiled
227            const String* mSource;
228            /// name of the source to be compiled
229            String mSourceName;
230            size_t mEndOfSource;
231
232            size_t mCurrentLine; /// current line number in source being tokenized
233        size_t mCharPos;     /// position in current line in source being tokenized
234
235            /// storage container for constants defined in source
236        /// container uses Token index as a key associated with a float constant
237            std::map<size_t, float> mConstants;
238            /// storage container for string labels defined in source
239        /// container uses Token index as a key associated with a label
240        std::map<size_t, String> mLabels;
241        /// flag indicates when a label is being parsed.
242        /// It gets set false when a terminal token not of _character_ is encountered
243        bool mLabelIsActive;
244        /// the key of the active label being built during pass 1.
245        /// a new key is calculated when mLabelIsActive switches from false to true
246        size_t mActiveLabelKey;
247        /// flag being true indicates that spaces are not to be skipped
248        /// automatically gets set to false when mLabelIsActive goes to false
249        bool mNoSpaceSkip;
250        /// if flag is true then next terminal token is not added to token que if found
251        /// but does effect rule path flow
252        bool mNoTerminalToken;
253
254            /// Active Contexts pattern used in pass 1 to determine which tokens are valid for a certain context
255            uint mActiveContexts;
256
257            /** perform pass 1 of compile process
258                    scans source for lexemes that can be tokenized and then
259                    performs general semantic and context verification on each lexeme before it is tokenized.
260                    A tokenized instruction list is built to be used by Pass 2.
261            A rule path can trigger Pass 2 execution if enough tokens have been generated in Pass 1.
262            Pass 1 will then pass control to pass 2 temporarily until the current tokens have been consumed.
263
264            */
265            bool doPass1();
266
267            /** performs Pass 2 of compile process which is execution of the tokens
268            @remark
269                    Pass 2 takes the token instructions generated in Pass 1 and
270                    builds the application specific instructions along with verifying
271                    symantic and context rules that could not be checked in Pass 1.
272        @par
273            Pass 2 execution consumes tokens and moves the Pass 2 token instruction position towards the end
274            of the token container.  Token execution can insert new tokens into the token container.
275            */
276            bool doPass2();
277
278        /** execute the action associated with the token pointed to by the Pass 2 token instruction position.
279            Its upto the child class to implement how it will associate a token key with and action.
280            Actions should be placed at positions withing the BNF grammer (instruction que) that indicate
281            enough tokens exist for pass 2 processing to take place.
282        */
283        virtual void executeTokenAction(const size_t tokenID) = 0;
284        /** setup client token definitions.  Gets called when BNF grammer is being setup.
285        */
286        virtual void setupTokenDefinitions(void) = 0;
287        /** Gets the next token from the instruction que.  If an unkown token is found then an exception is raised but
288            the instruction pointer is still moved passed the unknown token.  The subclass should catch the exception,
289            provide an error message, and attempt recovery.
290
291        @param expectedTokenID if set then if tokenID does not match then an exception is raised
292        */
293        const TokenInst& getNextToken(const size_t expectedTokenID = 0);
294        /** Gets the current token from the instruction que.
295        */
296        const TokenInst& getCurrentToken(void);
297        /** If a next token instruction exist then test if its token ID matches.
298            This method is usefull for peeking ahead during pass 2 to see if a certain
299            token exists.
300        @param expectedTokenID is the ID of the token to match.
301        */
302        bool testNextTokenID(const size_t expectedTokenID);
303        /**
304        */
305        void replaceToken(void);
306        /** Gets the next token's associated floating point value in the instruction que that was parsed from the
307            text source.  If an unkown token is found or no associated value was found then an exception is raised but
308            the instruction pointer is still moved passed the unknown token.  The subclass should catch the exception,
309            provide an error message, and attempt recovery.
310        */
311        float getNextTokenValue(void);
312        /** Gets the next token's associated text label in the instruction que that was parsed from the
313            text source.  If an unkown token is found or no associated label was found then an exception is raised but
314            the instruction pointer is still moved passed the unknown token.  The subclass should catch the exception,
315            provide an error message, and attempt recovery.
316        */
317        const String& getNextTokenLabel(void);
318        /** Gets the number of tokens waiting in the instruction que that need to be processed by an token action in pass 2.
319        */
320        size_t getPass2TokenQueCount(void) const;
321        /** Get the number of tokens not processed by action token.
322            Client Actions should use this method to retreive the number of parameters(tokens)
323            remaining to be processed in the action.
324        */
325        size_t getRemainingTokensForAction(void) const;
326
327        /** Add a lexeme token association.  The backend compiler uses the associations between lexeme
328            and token when building the rule base from the BNF script so all associations must  be done
329            prior to compiling a source.
330        @param lexeme is the name of the token and use when parsing the source to determin a match for a token.
331        @param token is the ID associated with the lexeme
332        @param hasAction must be set true if the client wants an action triggered when this token is generated
333        @param caseSensitive should be set true if lexeme match should use case sensitivity
334        */
335        void addLexemeToken(const String& lexeme, const size_t token, const bool hasAction = false, const bool caseSensitive = false);
336
337        /** sets up the parser rules for the client based on the BNF Grammer text passed in.
338            Raises an exception if the grammer did not compile successfully.  This method gets called
339            when a call to compile occurs and no compiled BNF grammer exists, otherwise nothing will happen since the compiler has no rules to work
340            with.  The grammer only needs to be set once during the lifetime of the compiler unless the
341            grammer changes.
342            BNF Grammer rules are cached once the BNF grammer source is compiled.
343            The client should never have to call this method directly.
344        */
345        void setClientBNFGrammer(void);
346
347
348
349        /// find the eol charater
350            void findEOL();
351
352            /** check to see if the text at the present position in the source is a numerical constant
353            @param fvalue is a reference that will receive the float value that is in the source
354            @param charsize reference to receive number of characters that make of the value in the source
355            @return
356                    true if characters form a valid float representation
357                    false if a number value could not be extracted
358            */
359            bool isFloatValue(float& fvalue, size_t& charsize) const;
360
361        /** Check if source at current position is supposed to be a user defined character label.
362        A new label is processed when previous operation was not _character_ otherwise the processed
363        character (if match was found) is added to the current label.  This allows _character_ operations
364        to be chained together to form a crude regular expression to build a label.
365            @param rulepathIDX index into rule path database of token to validate.
366            @return
367                    true if token was found for character label.
368        */
369        bool isCharacterLabel(const size_t rulepathIDX);
370            /** check to see if the text is in the lexeme text library
371            @param lexeme points to begining of text where a lexem token might exist
372            @param caseSensitive set to true if match should be case sensitive
373            @return
374                    true if a matching token could be found in the token type library
375                    false if could not be tokenized
376            */
377            bool isLexemeMatch(const String& lexeme, const bool caseSensitive) const;
378            /// position to the next possible valid sysmbol
379            bool positionToNextLexeme();
380            /** process input source text using rulepath to determine allowed tokens
381            @remarks
382                    the method is reentrant and recursive
383                    if a non-terminal token is encountered in the current rule path then the method is
384                    called using the new rule path referenced by the non-terminal token
385                    Tokens can have the following operation states which effects the flow path of the rule
386                            RULE: defines a rule path for the non-terminal token
387                            AND: the token is required for the rule to pass
388                            OR: if the previous tokens failed then try these ones
389                            OPTIONAL: the token is optional and does not cause the rule to fail if the token is not found
390                            REPEAT: the token is required but there can be more than one in a sequence
391                DATA: Used by a previous token ie for character sets
392                NOTTEST: performs negative lookahead ie make sure the next token is not of a certain type
393                            END: end of the rule path - the method returns the succuss of the rule
394
395            @param rulepathIDX index into an array of Token Rules that define a rule path to be processed
396            @return
397                    true if rule passed - all required tokens found
398                    false if one or more tokens required to complete the rule were not found
399            */
400            bool processRulePath( size_t rulepathIDX);
401
402
403            /** setup ActiveContexts - should be called by subclass to setup initial language contexts
404        */
405            void setActiveContexts(const uint contexts){ mActiveContexts = contexts; }
406
407            /// comment specifiers are hard coded
408            void skipComments();
409
410            /// find end of line marker and move past it
411            void skipEOL();
412
413            /// skip all the white space which includes spaces and tabs
414            void skipWhiteSpace();
415
416
417            /** check if current position in source has the lexeme text equivalent to the TokenID
418            @param rulepathIDX index into rule path database of token to validate
419            @param activeRuleID index of non-terminal rule that generated the token
420            @return
421                    true if token was found
422                    false if token lexeme text does not match the source text
423                    if token is non-terminal then processRulePath is called
424            */
425            bool ValidateToken(const size_t rulepathIDX, const size_t activeRuleID);
426
427            /** scan through all the rules and initialize token definition with index to rules for non-terminal tokens.
428            Gets called when internal grammer is being verified or after client grammer has been parsed.
429        @param grammerName is the name of the grammer the token rules are for
430            */
431            void verifyTokenRuleLinks(const String& grammerName);
432            /** Checks the last token instruction and if it has an action then it triggers the action of the previously
433            found token having an action.
434            */
435            void checkTokenActionTrigger(void);
436            /** Get the text representation of the rule path.  This is a good way to way to visually verify
437            that the BNF grammer did compile correctly.
438            @param ruleID is the index into the rule path.
439            */
440            String getBNFGrammerTextFromRulePath(size_t ruleID);
441
442
443    private:
444        // used for interpreting BNF script
445        // keep it as static so that only one structure is created
446        // no matter how many times this class is instantiated.
447        static TokenState mBNFTokenState;
448        // maintain a map of BNF grammer
449        typedef std::map<String, TokenState> TokenStateContainer;
450        static TokenStateContainer mClientTokenStates;
451        /// if a previous token action was setup then activate it now
452        void activatePreviousTokenAction(void);
453        /// initialize token definitions and rule paths
454        void initBNFCompiler(void);
455        /// Convert BNF grammer token que created in pass 1 into a BNF rule path
456        void buildClientBNFRulePaths(void);
457        /// modify the last rule in the container. An end operation is added to the rule path.
458        void modifyLastRule(const OperationType pendingRuleOp, const size_t tokenID);
459        /** get the token ID for a lexeme in the client state. If the lexeme is not found then
460            it is added to the map and definition container and a new tokenID created.
461        @return the ID of the token.
462        */
463        size_t getClientLexemeTokenID(const String& lexeme, const bool isCaseSensitive = false);
464        /// Extract a Non Terminal identifier from the token que
465        void extractNonTerminal(const OperationType pendingRuleOp);
466        /// Extract a Terminal lexeme from the token que and add to current rule expression
467        void extractTerminal(const OperationType pendingRuleOp, const bool notoken = false);
468        /// Extract a set from the token que and add to current rule expression
469        void extractSet(const OperationType pendingRuleOp);
470        /// Extract a numeric constant from the token que and add it to the current rule expression
471        void extractNumericConstant(const OperationType pendingRuleOp);
472        String getLexemeText(size_t& ruleID);
473
474    public:
475
476            /// constructor
477            Compiler2Pass();
478        virtual ~Compiler2Pass() {}
479
480            /** compile the source - performs 2 passes.
481                    First pass is to tokinize, check semantics and context.
482                    The second pass is performed by using tokens to look up function implementors and executing
483            them which convert tokens to application specific instructions.
484            @remark
485                    Pass 2 only gets executed if Pass 1 has built enough tokens to complete a rule path and found no errors
486            @param source a pointer to the source text to be compiled
487            @return
488                    true if Pass 1 and Pass 2 are successfull
489                    false if any errors occur in Pass 1 or Pass 2
490            */
491            bool compile(const String& source, const String& sourceName);
492        /** gets BNF Grammer.  Gets called when BNF grammer has to be compiled for the first time.
493        */
494        virtual const String& getClientBNFGrammer(void) = 0;
495
496        /** get the name of the BNF grammer.
497        */
498        virtual const String& getClientGrammerName(void) = 0;
499
500    };
501
502}
503
504#endif
505
Note: See TracBrowser for help on using the repository browser.