| 1 |
// Scintilla source code edit control |
| 2 |
/** @file LexCOBOL.cxx |
| 3 |
** Lexer for COBOL |
| 4 |
** Based on LexPascal.cxx |
| 5 |
** Written by Laurent le Tynevez |
| 6 |
** Updated by Simon Steele <[email protected]> September 2002 |
| 7 |
** Updated by Mathias Rauen <[email protected]> May 2003 (Delphi adjustments) |
| 8 |
** Updated by Rod Falck, Aug 2006 Converted to COBOL |
| 9 |
**/ |
| 10 |
|
| 11 |
#include <cstdlib> |
| 12 |
#include <cassert> |
| 13 |
#include <cstring> |
| 14 |
#include <cctype> |
| 15 |
#include <cstdio> |
| 16 |
|
| 17 |
#include <string> |
| 18 |
#include <string_view> |
| 19 |
|
| 20 |
#include "ILexer.h" |
| 21 |
#include "Scintilla.h" |
| 22 |
#include "SciLexer.h" |
| 23 |
|
| 24 |
#include "WordList.h" |
| 25 |
#include "LexAccessor.h" |
| 26 |
#include "Accessor.h" |
| 27 |
#include "StyleContext.h" |
| 28 |
#include "CharacterSet.h" |
| 29 |
#include "LexerModule.h" |
| 30 |
|
| 31 |
using namespace Lexilla; |
| 32 |
|
| 33 |
#define IN_DIVISION 0x01 |
| 34 |
#define IN_DECLARATIVES 0x02 |
| 35 |
#define IN_SECTION 0x04 |
| 36 |
#define IN_PARAGRAPH 0x08 |
| 37 |
#define IN_FLAGS 0xF |
| 38 |
#define NOT_HEADER 0x10 |
| 39 |
|
| 40 |
namespace { |
| 41 |
|
| 42 |
bool isCOBOLoperator(char ch) |
| 43 |
{ |
| 44 |
return isoperator(ch); |
| 45 |
} |
| 46 |
|
| 47 |
bool isCOBOLwordchar(char ch) |
| 48 |
{ |
| 49 |
return IsASCII(ch) && (isalnum(ch) || ch == '-'); |
| 50 |
|
| 51 |
} |
| 52 |
|
| 53 |
bool isCOBOLwordstart(char ch) |
| 54 |
{ |
| 55 |
return IsASCII(ch) && isalnum(ch); |
| 56 |
} |
| 57 |
|
| 58 |
int CountBits(int nBits) |
| 59 |
{ |
| 60 |
int count = 0; |
| 61 |
for (int i = 0; i < 32; ++i) |
| 62 |
{ |
| 63 |
count += nBits & 1; |
| 64 |
nBits >>= 1; |
| 65 |
} |
| 66 |
return count; |
| 67 |
} |
| 68 |
|
| 69 |
void getRange(Sci_PositionU start, |
| 70 |
Sci_PositionU end, |
| 71 |
Accessor &styler, |
| 72 |
char *s, |
| 73 |
Sci_PositionU len) { |
| 74 |
Sci_PositionU i = 0; |
| 75 |
while ((i < end - start + 1) && (i < len-1)) { |
| 76 |
s[i] = static_cast<char>(tolower(styler[start + i])); |
| 77 |
i++; |
| 78 |
} |
| 79 |
s[i] = '\0'; |
| 80 |
} |
| 81 |
|
| 82 |
void ColourTo(Accessor &styler, Sci_PositionU end, unsigned int attr) { |
| 83 |
styler.ColourTo(end, attr); |
| 84 |
} |
| 85 |
|
| 86 |
|
| 87 |
int classifyWordCOBOL(Sci_PositionU start, Sci_PositionU end, /*WordList &keywords*/WordList *keywordlists[], Accessor &styler, int nContainment, bool *bAarea) { |
| 88 |
int ret = 0; |
| 89 |
|
| 90 |
char s[100]; |
| 91 |
s[0] = '\0'; |
| 92 |
s[1] = '\0'; |
| 93 |
getRange(start, end, styler, s, sizeof(s)); |
| 94 |
|
| 95 |
int chAttr = SCE_COBOL_IDENTIFIER; |
| 96 |
if (isdigit(s[0]) || (s[0] == '.') || (s[0] == 'v')) { |
| 97 |
chAttr = SCE_COBOL_NUMBER; |
| 98 |
char *p = s + 1; |
| 99 |
while (*p) { |
| 100 |
if ((!isdigit(*p) && (*p) != 'v') && isCOBOLwordchar(*p)) { |
| 101 |
chAttr = SCE_COBOL_IDENTIFIER; |
| 102 |
break; |
| 103 |
} |
| 104 |
++p; |
| 105 |
} |
| 106 |
} |
| 107 |
if (chAttr == SCE_COBOL_IDENTIFIER) { |
| 108 |
WordList& a_keywords = *keywordlists[0]; |
| 109 |
WordList& b_keywords = *keywordlists[1]; |
| 110 |
WordList& c_keywords = *keywordlists[2]; |
| 111 |
|
| 112 |
if (a_keywords.InList(s)) { |
| 113 |
chAttr = SCE_COBOL_WORD; |
| 114 |
} |
| 115 |
else if (b_keywords.InList(s)) { |
| 116 |
chAttr = SCE_COBOL_WORD2; |
| 117 |
} |
| 118 |
else if (c_keywords.InList(s)) { |
| 119 |
chAttr = SCE_COBOL_WORD3; |
| 120 |
} |
| 121 |
} |
| 122 |
if (*bAarea) { |
| 123 |
if (strcmp(s, "division") == 0) { |
| 124 |
ret = IN_DIVISION; |
| 125 |
// we've determined the containment, anything else is just ignored for those purposes |
| 126 |
*bAarea = false; |
| 127 |
} else if (strcmp(s, "declaratives") == 0) { |
| 128 |
ret = IN_DIVISION | IN_DECLARATIVES; |
| 129 |
if (nContainment & IN_DECLARATIVES) |
| 130 |
ret |= NOT_HEADER | IN_SECTION; |
| 131 |
// we've determined the containment, anything else is just ignored for those purposes |
| 132 |
*bAarea = false; |
| 133 |
} else if (strcmp(s, "section") == 0) { |
| 134 |
ret = (nContainment &~ IN_PARAGRAPH) | IN_SECTION; |
| 135 |
// we've determined the containment, anything else is just ignored for those purposes |
| 136 |
*bAarea = false; |
| 137 |
} else if (strcmp(s, "end") == 0 && (nContainment & IN_DECLARATIVES)) { |
| 138 |
ret = IN_DIVISION | IN_DECLARATIVES | IN_SECTION | NOT_HEADER; |
| 139 |
} else { |
| 140 |
ret = nContainment | IN_PARAGRAPH; |
| 141 |
} |
| 142 |
} |
| 143 |
ColourTo(styler, end, chAttr); |
| 144 |
return ret; |
| 145 |
} |
| 146 |
|
| 147 |
void ColouriseCOBOLDoc(Sci_PositionU startPos, Sci_Position length, int initStyle, WordList *keywordlists[], |
| 148 |
Accessor &styler) { |
| 149 |
|
| 150 |
styler.StartAt(startPos); |
| 151 |
|
| 152 |
int state = initStyle; |
| 153 |
if (state == SCE_COBOL_CHARACTER) // Does not leak onto next line |
| 154 |
state = SCE_COBOL_DEFAULT; |
| 155 |
char chPrev = ' '; |
| 156 |
char chNext = styler[startPos]; |
| 157 |
Sci_PositionU lengthDoc = startPos + length; |
| 158 |
|
| 159 |
int nContainment; |
| 160 |
|
| 161 |
Sci_Position currentLine = styler.GetLine(startPos); |
| 162 |
if (currentLine > 0) { |
| 163 |
styler.SetLineState(currentLine, styler.GetLineState(currentLine-1)); |
| 164 |
nContainment = styler.GetLineState(currentLine); |
| 165 |
nContainment &= ~NOT_HEADER; |
| 166 |
} else { |
| 167 |
styler.SetLineState(currentLine, 0); |
| 168 |
nContainment = 0; |
| 169 |
} |
| 170 |
|
| 171 |
styler.StartSegment(startPos); |
| 172 |
bool bNewLine = true; |
| 173 |
bool bAarea = !isspacechar(chNext); |
| 174 |
int column = 0; |
| 175 |
for (Sci_PositionU i = startPos; i < lengthDoc; i++) { |
| 176 |
char ch = chNext; |
| 177 |
|
| 178 |
chNext = styler.SafeGetCharAt(i + 1); |
| 179 |
|
| 180 |
++column; |
| 181 |
|
| 182 |
if (bNewLine) { |
| 183 |
column = 0; |
| 184 |
} |
| 185 |
if (column <= 1 && !bAarea) { |
| 186 |
bAarea = !isspacechar(ch); |
| 187 |
} |
| 188 |
bool bSetNewLine = false; |
| 189 |
if ((ch == '\r' && chNext != '\n') || (ch == '\n')) { |
| 190 |
// Trigger on CR only (Mac style) or either on LF from CR+LF (Dos/Win) or on LF alone (Unix) |
| 191 |
// Avoid triggering two times on Dos/Win |
| 192 |
// End of line |
| 193 |
if (state == SCE_COBOL_CHARACTER) { |
| 194 |
ColourTo(styler, i, state); |
| 195 |
state = SCE_COBOL_DEFAULT; |
| 196 |
} |
| 197 |
styler.SetLineState(currentLine, nContainment); |
| 198 |
currentLine++; |
| 199 |
bSetNewLine = true; |
| 200 |
if (nContainment & NOT_HEADER) |
| 201 |
nContainment &= ~(NOT_HEADER | IN_DECLARATIVES | IN_SECTION); |
| 202 |
} |
| 203 |
|
| 204 |
if (styler.IsLeadByte(ch)) { |
| 205 |
chNext = styler.SafeGetCharAt(i + 2); |
| 206 |
chPrev = ' '; |
| 207 |
i += 1; |
| 208 |
continue; |
| 209 |
} |
| 210 |
|
| 211 |
if (state == SCE_COBOL_DEFAULT) { |
| 212 |
if (isCOBOLwordstart(ch) || (ch == '$' && IsASCII(chNext) && isalpha(chNext))) { |
| 213 |
ColourTo(styler, i-1, state); |
| 214 |
state = SCE_COBOL_IDENTIFIER; |
| 215 |
} else if (column == 6 && (ch == '*' || ch == '/')) { |
| 216 |
// Cobol comment line: asterisk in column 7. |
| 217 |
ColourTo(styler, i-1, state); |
| 218 |
state = SCE_COBOL_COMMENTLINE; |
| 219 |
} else if (ch == '*' && chNext == '>') { |
| 220 |
// Cobol inline comment: asterisk, followed by greater than. |
| 221 |
ColourTo(styler, i-1, state); |
| 222 |
state = SCE_COBOL_COMMENTLINE; |
| 223 |
} else if (column == 0 && ch == '*' && chNext != '*') { |
| 224 |
ColourTo(styler, i-1, state); |
| 225 |
state = SCE_COBOL_COMMENTLINE; |
| 226 |
} else if (column == 0 && ch == '/' && chNext != '*') { |
| 227 |
ColourTo(styler, i-1, state); |
| 228 |
state = SCE_COBOL_COMMENTLINE; |
| 229 |
} else if (column == 0 && ch == '*' && chNext == '*') { |
| 230 |
ColourTo(styler, i-1, state); |
| 231 |
state = SCE_COBOL_COMMENTDOC; |
| 232 |
} else if (column == 0 && ch == '/' && chNext == '*') { |
| 233 |
ColourTo(styler, i-1, state); |
| 234 |
state = SCE_COBOL_COMMENTDOC; |
| 235 |
} else if (ch == '"') { |
| 236 |
ColourTo(styler, i-1, state); |
| 237 |
state = SCE_COBOL_STRING; |
| 238 |
} else if (ch == '\'') { |
| 239 |
ColourTo(styler, i-1, state); |
| 240 |
state = SCE_COBOL_CHARACTER; |
| 241 |
} else if (ch == '?' && column == 0) { |
| 242 |
ColourTo(styler, i-1, state); |
| 243 |
state = SCE_COBOL_PREPROCESSOR; |
| 244 |
} else if (isCOBOLoperator(ch)) { |
| 245 |
ColourTo(styler, i-1, state); |
| 246 |
ColourTo(styler, i, SCE_COBOL_OPERATOR); |
| 247 |
} |
| 248 |
} else if (state == SCE_COBOL_IDENTIFIER) { |
| 249 |
if (!isCOBOLwordchar(ch)) { |
| 250 |
int lStateChange = classifyWordCOBOL(styler.GetStartSegment(), i - 1, keywordlists, styler, nContainment, &bAarea); |
| 251 |
|
| 252 |
if(lStateChange != 0) { |
| 253 |
styler.SetLineState(currentLine, lStateChange); |
| 254 |
nContainment = lStateChange; |
| 255 |
} |
| 256 |
|
| 257 |
state = SCE_COBOL_DEFAULT; |
| 258 |
chNext = styler.SafeGetCharAt(i + 1); |
| 259 |
if (column == 6 && (ch == '*' || ch == '/')) { |
| 260 |
state = SCE_COBOL_COMMENTLINE; |
| 261 |
} else if (ch == '"') { |
| 262 |
state = SCE_COBOL_STRING; |
| 263 |
} else if (ch == '\'') { |
| 264 |
state = SCE_COBOL_CHARACTER; |
| 265 |
} else if (isCOBOLoperator(ch)) { |
| 266 |
ColourTo(styler, i, SCE_COBOL_OPERATOR); |
| 267 |
} |
| 268 |
} |
| 269 |
} else { |
| 270 |
if (state == SCE_COBOL_PREPROCESSOR) { |
| 271 |
if ((ch == '\r' || ch == '\n') && !(chPrev == '\\' || chPrev == '\r')) { |
| 272 |
ColourTo(styler, i-1, state); |
| 273 |
state = SCE_COBOL_DEFAULT; |
| 274 |
} |
| 275 |
} else if (state == SCE_COBOL_COMMENT) { |
| 276 |
if (ch == '\r' || ch == '\n') { |
| 277 |
ColourTo(styler, i-1, state); |
| 278 |
state = SCE_COBOL_DEFAULT; |
| 279 |
} |
| 280 |
} else if (state == SCE_COBOL_COMMENTDOC) { |
| 281 |
if (ch == '\r' || ch == '\n') { |
| 282 |
if (((i > styler.GetStartSegment() + 2) || ( |
| 283 |
(initStyle == SCE_COBOL_COMMENTDOC) && |
| 284 |
(styler.GetStartSegment() == static_cast<Sci_PositionU>(startPos))))) { |
| 285 |
ColourTo(styler, i-1, state); |
| 286 |
state = SCE_COBOL_DEFAULT; |
| 287 |
} |
| 288 |
} |
| 289 |
} else if (state == SCE_COBOL_COMMENTLINE) { |
| 290 |
if (ch == '\r' || ch == '\n') { |
| 291 |
ColourTo(styler, i-1, state); |
| 292 |
state = SCE_COBOL_DEFAULT; |
| 293 |
} |
| 294 |
} else if (state == SCE_COBOL_STRING) { |
| 295 |
if (ch == '"') { |
| 296 |
ColourTo(styler, i, state); |
| 297 |
state = SCE_COBOL_DEFAULT; |
| 298 |
} else if (ch == '\r' || ch == '\n') { |
| 299 |
ColourTo(styler, i-1, state); |
| 300 |
state = SCE_COBOL_DEFAULT; |
| 301 |
} |
| 302 |
} else if (state == SCE_COBOL_CHARACTER) { |
| 303 |
if (ch == '\'') { |
| 304 |
ColourTo(styler, i, state); |
| 305 |
state = SCE_COBOL_DEFAULT; |
| 306 |
} |
| 307 |
} |
| 308 |
} |
| 309 |
chPrev = ch; |
| 310 |
bNewLine = bSetNewLine; |
| 311 |
if (bNewLine) |
| 312 |
{ |
| 313 |
bAarea = false; |
| 314 |
} |
| 315 |
} |
| 316 |
ColourTo(styler, lengthDoc - 1, state); |
| 317 |
} |
| 318 |
|
| 319 |
void FoldCOBOLDoc(Sci_PositionU startPos, Sci_Position length, int, WordList *[], |
| 320 |
Accessor &styler) { |
| 321 |
bool foldCompact = styler.GetPropertyInt("fold.compact", 1) != 0; |
| 322 |
Sci_PositionU endPos = startPos + length; |
| 323 |
int visibleChars = 0; |
| 324 |
Sci_Position lineCurrent = styler.GetLine(startPos); |
| 325 |
int levelPrev = lineCurrent > 0 ? styler.LevelAt(lineCurrent - 1) & SC_FOLDLEVELNUMBERMASK : 0xFFF; |
| 326 |
char chNext = styler[startPos]; |
| 327 |
|
| 328 |
bool bNewLine = true; |
| 329 |
bool bAarea = !isspacechar(chNext); |
| 330 |
int column = 0; |
| 331 |
bool bComment = false; |
| 332 |
for (Sci_PositionU i = startPos; i < endPos; i++) { |
| 333 |
char ch = chNext; |
| 334 |
chNext = styler.SafeGetCharAt(i + 1); |
| 335 |
++column; |
| 336 |
|
| 337 |
if (bNewLine) { |
| 338 |
column = 0; |
| 339 |
bComment = (ch == '*' || ch == '/' || ch == '?'); |
| 340 |
} |
| 341 |
if (column <= 1 && !bAarea) { |
| 342 |
bAarea = !isspacechar(ch); |
| 343 |
} |
| 344 |
bool atEOL = (ch == '\r' && chNext != '\n') || (ch == '\n'); |
| 345 |
if (atEOL) { |
| 346 |
int nContainment = styler.GetLineState(lineCurrent); |
| 347 |
int lev = CountBits(nContainment & IN_FLAGS) | SC_FOLDLEVELBASE; |
| 348 |
if (bAarea && !bComment) |
| 349 |
--lev; |
| 350 |
if (visibleChars == 0 && foldCompact) |
| 351 |
lev |= SC_FOLDLEVELWHITEFLAG; |
| 352 |
if ((bAarea) && (visibleChars > 0) && !(nContainment & NOT_HEADER) && !bComment) |
| 353 |
lev |= SC_FOLDLEVELHEADERFLAG; |
| 354 |
if (lev != styler.LevelAt(lineCurrent)) { |
| 355 |
styler.SetLevel(lineCurrent, lev); |
| 356 |
} |
| 357 |
if ((lev & SC_FOLDLEVELNUMBERMASK) <= (levelPrev & SC_FOLDLEVELNUMBERMASK)) { |
| 358 |
// this level is at the same level or less than the previous line |
| 359 |
// therefore these is nothing for the previous header to collapse, so remove the header |
| 360 |
styler.SetLevel(lineCurrent - 1, levelPrev & ~SC_FOLDLEVELHEADERFLAG); |
| 361 |
} |
| 362 |
levelPrev = lev; |
| 363 |
visibleChars = 0; |
| 364 |
bAarea = false; |
| 365 |
bNewLine = true; |
| 366 |
lineCurrent++; |
| 367 |
} else { |
| 368 |
bNewLine = false; |
| 369 |
} |
| 370 |
|
| 371 |
|
| 372 |
if (!isspacechar(ch)) |
| 373 |
visibleChars++; |
| 374 |
} |
| 375 |
|
| 376 |
// Fill in the real level of the next line, keeping the current flags as they will be filled in later |
| 377 |
int flagsNext = styler.LevelAt(lineCurrent) & ~SC_FOLDLEVELNUMBERMASK; |
| 378 |
styler.SetLevel(lineCurrent, levelPrev | flagsNext); |
| 379 |
} |
| 380 |
|
| 381 |
const char * const COBOLWordListDesc[] = { |
| 382 |
"A Keywords", |
| 383 |
"B Keywords", |
| 384 |
"Extended Keywords", |
| 385 |
nullptr |
| 386 |
}; |
| 387 |
|
| 388 |
} |
| 389 |
|
| 390 |
extern const LexerModule lmCOBOL(SCLEX_COBOL, ColouriseCOBOLDoc, "COBOL", FoldCOBOLDoc, COBOLWordListDesc); |