You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
1090 lines
37 KiB
1090 lines
37 KiB
// Scintilla source code edit control |
|
/** @file LexPython.cxx |
|
** Lexer for Python. |
|
**/ |
|
// Copyright 1998-2002 by Neil Hodgson <neilh@scintilla.org> |
|
// The License.txt file describes the conditions under which this software may be distributed. |
|
|
|
#include <cstdlib> |
|
#include <cassert> |
|
#include <cstring> |
|
|
|
#include <string> |
|
#include <string_view> |
|
#include <vector> |
|
#include <map> |
|
#include <algorithm> |
|
#include <functional> |
|
|
|
#include "ILexer.h" |
|
#include "Scintilla.h" |
|
#include "SciLexer.h" |
|
|
|
#include "WordList.h" |
|
#include "LexAccessor.h" |
|
#include "Accessor.h" |
|
#include "StyleContext.h" |
|
#include "CharacterSet.h" |
|
#include "CharacterCategory.h" |
|
#include "LexerModule.h" |
|
#include "OptionSet.h" |
|
#include "SubStyles.h" |
|
#include "DefaultLexer.h" |
|
|
|
using namespace Scintilla; |
|
using namespace Lexilla; |
|
|
|
namespace { |
|
// Use an unnamed namespace to protect the functions and classes from name conflicts |
|
|
|
/* Notes on f-strings: f-strings are strings prefixed with f (e.g. f'') that may |
|
have arbitrary expressions in {}. The tokens in the expressions are lexed as if |
|
they were outside of any string. Expressions may contain { and } characters as |
|
long as there is a closing } for every {, may be 2+ lines in a triple quoted |
|
string, and may have a formatting specifier following a ! or :, but both ! |
|
and : are valid inside of a bracketed expression and != is a valid |
|
expression token even outside of a bracketed expression. |
|
|
|
When in an f-string expression, the lexer keeps track of the state value of |
|
the f-string and the nesting count for the expression (# of [, (, { seen - # of |
|
}, ), ] seen). f-strings may be nested (e.g. f'{ a + f"{1+2}"') so a stack of |
|
states and nesting counts is kept. If a f-string expression continues beyond |
|
the end of a line, this stack is saved in a std::map that maps a line number to |
|
the stack at the end of that line. std::vector is used for the stack. |
|
|
|
The PEP for f-strings is at https://www.python.org/dev/peps/pep-0498/ |
|
*/ |
|
struct SingleFStringExpState { |
|
int state; |
|
int nestingCount; |
|
}; |
|
|
|
/* kwCDef, kwCTypeName only used for Cython */ |
|
enum kwType { kwOther, kwClass, kwDef, kwImport, kwCDef, kwCTypeName, kwCPDef }; |
|
|
|
enum literalsAllowed { litNone = 0, litU = 1, litB = 2, litF = 4 }; |
|
|
|
constexpr int indicatorWhitespace = 1; |
|
|
|
bool IsPyComment(Accessor &styler, Sci_Position pos, Sci_Position len) { |
|
return len > 0 && styler[pos] == '#'; |
|
} |
|
|
|
constexpr bool IsPyStringTypeChar(int ch, literalsAllowed allowed) noexcept { |
|
return |
|
((allowed & litB) && (ch == 'b' || ch == 'B')) || |
|
((allowed & litU) && (ch == 'u' || ch == 'U')) || |
|
((allowed & litF) && (ch == 'f' || ch == 'F')); |
|
} |
|
|
|
constexpr bool IsQuote(int ch) { |
|
return AnyOf(ch, '"', '\''); |
|
} |
|
|
|
constexpr bool IsRawPrefix(int ch) { |
|
return AnyOf(ch, 'r', 'R'); |
|
} |
|
|
|
bool IsPyStringStart(int ch, int chNext, int chNext2, literalsAllowed allowed) noexcept { |
|
// To cover both python2 and python3 lex character prefixes as -- |
|
// ur'' is a string, but ru'' is not |
|
// fr'', rf'', br'', rb'' are all strings |
|
if (IsQuote(ch)) |
|
return true; |
|
if (IsPyStringTypeChar(ch, allowed)) { |
|
if (IsQuote(chNext)) |
|
return true; |
|
if (IsRawPrefix(chNext) && IsQuote(chNext2)) |
|
return true; |
|
} |
|
if (IsRawPrefix(ch)) { |
|
if (IsQuote(chNext)) |
|
return true; |
|
if (IsPyStringTypeChar(chNext, allowed) && !AnyOf(chNext, 'u', 'U') && IsQuote(chNext2)) |
|
return true; |
|
} |
|
|
|
return false; |
|
} |
|
|
|
constexpr bool IsPyFStringState(int st) noexcept { |
|
return AnyOf(st, SCE_P_FCHARACTER, SCE_P_FSTRING, SCE_P_FTRIPLE, SCE_P_FTRIPLEDOUBLE); |
|
} |
|
|
|
constexpr bool IsPySingleQuoteStringState(int st) noexcept { |
|
return AnyOf(st, SCE_P_CHARACTER, SCE_P_STRING, SCE_P_FCHARACTER, SCE_P_FSTRING); |
|
} |
|
|
|
constexpr bool IsPyTripleQuoteStringState(int st) noexcept { |
|
return AnyOf(st, SCE_P_TRIPLE, SCE_P_TRIPLEDOUBLE, SCE_P_FTRIPLE, SCE_P_FTRIPLEDOUBLE); |
|
} |
|
|
|
char GetPyStringQuoteChar(int st) noexcept { |
|
if (AnyOf(st, SCE_P_CHARACTER, SCE_P_FCHARACTER, SCE_P_TRIPLE, SCE_P_FTRIPLE)) |
|
return '\''; |
|
if (AnyOf(st, SCE_P_STRING, SCE_P_FSTRING, SCE_P_TRIPLEDOUBLE, SCE_P_FTRIPLEDOUBLE)) |
|
return '"'; |
|
|
|
return '\0'; |
|
} |
|
|
|
void PushStateToStack(int state, std::vector<SingleFStringExpState> &stack, SingleFStringExpState *¤tFStringExp) { |
|
const SingleFStringExpState single = {state, 0}; |
|
stack.push_back(single); |
|
|
|
currentFStringExp = &stack.back(); |
|
} |
|
|
|
int PopFromStateStack(std::vector<SingleFStringExpState> &stack, SingleFStringExpState *¤tFStringExp) noexcept { |
|
int state = 0; |
|
|
|
if (!stack.empty()) { |
|
state = stack.back().state; |
|
stack.pop_back(); |
|
} |
|
|
|
if (stack.empty()) { |
|
currentFStringExp = nullptr; |
|
} else { |
|
currentFStringExp = &stack.back(); |
|
} |
|
|
|
return state; |
|
} |
|
|
|
/* Return the state to use for the string starting at i; *nextIndex will be set to the first index following the quote(s) */ |
|
int GetPyStringState(Accessor &styler, Sci_Position i, Sci_PositionU *nextIndex, literalsAllowed allowed) { |
|
char ch = styler.SafeGetCharAt(i); |
|
char chNext = styler.SafeGetCharAt(i + 1); |
|
bool isFString = false; |
|
|
|
// Advance beyond r, a type char, or both (in either order) |
|
// Note that this depends on IsPyStringStart to enforce ru'' not being a string |
|
if (IsRawPrefix(ch)) { |
|
i++; |
|
if (IsPyStringTypeChar(chNext, allowed)) { |
|
if (AnyOf(chNext, 'f', 'F')) |
|
isFString = true; |
|
i++; |
|
} |
|
ch = styler.SafeGetCharAt(i); |
|
chNext = styler.SafeGetCharAt(i + 1); |
|
} else if (IsPyStringTypeChar(ch, allowed)) { |
|
if (AnyOf(ch, 'f', 'F')) |
|
isFString = true; |
|
if (IsRawPrefix(chNext)) |
|
i += 2; |
|
else |
|
i += 1; |
|
ch = styler.SafeGetCharAt(i); |
|
chNext = styler.SafeGetCharAt(i + 1); |
|
} |
|
|
|
// ch and i will be the first quote |
|
if (!IsQuote(ch)) { |
|
*nextIndex = i + 1; |
|
return SCE_P_DEFAULT; |
|
} |
|
|
|
if (ch == chNext && ch == styler.SafeGetCharAt(i + 2)) { |
|
*nextIndex = i + 3; |
|
|
|
if (ch == '"') |
|
return (isFString ? SCE_P_FTRIPLEDOUBLE : SCE_P_TRIPLEDOUBLE); |
|
else |
|
return (isFString ? SCE_P_FTRIPLE : SCE_P_TRIPLE); |
|
} else { |
|
*nextIndex = i + 1; |
|
|
|
if (ch == '"') |
|
return (isFString ? SCE_P_FSTRING : SCE_P_STRING); |
|
else |
|
return (isFString ? SCE_P_FCHARACTER : SCE_P_CHARACTER); |
|
} |
|
} |
|
|
|
bool IsAWordChar(int ch, bool unicodeIdentifiers) noexcept { |
|
if (IsASCII(ch)) |
|
return (IsAlphaNumeric(ch) || ch == '.' || ch == '_'); |
|
|
|
if (!unicodeIdentifiers) |
|
return false; |
|
|
|
// Python uses the XID_Continue set from Unicode data |
|
return IsXidContinue(ch); |
|
} |
|
|
|
bool IsAWordStart(int ch, bool unicodeIdentifiers) noexcept { |
|
if (IsASCII(ch)) |
|
return (IsUpperOrLowerCase(ch) || ch == '_'); |
|
|
|
if (!unicodeIdentifiers) |
|
return false; |
|
|
|
// Python uses the XID_Start set from Unicode data |
|
return IsXidStart(ch); |
|
} |
|
|
|
bool IsFirstNonWhitespace(Sci_Position pos, Accessor &styler) { |
|
const Sci_Position line = styler.GetLine(pos); |
|
const Sci_Position start_pos = styler.LineStart(line); |
|
for (Sci_Position i = start_pos; i < pos; i++) { |
|
const char ch = styler[i]; |
|
if (!IsASpaceOrTab(ch)) |
|
return false; |
|
} |
|
return true; |
|
} |
|
|
|
unsigned char GetNextNonWhitespaceChar(Accessor &styler, Sci_PositionU pos, Sci_PositionU maxPos, Sci_PositionU *charPosPtr = nullptr) { |
|
while (pos < maxPos) { |
|
const unsigned char ch = styler.SafeGetCharAt(pos, '\0'); |
|
if (!AnyOf(ch, ' ', '\t', '\n', '\r')) { |
|
if (charPosPtr != nullptr) { |
|
*charPosPtr = pos; |
|
} |
|
return ch; |
|
} |
|
pos++; |
|
} |
|
|
|
return '\0'; |
|
} |
|
|
|
// Returns whether symbol is "match" or "case" and it is an identifier & |
|
// not a keyword. Does not require the line to end with : so "match\n" |
|
// and "match (x)\n" return false because match could be a keyword once |
|
// more text is added |
|
bool IsMatchOrCaseIdentifier(const StyleContext &sc, Accessor &styler, const char *symbol) { |
|
if (strcmp(symbol, "match") != 0 && strcmp(symbol, "case") != 0) { |
|
return false; |
|
} |
|
|
|
if (!IsFirstNonWhitespace(sc.currentPos - strlen(symbol), styler)) { |
|
return true; |
|
} |
|
|
|
// The match keyword can be followed by an expression; the case keyword |
|
// is a bit more restrictive but not much. Here, we look for the next |
|
// char and return false if the char cannot start an expression; for '.' |
|
// we look at the following char to see if it's a digit because .1 is a |
|
// number |
|
Sci_PositionU nextCharPos = 0; |
|
const unsigned char nextChar = GetNextNonWhitespaceChar(styler, sc.currentPos, sc.lineEnd, &nextCharPos); |
|
if (nextChar == '=' || nextChar == '#') { |
|
return true; |
|
} |
|
if (nextChar == '.' && nextCharPos >= sc.currentPos) { |
|
const unsigned char followingChar = GetNextNonWhitespaceChar(styler, nextCharPos+1, sc.lineEnd); |
|
if (!IsADigit(followingChar)) { |
|
return true; |
|
} |
|
} |
|
|
|
return false; |
|
} |
|
|
|
// Options used for LexerPython |
|
struct OptionsPython { |
|
int whingeLevel = 0; |
|
bool base2or8Literals = true; |
|
bool stringsU = true; |
|
bool stringsB = true; |
|
bool stringsF = true; |
|
bool stringsOverNewline = false; |
|
bool keywords2NoSubIdentifiers = false; |
|
bool fold = false; |
|
bool foldQuotes = false; |
|
bool foldCompact = false; |
|
bool unicodeIdentifiers = true; |
|
int identifierAttributes = 0; |
|
int decoratorAttributes = 0; |
|
bool pep701StringsF = true; |
|
|
|
[[nodiscard]] literalsAllowed AllowedLiterals() const noexcept { |
|
literalsAllowed allowedLiterals = stringsU ? litU : litNone; |
|
if (stringsB) |
|
allowedLiterals = static_cast<literalsAllowed>(allowedLiterals | litB); |
|
if (stringsF) |
|
allowedLiterals = static_cast<literalsAllowed>(allowedLiterals | litF); |
|
return allowedLiterals; |
|
} |
|
}; |
|
|
|
const char *const pythonWordListDesc[] = { |
|
"Keywords", |
|
"Highlighted identifiers", |
|
nullptr |
|
}; |
|
|
|
struct OptionSetPython : public OptionSet<OptionsPython> { |
|
OptionSetPython() { |
|
DefineProperty("tab.timmy.whinge.level", &OptionsPython::whingeLevel, |
|
"For Python code, checks whether indenting is consistent. " |
|
"The default, 0 turns off indentation checking, " |
|
"1 checks whether each line is potentially inconsistent with the previous line, " |
|
"2 checks whether any space characters occur before a tab character in the indentation, " |
|
"3 checks whether any spaces are in the indentation, and " |
|
"4 checks for any tab characters in the indentation. " |
|
"1 is a good level to use."); |
|
|
|
DefineProperty("lexer.python.literals.binary", &OptionsPython::base2or8Literals, |
|
"Set to 0 to not recognise Python 3 binary and octal literals: 0b1011 0o712."); |
|
|
|
DefineProperty("lexer.python.strings.u", &OptionsPython::stringsU, |
|
"Set to 0 to not recognise Python Unicode literals u\"x\" as used before Python 3."); |
|
|
|
DefineProperty("lexer.python.strings.b", &OptionsPython::stringsB, |
|
"Set to 0 to not recognise Python 3 bytes literals b\"x\"."); |
|
|
|
DefineProperty("lexer.python.strings.f", &OptionsPython::stringsF, |
|
"Set to 0 to not recognise Python 3.6 f-string literals f\"var={var}\"."); |
|
|
|
DefineProperty("lexer.python.strings.f.pep.701", &OptionsPython::pep701StringsF, |
|
"Set to 0 to use pre-PEP 701 / Python 3.12 f-string lexing."); |
|
|
|
DefineProperty("lexer.python.strings.over.newline", &OptionsPython::stringsOverNewline, |
|
"Set to 1 to allow strings to span newline characters."); |
|
|
|
DefineProperty("lexer.python.keywords2.no.sub.identifiers", &OptionsPython::keywords2NoSubIdentifiers, |
|
"When enabled, it will not style keywords2 items that are used as a sub-identifier. " |
|
"Example: when set, will not highlight \"foo.open\" when \"open\" is a keywords2 item."); |
|
|
|
DefineProperty("fold", &OptionsPython::fold); |
|
|
|
DefineProperty("fold.quotes.python", &OptionsPython::foldQuotes, |
|
"This option enables folding multi-line quoted strings when using the Python lexer."); |
|
|
|
DefineProperty("fold.compact", &OptionsPython::foldCompact); |
|
|
|
DefineProperty("lexer.python.unicode.identifiers", &OptionsPython::unicodeIdentifiers, |
|
"Set to 0 to not recognise Python 3 Unicode identifiers."); |
|
|
|
DefineProperty("lexer.python.identifier.attributes", &OptionsPython::identifierAttributes, |
|
"Set to 1 to recognise Python identifier attributes."); |
|
|
|
DefineProperty("lexer.python.decorator.attributes", &OptionsPython::decoratorAttributes, |
|
"Set to 1 to recognise Python decorator attributes."); |
|
|
|
DefineWordListSets(pythonWordListDesc); |
|
} |
|
}; |
|
|
|
const char styleSubable[] = { SCE_P_IDENTIFIER, 0 }; |
|
|
|
const LexicalClass lexicalClasses[] = { |
|
// Lexer Python SCLEX_PYTHON SCE_P_: |
|
0, "SCE_P_DEFAULT", "default", "White space", |
|
1, "SCE_P_COMMENTLINE", "comment line", "Comment", |
|
2, "SCE_P_NUMBER", "literal numeric", "Number", |
|
3, "SCE_P_STRING", "literal string", "String", |
|
4, "SCE_P_CHARACTER", "literal string", "Single quoted string", |
|
5, "SCE_P_WORD", "keyword", "Keyword", |
|
6, "SCE_P_TRIPLE", "literal string", "Triple quotes", |
|
7, "SCE_P_TRIPLEDOUBLE", "literal string", "Triple double quotes", |
|
8, "SCE_P_CLASSNAME", "identifier", "Class name definition", |
|
9, "SCE_P_DEFNAME", "identifier", "Function or method name definition", |
|
10, "SCE_P_OPERATOR", "operator", "Operators", |
|
11, "SCE_P_IDENTIFIER", "identifier", "Identifiers", |
|
12, "SCE_P_COMMENTBLOCK", "comment", "Comment-blocks", |
|
13, "SCE_P_STRINGEOL", "error literal string", "End of line where string is not closed", |
|
14, "SCE_P_WORD2", "identifier", "Highlighted identifiers", |
|
15, "SCE_P_DECORATOR", "preprocessor", "Decorators", |
|
16, "SCE_P_FSTRING", "literal string interpolated", "F-String", |
|
17, "SCE_P_FCHARACTER", "literal string interpolated", "Single quoted f-string", |
|
18, "SCE_P_FTRIPLE", "literal string interpolated", "Triple quoted f-string", |
|
19, "SCE_P_FTRIPLEDOUBLE", "literal string interpolated", "Triple double quoted f-string", |
|
20, "SCE_P_ATTRIBUTE", "identifier", "Attribute of identifier", |
|
}; |
|
|
|
class LexerPython : public DefaultLexer { |
|
WordList keywords; |
|
WordList keywords2; |
|
OptionsPython options; |
|
OptionSetPython osPython; |
|
enum { ssIdentifier }; |
|
SubStyles subStyles{styleSubable}; |
|
std::map<Sci_Position, std::vector<SingleFStringExpState> > ftripleStateAtEol; |
|
public: |
|
explicit LexerPython() : |
|
DefaultLexer("python", SCLEX_PYTHON, lexicalClasses, std::size(lexicalClasses)) { |
|
} |
|
~LexerPython() override = default; |
|
void SCI_METHOD Release() override { |
|
delete this; |
|
} |
|
int SCI_METHOD Version() const override { |
|
return lvRelease5; |
|
} |
|
const char *SCI_METHOD PropertyNames() override { |
|
return osPython.PropertyNames(); |
|
} |
|
int SCI_METHOD PropertyType(const char *name) override { |
|
return osPython.PropertyType(name); |
|
} |
|
const char *SCI_METHOD DescribeProperty(const char *name) override { |
|
return osPython.DescribeProperty(name); |
|
} |
|
Sci_Position SCI_METHOD PropertySet(const char *key, const char *val) override; |
|
const char *SCI_METHOD PropertyGet(const char *key) override { |
|
return osPython.PropertyGet(key); |
|
} |
|
const char *SCI_METHOD DescribeWordListSets() override { |
|
return osPython.DescribeWordListSets(); |
|
} |
|
Sci_Position SCI_METHOD WordListSet(int n, const char *wl) override; |
|
void SCI_METHOD Lex(Sci_PositionU startPos, Sci_Position length, int initStyle, IDocument *pAccess) override; |
|
void SCI_METHOD Fold(Sci_PositionU startPos, Sci_Position length, int initStyle, IDocument *pAccess) override; |
|
|
|
void *SCI_METHOD PrivateCall(int, void *) override { |
|
return nullptr; |
|
} |
|
|
|
int SCI_METHOD LineEndTypesSupported() override { |
|
return SC_LINE_END_TYPE_UNICODE; |
|
} |
|
|
|
int SCI_METHOD AllocateSubStyles(int styleBase, int numberStyles) override { |
|
return subStyles.Allocate(styleBase, numberStyles); |
|
} |
|
int SCI_METHOD SubStylesStart(int styleBase) override { |
|
return subStyles.Start(styleBase); |
|
} |
|
int SCI_METHOD SubStylesLength(int styleBase) override { |
|
return subStyles.Length(styleBase); |
|
} |
|
int SCI_METHOD StyleFromSubStyle(int subStyle) override { |
|
const int styleBase = subStyles.BaseStyle(subStyle); |
|
return styleBase; |
|
} |
|
int SCI_METHOD PrimaryStyleFromStyle(int style) override { |
|
return style; |
|
} |
|
void SCI_METHOD FreeSubStyles() override { |
|
subStyles.Free(); |
|
} |
|
void SCI_METHOD SetIdentifiers(int style, const char *identifiers) override { |
|
subStyles.SetIdentifiers(style, identifiers); |
|
} |
|
int SCI_METHOD DistanceToSecondaryStyles() override { |
|
return 0; |
|
} |
|
const char *SCI_METHOD GetSubStyleBases() override { |
|
return styleSubable; |
|
} |
|
|
|
static ILexer5 *LexerFactoryPython() { |
|
return new LexerPython(); |
|
} |
|
|
|
private: |
|
void ProcessLineEnd(StyleContext &sc, std::vector<SingleFStringExpState> &fstringStateStack, SingleFStringExpState *¤tFStringExp, bool &inContinuedString); |
|
}; |
|
|
|
Sci_Position SCI_METHOD LexerPython::PropertySet(const char *key, const char *val) { |
|
if (osPython.PropertySet(&options, key, val)) { |
|
return 0; |
|
} |
|
return -1; |
|
} |
|
|
|
Sci_Position SCI_METHOD LexerPython::WordListSet(int n, const char *wl) { |
|
WordList *wordListN = nullptr; |
|
switch (n) { |
|
case 0: |
|
wordListN = &keywords; |
|
break; |
|
case 1: |
|
wordListN = &keywords2; |
|
break; |
|
default: |
|
break; |
|
} |
|
Sci_Position firstModification = -1; |
|
if (wordListN) { |
|
if (wordListN->Set(wl)) { |
|
firstModification = 0; |
|
} |
|
} |
|
return firstModification; |
|
} |
|
|
|
void LexerPython::ProcessLineEnd(StyleContext &sc, std::vector<SingleFStringExpState> &fstringStateStack, SingleFStringExpState *¤tFStringExp, bool &inContinuedString) { |
|
// Before pep 701 single quote f-string's could not continue to a 2nd+ line |
|
// Post pep 701, they can continue both with a trailing \ and if a { field is |
|
// not ended with a } |
|
if (!options.pep701StringsF) { |
|
long deepestSingleStateIndex = -1; |
|
unsigned long i; |
|
|
|
// Find the deepest single quote state because that string will end; no \ continuation in f-string |
|
for (i = 0; i < fstringStateStack.size(); i++) { |
|
if (IsPySingleQuoteStringState(fstringStateStack[i].state)) { |
|
deepestSingleStateIndex = i; |
|
break; |
|
} |
|
} |
|
|
|
if (deepestSingleStateIndex != -1) { |
|
sc.SetState(fstringStateStack[deepestSingleStateIndex].state); |
|
while (fstringStateStack.size() > static_cast<unsigned long>(deepestSingleStateIndex)) { |
|
PopFromStateStack(fstringStateStack, currentFStringExp); |
|
} |
|
} |
|
} |
|
|
|
if (!fstringStateStack.empty()) { |
|
std::pair<Sci_Position, std::vector<SingleFStringExpState> > val; |
|
val.first = sc.currentLine; |
|
val.second = fstringStateStack; |
|
|
|
ftripleStateAtEol.insert(val); |
|
} |
|
|
|
if ((sc.state == SCE_P_DEFAULT) |
|
|| IsPyTripleQuoteStringState(sc.state)) { |
|
// 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.SetState(sc.state); |
|
} |
|
if (IsPySingleQuoteStringState(sc.state)) { |
|
if (inContinuedString || options.stringsOverNewline) { |
|
inContinuedString = false; |
|
} else { |
|
sc.ChangeState(SCE_P_STRINGEOL); |
|
sc.ForwardSetState(SCE_P_DEFAULT); |
|
} |
|
} |
|
} |
|
|
|
void SCI_METHOD LexerPython::Lex(Sci_PositionU startPos, Sci_Position length, int initStyle, IDocument *pAccess) { |
|
Accessor styler(pAccess, nullptr); |
|
|
|
// Track whether in f-string expression; vector is used for a stack to |
|
// handle nested f-strings such as f"""{f'''{f"{f'{1}'}"}'''}""" |
|
std::vector<SingleFStringExpState> fstringStateStack; |
|
SingleFStringExpState *currentFStringExp = nullptr; |
|
|
|
const Sci_Position endPos = startPos + length; |
|
|
|
// Backtrack to previous line in case need to fix its tab whinging |
|
Sci_Position lineCurrent = styler.GetLine(startPos); |
|
if (startPos > 0) { |
|
if (lineCurrent > 0) { |
|
lineCurrent--; |
|
// Look for backslash-continued lines |
|
while (lineCurrent > 0) { |
|
const Sci_Position eolPos = styler.LineStart(lineCurrent) - 1; |
|
const int eolStyle = styler.StyleIndexAt(eolPos); |
|
if (AnyOf(eolStyle, SCE_P_STRING, SCE_P_CHARACTER, SCE_P_STRINGEOL)) { |
|
lineCurrent -= 1; |
|
} else { |
|
break; |
|
} |
|
} |
|
startPos = styler.LineStart(lineCurrent); |
|
} |
|
initStyle = startPos == 0 ? SCE_P_DEFAULT : styler.StyleIndexAt(startPos - 1); |
|
} |
|
|
|
const literalsAllowed allowedLiterals = options.AllowedLiterals(); |
|
|
|
if (initStyle == SCE_P_STRINGEOL) { |
|
initStyle = SCE_P_DEFAULT; |
|
} |
|
|
|
// Set up fstate stack from last line and remove any subsequent ftriple at eol states |
|
std::map<Sci_Position, std::vector<SingleFStringExpState> >::iterator it; |
|
it = ftripleStateAtEol.find(lineCurrent - 1); |
|
if (it != ftripleStateAtEol.end() && !it->second.empty()) { |
|
fstringStateStack = it->second; |
|
currentFStringExp = &fstringStateStack.back(); |
|
} |
|
it = ftripleStateAtEol.lower_bound(lineCurrent); |
|
if (it != ftripleStateAtEol.end()) { |
|
ftripleStateAtEol.erase(it, ftripleStateAtEol.end()); |
|
} |
|
|
|
kwType kwLast = kwOther; |
|
int spaceFlags = 0; |
|
styler.IndentAmount(lineCurrent, &spaceFlags, IsPyComment); |
|
bool base_n_number = false; |
|
|
|
const WordClassifier &classifierIdentifiers = subStyles.Classifier(SCE_P_IDENTIFIER); |
|
|
|
StyleContext sc(startPos, endPos - startPos, initStyle, styler); |
|
|
|
bool indentGood = true; |
|
Sci_Position startIndicator = sc.currentPos; |
|
bool inContinuedString = false; |
|
|
|
for (; sc.More(); sc.Forward()) { |
|
|
|
if (sc.atLineStart) { |
|
styler.IndentAmount(lineCurrent, &spaceFlags, IsPyComment); |
|
indentGood = true; |
|
if (options.whingeLevel == 1) { |
|
indentGood = (spaceFlags & wsInconsistent) == 0; |
|
} else if (options.whingeLevel == 2) { |
|
indentGood = (spaceFlags & wsSpaceTab) == 0; |
|
} else if (options.whingeLevel == 3) { |
|
indentGood = (spaceFlags & wsSpace) == 0; |
|
} else if (options.whingeLevel == 4) { |
|
indentGood = (spaceFlags & wsTab) == 0; |
|
} |
|
if (!indentGood) { |
|
styler.IndicatorFill(startIndicator, sc.currentPos, indicatorWhitespace, 0); |
|
startIndicator = sc.currentPos; |
|
} |
|
} |
|
|
|
if (sc.atLineEnd) { |
|
ProcessLineEnd(sc, fstringStateStack, currentFStringExp, inContinuedString); |
|
lineCurrent++; |
|
if (!sc.More()) |
|
break; |
|
} |
|
|
|
bool needEOLCheck = false; |
|
|
|
|
|
if (sc.state == SCE_P_OPERATOR) { |
|
kwLast = kwOther; |
|
sc.SetState(SCE_P_DEFAULT); |
|
} else if (sc.state == SCE_P_NUMBER) { |
|
if (!IsAWordChar(sc.ch, false) && |
|
!(!base_n_number && ((sc.ch == '+' || sc.ch == '-') && (sc.chPrev == 'e' || sc.chPrev == 'E')))) { |
|
sc.SetState(SCE_P_DEFAULT); |
|
} |
|
} else if (sc.state == SCE_P_IDENTIFIER) { |
|
if ((sc.ch == '.') || (!IsAWordChar(sc.ch, options.unicodeIdentifiers))) { |
|
char identifier[100]; |
|
sc.GetCurrent(identifier, sizeof(identifier)); |
|
int style = SCE_P_IDENTIFIER; |
|
if ((kwLast == kwImport) && (strcmp(identifier, "as") == 0)) { |
|
style = SCE_P_WORD; |
|
} else if (keywords.InList(identifier) && !IsMatchOrCaseIdentifier(sc, styler, identifier)) { |
|
style = SCE_P_WORD; |
|
} else if (kwLast == kwClass) { |
|
style = SCE_P_CLASSNAME; |
|
} else if (kwLast == kwDef) { |
|
style = SCE_P_DEFNAME; |
|
} else if (kwLast == kwCDef || kwLast == kwCPDef) { |
|
Sci_Position pos = sc.currentPos; |
|
unsigned char ch = styler.SafeGetCharAt(pos, '\0'); |
|
while (ch != '\0') { |
|
if (ch == '(') { |
|
style = SCE_P_DEFNAME; |
|
break; |
|
} else if (ch == ':') { |
|
style = SCE_P_CLASSNAME; |
|
break; |
|
} else if (ch == ' ' || ch == '\t' || ch == '\n' || ch == '\r') { |
|
pos++; |
|
ch = styler.SafeGetCharAt(pos, '\0'); |
|
} else { |
|
break; |
|
} |
|
} |
|
} else if (keywords2.InList(identifier)) { |
|
if (options.keywords2NoSubIdentifiers) { |
|
// We don't want to highlight keywords2 |
|
// that are used as a sub-identifier, |
|
// i.e. not open in "foo.open". |
|
const Sci_Position pos = styler.GetStartSegment() - 1; |
|
if (pos < 0 || (styler.SafeGetCharAt(pos, '\0') != '.')) |
|
style = SCE_P_WORD2; |
|
} else { |
|
style = SCE_P_WORD2; |
|
} |
|
} else { |
|
const int subStyle = classifierIdentifiers.ValueFor(identifier); |
|
if (subStyle >= 0) { |
|
style = subStyle; |
|
} |
|
if (options.identifierAttributes > 0 || options.decoratorAttributes > 0) { |
|
// Does the user even want attributes styled? |
|
Sci_Position pos = styler.GetStartSegment() - 1; |
|
unsigned char ch = styler.SafeGetCharAt(pos, '\0'); |
|
while (ch != '\0' && (ch == '.' || ch == ' ' || ch == '\\' || ch == '\t' || ch == '\n' || ch == '\r')) { |
|
// Backwards search for a . while only allowing certain valid characters |
|
if (IsAWordChar(ch, options.unicodeIdentifiers)) { |
|
break; |
|
} |
|
pos--; |
|
ch = styler.SafeGetCharAt(pos, '\0'); |
|
} |
|
if (pos < 0 || ch == '.') { |
|
// Is this an attribute we could style? if it is, do as asked |
|
bool isComment = false; |
|
bool isDecoratorAttribute = false; |
|
const Sci_Position attrLine = styler.GetLine(pos); |
|
for (Sci_Position i = styler.LineStart(attrLine); i < pos; i++) { |
|
const char attrCh = styler[i]; |
|
if (attrCh == '@') |
|
isDecoratorAttribute = true; |
|
if (attrCh == '#') |
|
isComment = true; |
|
// Detect if this attribute belongs to a decorator |
|
if (!IsASpaceOrTab(ch)) |
|
break; |
|
} |
|
if (((isDecoratorAttribute) && (!isComment)) && (((options.decoratorAttributes == 1) && (style == SCE_P_IDENTIFIER)) || (options.decoratorAttributes == 2))) { |
|
// Style decorator attributes as decorators but respect already styled identifiers (unless requested to ignore already styled identifiers) |
|
style = SCE_P_DECORATOR; |
|
} |
|
if (((!isDecoratorAttribute) && (!isComment)) && (((options.identifierAttributes == 1) && (style == SCE_P_IDENTIFIER)) || (options.identifierAttributes == 2))) { |
|
// Style attributes and ignore decorator attributes but respect already styled identifiers (unless requested to ignore already styled identifiers) |
|
style = SCE_P_ATTRIBUTE; |
|
} |
|
} |
|
} |
|
} |
|
sc.ChangeState(style); |
|
sc.SetState(SCE_P_DEFAULT); |
|
if (style == SCE_P_WORD) { |
|
if (0 == strcmp(identifier, "class")) |
|
kwLast = kwClass; |
|
else if (0 == strcmp(identifier, "def")) |
|
kwLast = kwDef; |
|
else if (0 == strcmp(identifier, "import")) |
|
kwLast = kwImport; |
|
else if (0 == strcmp(identifier, "cdef")) |
|
kwLast = kwCDef; |
|
else if (0 == strcmp(identifier, "cpdef")) |
|
kwLast = kwCPDef; |
|
else if (0 == strcmp(identifier, "cimport")) |
|
kwLast = kwImport; |
|
else if (kwLast != kwCDef && kwLast != kwCPDef) |
|
kwLast = kwOther; |
|
} else if (kwLast != kwCDef && kwLast != kwCPDef) { |
|
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 (!IsAWordStart(sc.ch, options.unicodeIdentifiers)) { |
|
sc.SetState(SCE_P_DEFAULT); |
|
} |
|
} else if (IsPySingleQuoteStringState(sc.state)) { |
|
if (sc.ch == '\\') { |
|
if ((sc.chNext == '\r') && (sc.GetRelative(2) == '\n')) { |
|
sc.Forward(); |
|
} |
|
if (sc.chNext == '\n' || sc.chNext == '\r') { |
|
inContinuedString = true; |
|
} else { |
|
// Don't roll over the newline. |
|
sc.Forward(); |
|
} |
|
} else if (sc.ch == GetPyStringQuoteChar(sc.state)) { |
|
sc.ForwardSetState(SCE_P_DEFAULT); |
|
needEOLCheck = true; |
|
} |
|
} else if ((sc.state == SCE_P_TRIPLE) || (sc.state == SCE_P_FTRIPLE)) { |
|
if (sc.ch == '\\') { |
|
sc.Forward(); |
|
} else if (sc.Match(R"(''')")) { |
|
sc.Forward(); |
|
sc.Forward(); |
|
sc.ForwardSetState(SCE_P_DEFAULT); |
|
needEOLCheck = true; |
|
} |
|
} else if ((sc.state == SCE_P_TRIPLEDOUBLE) || (sc.state == SCE_P_FTRIPLEDOUBLE)) { |
|
if (sc.ch == '\\') { |
|
sc.Forward(); |
|
} else if (sc.Match(R"(""")")) { |
|
sc.Forward(); |
|
sc.Forward(); |
|
sc.ForwardSetState(SCE_P_DEFAULT); |
|
needEOLCheck = true; |
|
} |
|
} |
|
|
|
// Note if used and not if else because string states also match |
|
// some of the above clauses |
|
if (IsPyFStringState(sc.state) && sc.ch == '{') { |
|
if (sc.chNext == '{') { |
|
sc.Forward(); |
|
} else { |
|
PushStateToStack(sc.state, fstringStateStack, currentFStringExp); |
|
sc.ForwardSetState(SCE_P_DEFAULT); |
|
} |
|
needEOLCheck = true; |
|
} |
|
|
|
// If in an f-string expression and pre pep 701 lexing is used, |
|
// check for the ending quote(s) and end f-string to handle |
|
// syntactically incorrect cases like f'{' and f"""{""". Post |
|
// pep 701, a quote may appear in a { } field so cases like |
|
// f"n = {":".join(seq)}" is valid |
|
if (!options.pep701StringsF && !fstringStateStack.empty() && IsQuote(sc.ch)) { |
|
long matching_stack_i = -1; |
|
for (unsigned long stack_i = 0; stack_i < fstringStateStack.size() && matching_stack_i == -1; stack_i++) { |
|
const int stack_state = fstringStateStack[stack_i].state; |
|
const char quote = GetPyStringQuoteChar(stack_state); |
|
if (sc.ch == quote) { |
|
if (IsPySingleQuoteStringState(stack_state)) { |
|
matching_stack_i = stack_i; |
|
} else if (quote == '"' ? sc.Match(R"(""")") : sc.Match("'''")) { |
|
matching_stack_i = stack_i; |
|
} |
|
} |
|
} |
|
|
|
if (matching_stack_i != -1) { |
|
sc.SetState(fstringStateStack[matching_stack_i].state); |
|
if (IsPyTripleQuoteStringState(fstringStateStack[matching_stack_i].state)) { |
|
sc.Forward(); |
|
sc.Forward(); |
|
} |
|
sc.ForwardSetState(SCE_P_DEFAULT); |
|
needEOLCheck = true; |
|
|
|
while (fstringStateStack.size() > static_cast<unsigned long>(matching_stack_i)) { |
|
PopFromStateStack(fstringStateStack, currentFStringExp); |
|
} |
|
} |
|
} |
|
// End of code to find the end of a state |
|
|
|
if (!indentGood && !IsASpaceOrTab(sc.ch)) { |
|
styler.IndicatorFill(startIndicator, sc.currentPos, indicatorWhitespace, 1); |
|
startIndicator = sc.currentPos; |
|
indentGood = true; |
|
} |
|
|
|
// One cdef or cpdef line, clear kwLast only at end of line |
|
if ((kwLast == kwCDef || kwLast == kwCPDef) && sc.atLineEnd) { |
|
kwLast = kwOther; |
|
} |
|
|
|
// State exit code may have moved on to end of line |
|
if (needEOLCheck && sc.atLineEnd) { |
|
ProcessLineEnd(sc, fstringStateStack, currentFStringExp, inContinuedString); |
|
lineCurrent++; |
|
styler.IndentAmount(lineCurrent, &spaceFlags, IsPyComment); |
|
if (!sc.More()) |
|
break; |
|
} |
|
|
|
// If in f-string expression, check for }, :, ! to resume f-string state or update nesting count |
|
if (currentFStringExp && !IsPySingleQuoteStringState(sc.state) && !IsPyTripleQuoteStringState(sc.state)) { |
|
if (currentFStringExp->nestingCount == 0 && (sc.ch == '}' || sc.ch == ':' || (sc.ch == '!' && sc.chNext != '='))) { |
|
sc.SetState(PopFromStateStack(fstringStateStack, currentFStringExp)); |
|
} else { |
|
if (sc.ch == '{' || sc.ch == '[' || sc.ch == '(') { |
|
currentFStringExp->nestingCount++; |
|
} else if (sc.ch == '}' || sc.ch == ']' || sc.ch == ')') { |
|
currentFStringExp->nestingCount--; |
|
} |
|
} |
|
} |
|
|
|
// Check for a new state starting character |
|
if (sc.state == SCE_P_DEFAULT) { |
|
if (IsADigit(sc.ch) || (sc.ch == '.' && IsADigit(sc.chNext))) { |
|
if (sc.ch == '0' && (sc.chNext == 'x' || sc.chNext == 'X')) { |
|
base_n_number = true; |
|
sc.SetState(SCE_P_NUMBER); |
|
} else if (sc.ch == '0' && |
|
(sc.chNext == 'o' || sc.chNext == 'O' || sc.chNext == 'b' || sc.chNext == 'B')) { |
|
if (options.base2or8Literals) { |
|
base_n_number = true; |
|
sc.SetState(SCE_P_NUMBER); |
|
} else { |
|
sc.SetState(SCE_P_NUMBER); |
|
sc.ForwardSetState(SCE_P_IDENTIFIER); |
|
} |
|
} else { |
|
base_n_number = false; |
|
sc.SetState(SCE_P_NUMBER); |
|
} |
|
} else if (isoperator(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 (sc.ch == '@') { |
|
if (IsFirstNonWhitespace(sc.currentPos, styler)) |
|
sc.SetState(SCE_P_DECORATOR); |
|
else |
|
sc.SetState(SCE_P_OPERATOR); |
|
} else if (IsPyStringStart(sc.ch, sc.chNext, sc.GetRelative(2), allowedLiterals)) { |
|
Sci_PositionU nextIndex = 0; |
|
sc.SetState(GetPyStringState(styler, sc.currentPos, &nextIndex, allowedLiterals)); |
|
while (nextIndex > (sc.currentPos + 1) && sc.More()) { |
|
sc.Forward(); |
|
} |
|
} else if (IsAWordStart(sc.ch, options.unicodeIdentifiers)) { |
|
sc.SetState(SCE_P_IDENTIFIER); |
|
} |
|
} |
|
} |
|
styler.IndicatorFill(startIndicator, sc.currentPos, indicatorWhitespace, 0); |
|
sc.Complete(); |
|
} |
|
|
|
bool IsCommentLine(Sci_Position line, Accessor &styler) { |
|
const Sci_Position pos = styler.LineStart(line); |
|
const Sci_Position eol_pos = styler.LineStart(line + 1) - 1; |
|
for (Sci_Position i = pos; i < eol_pos; i++) { |
|
const char ch = styler[i]; |
|
if (ch == '#') |
|
return true; |
|
if (!IsASpaceOrTab(ch)) |
|
return false; |
|
} |
|
return false; |
|
} |
|
|
|
bool IsQuoteLine(Sci_Position line, const Accessor &styler) { |
|
const int style = styler.StyleIndexAt(styler.LineStart(line)); |
|
return IsPyTripleQuoteStringState(style); |
|
} |
|
|
|
|
|
void SCI_METHOD LexerPython::Fold(Sci_PositionU startPos, Sci_Position length, int /*initStyle - unused*/, IDocument *pAccess) { |
|
if (!options.fold) |
|
return; |
|
|
|
Accessor styler(pAccess, nullptr); |
|
|
|
const Sci_Position maxPos = startPos + length; |
|
const Sci_Position maxLines = (maxPos == styler.Length()) ? styler.GetLine(maxPos) : styler.GetLine(maxPos - 1); // Requested last line |
|
const Sci_Position docLines = styler.GetLine(styler.Length()); // Available last line |
|
|
|
// 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 |
|
// at least one line in all cases) |
|
int spaceFlags = 0; |
|
Sci_Position lineCurrent = styler.GetLine(startPos); |
|
int indentCurrent = styler.IndentAmount(lineCurrent, &spaceFlags, nullptr); |
|
while (lineCurrent > 0) { |
|
lineCurrent--; |
|
indentCurrent = styler.IndentAmount(lineCurrent, &spaceFlags, nullptr); |
|
if (!(indentCurrent & SC_FOLDLEVELWHITEFLAG) && |
|
(!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; |
|
if (lineCurrent >= 1) |
|
prev_state = styler.StyleIndexAt(startPos - 1); |
|
int prevQuote = options.foldQuotes && IsPyTripleQuoteStringState(prev_state); |
|
|
|
// Process all characters to end of requested range or end of any triple quote |
|
//that hangs over the end of the range. Cap processing in all cases |
|
// to end of document (in case of unclosed quote at end). |
|
while ((lineCurrent <= docLines) && ((lineCurrent <= maxLines) || prevQuote)) { |
|
|
|
// Gather info |
|
int lev = indentCurrent; |
|
Sci_Position lineNext = lineCurrent + 1; |
|
int indentNext = indentCurrent; |
|
int quote = false; |
|
if (lineNext <= docLines) { |
|
// Information about next line is only available if not at end of document |
|
indentNext = styler.IndentAmount(lineNext, &spaceFlags, nullptr); |
|
const Sci_Position lookAtPos = (styler.LineStart(lineNext) == styler.Length()) ? styler.Length() - 1 : styler.LineStart(lineNext); |
|
const int style = styler.StyleIndexAt(lookAtPos); |
|
quote = options.foldQuotes && IsPyTripleQuoteStringState(style); |
|
} |
|
const bool quote_start = (quote && !prevQuote); |
|
const bool quote_continue = (quote && prevQuote); |
|
if (!quote || !prevQuote) |
|
indentCurrentLevel = indentCurrent & SC_FOLDLEVELNUMBERMASK; |
|
if (quote) |
|
indentNext = indentCurrentLevel; |
|
if (indentNext & SC_FOLDLEVELWHITEFLAG) |
|
indentNext = SC_FOLDLEVELWHITEFLAG | indentCurrentLevel; |
|
|
|
if (quote_start) { |
|
// Place fold point at start of triple quoted string |
|
lev |= SC_FOLDLEVELHEADERFLAG; |
|
} else if (quote_continue || prevQuote) { |
|
// Add level to rest of lines in the string |
|
lev = lev + 1; |
|
} |
|
|
|
// 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. If comments end file, use the min |
|
// comment indent as the level after |
|
|
|
int minCommentLevel = indentCurrentLevel; |
|
while (!quote && |
|
(lineNext < docLines) && |
|
((indentNext & SC_FOLDLEVELWHITEFLAG) || (IsCommentLine(lineNext, styler)))) { |
|
|
|
if (IsCommentLine(lineNext, styler) && indentNext < minCommentLevel) { |
|
minCommentLevel = indentNext; |
|
} |
|
|
|
lineNext++; |
|
indentNext = styler.IndentAmount(lineNext, &spaceFlags, nullptr); |
|
} |
|
|
|
const int levelAfterComments = ((lineNext < docLines) ? indentNext & SC_FOLDLEVELNUMBERMASK : minCommentLevel); |
|
const int levelBeforeComments = std::max(indentCurrentLevel, levelAfterComments); |
|
|
|
// Now set all the indent levels on the lines we skipped |
|
// 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 |
|
|
|
Sci_Position skipLine = lineNext; |
|
int skipLevel = levelAfterComments; |
|
|
|
while (--skipLine > lineCurrent) { |
|
const int skipLineIndent = styler.IndentAmount(skipLine, &spaceFlags, nullptr); |
|
|
|
if (options.foldCompact) { |
|
if ((skipLineIndent & SC_FOLDLEVELNUMBERMASK) > levelAfterComments) |
|
skipLevel = levelBeforeComments; |
|
|
|
const int whiteFlag = skipLineIndent & SC_FOLDLEVELWHITEFLAG; |
|
|
|
styler.SetLevel(skipLine, skipLevel | whiteFlag); |
|
} else { |
|
if ((skipLineIndent & SC_FOLDLEVELNUMBERMASK) > levelAfterComments && |
|
!(skipLineIndent & SC_FOLDLEVELWHITEFLAG) && |
|
!IsCommentLine(skipLine, styler)) |
|
skipLevel = levelBeforeComments; |
|
|
|
styler.SetLevel(skipLine, skipLevel); |
|
} |
|
} |
|
|
|
// Set fold header on non-quote line |
|
if (!quote && !(indentCurrent & SC_FOLDLEVELWHITEFLAG)) { |
|
if ((indentCurrent & SC_FOLDLEVELNUMBERMASK) < (indentNext & SC_FOLDLEVELNUMBERMASK)) |
|
lev |= SC_FOLDLEVELHEADERFLAG; |
|
} |
|
|
|
// Keep track of triple quote state of previous line |
|
prevQuote = quote; |
|
|
|
// Set fold level for this line and move to next line |
|
styler.SetLevel(lineCurrent, options.foldCompact ? lev : lev & ~SC_FOLDLEVELWHITEFLAG); |
|
indentCurrent = indentNext; |
|
lineCurrent = lineNext; |
|
} |
|
|
|
// NOTE: Cannot set level of last line here because indentCurrent doesn't have |
|
// header flag set; the loop above is crafted to take care of this case! |
|
//styler.SetLevel(lineCurrent, indentCurrent); |
|
} |
|
|
|
} |
|
|
|
extern const LexerModule lmPython(SCLEX_PYTHON, LexerPython::LexerFactoryPython, "python", |
|
pythonWordListDesc);
|
|
|