/** @file LexPython.cxx
** Lexer for Python.
**/
-// Copyright 1998-2001 by Neil Hodgson <neilh@scintilla.org>
+// Copyright 1998-2002 by Neil Hodgson <neilh@scintilla.org>
// The License.txt file describes the conditions under which this software may be distributed.
#include <stdlib.h>
#include "Scintilla.h"
#include "SciLexer.h"
+#ifdef SCI_NAMESPACE
+using namespace Scintilla;
+#endif
+
enum kwType { kwOther, kwClass, kwDef, kwImport };
+static const int indicatorWhitespace = 1;
static bool IsPyComment(Accessor &styler, int pos, int len) {
- return len>0 && styler[pos]=='#';
+ return len > 0 && styler[pos] == '#';
}
static bool IsPyStringStart(int ch, int chNext, int chNext2) {
}
/* Return the state to use for the string starting at i; *nextIndex will be set to the first index following the quote(s) */
-static int GetPyStringState(Accessor &styler, int i, int *nextIndex) {
+static int GetPyStringState(Accessor &styler, int i, unsigned int *nextIndex) {
char ch = styler.SafeGetCharAt(i);
char chNext = styler.SafeGetCharAt(i + 1);
i++;
ch = styler.SafeGetCharAt(i);
chNext = styler.SafeGetCharAt(i + 1);
- }
- else if (ch == 'u' || ch == 'U') {
+ } else if (ch == 'u' || ch == 'U') {
if (chNext == 'r' || chNext == 'R')
i += 2;
else
}
}
-inline bool IsAWordChar(int ch) {
+static inline bool IsAWordChar(int ch) {
return (ch < 0x80) && (isalnum(ch) || ch == '.' || ch == '_');
}
-inline bool IsAWordStart(int ch) {
+static inline bool IsAWordStart(int ch) {
return (ch < 0x80) && (isalnum(ch) || ch == '_');
}
static void ColourisePyDoc(unsigned int startPos, int length, int initStyle,
- WordList *keywordlists[], Accessor &styler) {
+ WordList *keywordlists[], Accessor &styler) {
int endPos = startPos + length;
int lineCurrent = styler.GetLine(startPos);
if (startPos > 0) {
if (lineCurrent > 0) {
- startPos = styler.LineStart(lineCurrent-1);
+ lineCurrent--;
+ startPos = styler.LineStart(lineCurrent);
if (startPos == 0)
initStyle = SCE_P_DEFAULT;
else
- initStyle = styler.StyleAt(startPos-1);
+ initStyle = styler.StyleAt(startPos - 1);
}
}
WordList &keywords = *keywordlists[0];
+ WordList &keywords2 = *keywordlists[1];
const int whingeLevel = styler.GetPropertyInt("tab.timmy.whinge.level");
kwType kwLast = kwOther;
int spaceFlags = 0;
styler.IndentAmount(lineCurrent, &spaceFlags, IsPyComment);
-
- // Python uses a different mask because bad indentation is marked by oring with 32
- StyleContext sc(startPos, endPos-startPos, initStyle, styler, 0x7f);
-
+ bool hexadecimal = false;
+
+ StyleContext sc(startPos, endPos - startPos, initStyle, styler);
+
+ bool indentGood = true;
+ int startIndicator = sc.currentPos;
+
for (; sc.More(); sc.Forward()) {
if (sc.atLineStart) {
- const char chBad = static_cast<char>(64);
- const char chGood = static_cast<char>(0);
- char chFlags = chGood;
+ styler.IndentAmount(lineCurrent, &spaceFlags, IsPyComment);
+ indentGood = true;
if (whingeLevel == 1) {
- chFlags = (spaceFlags & wsInconsistent) ? chBad : chGood;
+ indentGood = (spaceFlags & wsInconsistent) == 0;
} else if (whingeLevel == 2) {
- chFlags = (spaceFlags & wsSpaceTab) ? chBad : chGood;
+ indentGood = (spaceFlags & wsSpaceTab) == 0;
} else if (whingeLevel == 3) {
- chFlags = (spaceFlags & wsSpace) ? chBad : chGood;
+ indentGood = (spaceFlags & wsSpace) == 0;
} else if (whingeLevel == 4) {
- chFlags = (spaceFlags & wsTab) ? chBad : chGood;
+ indentGood = (spaceFlags & wsTab) == 0;
+ }
+ if (!indentGood) {
+ styler.IndicatorFill(startIndicator, sc.currentPos, indicatorWhitespace, 0);
+ startIndicator = sc.currentPos;
}
- styler.SetFlags(chFlags, static_cast<char>(sc.state));
}
if (sc.atLineEnd) {
- if ((sc.state == SCE_P_DEFAULT) ||
- (sc.state == SCE_P_TRIPLE) ||
- (sc.state == SCE_P_TRIPLEDOUBLE)) {
+ if ((sc.state == SCE_P_DEFAULT) ||
+ (sc.state == SCE_P_TRIPLE) ||
+ (sc.state == SCE_P_TRIPLEDOUBLE)) {
// Perform colourisation of white space and triple quoted strings at end of each line to allow
// tab marking to work inside white space and triple quoted strings
- sc.ForwardSetState(sc.state);
+ sc.SetState(sc.state);
}
lineCurrent++;
- styler.IndentAmount(lineCurrent, &spaceFlags, IsPyComment);
if ((sc.state == SCE_P_STRING) || (sc.state == SCE_P_CHARACTER)) {
sc.ChangeState(SCE_P_STRINGEOL);
sc.ForwardSetState(SCE_P_DEFAULT);
}
+ if (!sc.More())
+ break;
}
+ bool needEOLCheck = false;
+
// Check for a state end
if (sc.state == SCE_P_OPERATOR) {
kwLast = kwOther;
- sc.SetState(SCE_C_DEFAULT);
+ sc.SetState(SCE_P_DEFAULT);
} else if (sc.state == SCE_P_NUMBER) {
- if (!IsAWordChar(sc.ch)) {
+ if (!IsAWordChar(sc.ch) &&
+ !(!hexadecimal && ((sc.ch == '+' || sc.ch == '-') && (sc.chPrev == 'e' || sc.chPrev == 'E')))) {
sc.SetState(SCE_P_DEFAULT);
}
- } else if (sc.state == SCE_P_WORD) {
+ } else if (sc.state == SCE_P_IDENTIFIER) {
if ((sc.ch == '.') || (!IsAWordChar(sc.ch))) {
char s[100];
sc.GetCurrent(s, sizeof(s));
style = SCE_P_CLASSNAME;
} else if (kwLast == kwDef) {
style = SCE_P_DEFNAME;
+ } else if (keywords2.InList(s)) {
+ style = SCE_P_WORD2;
}
sc.ChangeState(style);
sc.SetState(SCE_P_DEFAULT);
kwLast = kwImport;
else
kwLast = kwOther;
- } else if (style == SCE_P_CLASSNAME) {
- kwLast = kwOther;
- } else if (style == SCE_P_DEFNAME) {
- kwLast = kwOther;
+ } else {
+ kwLast = kwOther;
}
}
} else if ((sc.state == SCE_P_COMMENTLINE) || (sc.state == SCE_P_COMMENTBLOCK)) {
if (sc.ch == '\r' || sc.ch == '\n') {
sc.SetState(SCE_P_DEFAULT);
}
+ } else if (sc.state == SCE_P_DECORATOR) {
+ if (!IsAWordChar(sc.ch)) {
+ sc.SetState(SCE_P_DEFAULT);
+ }
} else if ((sc.state == SCE_P_STRING) || (sc.state == SCE_P_CHARACTER)) {
if (sc.ch == '\\') {
if ((sc.chNext == '\r') && (sc.GetRelative(2) == '\n')) {
sc.Forward();
} else if ((sc.state == SCE_P_STRING) && (sc.ch == '\"')) {
sc.ForwardSetState(SCE_P_DEFAULT);
+ needEOLCheck = true;
} else if ((sc.state == SCE_P_CHARACTER) && (sc.ch == '\'')) {
sc.ForwardSetState(SCE_P_DEFAULT);
+ needEOLCheck = true;
}
} else if (sc.state == SCE_P_TRIPLE) {
if (sc.ch == '\\') {
sc.Forward();
sc.Forward();
sc.ForwardSetState(SCE_P_DEFAULT);
+ needEOLCheck = true;
}
} else if (sc.state == SCE_P_TRIPLEDOUBLE) {
if (sc.ch == '\\') {
sc.Forward();
sc.Forward();
sc.ForwardSetState(SCE_P_DEFAULT);
+ needEOLCheck = true;
}
}
+ if (!indentGood && !IsASpaceOrTab(sc.ch)) {
+ styler.IndicatorFill(startIndicator, sc.currentPos, indicatorWhitespace, 1);
+ startIndicator = sc.currentPos;
+ indentGood = true;
+ }
+
+ // State exit code may have moved on to end of line
+ if (needEOLCheck && sc.atLineEnd) {
+ lineCurrent++;
+ styler.IndentAmount(lineCurrent, &spaceFlags, IsPyComment);
+ if (!sc.More())
+ break;
+ }
+
// Check for a new state starting character
if (sc.state == SCE_P_DEFAULT) {
- if (isascii(sc.ch) && isoperator(static_cast<char>(sc.ch)) || sc.ch == '`') {
+ if (IsADigit(sc.ch) || (sc.ch == '.' && IsADigit(sc.chNext))) {
+ if (sc.ch == '0' && (sc.chNext == 'x' || sc.chNext == 'X')) {
+ hexadecimal = true;
+ } else {
+ hexadecimal = false;
+ }
+ sc.SetState(SCE_P_NUMBER);
+ } else if (isascii(sc.ch) && isoperator(static_cast<char>(sc.ch)) || sc.ch == '`') {
sc.SetState(SCE_P_OPERATOR);
} else if (sc.ch == '#') {
sc.SetState(sc.chNext == '#' ? SCE_P_COMMENTBLOCK : SCE_P_COMMENTLINE);
- } else if (IsADigit(sc.ch) || (sc.ch == '.' && IsADigit(sc.chNext))) {
- sc.SetState(SCE_P_NUMBER);
+ } else if (sc.ch == '@') {
+ sc.SetState(SCE_P_DECORATOR);
} else if (IsPyStringStart(sc.ch, sc.chNext, sc.GetRelative(2))) {
- int nextIndex = 0;
+ unsigned int nextIndex = 0;
sc.SetState(GetPyStringState(styler, sc.currentPos, &nextIndex));
- while (nextIndex > (sc.currentPos+1)) {
+ while (nextIndex > (sc.currentPos + 1) && sc.More()) {
sc.Forward();
}
} else if (IsAWordStart(sc.ch)) {
- sc.SetState(SCE_P_WORD);
+ sc.SetState(SCE_P_IDENTIFIER);
}
}
}
+ styler.IndicatorFill(startIndicator, sc.currentPos, indicatorWhitespace, 0);
sc.Complete();
}
static bool IsCommentLine(int line, Accessor &styler) {
int pos = styler.LineStart(line);
- int eol_pos = styler.LineStart(line+1) - 1;
+ int eol_pos = styler.LineStart(line + 1) - 1;
for (int i = pos; i < eol_pos; i++) {
char ch = styler[i];
if (ch == '#')
static bool IsQuoteLine(int line, Accessor &styler) {
int style = styler.StyleAt(styler.LineStart(line)) & 31;
- return ((style == SCE_P_TRIPLE) || (style== SCE_P_TRIPLEDOUBLE));
+ return ((style == SCE_P_TRIPLE) || (style == SCE_P_TRIPLEDOUBLE));
}
static void FoldPyDoc(unsigned int startPos, int length, int /*initStyle - unused*/,
- WordList *[], Accessor &styler) {
+ WordList *[], Accessor &styler) {
const int maxPos = startPos + length;
- const int maxLines = styler.GetLine(maxPos-1); // Requested last line
+ const int maxLines = styler.GetLine(maxPos - 1); // Requested last line
const int docLines = styler.GetLine(styler.Length() - 1); // Available last line
- const bool foldComment = styler.GetPropertyInt("fold.comment.python");
- const bool foldQuotes = styler.GetPropertyInt("fold.quotes.python");
-
+ const bool foldComment = styler.GetPropertyInt("fold.comment.python") != 0;
+ const bool foldQuotes = styler.GetPropertyInt("fold.quotes.python") != 0;
+
// Backtrack to previous non-blank line so we can determine indent level
// for any white space lines (needed esp. within triple quoted strings)
// and so we can fix any preceding fold level (which is why we go back
lineCurrent--;
indentCurrent = styler.IndentAmount(lineCurrent, &spaceFlags, NULL);
if (!(indentCurrent & SC_FOLDLEVELWHITEFLAG) &&
- (!IsCommentLine(lineCurrent, styler)) &&
- (!IsQuoteLine(lineCurrent, styler)))
+ (!IsCommentLine(lineCurrent, styler)) &&
+ (!IsQuoteLine(lineCurrent, styler)))
break;
}
int indentCurrentLevel = indentCurrent & SC_FOLDLEVELNUMBERMASK;
-
+
// Set up initial loop state
startPos = styler.LineStart(lineCurrent);
int prev_state = SCE_P_DEFAULT & 31;
if (lineCurrent >= 1)
- prev_state = styler.StyleAt(startPos-1) & 31;
+ prev_state = styler.StyleAt(startPos - 1) & 31;
int prevQuote = foldQuotes && ((prev_state == SCE_P_TRIPLE) || (prev_state == SCE_P_TRIPLEDOUBLE));
int prevComment = 0;
if (lineCurrent >= 1)
// Information about next line is only available if not at end of document
indentNext = styler.IndentAmount(lineNext, &spaceFlags, NULL);
int style = styler.StyleAt(styler.LineStart(lineNext)) & 31;
- quote = foldQuotes && ((style == SCE_P_TRIPLE) || (style== SCE_P_TRIPLEDOUBLE));
+ quote = foldQuotes && ((style == SCE_P_TRIPLE) || (style == SCE_P_TRIPLEDOUBLE));
}
const int quote_start = (quote && !prevQuote);
const int quote_continue = (quote && prevQuote);
const int comment = foldComment && IsCommentLine(lineCurrent, styler);
const int comment_start = (comment && !prevComment && (lineNext <= docLines) &&
- IsCommentLine(lineNext, styler) && (lev > SC_FOLDLEVELBASE));
+ IsCommentLine(lineNext, styler) && (lev > SC_FOLDLEVELBASE));
const int comment_continue = (comment && prevComment);
if ((!quote || !prevQuote) && !comment)
indentCurrentLevel = indentCurrent & SC_FOLDLEVELNUMBERMASK;
lev = lev + 1;
}
- // Skip past any blank lines for next indent level info; we skip also comments
- // starting in column 0 which effectively folds them into surrounding code rather
+ // Skip past any blank lines for next indent level info; we skip also
+ // comments (all comments, not just those starting in column 0)
+ // which effectively folds them into surrounding code rather
// than screwing up folding.
- const int saveIndentNext = indentNext;
+
while (!quote &&
- (lineNext < docLines) &&
- ((indentNext & SC_FOLDLEVELWHITEFLAG) ||
- (lineNext <= docLines && styler[styler.LineStart(lineNext)] == '#'))) {
+ (lineNext < docLines) &&
+ ((indentNext & SC_FOLDLEVELWHITEFLAG) ||
+ (lineNext <= docLines && IsCommentLine(lineNext, styler)))) {
lineNext++;
indentNext = styler.IndentAmount(lineNext, &spaceFlags, NULL);
}
- // Next compute max indent level of current line and next non-blank line.
- // This is the level to which we set all the intervening blank or comment lines.
- const int skip_level = Platform::Maximum(indentCurrentLevel,
- indentNext & SC_FOLDLEVELNUMBERMASK);
-
+ const int levelAfterComments = indentNext & SC_FOLDLEVELNUMBERMASK;
+ const int levelBeforeComments = Platform::Maximum(indentCurrentLevel,levelAfterComments);
+
// Now set all the indent levels on the lines we skipped
- int skipLine = lineCurrent + 1;
- int skipIndentNext = saveIndentNext;
- while (skipLine < lineNext) {
- int skipLineLevel = skip_level;
- if (skipIndentNext & SC_FOLDLEVELWHITEFLAG)
- skipLineLevel = SC_FOLDLEVELWHITEFLAG | skipLineLevel;
- styler.SetLevel(skipLine, skipLineLevel);
- skipLine++;
- skipIndentNext = styler.IndentAmount(skipLine, &spaceFlags, NULL);
+ // Do this from end to start. Once we encounter one line
+ // which is indented more than the line after the end of
+ // the comment-block, use the level of the block before
+
+ int skipLine = lineNext;
+ int skipLevel = levelAfterComments;
+
+ while (--skipLine > lineCurrent) {
+ int skipLineIndent = styler.IndentAmount(skipLine, &spaceFlags, NULL);
+
+ if ((skipLineIndent & SC_FOLDLEVELNUMBERMASK) > levelAfterComments)
+ skipLevel = levelBeforeComments;
+
+ int whiteFlag = skipLineIndent & SC_FOLDLEVELWHITEFLAG;
+
+ styler.SetLevel(skipLine, skipLevel | whiteFlag);
}
-
+
// Set fold header on non-quote/non-comment line
if (!quote && !comment && !(indentCurrent & SC_FOLDLEVELWHITEFLAG) ) {
if ((indentCurrent & SC_FOLDLEVELNUMBERMASK) < (indentNext & SC_FOLDLEVELNUMBERMASK))
//styler.SetLevel(lineCurrent, indentCurrent);
}
-LexerModule lmPython(SCLEX_PYTHON, ColourisePyDoc, "python", FoldPyDoc);
+static const char * const pythonWordListDesc[] = {
+ "Keywords",
+ "Highlighted identifiers",
+ 0
+};
+
+LexerModule lmPython(SCLEX_PYTHON, ColourisePyDoc, "python", FoldPyDoc,
+ pythonWordListDesc);