Skip to content

fix: PCRE2 interface - bug matching line-end ($) #5571

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
RaiKoHoff merged 1 commit into from
Mar 6, 2026
Merged

fix: PCRE2 interface - bug matching line-end ($) #5571

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
65 changes: 36 additions & 29 deletions scintilla/pcre2/scintilla/PCRE2RegExEngine.cxx
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -63,7 +63,7 @@ using namespace Scintilla::Internal;
// *** PCRE2 configuration ***
// ============================================================================

enum class EOLmode : int { UDEF = -1, CRLF = SC_EOL_CRLF, CR = SC_EOL_CR, LF = SC_EOL_LF };


// ============================================================================
// ============================================================================
Expand All @@ -76,14 +76,17 @@ class PCRE2RegExEngine : public RegexSearchBase
: m_CompileOptions(PCRE2_UTF | PCRE2_UCP | PCRE2_MULTILINE)
, m_CompiledPattern(nullptr)
, m_MatchData(nullptr)
, m_CompileContext(nullptr)
, m_MatchContext(nullptr)
, m_EOLmode(EOLmode::UDEF)
, m_RangeBeg(-1)
, m_RangeEnd(-1)
, m_ErrorInfo()
, m_MatchPos(-1)
, m_MatchLen(0)
{
m_CompileContext = pcre2_compile_context_create(nullptr);
pcre2_set_newline(m_CompileContext, PCRE2_NEWLINE_ANYCRLF);

m_MatchContext = pcre2_match_context_create(nullptr);
// Set match limits to prevent catastrophic backtracking
pcre2_set_match_limit(m_MatchContext, 10000000);
Expand All @@ -97,6 +100,10 @@ class PCRE2RegExEngine : public RegexSearchBase
pcre2_match_context_free(m_MatchContext);
m_MatchContext = nullptr;
}
if (m_CompileContext) {
pcre2_compile_context_free(m_CompileContext);
m_CompileContext = nullptr;
}
}

Sci::Position FindText(Document* doc, Sci::Position minPos, Sci::Position maxPos, const char* pattern,
Expand All @@ -108,27 +115,27 @@ class PCRE2RegExEngine : public RegexSearchBase

void clear();

std::string translateRegExpr(const std::string & regExprStr, bool wholeWord, bool wordStart, EndOfLine eolMode);
std::string translateRegExpr(const std::string & regExprStr, bool wholeWord, bool wordStart);

std::string convertReplExpr(const std::string & replStr);

private:

std::string m_RegExprStrg;

uint32_t m_CompileOptions;
pcre2_code* m_CompiledPattern;
pcre2_match_data* m_MatchData;
pcre2_match_context* m_MatchContext;
EOLmode m_EOLmode;
uint32_t m_CompileOptions;
pcre2_code* m_CompiledPattern;
pcre2_match_data* m_MatchData;
pcre2_compile_context* m_CompileContext;
pcre2_match_context* m_MatchContext;

Sci::Position m_RangeBeg;
Sci::Position m_RangeEnd;
Sci::Position m_RangeBeg;
Sci::Position m_RangeEnd;

char m_ErrorInfo[256];
char m_ErrorInfo[256];

Sci::Position m_MatchPos;
Sci::Position m_MatchLen;
Sci::Position m_MatchPos;
Sci::Position m_MatchLen;

public:
std::string m_SubstBuffer;
Expand Down Expand Up @@ -252,10 +259,9 @@ Sci::Position PCRE2RegExEngine::FindText(Document* doc, Sci::Position minPos, Sc
Sci::Position const rangeEnd = (findForward) ? maxPos : minPos;
//Sci::Position const rangeLen = (rangeEnd - rangeBeg);

EOLmode const eolMode = static_cast<EOLmode>(doc->eolMode);

// --- Build compile options ---
// PCRE2_MULTILINE: ^/$ match at line boundaries (Oniguruma's default behavior)
// PCRE2_MULTILINE: ^/$ match at line boundaries
// Newline convention (ANYCRLF) is set on m_CompileContext, not here
uint32_t compileOptions = PCRE2_UTF | PCRE2_UCP | PCRE2_MULTILINE;

if (!caseSensitive) {
Expand All @@ -265,18 +271,17 @@ Sci::Position PCRE2RegExEngine::FindText(Document* doc, Sci::Position minPos, Sc
compileOptions |= PCRE2_DOTALL; // Note: Oniguruma called this MULTILINE
}

std::string const sRegExprStrg = translateRegExpr(pattern, word, wordStart, doc->eolMode);
std::string const sRegExprStrg = translateRegExpr(pattern, word, wordStart);

bool const bReCompile = (m_CompiledPattern == nullptr) || (m_CompileOptions != compileOptions)
|| (m_RegExprStrg.compare(sRegExprStrg) != 0) || (m_EOLmode != eolMode);
|| (m_RegExprStrg.compare(sRegExprStrg) != 0);

if (bReCompile) {
clear();
m_RegExprStrg = sRegExprStrg;
m_CompileOptions = compileOptions;
m_RangeBeg = rangeBeg;
m_RangeEnd = rangeEnd;
m_EOLmode = eolMode;
m_ErrorInfo[0] = '\0';

try {
Expand All @@ -288,7 +293,7 @@ Sci::Position PCRE2RegExEngine::FindText(Document* doc, Sci::Position minPos, Sc
m_CompileOptions,
&errorcode,
&erroroffset,
nullptr // default compile context
m_CompileContext // newline convention set to ANYCRLF
);

if (!m_CompiledPattern) {
Expand Down Expand Up @@ -555,11 +560,8 @@ const char* PCRE2RegExEngine::SubstituteByPosition(Document* doc, const char* te
//
// private methods

std::string PCRE2RegExEngine::translateRegExpr(const std::string & regExprStr, bool wholeWord, bool wordStart,
EndOfLine eolMode)
std::string PCRE2RegExEngine::translateRegExpr(const std::string & regExprStr, bool wholeWord, bool wordStart)
{
UNREFERENCED_PARAMETER(eolMode);

std::string transRegExpr;

if (wholeWord || wordStart) { // push '\b' at the begin of regexpr
Expand Down Expand Up @@ -755,8 +757,11 @@ class SimplePCRE2Engine
: m_CompileOptions(0)
, m_CompiledPattern(nullptr)
, m_MatchData(nullptr)
, m_CompileContext(nullptr)
, m_ErrorInfo()
{
m_CompileContext = pcre2_compile_context_create(nullptr);
pcre2_set_newline(m_CompileContext, PCRE2_NEWLINE_ANYCRLF);
}

~SimplePCRE2Engine() noexcept
Expand All @@ -767,6 +772,9 @@ class SimplePCRE2Engine
if (m_CompiledPattern) {
pcre2_code_free(m_CompiledPattern);
}
if (m_CompileContext) {
pcre2_compile_context_free(m_CompileContext);
}
}

// non-copyable
Expand All @@ -781,6 +789,7 @@ class SimplePCRE2Engine
uint32_t m_CompileOptions;
pcre2_code* m_CompiledPattern;
pcre2_match_data* m_MatchData;
pcre2_compile_context* m_CompileContext;

char m_ErrorInfo[256];

Expand Down Expand Up @@ -832,7 +841,7 @@ ptrdiff_t SimplePCRE2Engine::Find(const char* pattern, const char* document, con
auto const patternLen = strlen(pattern);
auto const stringLen = strlen(document);

// Build compile options
// Build compile options (newline convention set on m_CompileContext)
uint32_t compileOptions = PCRE2_UTF | PCRE2_UCP | PCRE2_MULTILINE;
if (!caseSensitive) {
compileOptions |= PCRE2_CASELESS;
Expand Down Expand Up @@ -868,7 +877,7 @@ ptrdiff_t SimplePCRE2Engine::Find(const char* pattern, const char* document, con
m_CompileOptions,
&errorcode,
&erroroffset,
nullptr
m_CompileContext
);

if (!m_CompiledPattern) {
Expand Down Expand Up @@ -930,9 +939,7 @@ extern "C"
#ifdef SCINTILLA_DLL
__declspec(dllexport)
#endif
ptrdiff_t WINAPI RegExFind(const char *pchPattern, const char *pchText, const bool caseSensitive, const int eolMode, int *matchLen_out) {

UNREFERENCED_PARAMETER(eolMode);
ptrdiff_t WINAPI RegExFind(const char *pchPattern, const char *pchText, const bool caseSensitive, int *matchLen_out) {

// Static cached engine: pattern is compiled once, reused across calls.
// Only recompiles when pattern or options change.
Expand Down
3 changes: 2 additions & 1 deletion src/Edit.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2157,7 +2157,7 @@ void EditURLDecode(const bool isPathConvert)

// can URL be found by Hyperlink pattern matching ?
int matchLen = 0;
ptrdiff_t const pos = RegExFind(s_pUrlRegExA, pszUnescaped, false, SciCall_GetEOLMode(), &matchLen);
ptrdiff_t const pos = RegExFind(s_pUrlRegExA, pszUnescaped, false, &matchLen);
bool const bIsValidConversion = isPathConvert ? ((pos >= 0) && (cchUnescapedDec == matchLen)) : true;

if (bIsValidConversion) {
Expand Down Expand Up @@ -3117,6 +3117,7 @@ void EditCutLines(HWND hwnd, const bool bMSBehavSelEmpty)
bool const bIsLineEmpty = Sci_GetNetLineLength(Sci_GetCurrentLineNumber()) == 0;
UndoTransActionBegin();
if (SciCall_IsSelectionEmpty() && bMSBehavSelEmpty) {
//? SciCall_CutAllowLine(); - does it the same as CopyAllowLine() + LineDelete()?
SciCall_CopyAllowLine(); // (!) VisualStudio behavior
// On Windows, an extra "MSDEVLineSelect" marker is added to the clipboard
// which is then used in SCI_PASTE to paste the whole line before the current line.
Expand Down
4 changes: 2 additions & 2 deletions src/SciCall.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -104,8 +104,8 @@ LRESULT WINAPI Scintilla_DirectStatusFunction(HANDLE, UINT, WPARAM, LPARAM, LPIN
//=============================================================================

// PCRE2 RegEx search (exported from PCRE2RegExEngine.cxx)
ptrdiff_t WINAPI RegExFind(const char* pchPattern, const char* pchText,
const bool caseSensitive, const int eolMode, int *matchLen_out);
ptrdiff_t WINAPI RegExFind(const char* pchPattern, const char* pchText,
const bool caseSensitive, int *matchLen_out);

//=============================================================================

Expand Down
113 changes: 110 additions & 3 deletions src/Styles.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -227,6 +227,36 @@ typedef enum {
FW_IDX_ULTRADARK
} FW_IDX;

// Font Stretch
typedef struct _fntstrtch {
LPCWSTR const wname;
int const stretch;
} FONTSTRETCH_T;

static const FONTSTRETCH_T FontStretches[9] = {
{ L"ultracondensed", SC_STRETCH_ULTRA_CONDENSED }, // 0
{ L"extracondensed", SC_STRETCH_EXTRA_CONDENSED }, // 1
{ L"condensed", SC_STRETCH_CONDENSED }, // 2
{ L"semicondensed", SC_STRETCH_SEMI_CONDENSED }, // 3
{ L"normal", SC_STRETCH_NORMAL }, // 4 (default)
{ L"semiexpanded", SC_STRETCH_SEMI_EXPANDED }, // 5
{ L"expanded", SC_STRETCH_EXPANDED }, // 6
{ L"extraexpanded", SC_STRETCH_EXTRA_EXPANDED }, // 7
{ L"ultraexpanded", SC_STRETCH_ULTRA_EXPANDED }, // 8
};

typedef enum {
FS_IDX_ULTRACONDENSED = 0,
FS_IDX_EXTRACONDENSED,
FS_IDX_CONDENSED,
FS_IDX_SEMICONDENSED,
FS_IDX_NORMAL,
FS_IDX_SEMIEXPANDED,
FS_IDX_EXPANDED,
FS_IDX_EXTRAEXPANDED,
FS_IDX_ULTRAEXPANDED,
} FONTSTRETCH_IDX;

//// font quality
//#define Style_StrHasAttrNone(lpszStyle) Style_StrHasAttribute((lpszStyle), L"none")
//#define Style_StrHasAttrStdType(lpszStyle) Style_StrHasAttribute((lpszStyle), L"standard")
Expand Down Expand Up @@ -2514,7 +2544,7 @@ PEDITLEXER Style_RegExMatchLexer(LPCWSTR lpszFileName)
char regexpat[HUGE_BUFFER] = { '\0' };
WideCharToMultiByte(CP_UTF8, 0, f, (int)(e-f), regexpat, (int)COUNTOF(regexpat), NULL, NULL);

if (RegExFind(regexpat, chFilePath, false, SciCall_GetEOLMode(), NULL) >= 0) {
if (RegExFind(regexpat, chFilePath, false, NULL) >= 0) {
return g_pLexArray[iLex];
}
}
Expand Down Expand Up @@ -3378,6 +3408,50 @@ void Style_AppendWeightAttribute(LPWSTR lpszWeight, int cchSize, int fontWeight)
}


//=============================================================================
//
// Style_StrGetStretchValue()
//
bool Style_StrGetStretchValue(LPCWSTR lpszStyle, int* stretch)
{
int fontStretch = SC_STRETCH_NORMAL;
bool bFound = false;
for (int i = FS_IDX_ULTRACONDENSED; i <= FS_IDX_ULTRAEXPANDED; ++i) {
if (Style_StrHasAttribute(lpszStyle, FontStretches[i].wname)) {
fontStretch = FontStretches[i].stretch;
bFound = true;
break;
}
}
if (bFound) {
*stretch = fontStretch;
}
return bFound;
}


//=============================================================================
//
// Style_AppendStretchAttribute()
//
void Style_AppendStretchAttribute(LPWSTR lpszStyle, int cchSize, int fontStretch)
{
if (fontStretch == SC_STRETCH_NORMAL) {
return; // normal is default, no need to append
}
const WCHAR *pFontStretch = NULL;
for (int i = FS_IDX_ULTRACONDENSED; i <= FS_IDX_ULTRAEXPANDED; ++i) {
if (fontStretch == FontStretches[i].stretch) {
pFontStretch = FontStretches[i].wname;
break;
}
}
if (pFontStretch) {
AppendStyle(lpszStyle, cchSize, pFontStretch);
}
}


//=============================================================================
//
// Style_StrGetColor()
Expand Down Expand Up @@ -3599,6 +3673,26 @@ void Style_CopyStyles_IfNotDefined(LPCWSTR lpszStyleSrc, LPWSTR lpszStyleDest, i
AppendStyle(szTmpStyle, COUNTOF(szTmpStyle), pFontWeight);
}

// Font Stretch
const WCHAR *pFontStretch = NULL;
for (int idx = FS_IDX_ULTRACONDENSED; idx <= FS_IDX_ULTRAEXPANDED; ++idx) {
if (Style_StrHasAttribute(lpszStyleDest, FontStretches[idx].wname)) {
pFontStretch = FontStretches[idx].wname;
break;
}
}
if (!bIsFontDefInDestination && !pFontStretch) {
for (int idx = FS_IDX_ULTRACONDENSED; idx <= FS_IDX_ULTRAEXPANDED; ++idx) {
if (Style_StrHasAttribute(lpszStyleSrc, FontStretches[idx].wname)) {
pFontStretch = FontStretches[idx].wname;
break;
}
}
}
if (pFontStretch) {
AppendStyle(szTmpStyle, COUNTOF(szTmpStyle), pFontStretch);
}

if (Style_StrHasAttribute(lpszStyleDest, FontEffects[FE_ITALIC])) {
AppendStyle(szTmpStyle, COUNTOF(szTmpStyle), FontEffects[FE_ITALIC]);
} else if (!bIsFontDefInDestination && Style_StrHasAttribute(lpszStyleSrc, FontEffects[FE_ITALIC])) {
Expand Down Expand Up @@ -3818,7 +3912,8 @@ bool Style_SelectFont(HWND hwnd, LPWSTR lpszStyle, int cchStyle, LPCWSTR sLexerN
int const iFontHeight = PointSizeToFontHeight(fFontSize, hdc);
ReleaseDC(hwnd, hdc);

int const iFontStretch = 0; // with calculated automatically
int iFontStretch = SC_STRETCH_NORMAL;
Style_StrGetStretchValue(lpszStyle, &iFontStretch);
bool const bIsUnderline = Style_StrHasAttribute(lpszStyle, FontEffects[FE_UNDERLINE]);
bool const bIsStrikeout = Style_StrHasAttribute(lpszStyle, FontEffects[FE_STRIKEOUT]);

Expand All @@ -3835,7 +3930,7 @@ bool Style_SelectFont(HWND hwnd, LPWSTR lpszStyle, int cchStyle, LPCWSTR sLexerN
LOGFONT lf = { 0 };
lf.lfCharSet = (BYTE)iCharSet;
lf.lfHeight = iFontHeight;
lf.lfWidth = iFontStretch;
lf.lfWidth = 0; // let system calculate character width
lf.lfWeight = iFontWeight;
lf.lfItalic = (BYTE)(BOOL)bIsItalic;
lf.lfUnderline = (BYTE)(BOOL)bIsUnderline;
Expand Down Expand Up @@ -3942,6 +4037,11 @@ bool Style_SelectFont(HWND hwnd, LPWSTR lpszStyle, int cchStyle, LPCWSTR sLexerN
Style_AppendWeightAttribute(szNewStyle, COUNTOF(szNewStyle), lf.lfWeight);
}

// persist stretch (ChooseFont dialog doesn't modify it, so round-trip the original)
if (iFontStretch != SC_STRETCH_NORMAL) {
Style_AppendStretchAttribute(szNewStyle, COUNTOF(szNewStyle), iFontStretch);
}

if (lf.lfItalic) {
AppendStyle(szNewStyle, COUNTOF(szNewStyle), FontEffects[FE_ITALIC]);
}
Expand Down Expand Up @@ -4109,6 +4209,13 @@ void Style_SetStyles(HWND hwnd, const int iStyle, LPCWSTR lpszStyle, const float
SciCall_StyleSetWeight(iStyle, SC_WEIGHT_NORMAL);
}

// Font Stretch
if (Style_StrGetStretchValue(lpszStyle, &iValue)) {
SciCall_StyleSetStretch(iStyle, iValue);
} else if (bIsDefaultStyle) {
SciCall_StyleSetStretch(iStyle, SC_STRETCH_NORMAL);
}

// Italic
SciCall_StyleSetItalic(iStyle, Style_StrHasAttribute(lpszStyle, FontEffects[FE_ITALIC]));

Expand Down
Loading