?  search:
summary | log | graphiclog1 | graphiclog2 | commit | commitdiff | tree | refs | edit | fork
blame | history | raw | HEAD
Merge pull request #3 from Chris--A/typename-fix
[arduino-ctags.git] / flex.c
blob038ac4eb1861596aab1cf405bd1fb160163976a4
1 /*
2 * $Id: flex.c 666 2008-05-15 17:47:31Z dfishburn $
3 *
4 * Copyright (c) 2008, David Fishburn
5 *
6 * This source code is released for free distribution under the terms of the
7 * GNU General Public License.
8 *
9 * This module contains functions for generating tags for Adobe languages.
10 * There are a number of different ones, but this will begin with:
11 * Flex
12 * MXML files (*.mMacromedia XML)
13 * ActionScript files (*.as)
14 *
15 * Flex 3 language reference
16 * http://livedocs.adobe.com/flex/3/langref/index.html
17 */
18
19 /*
20 * INCLUDE FILES
21 */
22 #include "general.h" /* must always come first */
23 #include <ctype.h> /* to define isalpha () */
24 #include <setjmp.h>
25 #ifdef DEBUG
26 #include <stdio.h>
27 #endif
28
29 #include "debug.h"
30 #include "entry.h"
31 #include "keyword.h"
32 #include "parse.h"
33 #include "read.h"
34 #include "routines.h"
35 #include "vstring.h"
36
37 /*
38 * MACROS
39 */
40 #define isType(token,t) (boolean) ((token)->type == (t))
41 #define isKeyword(token,k) (boolean) ((token)->keyword == (k))
42
43 /*
44 * DATA DECLARATIONS
45 */
46
47 typedef enum eException { ExceptionNone, ExceptionEOF } exception_t;
48
49 /*
50 * Tracks class and function names already created
51 */
52 static stringList *ClassNames;
53 static stringList *FunctionNames;
54
55 /* Used to specify type of keyword.
56 */
57 typedef enum eKeywordId {
58 KEYWORD_NONE = -1,
59 KEYWORD_function,
60 KEYWORD_capital_function,
61 KEYWORD_object,
62 KEYWORD_capital_object,
63 KEYWORD_prototype,
64 KEYWORD_var,
65 KEYWORD_new,
66 KEYWORD_this,
67 KEYWORD_for,
68 KEYWORD_while,
69 KEYWORD_do,
70 KEYWORD_if,
71 KEYWORD_else,
72 KEYWORD_switch,
73 KEYWORD_try,
74 KEYWORD_catch,
75 KEYWORD_finally,
76 KEYWORD_public,
77 KEYWORD_private,
78 KEYWORD_static,
79 KEYWORD_class,
80 KEYWORD_id,
81 KEYWORD_script,
82 KEYWORD_cdata,
83 KEYWORD_mx,
84 KEYWORD_override
85 } keywordId;
86
87 /* Used to determine whether keyword is valid for the token language and
88 * what its ID is.
89 */
90 typedef struct sKeywordDesc {
91 const char *name;
92 keywordId id;
93 } keywordDesc;
94
95 typedef enum eTokenType {
96 TOKEN_UNDEFINED,
97 TOKEN_CHARACTER,
98 TOKEN_CLOSE_PAREN,
99 TOKEN_SEMICOLON,
100 TOKEN_COLON,
101 TOKEN_COMMA,
102 TOKEN_KEYWORD,
103 TOKEN_OPEN_PAREN,
104 TOKEN_OPERATOR,
105 TOKEN_IDENTIFIER,
106 TOKEN_STRING,
107 TOKEN_PERIOD,
108 TOKEN_OPEN_CURLY,
109 TOKEN_CLOSE_CURLY,
110 TOKEN_EQUAL_SIGN,
111 TOKEN_EXCLAMATION,
112 TOKEN_FORWARD_SLASH,
113 TOKEN_OPEN_SQUARE,
114 TOKEN_CLOSE_SQUARE,
115 TOKEN_OPEN_MXML,
116 TOKEN_CLOSE_MXML,
117 TOKEN_CLOSE_SGML,
118 TOKEN_LESS_THAN,
119 TOKEN_GREATER_THAN,
120 TOKEN_QUESTION_MARK
121 } tokenType;
122
123 typedef struct sTokenInfo {
124 tokenType type;
125 keywordId keyword;
126 vString * string;
127 vString * scope;
128 unsigned long lineNumber;
129 fpos_t filePosition;
130 int nestLevel;
131 boolean ignoreTag;
132 boolean isClass;
133 } tokenInfo;
134
135 /*
136 * DATA DEFINITIONS
137 */
138
139 static langType Lang_js;
140
141 static jmp_buf Exception;
142
143 typedef enum {
144 FLEXTAG_FUNCTION,
145 FLEXTAG_CLASS,
146 FLEXTAG_METHOD,
147 FLEXTAG_PROPERTY,
148 FLEXTAG_VARIABLE,
149 FLEXTAG_MXTAG,
150 FLEXTAG_COUNT
151 } flexKind;
152
153 static kindOption FlexKinds [] = {
154 { TRUE, 'f', "function", "functions" },
155 { TRUE, 'c', "class", "classes" },
156 { TRUE, 'm', "method", "methods" },
157 { TRUE, 'p', "property", "properties" },
158 { TRUE, 'v', "variable", "global variables" },
159 { TRUE, 'x', "mxtag", "mxtags" }
160 };
161
162 static const keywordDesc FlexKeywordTable [] = {
163 /* keyword keyword ID */
164 { "function", KEYWORD_function },
165 { "Function", KEYWORD_capital_function },
166 { "object", KEYWORD_object },
167 { "Object", KEYWORD_capital_object },
168 { "prototype", KEYWORD_prototype },
169 { "var", KEYWORD_var },
170 { "new", KEYWORD_new },
171 { "this", KEYWORD_this },
172 { "for", KEYWORD_for },
173 { "while", KEYWORD_while },
174 { "do", KEYWORD_do },
175 { "if", KEYWORD_if },
176 { "else", KEYWORD_else },
177 { "switch", KEYWORD_switch },
178 { "try", KEYWORD_try },
179 { "catch", KEYWORD_catch },
180 { "finally", KEYWORD_finally },
181 { "public", KEYWORD_public },
182 { "private", KEYWORD_private },
183 { "static", KEYWORD_static },
184 { "class", KEYWORD_class },
185 { "id", KEYWORD_id },
186 { "script", KEYWORD_script },
187 { "cdata", KEYWORD_cdata },
188 { "mx", KEYWORD_mx },
189 { "override", KEYWORD_override }
190 };
191
192 /*
193 * FUNCTION DEFINITIONS
194 */
195
196 /* Recursive functions */
197 static void parseFunction (tokenInfo *const token);
198 static boolean parseBlock (tokenInfo *const token, tokenInfo *const parent);
199 static boolean parseLine (tokenInfo *const token);
200 static boolean parseActionScript (tokenInfo *const token);
201
202 static boolean isIdentChar (const int c)
203 {
204 return (boolean)
205 (isalpha (c) || isdigit (c) || c == '$' ||
206 c == '@' || c == '_' || c == '#');
207 }
208
209 static void buildFlexKeywordHash (void)
210 {
211 const size_t count = sizeof (FlexKeywordTable) /
212 sizeof (FlexKeywordTable [0]);
213 size_t i;
214 for (i = 0 ; i < count ; ++i)
215 {
216 const keywordDesc* const p = &FlexKeywordTable [i];
217 addKeyword (p->name, Lang_js, (int) p->id);
218 }
219 }
220
221 static tokenInfo *newToken (void)
222 {
223 tokenInfo *const token = xMalloc (1, tokenInfo);
224
225 token->type = TOKEN_UNDEFINED;
226 token->keyword = KEYWORD_NONE;
227 token->string = vStringNew ();
228 token->scope = vStringNew ();
229 token->nestLevel = 0;
230 token->isClass = FALSE;
231 token->ignoreTag = FALSE;
232 token->lineNumber = getSourceLineNumber ();
233 token->filePosition = getInputFilePosition ();
234
235 return token;
236 }
237
238 static void deleteToken (tokenInfo *const token)
239 {
240 vStringDelete (token->string);
241 vStringDelete (token->scope);
242 eFree (token);
243 }
244
245 /*
246 * Tag generation functions
247 */
248
249 static void makeConstTag (tokenInfo *const token, const flexKind kind)
250 {
251 if (FlexKinds [kind].enabled && ! token->ignoreTag )
252 {
253 const char *const name = vStringValue (token->string);
254 tagEntryInfo e;
255 initTagEntry (&e, name);
256
257 e.lineNumber = token->lineNumber;
258 e.filePosition = token->filePosition;
259 e.kindName = FlexKinds [kind].name;
260 e.kind = FlexKinds [kind].letter;
261
262 makeTagEntry (&e);
263 }
264 }
265
266 static void makeFlexTag (tokenInfo *const token, flexKind kind)
267 {
268 vString * fulltag;
269
270 if (FlexKinds [kind].enabled && ! token->ignoreTag )
271 {
272 DebugStatement (
273 debugPrintf (DEBUG_PARSE
274 , "\n makeFlexTag start: token isClass:%d scope:%s name:%s\n"
275 , token->isClass
276 , vStringValue(token->scope)
277 , vStringValue(token->string)
278 );
279 );
280 if (kind == FLEXTAG_FUNCTION && token->isClass )
281 {
282 kind = FLEXTAG_METHOD;
283 }
284 /*
285 * If a scope has been added to the token, change the token
286 * string to include the scope when making the tag.
287 */
288 if ( vStringLength(token->scope) > 0 )
289 {
290 fulltag = vStringNew ();
291 vStringCopy(fulltag, token->scope);
292 vStringCatS (fulltag, ".");
293 vStringCatS (fulltag, vStringValue(token->string));
294 vStringTerminate(fulltag);
295 vStringCopy(token->string, fulltag);
296 vStringDelete (fulltag);
297 }
298 makeConstTag (token, kind);
299 }
300 }
301
302 static void makeClassTag (tokenInfo *const token)
303 {
304 vString * fulltag;
305
306 if ( ! token->ignoreTag )
307 {
308 fulltag = vStringNew ();
309 if (vStringLength (token->scope) > 0)
310 {
311 vStringCopy(fulltag, token->scope);
312 vStringCatS (fulltag, ".");
313 vStringCatS (fulltag, vStringValue(token->string));
314 }
315 else
316 {
317 vStringCopy(fulltag, token->string);
318 }
319 vStringTerminate(fulltag);
320 if ( ! stringListHas(ClassNames, vStringValue (fulltag)) )
321 {
322 stringListAdd (ClassNames, vStringNewCopy (fulltag));
323 makeFlexTag (token, FLEXTAG_CLASS);
324 }
325 vStringDelete (fulltag);
326 }
327 }
328
329 static void makeMXTag (tokenInfo *const token)
330 {
331 vString * fulltag;
332
333 if ( ! token->ignoreTag )
334 {
335 fulltag = vStringNew ();
336 if (vStringLength (token->scope) > 0)
337 {
338 vStringCopy(fulltag, token->scope);
339 vStringCatS (fulltag, ".");
340 vStringCatS (fulltag, vStringValue(token->string));
341 }
342 else
343 {
344 vStringCopy(fulltag, token->string);
345 }
346 vStringTerminate(fulltag);
347 makeFlexTag (token, FLEXTAG_MXTAG);
348 vStringDelete (fulltag);
349 }
350 }
351
352 static void makeFunctionTag (tokenInfo *const token)
353 {
354 vString * fulltag;
355
356 if ( ! token->ignoreTag )
357 {
358 fulltag = vStringNew ();
359 if (vStringLength (token->scope) > 0)
360 {
361 vStringCopy(fulltag, token->scope);
362 vStringCatS (fulltag, ".");
363 vStringCatS (fulltag, vStringValue(token->string));
364 }
365 else
366 {
367 vStringCopy(fulltag, token->string);
368 }
369 vStringTerminate(fulltag);
370 if ( ! stringListHas(FunctionNames, vStringValue (fulltag)) )
371 {
372 stringListAdd (FunctionNames, vStringNewCopy (fulltag));
373 makeFlexTag (token, FLEXTAG_FUNCTION);
374 }
375 vStringDelete (fulltag);
376 }
377 }
378
379 /*
380 * Parsing functions
381 */
382
383 static void parseString (vString *const string, const int delimiter)
384 {
385 boolean end = FALSE;
386 while (! end)
387 {
388 int c = fileGetc ();
389 if (c == EOF)
390 end = TRUE;
391 else if (c == '\\')
392 {
393 c = fileGetc(); /* This maybe a ' or ". */
394 vStringPut(string, c);
395 }
396 else if (c == delimiter)
397 end = TRUE;
398 else
399 vStringPut (string, c);
400 }
401 vStringTerminate (string);
402 }
403
404 /* Read a C identifier beginning with "firstChar" and places it into
405 * "name".
406 */
407 static void parseIdentifier (vString *const string, const int firstChar)
408 {
409 int c = firstChar;
410 Assert (isIdentChar (c));
411 do
412 {
413 vStringPut (string, c);
414 c = fileGetc ();
415 } while (isIdentChar (c));
416 vStringTerminate (string);
417 if (!isspace (c))
418 fileUngetc (c); /* unget non-identifier character */
419 }
420
421 static void readToken (tokenInfo *const token)
422 {
423 int c;
424
425 token->type = TOKEN_UNDEFINED;
426 token->keyword = KEYWORD_NONE;
427 vStringClear (token->string);
428
429 getNextChar:
430 do
431 {
432 c = fileGetc ();
433 token->lineNumber = getSourceLineNumber ();
434 token->filePosition = getInputFilePosition ();
435 }
436 while (c == '\t' || c == ' ' || c == '\n');
437
438 switch (c)
439 {
440 case EOF: longjmp (Exception, (int)ExceptionEOF); break;
441 case '(': token->type = TOKEN_OPEN_PAREN; break;
442 case ')': token->type = TOKEN_CLOSE_PAREN; break;
443 case ';': token->type = TOKEN_SEMICOLON; break;
444 case ',': token->type = TOKEN_COMMA; break;
445 case '.': token->type = TOKEN_PERIOD; break;
446 case ':': token->type = TOKEN_COLON; break;
447 case '{': token->type = TOKEN_OPEN_CURLY; break;
448 case '}': token->type = TOKEN_CLOSE_CURLY; break;
449 case '=': token->type = TOKEN_EQUAL_SIGN; break;
450 case '[': token->type = TOKEN_OPEN_SQUARE; break;
451 case ']': token->type = TOKEN_CLOSE_SQUARE; break;
452 case '?': token->type = TOKEN_QUESTION_MARK; break;
453
454 case '\'':
455 case '"':
456 token->type = TOKEN_STRING;
457 parseString (token->string, c);
458 token->lineNumber = getSourceLineNumber ();
459 token->filePosition = getInputFilePosition ();
460 break;
461
462 case '\\':
463 c = fileGetc ();
464 if (c != '\\' && c != '"' && !isspace (c))
465 fileUngetc (c);
466 token->type = TOKEN_CHARACTER;
467 token->lineNumber = getSourceLineNumber ();
468 token->filePosition = getInputFilePosition ();
469 break;
470
471 case '/':
472 {
473 int d = fileGetc ();
474 if ( (d != '*') && /* is this the start of a comment? */
475 (d != '/') && /* is a one line comment? */
476 (d != '>') ) /* is this a close XML tag? */
477 {
478 fileUngetc (d);
479 token->type = TOKEN_FORWARD_SLASH;
480 token->lineNumber = getSourceLineNumber ();
481 token->filePosition = getInputFilePosition ();
482 }
483 else
484 {
485 if (d == '*')
486 {
487 do
488 {
489 fileSkipToCharacter ('*');
490 c = fileGetc ();
491 if (c == '/')
492 break;
493 else
494 fileUngetc (c);
495 } while (c != EOF && c != '\0');
496 goto getNextChar;
497 }
498 else if (d == '/') /* is this the start of a comment? */
499 {
500 fileSkipToCharacter ('\n');
501 goto getNextChar;
502 }
503 else if (d == '>') /* is this the start of a comment? */
504 {
505 token->type = TOKEN_CLOSE_SGML;
506 token->lineNumber = getSourceLineNumber ();
507 token->filePosition = getInputFilePosition ();
508 }
509 }
510 break;
511 }
512
513 case '<':
514 {
515 /*
516 * An XML comment looks like this
517 * <!-- anything over multiple lines -->
518 */
519 int d = fileGetc ();
520
521 if ( (d != '!' ) && /* is this the start of a comment? */
522 (d != '/' ) && /* is this the start of a closing mx tag */
523 (d != 'm' ) ) /* is this the start of a mx tag */
524 {
525 fileUngetc (d);
526 token->type = TOKEN_LESS_THAN;
527 token->lineNumber = getSourceLineNumber ();
528 token->filePosition = getInputFilePosition ();
529
530 }
531 else
532 {
533 if (d == '!')
534 {
535 int e = fileGetc ();
536 if ( e != '-' ) /* is this the start of a comment? */
537 {
538 fileUngetc (e);
539 fileUngetc (d);
540 token->type = TOKEN_LESS_THAN;
541 token->lineNumber = getSourceLineNumber ();
542 token->filePosition = getInputFilePosition ();
543 }
544 else
545 {
546 if (e == '-')
547 {
548 int f = fileGetc ();
549 if ( f != '-' ) /* is this the start of a comment? */
550 {
551 fileUngetc (f);
552 fileUngetc (e);
553 fileUngetc (d);
554 token->type = TOKEN_LESS_THAN;
555 token->lineNumber = getSourceLineNumber ();
556 token->filePosition = getInputFilePosition ();
557 }
558 else
559 {
560 if (f == '-')
561 {
562 do
563 {
564 fileSkipToCharacter ('-');
565 c = fileGetc ();
566 if (c == '-')
567 {
568 d = fileGetc ();
569 if (d == '>')
570 break;
571 else
572 {
573 fileUngetc (d);
574 fileUngetc (c);
575 }
576 break;
577 }
578 else
579 fileUngetc (c);
580 } while (c != EOF && c != '\0');
581 goto getNextChar;
582 }
583 }
584 }
585 }
586 }
587 else if (d == 'm')
588 {
589 int e = fileGetc ();
590 if ( e != 'x' ) /* continuing an mx tag */
591 {
592 fileUngetc (e);
593 fileUngetc (d);
594 token->type = TOKEN_LESS_THAN;
595 token->lineNumber = getSourceLineNumber ();
596 token->filePosition = getInputFilePosition ();
597 }
598 else
599 {
600 if (e == 'x')
601 {
602 int f = fileGetc ();
603 if ( f != ':' ) /* is this the start of a comment? */
604 {
605 fileUngetc (f);
606 fileUngetc (e);
607 fileUngetc (d);
608 token->type = TOKEN_LESS_THAN;
609 token->lineNumber = getSourceLineNumber ();
610 token->filePosition = getInputFilePosition ();
611 }
612 else
613 {
614 if (f == ':')
615 {
616 token->type = TOKEN_OPEN_MXML;
617 token->lineNumber = getSourceLineNumber ();
618 token->filePosition = getInputFilePosition ();
619 }
620 }
621 }
622 }
623 }
624 else if (d == '/')
625 {
626 int e = fileGetc ();
627 if ( e != 'm' ) /* continuing an mx tag */
628 {
629 fileUngetc (e);
630 fileUngetc (d);
631 token->type = TOKEN_LESS_THAN;
632 token->lineNumber = getSourceLineNumber ();
633 token->filePosition = getInputFilePosition ();
634 }
635 else
636 {
637 int f = fileGetc ();
638 if ( f != 'x' ) /* continuing an mx tag */
639 {
640 fileUngetc (f);
641 fileUngetc (e);
642 token->type = TOKEN_LESS_THAN;
643 token->lineNumber = getSourceLineNumber ();
644 token->filePosition = getInputFilePosition ();
645 }
646 else
647 {
648 if (f == 'x')
649 {
650 int g = fileGetc ();
651 if ( g != ':' ) /* is this the start of a comment? */
652 {
653 fileUngetc (g);
654 fileUngetc (f);
655 fileUngetc (e);
656 token->type = TOKEN_LESS_THAN;
657 token->lineNumber = getSourceLineNumber ();
658 token->filePosition = getInputFilePosition ();
659 }
660 else
661 {
662 if (g == ':')
663 {
664 token->type = TOKEN_CLOSE_MXML;
665 token->lineNumber = getSourceLineNumber ();
666 token->filePosition = getInputFilePosition ();
667 }
668 }
669 }
670 }
671 }
672 }
673 }
674 break;
675 }
676
677 case '>':
678 token->type = TOKEN_GREATER_THAN;
679 token->lineNumber = getSourceLineNumber ();
680 token->filePosition = getInputFilePosition ();
681 break;
682
683 case '!':
684 token->type = TOKEN_EXCLAMATION;
685 /*token->lineNumber = getSourceLineNumber ();
686 token->filePosition = getInputFilePosition ();*/
687 break;
688
689 default:
690 if (! isIdentChar (c))
691 token->type = TOKEN_UNDEFINED;
692 else
693 {
694 parseIdentifier (token->string, c);
695 token->lineNumber = getSourceLineNumber ();
696 token->filePosition = getInputFilePosition ();
697 token->keyword = analyzeToken (token->string, Lang_js);
698 if (isKeyword (token, KEYWORD_NONE))
699 token->type = TOKEN_IDENTIFIER;
700 else
701 token->type = TOKEN_KEYWORD;
702 }
703 break;
704 }
705 }
706
707 static void copyToken (tokenInfo *const dest, tokenInfo *const src)
708 {
709 dest->nestLevel = src->nestLevel;
710 dest->lineNumber = src->lineNumber;
711 dest->filePosition = src->filePosition;
712 dest->type = src->type;
713 dest->keyword = src->keyword;
714 dest->isClass = src->isClass;
715 vStringCopy(dest->string, src->string);
716 vStringCopy(dest->scope, src->scope);
717 }
718
719 /*
720 * Token parsing functions
721 */
722
723 static void skipArgumentList (tokenInfo *const token)
724 {
725 int nest_level = 0;
726
727 /*
728 * Other databases can have arguments with fully declared
729 * datatypes:
730 * ( name varchar(30), text binary(10) )
731 * So we must check for nested open and closing parentheses
732 */
733
734 if (isType (token, TOKEN_OPEN_PAREN)) /* arguments? */
735 {
736 nest_level++;
737 while (! (isType (token, TOKEN_CLOSE_PAREN) && (nest_level == 0)))
738 {
739 readToken (token);
740 if (isType (token, TOKEN_OPEN_PAREN))
741 {
742 nest_level++;
743 }
744 if (isType (token, TOKEN_CLOSE_PAREN))
745 {
746 if (nest_level > 0)
747 {
748 nest_level--;
749 }
750 }
751 }
752 readToken (token);
753 }
754 }
755
756 static void skipArrayList (tokenInfo *const token)
757 {
758 int nest_level = 0;
759
760 /*
761 * Handle square brackets
762 * var name[1]
763 * So we must check for nested open and closing square brackets
764 */
765
766 if (isType (token, TOKEN_OPEN_SQUARE)) /* arguments? */
767 {
768 nest_level++;
769 while (! (isType (token, TOKEN_CLOSE_SQUARE) && (nest_level == 0)))
770 {
771 readToken (token);
772 if (isType (token, TOKEN_OPEN_SQUARE))
773 {
774 nest_level++;
775 }
776 if (isType (token, TOKEN_CLOSE_SQUARE))
777 {
778 if (nest_level > 0)
779 {
780 nest_level--;
781 }
782 }
783 }
784 readToken (token);
785 }
786 }
787
788 static void addContext (tokenInfo* const parent, const tokenInfo* const child)
789 {
790 if (vStringLength (parent->string) > 0)
791 {
792 vStringCatS (parent->string, ".");
793 }
794 vStringCatS (parent->string, vStringValue(child->string));
795 vStringTerminate(parent->string);
796 }
797
798 static void addToScope (tokenInfo* const token, vString* const extra)
799 {
800 if (vStringLength (token->scope) > 0)
801 {
802 vStringCatS (token->scope, ".");
803 }
804 vStringCatS (token->scope, vStringValue(extra));
805 vStringTerminate(token->scope);
806 }
807
808 /*
809 * Scanning functions
810 */
811
812 static void findCmdTerm (tokenInfo *const token)
813 {
814 /*
815 * Read until we find either a semicolon or closing brace.
816 * Any nested braces will be handled within.
817 */
818 while (! ( isType (token, TOKEN_SEMICOLON) ||
819 isType (token, TOKEN_CLOSE_CURLY) ) )
820 {
821 /* Handle nested blocks */
822 if ( isType (token, TOKEN_OPEN_CURLY))
823 {
824 parseBlock (token, token);
825 }
826 else if ( isType (token, TOKEN_OPEN_PAREN) )
827 {
828 skipArgumentList(token);
829 }
830 else
831 {
832 readToken (token);
833 }
834 }
835 }
836
837 static void parseSwitch (tokenInfo *const token)
838 {
839 /*
840 * switch (expression){
841 * case value1:
842 * statement;
843 * break;
844 * case value2:
845 * statement;
846 * break;
847 * default : statement;
848 * }
849 */
850
851 readToken (token);
852
853 if (isType (token, TOKEN_OPEN_PAREN))
854 {
855 skipArgumentList(token);
856 }
857
858 if (isType (token, TOKEN_OPEN_CURLY))
859 {
860 do
861 {
862 readToken (token);
863 } while (! (isType (token, TOKEN_CLOSE_SGML) ||
864 isType (token, TOKEN_CLOSE_MXML) ||
865 isType (token, TOKEN_CLOSE_CURLY) ||
866 isType (token, TOKEN_GREATER_THAN)) );
867 }
868
869 }
870
871 static void parseLoop (tokenInfo *const token)
872 {
873 /*
874 * Handles these statements
875 * for (x=0; x<3; x++)
876 * document.write("This text is repeated three times<br>");
877 *
878 * for (x=0; x<3; x++)
879 * {
880 * document.write("This text is repeated three times<br>");
881 * }
882 *
883 * while (number<5){
884 * document.write(number+"<br>");
885 * number++;
886 * }
887 *
888 * do{
889 * document.write(number+"<br>");
890 * number++;
891 * }
892 * while (number<5);
893 */
894
895 if (isKeyword (token, KEYWORD_for) || isKeyword (token, KEYWORD_while))
896 {
897 readToken(token);
898
899 if (isType (token, TOKEN_OPEN_PAREN))
900 {
901 /*
902 * Handle nameless functions, these will only
903 * be considered methods.
904 */
905 skipArgumentList(token);
906 }
907
908 if (isType (token, TOKEN_OPEN_CURLY))
909 {
910 /*
911 * This will be either a function or a class.
912 * We can only determine this by checking the body
913 * of the function. If we find a "this." we know
914 * it is a class, otherwise it is a function.
915 */
916 parseBlock (token, token);
917 }
918 else
919 {
920 parseLine(token);
921 }
922 }
923 else if (isKeyword (token, KEYWORD_do))
924 {
925 readToken(token);
926
927 if (isType (token, TOKEN_OPEN_CURLY))
928 {
929 /*
930 * This will be either a function or a class.
931 * We can only determine this by checking the body
932 * of the function. If we find a "this." we know
933 * it is a class, otherwise it is a function.
934 */
935 parseBlock (token, token);
936 }
937 else
938 {
939 parseLine(token);
940 }
941
942 readToken(token);
943
944 if (isKeyword (token, KEYWORD_while))
945 {
946 readToken(token);
947
948 if (isType (token, TOKEN_OPEN_PAREN))
949 {
950 /*
951 * Handle nameless functions, these will only
952 * be considered methods.
953 */
954 skipArgumentList(token);
955 }
956 }
957 }
958 }
959
960 static boolean parseIf (tokenInfo *const token)
961 {
962 boolean read_next_token = TRUE;
963 /*
964 * If statements have two forms
965 * if ( ... )
966 * one line;
967 *
968 * if ( ... )
969 * statement;
970 * else
971 * statement
972 *
973 * if ( ... ) {
974 * multiple;
975 * statements;
976 * }
977 *
978 *
979 * if ( ... ) {
980 * return elem
981 * }
982 *
983 * This example if correctly written, but the
984 * else contains only 1 statement without a terminator
985 * since the function finishes with the closing brace.
986 *
987 * function a(flag){
988 * if(flag)
989 * test(1);
990 * else
991 * test(2)
992 * }
993 *
994 * TODO: Deal with statements that can optional end
995 * without a semi-colon. Currently this messes up
996 * the parsing of blocks.
997 * Need to somehow detect this has happened, and either
998 * backup a token, or skip reading the next token if
999 * that is possible from all code locations.
1000 *
1001 */
1002
1003 readToken (token);
1004
1005 if (isKeyword (token, KEYWORD_if))
1006 {
1007 /*
1008 * Check for an "else if" and consume the "if"
1009 */
1010 readToken (token);
1011 }
1012
1013 if (isType (token, TOKEN_OPEN_PAREN))
1014 {
1015 /*
1016 * Handle nameless functions, these will only
1017 * be considered methods.
1018 */
1019 skipArgumentList(token);
1020 }
1021
1022 if (isType (token, TOKEN_OPEN_CURLY))
1023 {
1024 /*
1025 * This will be either a function or a class.
1026 * We can only determine this by checking the body
1027 * of the function. If we find a "this." we know
1028 * it is a class, otherwise it is a function.
1029 */
1030 parseBlock (token, token);
1031 }
1032 else
1033 {
1034 findCmdTerm (token);
1035
1036 /*
1037 * The IF could be followed by an ELSE statement.
1038 * This too could have two formats, a curly braced
1039 * multiline section, or another single line.
1040 */
1041
1042 if (isType (token, TOKEN_CLOSE_CURLY))
1043 {
1044 /*
1045 * This statement did not have a line terminator.
1046 */
1047 read_next_token = FALSE;
1048 }
1049 else
1050 {
1051 readToken (token);
1052
1053 if (isType (token, TOKEN_CLOSE_CURLY))
1054 {
1055 /*
1056 * This statement did not have a line terminator.
1057 */
1058 read_next_token = FALSE;
1059 }
1060 else
1061 {
1062 if (isKeyword (token, KEYWORD_else))
1063 read_next_token = parseIf (token);
1064 }
1065 }
1066 }
1067 return read_next_token;
1068 }
1069
1070 static void parseFunction (tokenInfo *const token)
1071 {
1072 tokenInfo *const name = newToken ();
1073
1074 /*
1075 * This deals with these formats
1076 * private static function ioErrorHandler( event:IOErrorEvent ):void {
1077 */
1078
1079 if ( isKeyword(token, KEYWORD_function) )
1080 {
1081 readToken (token);
1082 }
1083
1084 copyToken (name, token);
1085 /* Add scope in case this is an INNER function
1086 addToScope(name, token->scope);
1087 */
1088
1089 DebugStatement (
1090 debugPrintf (DEBUG_PARSE
1091 , "\n parseFunction: token isClass:%d scope:%s name:%s\n"
1092 , token->isClass
1093 , vStringValue(token->scope)
1094 , vStringValue(token->string)
1095 );
1096 );
1097 DebugStatement (
1098 debugPrintf (DEBUG_PARSE
1099 , "\n parseFunction: name isClass:%d scope:%s name:%s\n"
1100 , name->isClass
1101 , vStringValue(name->scope)
1102 , vStringValue(name->string)
1103 );
1104 );
1105
1106 readToken (token);
1107
1108 if ( isType (token, TOKEN_OPEN_PAREN) )
1109 skipArgumentList(token);
1110
1111 if ( isType (token, TOKEN_COLON) )
1112 {
1113 /*
1114 * function fname ():ReturnType
1115 */
1116 readToken (token);
1117 readToken (token);
1118 }
1119
1120 if ( isType (token, TOKEN_OPEN_CURLY) )
1121 {
1122 DebugStatement (
1123 debugPrintf (DEBUG_PARSE
1124 , "\n parseFunction end: name isClass:%d scope:%s name:%s\n"
1125 , name->isClass
1126 , vStringValue(name->scope)
1127 , vStringValue(name->string)
1128 );
1129 );
1130 parseBlock (token, name);
1131 DebugStatement (
1132 debugPrintf (DEBUG_PARSE
1133 , "\n parseFunction end2: token isClass:%d scope:%s name:%s\n"
1134 , token->isClass
1135 , vStringValue(token->scope)
1136 , vStringValue(token->string)
1137 );
1138 );
1139 DebugStatement (
1140 debugPrintf (DEBUG_PARSE
1141 , "\n parseFunction end2: token isClass:%d scope:%s name:%s\n"
1142 , token->isClass
1143 , vStringValue(token->scope)
1144 , vStringValue(token->string)
1145 );
1146 );
1147 DebugStatement (
1148 debugPrintf (DEBUG_PARSE
1149 , "\n parseFunction end3: name isClass:%d scope:%s name:%s\n"
1150 , name->isClass
1151 , vStringValue(name->scope)
1152 , vStringValue(name->string)
1153 );
1154 );
1155 makeFunctionTag (name);
1156 }
1157
1158 findCmdTerm (token);
1159
1160 deleteToken (name);
1161 }
1162
1163 static boolean parseBlock (tokenInfo *const token, tokenInfo *const parent)
1164 {
1165 boolean read_next_token = TRUE;
1166 vString * saveScope = vStringNew ();
1167
1168 vStringClear(saveScope);
1169 vStringCopy (saveScope, token->scope);
1170 token->nestLevel++;
1171 DebugStatement (
1172 debugPrintf (DEBUG_PARSE
1173 , "\n parseBlock start: token isClass:%d scope:%s name:%s\n"
1174 , token->isClass
1175 , vStringValue(token->scope)
1176 , vStringValue(token->string)
1177 );
1178 );
1179 /*
1180 * Make this routine a bit more forgiving.
1181 * If called on an open_curly advance it
1182 */
1183 if ( isType (token, TOKEN_OPEN_CURLY) &&
1184 isKeyword(token, KEYWORD_NONE) )
1185 readToken(token);
1186
1187 if (! isType (token, TOKEN_CLOSE_CURLY))
1188 {
1189 /*
1190 * Read until we find the closing brace,
1191 * any nested braces will be handled within
1192 */
1193 do
1194 {
1195 if (isType (token, TOKEN_OPEN_CURLY))
1196 {
1197 /* Handle nested blocks */
1198 parseBlock (token, parent);
1199 }
1200 else
1201 {
1202 /*
1203 * It is possible for a line to have no terminator
1204 * if the following line is a closing brace.
1205 * parseLine will detect this case and indicate
1206 * whether we should read an additional token.
1207 */
1208 read_next_token = parseLine (token);
1209 }
1210
1211 /*
1212 * Always read a new token unless we find a statement without
1213 * a ending terminator
1214 */
1215 if( read_next_token )
1216 readToken(token);
1217
1218 /*
1219 * If we find a statement without a terminator consider the
1220 * block finished, otherwise the stack will be off by one.
1221 */
1222 } while (! isType (token, TOKEN_CLOSE_CURLY) && read_next_token );
1223 }
1224
1225 vStringDelete(saveScope);
1226 token->nestLevel--;
1227
1228 DebugStatement (
1229 debugPrintf (DEBUG_PARSE
1230 , "\n parseBlock end: token isClass:%d scope:%s name:%s\n"
1231 , token->isClass
1232 , vStringValue(token->scope)
1233 , vStringValue(token->string)
1234 );
1235 );
1236 return FALSE;
1237 }
1238
1239 static void parseMethods (tokenInfo *const token, tokenInfo *const class)
1240 {
1241 tokenInfo *const name = newToken ();
1242
1243 /*
1244 * This deals with these formats
1245 * validProperty : 2,
1246 * validMethod : function(a,b) {}
1247 * 'validMethod2' : function(a,b) {}
1248 * container.dirtyTab = {'url': false, 'title':false, 'snapshot':false, '*': false}
1249 */
1250
1251 do
1252 {
1253 readToken (token);
1254 if (isType (token, TOKEN_STRING) || isKeyword(token, KEYWORD_NONE))
1255 {
1256 copyToken (name, token);
1257
1258 readToken (token);
1259 if ( isType (token, TOKEN_COLON) )
1260 {
1261 readToken (token);
1262 if ( isKeyword (token, KEYWORD_function) )
1263 {
1264 readToken (token);
1265 if ( isType (token, TOKEN_OPEN_PAREN) )
1266 {
1267 skipArgumentList(token);
1268 }
1269
1270 if (isType (token, TOKEN_OPEN_CURLY))
1271 {
1272 addToScope (name, class->string);
1273 makeFlexTag (name, FLEXTAG_METHOD);
1274 parseBlock (token, name);
1275
1276 /*
1277 * Read to the closing curly, check next
1278 * token, if a comma, we must loop again
1279 */
1280 readToken (token);
1281 }
1282 }
1283 else
1284 {
1285 addToScope (name, class->string);
1286 makeFlexTag (name, FLEXTAG_PROPERTY);
1287
1288 /*
1289 * Read the next token, if a comma
1290 * we must loop again
1291 */
1292 readToken (token);
1293 }
1294 }
1295 }
1296 } while ( isType(token, TOKEN_COMMA) );
1297
1298 findCmdTerm (token);
1299
1300 deleteToken (name);
1301 }
1302
1303 static boolean parseVar (tokenInfo *const token, boolean is_public)
1304 {
1305 tokenInfo *const name = newToken ();
1306 tokenInfo *const secondary_name = newToken ();
1307 vString * saveScope = vStringNew ();
1308 boolean is_terminated = TRUE;
1309
1310 vStringClear(saveScope);
1311 vStringCopy (saveScope, token->scope);
1312 /*
1313 * Variables are defined as:
1314 * private static var lastFaultMessage:Date = new Date( 0 );
1315 * private static var webRequests:ArrayCollection = new ArrayCollection();
1316 */
1317
1318 if ( isKeyword(token, KEYWORD_var) )
1319 {
1320 readToken(token);
1321 }
1322
1323 /* Variable name */
1324 copyToken (name, token);
1325 readToken(token);
1326
1327 if ( isType (token, TOKEN_COLON) )
1328 {
1329 /*
1330 * var vname ():DataType = new Date();
1331 * var vname ():DataType;
1332 */
1333 readToken (token);
1334 readToken (token);
1335 }
1336
1337 while (! isType (token, TOKEN_SEMICOLON) )
1338 {
1339 readToken (token);
1340 }
1341
1342 if ( isType (token, TOKEN_SEMICOLON) )
1343 {
1344 /*
1345 * Only create variables for global scope
1346 */
1347 /* if ( token->nestLevel == 0 && is_global ) */
1348 if ( is_public )
1349 {
1350 if (isType (token, TOKEN_SEMICOLON))
1351 makeFlexTag (name, FLEXTAG_VARIABLE);
1352 }
1353 }
1354
1355 vStringCopy(token->scope, saveScope);
1356 deleteToken (name);
1357 deleteToken (secondary_name);
1358 vStringDelete(saveScope);
1359
1360 return is_terminated;
1361 }
1362
1363 static boolean parseClass (tokenInfo *const token)
1364 {
1365 tokenInfo *const name = newToken ();
1366 vString * saveScope = vStringNew ();
1367 boolean saveIsClass = token->isClass;
1368
1369 vStringClear(saveScope);
1370 vStringCopy (saveScope, token->scope);
1371 /*
1372 * Variables are defined as:
1373 * private static var lastFaultMessage:Date = new Date( 0 );
1374 * private static var webRequests:ArrayCollection = new ArrayCollection();
1375 */
1376
1377 if ( isKeyword(token, KEYWORD_class) )
1378 {
1379 readToken(token);
1380 }
1381
1382 token->isClass = TRUE;
1383 /* Add class name to scope */
1384 addToScope(token, token->string);
1385 /* Class name */
1386 copyToken (name, token);
1387 readToken(token);
1388
1389 DebugStatement (
1390 debugPrintf (DEBUG_PARSE
1391 , "\n parseClass start: token isClass:%d scope:%s name:%s\n"
1392 , token->isClass
1393 , vStringValue(token->scope)
1394 , vStringValue(token->string)
1395 );
1396 );
1397 if ( isType (token, TOKEN_OPEN_CURLY) )
1398 {
1399 makeClassTag (name);
1400 parseBlock (token, name);
1401 }
1402
1403 DebugStatement (
1404 debugPrintf (DEBUG_PARSE
1405 , "\n parseClass end: token isClass:%d scope:%s name:%s\n"
1406 , token->isClass
1407 , vStringValue(token->scope)
1408 , vStringValue(token->string)
1409 );
1410 );
1411 vStringCopy(token->scope, saveScope);
1412 token->isClass = saveIsClass;
1413 deleteToken (name);
1414 vStringDelete(saveScope);
1415
1416 return TRUE;
1417 }
1418
1419 static boolean parseStatement (tokenInfo *const token)
1420 {
1421 tokenInfo *const name = newToken ();
1422 tokenInfo *const secondary_name = newToken ();
1423 vString * saveScope = vStringNew ();
1424 boolean is_public = FALSE;
1425 boolean is_class = FALSE;
1426 boolean is_terminated = TRUE;
1427 boolean is_global = FALSE;
1428 boolean is_prototype = FALSE;
1429 vString * fulltag;
1430
1431 vStringClear(saveScope);
1432 vStringCopy (saveScope, token->scope);
1433 DebugStatement (
1434 debugPrintf (DEBUG_PARSE
1435 , "\n parseStatement: token isClass:%d scope:%s name:%s\n"
1436 , token->isClass
1437 , vStringValue(token->scope)
1438 , vStringValue(token->string)
1439 );
1440 );
1441 /*
1442 * Functions can be named or unnamed.
1443 * This deals with these formats:
1444 * Function
1445 * validFunctionOne = function(a,b) {}
1446 * testlib.validFunctionFive = function(a,b) {}
1447 * var innerThree = function(a,b) {}
1448 * var innerFour = (a,b) {}
1449 * var D2 = secondary_fcn_name(a,b) {}
1450 * var D3 = new Function("a", "b", "return a+b;");
1451 * Class
1452 * testlib.extras.ValidClassOne = function(a,b) {
1453 * this.a = a;
1454 * }
1455 * Class Methods
1456 * testlib.extras.ValidClassOne.prototype = {
1457 * 'validMethodOne' : function(a,b) {},
1458 * 'validMethodTwo' : function(a,b) {}
1459 * }
1460 * ValidClassTwo = function ()
1461 * {
1462 * this.validMethodThree = function() {}
1463 * // unnamed method
1464 * this.validMethodFour = () {}
1465 * }
1466 * Database.prototype.validMethodThree = Database_getTodaysDate;
1467 */
1468
1469 if ( isKeyword(token, KEYWORD_public) )
1470 {
1471 is_public = TRUE;
1472 readToken(token);
1473 }
1474
1475 if ( isKeyword(token, KEYWORD_private) )
1476 {
1477 readToken(token);
1478 }
1479
1480 if ( isKeyword(token, KEYWORD_static) )
1481 {
1482 readToken(token);
1483 }
1484
1485 if (isType(token, TOKEN_KEYWORD))
1486 {
1487 switch (token->keyword)
1488 {
1489 case KEYWORD_for:
1490 case KEYWORD_while:
1491 case KEYWORD_do:
1492 parseLoop (token);
1493 break;
1494 case KEYWORD_if:
1495 case KEYWORD_else:
1496 case KEYWORD_try:
1497 case KEYWORD_catch:
1498 case KEYWORD_finally:
1499 /* Common semantics */
1500 is_terminated = parseIf (token);
1501 break;
1502 case KEYWORD_switch:
1503 parseSwitch (token);
1504 break;
1505 case KEYWORD_class:
1506 parseClass (token);
1507 return is_terminated;
1508 break;
1509 case KEYWORD_function:
1510 parseFunction (token);
1511 return is_terminated;
1512 break;
1513 case KEYWORD_var:
1514 parseVar (token, is_public);
1515 return is_terminated;
1516 break;
1517 default:
1518 readToken(token);
1519 break;
1520 }
1521 }
1522
1523 copyToken (name, token);
1524
1525 while (! isType (token, TOKEN_CLOSE_CURLY) &&
1526 ! isType (token, TOKEN_SEMICOLON) &&
1527 ! isType (token, TOKEN_EQUAL_SIGN) )
1528 {
1529 /* Potentially the name of the function */
1530 readToken (token);
1531 if (isType (token, TOKEN_PERIOD))
1532 {
1533 /*
1534 * Cannot be a global variable is it has dot references in the name
1535 */
1536 is_global = FALSE;
1537 do
1538 {
1539 readToken (token);
1540 if ( isKeyword(token, KEYWORD_NONE) )
1541 {
1542 if ( is_class )
1543 {
1544 vStringCopy(saveScope, token->scope);
1545 addToScope(token, name->string);
1546 }
1547 else
1548 addContext (name, token);
1549 }
1550 else if ( isKeyword(token, KEYWORD_prototype) )
1551 {
1552 /*
1553 * When we reach the "prototype" tag, we infer:
1554 * "BindAgent" is a class
1555 * "build" is a method
1556 *
1557 * function BindAgent( repeatableIdName, newParentIdName ) {
1558 * }
1559 *
1560 * CASE 1
1561 * Specified function name: "build"
1562 * BindAgent.prototype.build = function( mode ) {
1563 * ignore everything within this function
1564 * }
1565 *
1566 * CASE 2
1567 * Prototype listing
1568 * ValidClassOne.prototype = {
1569 * 'validMethodOne' : function(a,b) {},
1570 * 'validMethodTwo' : function(a,b) {}
1571 * }
1572 *
1573 */
1574 makeClassTag (name);
1575 is_class = TRUE;
1576 is_prototype = TRUE;
1577
1578 /*
1579 * There should a ".function_name" next.
1580 */
1581 readToken (token);
1582 if (isType (token, TOKEN_PERIOD))
1583 {
1584 /*
1585 * Handle CASE 1
1586 */
1587 readToken (token);
1588 if ( isKeyword(token, KEYWORD_NONE) )
1589 {
1590 vStringCopy(saveScope, token->scope);
1591 addToScope(token, name->string);
1592
1593 makeFlexTag (token, FLEXTAG_METHOD);
1594 /*
1595 * We can read until the end of the block / statement.
1596 * We need to correctly parse any nested blocks, but
1597 * we do NOT want to create any tags based on what is
1598 * within the blocks.
1599 */
1600 token->ignoreTag = TRUE;
1601 /*
1602 * Find to the end of the statement
1603 */
1604 findCmdTerm (token);
1605 token->ignoreTag = FALSE;
1606 is_terminated = TRUE;
1607 goto cleanUp;
1608 }
1609 }
1610 else if (isType (token, TOKEN_EQUAL_SIGN))
1611 {
1612 readToken (token);
1613 if (isType (token, TOKEN_OPEN_CURLY))
1614 {
1615 /*
1616 * Handle CASE 2
1617 *
1618 * Creates tags for each of these class methods
1619 * ValidClassOne.prototype = {
1620 * 'validMethodOne' : function(a,b) {},
1621 * 'validMethodTwo' : function(a,b) {}
1622 * }
1623 */
1624 parseMethods(token, name);
1625 /*
1626 * Find to the end of the statement
1627 */
1628 findCmdTerm (token);
1629 token->ignoreTag = FALSE;
1630 is_terminated = TRUE;
1631 goto cleanUp;
1632 }
1633 }
1634 }
1635 readToken (token);
1636 } while (isType (token, TOKEN_PERIOD));
1637 }
1638
1639 if ( isType (token, TOKEN_OPEN_PAREN) )
1640 skipArgumentList(token);
1641
1642 if ( isType (token, TOKEN_COLON) )
1643 {
1644 /*
1645 * Functions are of this form:
1646 * function fname ():ReturnType {
1647 */
1648 readToken (token);
1649 readToken (token);
1650 }
1651
1652 if ( isType (token, TOKEN_OPEN_SQUARE) )
1653 skipArrayList(token);
1654
1655 }
1656
1657 if ( isType (token, TOKEN_CLOSE_CURLY) )
1658 {
1659 /*
1660 * Reaching this section without having
1661 * processed an open curly brace indicates
1662 * the statement is most likely not terminated.
1663 */
1664 is_terminated = FALSE;
1665 goto cleanUp;
1666 }
1667
1668 if ( isType (token, TOKEN_SEMICOLON) )
1669 {
1670 /*
1671 * Only create variables for global scope
1672 */
1673 if ( token->nestLevel == 0 && is_global )
1674 {
1675 /*
1676 * Handles this syntax:
1677 * var g_var2;
1678 */
1679 if (isType (token, TOKEN_SEMICOLON))
1680 makeFlexTag (name, FLEXTAG_VARIABLE);
1681 }
1682 /*
1683 * Statement has ended.
1684 * This deals with calls to functions, like:
1685 * alert(..);
1686 */
1687 goto cleanUp;
1688 }
1689
1690 if ( isType (token, TOKEN_EQUAL_SIGN) )
1691 {
1692 readToken (token);
1693
1694 if ( isKeyword (token, KEYWORD_function) )
1695 {
1696 readToken (token);
1697
1698 if ( isKeyword (token, KEYWORD_NONE) &&
1699 ! isType (token, TOKEN_OPEN_PAREN) )
1700 {
1701 /*
1702 * Functions of this format:
1703 * var D2A = function theAdd(a, b)
1704 * {
1705 * return a+b;
1706 * }
1707 * Are really two separate defined functions and
1708 * can be referenced in two ways:
1709 * alert( D2A(1,2) ); // produces 3
1710 * alert( theAdd(1,2) ); // also produces 3
1711 * So it must have two tags:
1712 * D2A
1713 * theAdd
1714 * Save the reference to the name for later use, once
1715 * we have established this is a valid function we will
1716 * create the secondary reference to it.
1717 */
1718 copyToken (secondary_name, token);
1719 readToken (token);
1720 }
1721
1722 if ( isType (token, TOKEN_OPEN_PAREN) )
1723 skipArgumentList(token);
1724
1725 if (isType (token, TOKEN_OPEN_CURLY))
1726 {
1727 /*
1728 * This will be either a function or a class.
1729 * We can only determine this by checking the body
1730 * of the function. If we find a "this." we know
1731 * it is a class, otherwise it is a function.
1732 */
1733 if ( token->isClass )
1734 {
1735 makeFlexTag (name, FLEXTAG_METHOD);
1736 if ( vStringLength(secondary_name->string) > 0 )
1737 makeFunctionTag (secondary_name);
1738 parseBlock (token, name);
1739 }
1740 else
1741 {
1742 parseBlock (token, name);
1743 makeFunctionTag (name);
1744
1745 if ( vStringLength(secondary_name->string) > 0 )
1746 makeFunctionTag (secondary_name);
1747
1748 /*
1749 * Find to the end of the statement
1750 */
1751 goto cleanUp;
1752 }
1753 }
1754 }
1755 else if (isType (token, TOKEN_OPEN_PAREN))
1756 {
1757 /*
1758 * Handle nameless functions
1759 * this.method_name = () {}
1760 */
1761 skipArgumentList(token);
1762
1763 if (isType (token, TOKEN_OPEN_CURLY))
1764 {
1765 /*
1766 * Nameless functions are only setup as methods.
1767 */
1768 makeFlexTag (name, FLEXTAG_METHOD);
1769 parseBlock (token, name);
1770 }
1771 }
1772 else if (isType (token, TOKEN_OPEN_CURLY))
1773 {
1774 /*
1775 * Creates tags for each of these class methods
1776 * ValidClassOne.prototype = {
1777 * 'validMethodOne' : function(a,b) {},
1778 * 'validMethodTwo' : function(a,b) {}
1779 * }
1780 */
1781 parseMethods(token, name);
1782 if (isType (token, TOKEN_CLOSE_CURLY))
1783 {
1784 /*
1785 * Assume the closing parenthesis terminates
1786 * this statements.
1787 */
1788 is_terminated = TRUE;
1789 }
1790 }
1791 else if (isKeyword (token, KEYWORD_new))
1792 {
1793 readToken (token);
1794 if ( isKeyword (token, KEYWORD_function) ||
1795 isKeyword (token, KEYWORD_capital_function) ||
1796 isKeyword (token, KEYWORD_object) ||
1797 isKeyword (token, KEYWORD_capital_object) )
1798 {
1799 if ( isKeyword (token, KEYWORD_object) ||
1800 isKeyword (token, KEYWORD_capital_object) )
1801 is_class = TRUE;
1802
1803 readToken (token);
1804 if ( isType (token, TOKEN_OPEN_PAREN) )
1805 skipArgumentList(token);
1806
1807 if (isType (token, TOKEN_SEMICOLON))
1808 {
1809 if ( token->nestLevel == 0 )
1810 {
1811 if ( is_class )
1812 {
1813 makeClassTag (name);
1814 } else {
1815 makeFunctionTag (name);
1816 }
1817 }
1818 }
1819 }
1820 }
1821 else if (isKeyword (token, KEYWORD_NONE))
1822 {
1823 /*
1824 * Only create variables for global scope
1825 */
1826 if ( token->nestLevel == 0 && is_global )
1827 {
1828 /*
1829 * A pointer can be created to the function.
1830 * If we recognize the function/class name ignore the variable.
1831 * This format looks identical to a variable definition.
1832 * A variable defined outside of a block is considered
1833 * a global variable:
1834 * var g_var1 = 1;
1835 * var g_var2;
1836 * This is not a global variable:
1837 * var g_var = function;
1838 * This is a global variable:
1839 * var g_var = different_var_name;
1840 */
1841 fulltag = vStringNew ();
1842 if (vStringLength (token->scope) > 0)
1843 {
1844 vStringCopy(fulltag, token->scope);
1845 vStringCatS (fulltag, ".");
1846 vStringCatS (fulltag, vStringValue(token->string));
1847 }
1848 else
1849 {
1850 vStringCopy(fulltag, token->string);
1851 }
1852 vStringTerminate(fulltag);
1853 if ( ! stringListHas(FunctionNames, vStringValue (fulltag)) &&
1854 ! stringListHas(ClassNames, vStringValue (fulltag)) )
1855 {
1856 findCmdTerm (token);
1857 if (isType (token, TOKEN_SEMICOLON))
1858 makeFlexTag (name, FLEXTAG_VARIABLE);
1859 }
1860 vStringDelete (fulltag);
1861 }
1862 }
1863 }
1864 findCmdTerm (token);
1865
1866 /*
1867 * Statements can be optionally terminated in the case of
1868 * statement prior to a close curly brace as in the
1869 * document.write line below:
1870 *
1871 * function checkForUpdate() {
1872 * if( 1==1 ) {
1873 * document.write("hello from checkForUpdate<br>")
1874 * }
1875 * return 1;
1876 * }
1877 */
1878 if ( ! is_terminated && isType (token, TOKEN_CLOSE_CURLY))
1879 is_terminated = FALSE;
1880
1881
1882 cleanUp:
1883 vStringCopy(token->scope, saveScope);
1884 deleteToken (name);
1885 deleteToken (secondary_name);
1886 vStringDelete(saveScope);
1887
1888 return is_terminated;
1889 }
1890
1891 static boolean parseLine (tokenInfo *const token)
1892 {
1893 boolean is_terminated = TRUE;
1894 /*
1895 * Detect the common statements, if, while, for, do, ...
1896 * This is necessary since the last statement within a block "{}"
1897 * can be optionally terminated.
1898 *
1899 * If the statement is not terminated, we need to tell
1900 * the calling routine to prevent reading an additional token
1901 * looking for the end of the statement.
1902 */
1903
1904 if (isType(token, TOKEN_KEYWORD))
1905 {
1906 switch (token->keyword)
1907 {
1908 case KEYWORD_for:
1909 case KEYWORD_while:
1910 case KEYWORD_do:
1911 parseLoop (token);
1912 break;
1913 case KEYWORD_if:
1914 case KEYWORD_else:
1915 case KEYWORD_try:
1916 case KEYWORD_catch:
1917 case KEYWORD_finally:
1918 /* Common semantics */
1919 is_terminated = parseIf (token);
1920 break;
1921 case KEYWORD_switch:
1922 parseSwitch (token);
1923 break;
1924 default:
1925 parseStatement (token);
1926 break;
1927 }
1928 }
1929 else
1930 {
1931 /*
1932 * Special case where single line statements may not be
1933 * SEMICOLON terminated. parseBlock needs to know this
1934 * so that it does not read the next token.
1935 */
1936 is_terminated = parseStatement (token);
1937 }
1938 return is_terminated;
1939 }
1940
1941 static boolean parseCDATA (tokenInfo *const token)
1942 {
1943 if (isType (token, TOKEN_LESS_THAN))
1944 {
1945 /*
1946 * Handle these tags
1947 * <![CDATA[
1948 * ...
1949 * ]]>
1950 */
1951 readToken (token);
1952 if (isType (token, TOKEN_EXCLAMATION))
1953 {
1954 /*
1955 * Not sure why I had to comment these out, but I did.
1956 * readToken (token);
1957 * if (isType (token, TOKEN_OPEN_SQUARE))
1958 * {
1959 */
1960 readToken (token);
1961 if (isKeyword (token, KEYWORD_cdata))
1962 {
1963 readToken (token);
1964 if (isType (token, TOKEN_OPEN_SQUARE))
1965 {
1966 parseActionScript (token);
1967 if (isType (token, TOKEN_CLOSE_SQUARE))
1968 {
1969 readToken (token);
1970 if (isType (token, TOKEN_CLOSE_SQUARE))
1971 {
1972 readToken (token);
1973 }
1974 }
1975 }
1976 }
1977 /*} Not sure */
1978 }
1979 }
1980 else
1981 {
1982 parseActionScript (token);
1983 }
1984 return TRUE;
1985 }
1986
1987 static boolean parseMXML (tokenInfo *const token)
1988 {
1989 tokenInfo *const name = newToken ();
1990 tokenInfo *const type = newToken ();
1991 /*
1992 * Detect the common statements, if, while, for, do, ...
1993 * This is necessary since the last statement within a block "{}"
1994 * can be optionally terminated.
1995 *
1996 * If the statement is not terminated, we need to tell
1997 * the calling routine to prevent reading an additional token
1998 * looking for the end of the statement.
1999 */
2000
2001 readToken (token);
2002
2003 if (isKeyword (token, KEYWORD_script))
2004 {
2005 /*
2006 * These tags can be of this form:
2007 * <mx:Script src="filename.as" />
2008 */
2009 do
2010 {
2011 readToken (token);
2012 } while (! (isType (token, TOKEN_CLOSE_SGML) ||
2013 isType (token, TOKEN_CLOSE_MXML) ||
2014 isType (token, TOKEN_GREATER_THAN)) );
2015
2016 if (isType (token, TOKEN_CLOSE_MXML))
2017 {
2018 /*
2019 * We have found a </mx:type> tag
2020 * Finish reading the "type" and ">"
2021 */
2022 readToken (token);
2023 readToken (token);
2024 goto cleanUp;
2025 }
2026 if (isType (token, TOKEN_CLOSE_SGML))
2027 {
2028 /*
2029 * We have found a <mx:Script src="filename.as" />
2030 */
2031 goto cleanUp;
2032 }
2033
2034 /*
2035 * This is a beginning of an embedded script.
2036 * These typically are of this format:
2037 * <mx:Script>
2038 * <![CDATA[
2039 * ... ActionScript ...
2040 * ]]>
2041 * </mx:Script>
2042 */
2043 readToken (token);
2044 parseCDATA (token);
2045
2046 readToken (token);
2047 if (isType (token, TOKEN_CLOSE_MXML))
2048 {
2049 /*
2050 * We have found a </mx:type> tag
2051 * Finish reading the "type" and ">"
2052 */
2053 readToken (token);
2054 readToken (token);
2055 }
2056 goto cleanUp;
2057 }
2058
2059 copyToken (type, token);
2060
2061 readToken (token);
2062 do
2063 {
2064 if (isType (token, TOKEN_OPEN_MXML))
2065 {
2066 parseMXML (token);
2067 }
2068 else if (isKeyword (token, KEYWORD_id))
2069 {
2070 /* = */
2071 readToken (token);
2072 readToken (token);
2073
2074 copyToken (name, token);
2075 addToScope (name, type->string);
2076 makeMXTag (name);
2077 }
2078 readToken (token);
2079 } while (! (isType (token, TOKEN_CLOSE_SGML) || isType (token, TOKEN_CLOSE_MXML)) );
2080
2081 if (isType (token, TOKEN_CLOSE_MXML))
2082 {
2083 /*
2084 * We have found a </mx:type> tag
2085 * Finish reading the "type" and ">"
2086 */
2087 readToken (token);
2088 readToken (token);
2089 }
2090
2091 cleanUp:
2092 deleteToken (name);
2093 deleteToken (type);
2094 return TRUE;
2095 }
2096
2097 static boolean parseActionScript (tokenInfo *const token)
2098 {
2099 do
2100 {
2101 readToken (token);
2102
2103 if (isType (token, TOKEN_LESS_THAN))
2104 {
2105 /*
2106 * Handle these tags
2107 * <![CDATA[
2108 * ...
2109 * ]]>
2110 */
2111 readToken (token);
2112 if (isType (token, TOKEN_EQUAL_SIGN))
2113 {
2114 if (isType (token, TOKEN_OPEN_SQUARE))
2115 {
2116 readToken (token);
2117 if (isKeyword (token, KEYWORD_cdata))
2118 {
2119 readToken (token);
2120 }
2121 }
2122 }
2123 }
2124 if (isType (token, TOKEN_CLOSE_SQUARE))
2125 {
2126 /*
2127 * Handle these tags
2128 * <![CDATA[
2129 * ...
2130 * ]]>
2131 */
2132 readToken (token);
2133 if (isType (token, TOKEN_CLOSE_SQUARE))
2134 {
2135 readToken (token);
2136 if (isType (token, TOKEN_GREATER_THAN))
2137 {
2138 return TRUE;
2139 }
2140 }
2141 }
2142 else if (isType (token, TOKEN_CLOSE_MXML))
2143 {
2144 /*
2145 * Read the Script> tags
2146 */
2147 readToken (token);
2148 readToken (token);
2149 return TRUE;
2150 }
2151 else if (isType (token, TOKEN_OPEN_MXML))
2152 {
2153 parseMXML (token);
2154 }
2155 else
2156 {
2157 if (isType(token, TOKEN_KEYWORD))
2158 {
2159 if (isKeyword (token, KEYWORD_private) ||
2160 isKeyword (token, KEYWORD_public) ||
2161 isKeyword (token, KEYWORD_override) )
2162 {
2163 /*
2164 * Methods can be defined as:
2165 * private function f_name
2166 * public override function f_name
2167 * override private function f_name
2168 * Ignore these keywords if present.
2169 */
2170 readToken (token);
2171 }
2172 if (isKeyword (token, KEYWORD_private) ||
2173 isKeyword (token, KEYWORD_public) ||
2174 isKeyword (token, KEYWORD_override) )
2175 {
2176 /*
2177 * Methods can be defined as:
2178 * private function f_name
2179 * public override function f_name
2180 * override private function f_name
2181 * Ignore these keywords if present.
2182 */
2183 readToken (token);
2184 }
2185
2186 switch (token->keyword)
2187 {
2188 case KEYWORD_function: parseFunction (token); break;
2189 default: parseLine (token); break;
2190 }
2191 }
2192 else
2193 {
2194 parseLine (token);
2195 }
2196 }
2197 } while (TRUE);
2198 }
2199
2200 static void parseFlexFile (tokenInfo *const token)
2201 {
2202 do
2203 {
2204 readToken (token);
2205
2206 if (isType (token, TOKEN_OPEN_MXML))
2207 {
2208 parseMXML (token);
2209 }
2210 else if (isType (token, TOKEN_LESS_THAN))
2211 {
2212 readToken (token);
2213 if (isType (token, TOKEN_QUESTION_MARK))
2214 {
2215 /*
2216 * <?xml version="1.0" encoding="utf-8"?>
2217 */
2218 readToken (token);
2219 while (! isType (token, TOKEN_QUESTION_MARK) )
2220 {
2221 readToken (token);
2222 }
2223 readToken (token);
2224 }
2225 else if (isKeyword (token, KEYWORD_NONE))
2226 {
2227 /*
2228 * This is a simple XML tag, read until the closing statement
2229 * <something .... >
2230 * </something>
2231 */
2232 readToken (token);
2233 while (! isType (token, TOKEN_GREATER_THAN) )
2234 {
2235 readToken (token);
2236 }
2237 }
2238 }
2239 else
2240 {
2241 parseActionScript (token);
2242 }
2243 } while (TRUE);
2244 }
2245
2246 static void initialize (const langType language)
2247 {
2248 Assert (sizeof (FlexKinds) / sizeof (FlexKinds [0]) == FLEXTAG_COUNT);
2249 Lang_js = language;
2250 buildFlexKeywordHash ();
2251 }
2252
2253 static void findFlexTags (void)
2254 {
2255 tokenInfo *const token = newToken ();
2256 exception_t exception;
2257
2258 ClassNames = stringListNew ();
2259 FunctionNames = stringListNew ();
2260
2261 exception = (exception_t) (setjmp (Exception));
2262 while (exception == ExceptionNone)
2263 parseFlexFile (token);
2264
2265 stringListDelete (ClassNames);
2266 stringListDelete (FunctionNames);
2267 ClassNames = NULL;
2268 FunctionNames = NULL;
2269 deleteToken (token);
2270 }
2271
2272 /* Create parser definition stucture */
2273 extern parserDefinition* FlexParser (void)
2274 {
2275 static const char *const extensions [] = { "as", "mxml", NULL };
2276 parserDefinition *const def = parserNew ("Flex");
2277 def->extensions = extensions;
2278 /*
2279 * New definitions for parsing instead of regex
2280 */
2281 def->kinds = FlexKinds;
2282 def->kindCount = KIND_COUNT (FlexKinds);
2283 def->parser = findFlexTags;
2284 def->initialize = initialize;
2285
2286 return def;
2287 }
2288 /* vi:set tabstop=4 shiftwidth=4 noexpandtab: */