mirror of
https://git.do.srb2.org/STJr/UltimateZoneBuilder.git
synced 2024-11-30 15:41:30 +00:00
231 lines
6.9 KiB
C++
231 lines
6.9 KiB
C++
|
// Scintilla source code edit control
|
||
|
/** @file LexFlagShip.cxx
|
||
|
** Lexer for FlagShip
|
||
|
** (Syntactically compatible to other XBase dialects, like dBase, Clipper, Fox etc.)
|
||
|
**/
|
||
|
// Copyright 2005 by Randy Butler
|
||
|
// Copyright 1998-2003 by Neil Hodgson <neilh@scintilla.org>
|
||
|
// The License.txt file describes the conditions under which this software may be distributed.
|
||
|
|
||
|
#include <stdlib.h>
|
||
|
#include <string.h>
|
||
|
#include <ctype.h>
|
||
|
#include <stdio.h>
|
||
|
#include <stdarg.h>
|
||
|
|
||
|
#include "Platform.h"
|
||
|
|
||
|
#include "PropSet.h"
|
||
|
#include "Accessor.h"
|
||
|
#include "StyleContext.h"
|
||
|
#include "KeyWords.h"
|
||
|
#include "Scintilla.h"
|
||
|
#include "SciLexer.h"
|
||
|
|
||
|
#ifdef SCI_NAMESPACE
|
||
|
using namespace Scintilla;
|
||
|
#endif
|
||
|
|
||
|
static bool IsFlagShipComment(Accessor &styler, int pos, int len) {
|
||
|
return len>0 && styler[pos]=='\'';
|
||
|
}
|
||
|
|
||
|
static inline bool IsTypeCharacter(int ch) {
|
||
|
return ch == '%' || ch == '&' || ch == '@' || ch == '!' || ch == '#' || ch == '$';
|
||
|
}
|
||
|
|
||
|
// Extended to accept accented characters
|
||
|
static inline bool IsAWordChar(int ch) {
|
||
|
return ch >= 0x80 ||
|
||
|
(isalnum(ch) || ch == '.' || ch == '_');
|
||
|
}
|
||
|
|
||
|
static inline bool IsAWordStart(int ch) {
|
||
|
return ch >= 0x80 ||
|
||
|
(isalnum(ch) || ch == '_');
|
||
|
}
|
||
|
|
||
|
static inline bool IsADateCharacter(const int ch) {
|
||
|
return (ch < 0x80) &&
|
||
|
(isalnum(ch) || ch == '|' || ch == '-' || ch == '/' || ch == ':' || ch == ' ' || ch == '\t');
|
||
|
}
|
||
|
|
||
|
|
||
|
static void ColouriseFlagShipDoc(unsigned int startPos, int length, int initStyle,
|
||
|
WordList *keywordlists[], Accessor &styler) {
|
||
|
|
||
|
//bool FSScriptSyntax = true;
|
||
|
WordList &keywords = *keywordlists[0];
|
||
|
WordList &keywords2 = *keywordlists[1];
|
||
|
WordList &keywords3 = *keywordlists[2];
|
||
|
WordList &keywords4 = *keywordlists[3];
|
||
|
|
||
|
styler.StartAt(startPos);
|
||
|
|
||
|
int visibleChars = 0;
|
||
|
|
||
|
StyleContext sc(startPos, length, initStyle, styler);
|
||
|
|
||
|
for (; sc.More(); sc.Forward()) {
|
||
|
|
||
|
if (sc.state == SCE_FS_OPERATOR) {
|
||
|
sc.SetState(SCE_FS_DEFAULT);
|
||
|
} else if (sc.state == SCE_FS_IDENTIFIER) {
|
||
|
if (!IsAWordChar(sc.ch)) {
|
||
|
char s[100];
|
||
|
sc.GetCurrentLowered(s, sizeof(s));
|
||
|
if (keywords.InList(s)) {
|
||
|
sc.ChangeState(SCE_FS_KEYWORD);
|
||
|
} else if (keywords2.InList(s)) {
|
||
|
sc.ChangeState(SCE_FS_KEYWORD2);
|
||
|
} else if (keywords3.InList(s)) {
|
||
|
sc.ChangeState(SCE_FS_KEYWORD3);
|
||
|
} else if (keywords4.InList(s)) {
|
||
|
sc.ChangeState(SCE_FS_KEYWORD4);
|
||
|
}// Else, it is really an identifier...
|
||
|
sc.SetState(SCE_FS_DEFAULT);
|
||
|
}
|
||
|
} else if (sc.state == SCE_FS_NUMBER) {
|
||
|
if (!IsAWordChar(sc.ch)) {
|
||
|
sc.SetState(SCE_FS_DEFAULT);
|
||
|
}
|
||
|
} else if (sc.state == SCE_FS_STRING) {
|
||
|
// VB doubles quotes to preserve them, so just end this string
|
||
|
// state now as a following quote will start again
|
||
|
if (sc.ch == '\"') {
|
||
|
if (tolower(sc.chNext) == 'c') {
|
||
|
sc.Forward();
|
||
|
}
|
||
|
sc.ForwardSetState(SCE_FS_DEFAULT);
|
||
|
} else if (sc.atLineEnd) {
|
||
|
sc.ChangeState(SCE_FS_STRINGEOL);
|
||
|
sc.ForwardSetState(SCE_FS_DEFAULT);
|
||
|
}
|
||
|
} else if (sc.state == SCE_FS_COMMENT) {
|
||
|
if (sc.Match('*', '/')) { // new code
|
||
|
sc.Forward();
|
||
|
sc.ForwardSetState(SCE_FS_DEFAULT);
|
||
|
//if (sc.atLineEnd) { // old code
|
||
|
// sc.SetState(SCE_FS_DEFAULT);
|
||
|
}
|
||
|
} else if (sc.state == SCE_FS_COMMENTLINE) { //new code
|
||
|
if (sc.ch == '\r' || sc.ch == '\n') {
|
||
|
sc.SetState(SCE_FS_DEFAULT);
|
||
|
visibleChars = 0;
|
||
|
}
|
||
|
} else if (sc.state == SCE_FS_PREPROCESSOR) {
|
||
|
if (sc.atLineEnd) {
|
||
|
sc.SetState(SCE_FS_DEFAULT);
|
||
|
}
|
||
|
} else if (sc.state == SCE_FS_DATE) {
|
||
|
if (sc.ch == '#' || !IsADateCharacter(sc.chNext)) {
|
||
|
sc.ForwardSetState(SCE_FS_DEFAULT);
|
||
|
}
|
||
|
}
|
||
|
|
||
|
// Determine if a new state should be entered.
|
||
|
if (sc.state == SCE_FS_DEFAULT) {
|
||
|
if (sc.Match('/', '*')) { // New code
|
||
|
sc.SetState(SCE_FS_COMMENT);
|
||
|
sc.Forward(); // Eat the * so it isn't used for the end of the comment
|
||
|
//if (sc.ch == '\'') { // Old code
|
||
|
// sc.SetState(SCE_FS_COMMENT); // old code
|
||
|
} else if (sc.Match('/', '/')) { // New code
|
||
|
sc.SetState(SCE_FS_COMMENTLINE);
|
||
|
} else if (sc.ch == '\"') {
|
||
|
sc.SetState(SCE_FS_STRING);
|
||
|
} else if (sc.ch == '#' && visibleChars == 0) {
|
||
|
// Preprocessor commands are alone on their line
|
||
|
sc.SetState(SCE_FS_PREPROCESSOR);
|
||
|
} else if (sc.ch == '#') {
|
||
|
int n = 1;
|
||
|
int chSeek = ' ';
|
||
|
while ((n < 100) && (chSeek == ' ' || chSeek == '\t')) {
|
||
|
chSeek = sc.GetRelative(n);
|
||
|
n++;
|
||
|
}
|
||
|
if (IsADigit(chSeek)) {
|
||
|
sc.SetState(SCE_FS_DATE);
|
||
|
} else {
|
||
|
sc.SetState(SCE_FS_OPERATOR);
|
||
|
}
|
||
|
} else if (sc.ch == '&' && tolower(sc.chNext) == 'h') {
|
||
|
sc.SetState(SCE_FS_NUMBER);
|
||
|
} else if (sc.ch == '&' && tolower(sc.chNext) == 'o') {
|
||
|
sc.SetState(SCE_FS_NUMBER);
|
||
|
} else if (IsADigit(sc.ch) || (sc.ch == '.' && IsADigit(sc.chNext))) {
|
||
|
sc.SetState(SCE_FS_NUMBER);
|
||
|
} else if (IsAWordStart(sc.ch) || (sc.ch == '[')) {
|
||
|
sc.SetState(SCE_FS_IDENTIFIER);
|
||
|
} else if (isoperator(static_cast<char>(sc.ch)) || (sc.ch == '\\')) {
|
||
|
sc.SetState(SCE_FS_OPERATOR);
|
||
|
}
|
||
|
}
|
||
|
|
||
|
if (sc.atLineEnd) {
|
||
|
visibleChars = 0;
|
||
|
}
|
||
|
if (!IsASpace(sc.ch)) {
|
||
|
visibleChars++;
|
||
|
}
|
||
|
}
|
||
|
sc.Complete();
|
||
|
}
|
||
|
|
||
|
static void FoldFlagShipDoc(unsigned int startPos, int length, int,
|
||
|
WordList *[], Accessor &styler) {
|
||
|
|
||
|
int endPos = startPos + length;
|
||
|
|
||
|
// Backtrack to previous line in case need to fix its fold status
|
||
|
int lineCurrent = styler.GetLine(startPos);
|
||
|
if (startPos > 0) {
|
||
|
if (lineCurrent > 0) {
|
||
|
lineCurrent--;
|
||
|
startPos = styler.LineStart(lineCurrent);
|
||
|
}
|
||
|
}
|
||
|
int spaceFlags = 0;
|
||
|
int indentCurrent = styler.IndentAmount(lineCurrent, &spaceFlags, IsFlagShipComment);
|
||
|
char chNext = styler[startPos];
|
||
|
for (int i = startPos; i < endPos; i++) {
|
||
|
char ch = chNext;
|
||
|
chNext = styler.SafeGetCharAt(i + 1);
|
||
|
|
||
|
if ((ch == '\r' && chNext != '\n') || (ch == '\n') || (i == endPos)) {
|
||
|
int lev = indentCurrent;
|
||
|
int indentNext = styler.IndentAmount(lineCurrent + 1, &spaceFlags, IsFlagShipComment);
|
||
|
if (!(indentCurrent & SC_FOLDLEVELWHITEFLAG)) {
|
||
|
// Only non whitespace lines can be headers
|
||
|
if ((indentCurrent & SC_FOLDLEVELNUMBERMASK) < (indentNext & SC_FOLDLEVELNUMBERMASK)) {
|
||
|
lev |= SC_FOLDLEVELHEADERFLAG;
|
||
|
} else if (indentNext & SC_FOLDLEVELWHITEFLAG) {
|
||
|
// Line after is blank so check the next - maybe should continue further?
|
||
|
int spaceFlags2 = 0;
|
||
|
int indentNext2 = styler.IndentAmount(lineCurrent + 2, &spaceFlags2, IsFlagShipComment);
|
||
|
if ((indentCurrent & SC_FOLDLEVELNUMBERMASK) < (indentNext2 & SC_FOLDLEVELNUMBERMASK)) {
|
||
|
lev |= SC_FOLDLEVELHEADERFLAG;
|
||
|
}
|
||
|
}
|
||
|
}
|
||
|
indentCurrent = indentNext;
|
||
|
styler.SetLevel(lineCurrent, lev);
|
||
|
lineCurrent++;
|
||
|
}
|
||
|
}
|
||
|
}
|
||
|
|
||
|
|
||
|
static const char * const FSWordListDesc[] = {
|
||
|
"Keywords",
|
||
|
"functions",
|
||
|
"user2",
|
||
|
"user3",
|
||
|
0
|
||
|
};
|
||
|
|
||
|
LexerModule lmFlagShip(SCLEX_FLAGSHIP, ColouriseFlagShipDoc, "flagship", FoldFlagShipDoc, FSWordListDesc);
|
||
|
|
||
|
|
||
|
|