1 // Scintilla source code edit control
2 /** @file LexECL.cxx
3  ** Lexer for ECL.
4  **/
5 // Copyright 1998-2001 by Neil Hodgson <neilh@scintilla.org>
6 // The License.txt file describes the conditions under which this software may be distributed.
7 
8 #include <stdlib.h>
9 #include <string.h>
10 #include <ctype.h>
11 #include <stdio.h>
12 #include <stdarg.h>
13 #include <assert.h>
14 
15 #ifdef _MSC_VER
16 #pragma warning(disable: 4786)
17 #endif
18 #ifdef __BORLANDC__
19 // Borland C++ displays warnings in vector header without this
20 #pragma option -w-ccc -w-rch
21 #endif
22 
23 #include <string>
24 #include <vector>
25 #include <map>
26 #include <algorithm>
27 
28 #include "ILexer.h"
29 #include "Scintilla.h"
30 #include "SciLexer.h"
31 
32 #include "PropSetSimple.h"
33 #include "WordList.h"
34 #include "LexAccessor.h"
35 #include "Accessor.h"
36 #include "StyleContext.h"
37 #include "CharacterSet.h"
38 #include "LexerModule.h"
39 #include "OptionSet.h"
40 
41 #define SET_LOWER "abcdefghijklmnopqrstuvwxyz"
42 #define SET_UPPER "ABCDEFGHIJKLMNOPQRSTUVWXYZ"
43 #define SET_DIGITS "0123456789"
44 
45 #ifdef SCI_NAMESPACE
46 using namespace Scintilla;
47 #endif
48 
IsSpaceEquiv(int state)49 static bool IsSpaceEquiv(int state) {
50 	return (state <= SCE_ECL_COMMENTDOC) ||
51 		// including SCE_ECL_DEFAULT, SCE_ECL_COMMENT, SCE_ECL_COMMENTLINE
52 		(state == SCE_ECL_COMMENTLINEDOC) || (state == SCE_ECL_COMMENTDOCKEYWORD) ||
53 		(state == SCE_ECL_COMMENTDOCKEYWORDERROR);
54 }
55 
ColouriseEclDoc(unsigned int startPos,int length,int initStyle,WordList * keywordlists[],Accessor & styler)56 static void ColouriseEclDoc(unsigned int startPos, int length, int initStyle, WordList *keywordlists[],
57                             Accessor &styler) {
58 	WordList &keywords0 = *keywordlists[0];
59 	WordList &keywords1 = *keywordlists[1];
60 	WordList &keywords2 = *keywordlists[2];
61 	WordList &keywords3 = *keywordlists[3]; //Value Types
62 	WordList &keywords4 = *keywordlists[4];
63 	WordList &keywords5 = *keywordlists[5];
64 	WordList &keywords6 = *keywordlists[6];	//Javadoc Tags
65 	WordList cplusplus;
66 	cplusplus.Set("beginc endc");
67 
68 	bool stylingWithinPreprocessor = false;
69 
70 	CharacterSet setOKBeforeRE(CharacterSet::setNone, "(=,");
71 	CharacterSet setDoxygen(CharacterSet::setLower, "$@\\&<>#{}[]");
72 	CharacterSet setWordStart(CharacterSet::setAlpha, "_", 0x80, true);
73 	CharacterSet setWord(CharacterSet::setAlphaNum, "._", 0x80, true);
74 	CharacterSet setQualified(CharacterSet::setNone, "uUxX");
75 
76 	int chPrevNonWhite = ' ';
77 	int visibleChars = 0;
78 	bool lastWordWasUUID = false;
79 	int styleBeforeDCKeyword = SCE_ECL_DEFAULT;
80 	bool continuationLine = false;
81 
82 	if (initStyle == SCE_ECL_PREPROCESSOR) {
83 		// Set continuationLine if last character of previous line is '\'
84 		int lineCurrent = styler.GetLine(startPos);
85 		if (lineCurrent > 0) {
86 			int chBack = styler.SafeGetCharAt(startPos-1, 0);
87 			int chBack2 = styler.SafeGetCharAt(startPos-2, 0);
88 			int lineEndChar = '!';
89 			if (chBack2 == '\r' && chBack == '\n') {
90 				lineEndChar = styler.SafeGetCharAt(startPos-3, 0);
91 			} else if (chBack == '\n' || chBack == '\r') {
92 				lineEndChar = chBack2;
93 			}
94 			continuationLine = lineEndChar == '\\';
95 		}
96 	}
97 
98 	// look back to set chPrevNonWhite properly for better regex colouring
99 	if (startPos > 0) {
100 		int back = startPos;
101 		while (--back && IsSpaceEquiv(styler.StyleAt(back)))
102 			;
103 		if (styler.StyleAt(back) == SCE_ECL_OPERATOR) {
104 			chPrevNonWhite = styler.SafeGetCharAt(back);
105 		}
106 	}
107 
108 	StyleContext sc(startPos, length, initStyle, styler);
109 
110 	for (; sc.More(); sc.Forward()) {
111 		if (sc.atLineStart) {
112 			if (sc.state == SCE_ECL_STRING) {
113 				// Prevent SCE_ECL_STRINGEOL from leaking back to previous line which
114 				// ends with a line continuation by locking in the state upto this position.
115 				sc.SetState(SCE_ECL_STRING);
116 			}
117 			// Reset states to begining of colourise so no surprises
118 			// if different sets of lines lexed.
119 			visibleChars = 0;
120 			lastWordWasUUID = false;
121 		}
122 
123 		// Handle line continuation generically.
124 		if (sc.ch == '\\') {
125 			if (sc.chNext == '\n' || sc.chNext == '\r') {
126 				sc.Forward();
127 				if (sc.ch == '\r' && sc.chNext == '\n') {
128 					sc.Forward();
129 				}
130 				continuationLine = true;
131 				continue;
132 			}
133 		}
134 
135 		// Determine if the current state should terminate.
136 		switch (sc.state) {
137 			case SCE_ECL_ADDED:
138 			case SCE_ECL_DELETED:
139 			case SCE_ECL_CHANGED:
140 			case SCE_ECL_MOVED:
141 			if (sc.atLineStart)
142 					sc.SetState(SCE_ECL_DEFAULT);
143 				break;
144 			case SCE_ECL_OPERATOR:
145 				sc.SetState(SCE_ECL_DEFAULT);
146 				break;
147 			case SCE_ECL_NUMBER:
148 				// We accept almost anything because of hex. and number suffixes
149 				if (!setWord.Contains(sc.ch)) {
150 					sc.SetState(SCE_ECL_DEFAULT);
151 				}
152 				break;
153 			case SCE_ECL_IDENTIFIER:
154 				if (!setWord.Contains(sc.ch) || (sc.ch == '.')) {
155 					char s[1000];
156 					sc.GetCurrentLowered(s, sizeof(s));
157 					if (keywords0.InList(s)) {
158 						lastWordWasUUID = strcmp(s, "uuid") == 0;
159 						sc.ChangeState(SCE_ECL_WORD0);
160 					} else if (keywords1.InList(s)) {
161 						sc.ChangeState(SCE_ECL_WORD1);
162 					} else if (keywords2.InList(s)) {
163 						sc.ChangeState(SCE_ECL_WORD2);
164 					} else if (keywords4.InList(s)) {
165 						sc.ChangeState(SCE_ECL_WORD4);
166 					} else if (keywords5.InList(s)) {
167 						sc.ChangeState(SCE_ECL_WORD5);
168 					}
169 					else	//Data types are of from KEYWORD##
170 					{
171 						int i = static_cast<int>(strlen(s)) - 1;
172 						while(i >= 0 && (isdigit(s[i]) || s[i] == '_'))
173 							--i;
174 
175 						char s2[1000];
176 						strncpy(s2, s, i + 1);
177 						s2[i + 1] = 0;
178 						if (keywords3.InList(s2)) {
179 							sc.ChangeState(SCE_ECL_WORD3);
180 						}
181 					}
182 					sc.SetState(SCE_ECL_DEFAULT);
183 				}
184 				break;
185 			case SCE_ECL_PREPROCESSOR:
186 				if (sc.atLineStart && !continuationLine) {
187 					sc.SetState(SCE_ECL_DEFAULT);
188 				} else if (stylingWithinPreprocessor) {
189 					if (IsASpace(sc.ch)) {
190 						sc.SetState(SCE_ECL_DEFAULT);
191 					}
192 				} else {
193 					if (sc.Match('/', '*') || sc.Match('/', '/')) {
194 						sc.SetState(SCE_ECL_DEFAULT);
195 					}
196 				}
197 				break;
198 			case SCE_ECL_COMMENT:
199 				if (sc.Match('*', '/')) {
200 					sc.Forward();
201 					sc.ForwardSetState(SCE_ECL_DEFAULT);
202 				}
203 				break;
204 			case SCE_ECL_COMMENTDOC:
205 				if (sc.Match('*', '/')) {
206 					sc.Forward();
207 					sc.ForwardSetState(SCE_ECL_DEFAULT);
208 				} else if (sc.ch == '@' || sc.ch == '\\') { // JavaDoc and Doxygen support
209 					// Verify that we have the conditions to mark a comment-doc-keyword
210 					if ((IsASpace(sc.chPrev) || sc.chPrev == '*') && (!IsASpace(sc.chNext))) {
211 						styleBeforeDCKeyword = SCE_ECL_COMMENTDOC;
212 						sc.SetState(SCE_ECL_COMMENTDOCKEYWORD);
213 					}
214 				}
215 				break;
216 			case SCE_ECL_COMMENTLINE:
217 				if (sc.atLineStart) {
218 					sc.SetState(SCE_ECL_DEFAULT);
219 				}
220 				break;
221 			case SCE_ECL_COMMENTLINEDOC:
222 				if (sc.atLineStart) {
223 					sc.SetState(SCE_ECL_DEFAULT);
224 				} else if (sc.ch == '@' || sc.ch == '\\') { // JavaDoc and Doxygen support
225 					// Verify that we have the conditions to mark a comment-doc-keyword
226 					if ((IsASpace(sc.chPrev) || sc.chPrev == '/' || sc.chPrev == '!') && (!IsASpace(sc.chNext))) {
227 						styleBeforeDCKeyword = SCE_ECL_COMMENTLINEDOC;
228 						sc.SetState(SCE_ECL_COMMENTDOCKEYWORD);
229 					}
230 				}
231 				break;
232 			case SCE_ECL_COMMENTDOCKEYWORD:
233 				if ((styleBeforeDCKeyword == SCE_ECL_COMMENTDOC) && sc.Match('*', '/')) {
234 					sc.ChangeState(SCE_ECL_COMMENTDOCKEYWORDERROR);
235 					sc.Forward();
236 					sc.ForwardSetState(SCE_ECL_DEFAULT);
237 				} else if (!setDoxygen.Contains(sc.ch)) {
238 					char s[1000];
239 					sc.GetCurrentLowered(s, sizeof(s));
240 					if (!IsASpace(sc.ch) || !keywords6.InList(s+1)) {
241 						sc.ChangeState(SCE_ECL_COMMENTDOCKEYWORDERROR);
242 					}
243 					sc.SetState(styleBeforeDCKeyword);
244 				}
245 				break;
246 			case SCE_ECL_STRING:
247 				if (sc.atLineEnd) {
248 					sc.ChangeState(SCE_ECL_STRINGEOL);
249 				} else if (sc.ch == '\\') {
250 					if (sc.chNext == '\"' || sc.chNext == '\'' || sc.chNext == '\\') {
251 						sc.Forward();
252 					}
253 				} else if (sc.ch == '\"') {
254 					sc.ForwardSetState(SCE_ECL_DEFAULT);
255 				}
256 				break;
257 			case SCE_ECL_CHARACTER:
258 				if (sc.atLineEnd) {
259 					sc.ChangeState(SCE_ECL_STRINGEOL);
260 				} else if (sc.ch == '\\') {
261 					if (sc.chNext == '\"' || sc.chNext == '\'' || sc.chNext == '\\') {
262 						sc.Forward();
263 					}
264 				} else if (sc.ch == '\'') {
265 					sc.ForwardSetState(SCE_ECL_DEFAULT);
266 				}
267 				break;
268 			case SCE_ECL_REGEX:
269 				if (sc.atLineStart) {
270 					sc.SetState(SCE_ECL_DEFAULT);
271 				} else if (sc.ch == '/') {
272 					sc.Forward();
273 					while ((sc.ch < 0x80) && islower(sc.ch))
274 						sc.Forward();    // gobble regex flags
275 					sc.SetState(SCE_ECL_DEFAULT);
276 				} else if (sc.ch == '\\') {
277 					// Gobble up the quoted character
278 					if (sc.chNext == '\\' || sc.chNext == '/') {
279 						sc.Forward();
280 					}
281 				}
282 				break;
283 			case SCE_ECL_STRINGEOL:
284 				if (sc.atLineStart) {
285 					sc.SetState(SCE_ECL_DEFAULT);
286 				}
287 				break;
288 			case SCE_ECL_VERBATIM:
289 				if (sc.ch == '\"') {
290 					if (sc.chNext == '\"') {
291 						sc.Forward();
292 					} else {
293 						sc.ForwardSetState(SCE_ECL_DEFAULT);
294 					}
295 				}
296 				break;
297 			case SCE_ECL_UUID:
298 				if (sc.ch == '\r' || sc.ch == '\n' || sc.ch == ')') {
299 					sc.SetState(SCE_ECL_DEFAULT);
300 				}
301 				break;
302 		}
303 
304 		// Determine if a new state should be entered.
305 		int lineCurrent = styler.GetLine(sc.currentPos);
306 		int lineState = styler.GetLineState(lineCurrent);
307 		if (sc.state == SCE_ECL_DEFAULT) {
308 			if (lineState) {
309 				sc.SetState(lineState);
310 			}
311 			else if (sc.Match('@', '\"')) {
312 				sc.SetState(SCE_ECL_VERBATIM);
313 				sc.Forward();
314 			} else if (setQualified.Contains(sc.ch) && sc.chNext == '\'') {
315 				sc.SetState(SCE_ECL_CHARACTER);
316 				sc.Forward();
317 			} else if (IsADigit(sc.ch) || (sc.ch == '.' && IsADigit(sc.chNext))) {
318 				if (lastWordWasUUID) {
319 					sc.SetState(SCE_ECL_UUID);
320 					lastWordWasUUID = false;
321 				} else {
322 					sc.SetState(SCE_ECL_NUMBER);
323 				}
324 			} else if (setWordStart.Contains(sc.ch) || (sc.ch == '@')) {
325 				if (lastWordWasUUID) {
326 					sc.SetState(SCE_ECL_UUID);
327 					lastWordWasUUID = false;
328 				} else {
329 					sc.SetState(SCE_ECL_IDENTIFIER);
330 				}
331 			} else if (sc.Match('/', '*')) {
332 				if (sc.Match("/**") || sc.Match("/*!")) {	// Support of Qt/Doxygen doc. style
333 					sc.SetState(SCE_ECL_COMMENTDOC);
334 				} else {
335 					sc.SetState(SCE_ECL_COMMENT);
336 				}
337 				sc.Forward();	// Eat the * so it isn't used for the end of the comment
338 			} else if (sc.Match('/', '/')) {
339 				if ((sc.Match("///") && !sc.Match("////")) || sc.Match("//!"))
340 					// Support of Qt/Doxygen doc. style
341 					sc.SetState(SCE_ECL_COMMENTLINEDOC);
342 				else
343 					sc.SetState(SCE_ECL_COMMENTLINE);
344 			} else if (sc.ch == '/' && setOKBeforeRE.Contains(chPrevNonWhite)) {
345 				sc.SetState(SCE_ECL_REGEX);	// JavaScript's RegEx
346 //			} else if (sc.ch == '\"') {
347 //				sc.SetState(SCE_ECL_STRING);
348 			} else if (sc.ch == '\'') {
349 				sc.SetState(SCE_ECL_CHARACTER);
350 			} else if (sc.ch == '#' && visibleChars == 0) {
351 				// Preprocessor commands are alone on their line
352 				sc.SetState(SCE_ECL_PREPROCESSOR);
353 				// Skip whitespace between # and preprocessor word
354 				do {
355 					sc.Forward();
356 				} while ((sc.ch == ' ' || sc.ch == '\t') && sc.More());
357 				if (sc.atLineEnd) {
358 					sc.SetState(SCE_ECL_DEFAULT);
359 				}
360 			} else if (isoperator(static_cast<char>(sc.ch))) {
361 				sc.SetState(SCE_ECL_OPERATOR);
362 			}
363 		}
364 
365 		if (!IsASpace(sc.ch) && !IsSpaceEquiv(sc.state)) {
366 			chPrevNonWhite = sc.ch;
367 			visibleChars++;
368 		}
369 		continuationLine = false;
370 	}
371 	sc.Complete();
372 
373 }
374 
IsStreamCommentStyle(int style)375 static bool IsStreamCommentStyle(int style) {
376 	return style == SCE_ECL_COMMENT ||
377 		style == SCE_ECL_COMMENTDOC ||
378 		style == SCE_ECL_COMMENTDOCKEYWORD ||
379 		style == SCE_ECL_COMMENTDOCKEYWORDERROR;
380 }
381 
MatchNoCase(Accessor & styler,unsigned int & pos,const char * s)382 bool MatchNoCase(Accessor & styler, unsigned int & pos, const char *s) {
383 	int i=0;
384 	for (; *s; i++) {
385 		char compare_char = tolower(*s);
386 		char styler_char = tolower(styler.SafeGetCharAt(pos+i));
387 		if (compare_char != styler_char)
388 			return false;
389 		s++;
390 	}
391 	pos+=i-1;
392 	return true;
393 }
394 
395 
396 // Store both the current line's fold level and the next lines in the
397 // level store to make it easy to pick up with each increment
398 // and to make it possible to fiddle the current level for "} else {".
FoldEclDoc(unsigned int startPos,int length,int initStyle,WordList * [],Accessor & styler)399 static void FoldEclDoc(unsigned int startPos, int length, int initStyle,
400 					   WordList *[], Accessor &styler) {
401 	bool foldComment = true;
402 	bool foldPreprocessor = true;
403 	bool foldCompact = true;
404 	bool foldAtElse = true;
405 	unsigned int endPos = startPos + length;
406 	int visibleChars = 0;
407 	int lineCurrent = styler.GetLine(startPos);
408 	int levelCurrent = SC_FOLDLEVELBASE;
409 	if (lineCurrent > 0)
410 		levelCurrent = styler.LevelAt(lineCurrent-1) >> 16;
411 	int levelMinCurrent = levelCurrent;
412 	int levelNext = levelCurrent;
413 	char chNext = styler[startPos];
414 	int styleNext = styler.StyleAt(startPos);
415 	int style = initStyle;
416 	for (unsigned int i = startPos; i < endPos; i++) {
417 		char ch = chNext;
418 		chNext = styler.SafeGetCharAt(i + 1);
419 		int stylePrev = style;
420 		style = styleNext;
421 		styleNext = styler.StyleAt(i + 1);
422 		bool atEOL = (ch == '\r' && chNext != '\n') || (ch == '\n');
423 		if (foldComment && IsStreamCommentStyle(style)) {
424 			if (!IsStreamCommentStyle(stylePrev) && (stylePrev != SCE_ECL_COMMENTLINEDOC)) {
425 				levelNext++;
426 			} else if (!IsStreamCommentStyle(styleNext) && (styleNext != SCE_ECL_COMMENTLINEDOC) && !atEOL) {
427 				// Comments don't end at end of line and the next character may be unstyled.
428 				levelNext--;
429 			}
430 		}
431 		if (foldComment && (style == SCE_ECL_COMMENTLINE)) {
432 			if ((ch == '/') && (chNext == '/')) {
433 				char chNext2 = styler.SafeGetCharAt(i + 2);
434 				if (chNext2 == '{') {
435 					levelNext++;
436 				} else if (chNext2 == '}') {
437 					levelNext--;
438 				}
439 			}
440 		}
441 		if (foldPreprocessor && (style == SCE_ECL_PREPROCESSOR)) {
442 			if (ch == '#') {
443 				unsigned int j = i + 1;
444 				while ((j < endPos) && IsASpaceOrTab(styler.SafeGetCharAt(j))) {
445 					j++;
446 				}
447 				if (MatchNoCase(styler, j, "region") || MatchNoCase(styler, j, "if")) {
448 					levelNext++;
449 				} else if (MatchNoCase(styler, j, "endregion") || MatchNoCase(styler, j, "end")) {
450 					levelNext--;
451 				}
452 			}
453 		}
454 		if (style == SCE_ECL_OPERATOR) {
455 			if (ch == '{') {
456 				// Measure the minimum before a '{' to allow
457 				// folding on "} else {"
458 				if (levelMinCurrent > levelNext) {
459 					levelMinCurrent = levelNext;
460 				}
461 				levelNext++;
462 			} else if (ch == '}') {
463 				levelNext--;
464 			}
465 		}
466 		if (style == SCE_ECL_WORD2) {
467 			if (MatchNoCase(styler, i, "record") || MatchNoCase(styler, i, "transform") || MatchNoCase(styler, i, "type") || MatchNoCase(styler, i, "function") ||
468 				MatchNoCase(styler, i, "module") || MatchNoCase(styler, i, "service") || MatchNoCase(styler, i, "interface") || MatchNoCase(styler, i, "ifblock") ||
469 				MatchNoCase(styler, i, "macro") || MatchNoCase(styler, i, "beginc++")) {
470 				levelNext++;
471 			} else if (MatchNoCase(styler, i, "endmacro") || MatchNoCase(styler, i, "endc++") || MatchNoCase(styler, i, "end")) {
472 				levelNext--;
473 			}
474 		}
475 		if (atEOL || (i == endPos-1)) {
476 			int levelUse = levelCurrent;
477 			if (foldAtElse) {
478 				levelUse = levelMinCurrent;
479 			}
480 			int lev = levelUse | levelNext << 16;
481 			if (visibleChars == 0 && foldCompact)
482 				lev |= SC_FOLDLEVELWHITEFLAG;
483 			if (levelUse < levelNext)
484 				lev |= SC_FOLDLEVELHEADERFLAG;
485 			if (lev != styler.LevelAt(lineCurrent)) {
486 				styler.SetLevel(lineCurrent, lev);
487 			}
488 			lineCurrent++;
489 			levelCurrent = levelNext;
490 			levelMinCurrent = levelCurrent;
491 			if (atEOL && (i == static_cast<unsigned int>(styler.Length()-1))) {
492 				// There is an empty line at end of file so give it same level and empty
493 				styler.SetLevel(lineCurrent, (levelCurrent | levelCurrent << 16) | SC_FOLDLEVELWHITEFLAG);
494 			}
495 			visibleChars = 0;
496 		}
497 		if (!IsASpace(ch))
498 			visibleChars++;
499 	}
500 }
501 
502 static const char * const EclWordListDesc[] = {
503 	"Keywords",
504 	0
505 };
506 
507 LexerModule lmECL(
508    SCLEX_ECL,
509    ColouriseEclDoc,
510    "ecl",
511    FoldEclDoc,
512    EclWordListDesc);
513