diff options
-rw-r--r-- | cocoa/ScintillaCocoa.h | 1 | ||||
-rw-r--r-- | cocoa/ScintillaView.mm | 1 | ||||
-rw-r--r-- | doc/ScintillaHistory.html | 4 | ||||
-rw-r--r-- | src/UniConversion.cxx | 23 | ||||
-rw-r--r-- | src/UniConversion.h | 1 |
5 files changed, 30 insertions, 0 deletions
diff --git a/cocoa/ScintillaCocoa.h b/cocoa/ScintillaCocoa.h index 4d5d5c023..1defaf220 100644 --- a/cocoa/ScintillaCocoa.h +++ b/cocoa/ScintillaCocoa.h @@ -48,6 +48,7 @@ #include "CaseFolder.h" #include "Document.h" #include "CaseConvert.h" +#include "UniConversion.h" #include "Selection.h" #include "PositionCache.h" #include "EditModel.h" diff --git a/cocoa/ScintillaView.mm b/cocoa/ScintillaView.mm index 107c421c3..61c471d6a 100644 --- a/cocoa/ScintillaView.mm +++ b/cocoa/ScintillaView.mm @@ -437,6 +437,7 @@ static NSCursor *cursorFromEnum(Window::Cursor cursor) [mOwner message: SCI_SETTARGETRANGE wParam: posRange.location lParam: NSMaxRange(posRange)]; std::string text([mOwner message: SCI_TARGETASUTF8] + 1, 0); [mOwner message: SCI_TARGETASUTF8 wParam: 0 lParam: reinterpret_cast<sptr_t>(&text[0])]; + text = FixInvalidUTF8(text); NSString *result = [NSString stringWithUTF8String: text.c_str()]; NSMutableAttributedString *asResult = [[[NSMutableAttributedString alloc] initWithString:result] autorelease]; diff --git a/doc/ScintillaHistory.html b/doc/ScintillaHistory.html index 47966f3c1..c63563c7f 100644 --- a/doc/ScintillaHistory.html +++ b/doc/ScintillaHistory.html @@ -537,6 +537,10 @@ <a href="http://sourceforge.net/p/scintilla/bugs/1881/">Bug #1881</a>. </li> <li> + Fix potential problems with IME on Cocoa when document contains invalid + UTF-8. + </li> + <li> Fix crash on Cocoa with OS X 10.9 due to accessibility API not available. <a href="http://sourceforge.net/p/scintilla/bugs/1915/">Bug #1915</a>. </li> diff --git a/src/UniConversion.cxx b/src/UniConversion.cxx index 4da9e102a..d0028d65e 100644 --- a/src/UniConversion.cxx +++ b/src/UniConversion.cxx @@ -8,6 +8,7 @@ #include <stdlib.h> #include <stdexcept> +#include <string> #include "UniConversion.h" @@ -304,6 +305,28 @@ int UTF8DrawBytes(const unsigned char *us, int len) { return (utf8StatusNext & UTF8MaskInvalid) ? 1 : (utf8StatusNext & UTF8MaskWidth); } +// Replace invalid bytes in UTF-8 with the replacement character +std::string FixInvalidUTF8(const std::string &text) { + std::string result; + const unsigned char *us = reinterpret_cast<const unsigned char *>(text.c_str()); + size_t remaining = text.size(); + while (remaining > 0) { + const int utf8Status = UTF8Classify(us, static_cast<int>(remaining)); + if (utf8Status & UTF8MaskInvalid) { + // Replacement character 0xFFFD = UTF8:"efbfbd". + result.append("\xef\xbf\xbd"); + us++; + remaining--; + } else { + const int len = utf8Status&UTF8MaskWidth; + result.append(reinterpret_cast<const char *>(us), len); + us += len; + remaining -= len; + } + } + return result; +} + #ifdef SCI_NAMESPACE } #endif diff --git a/src/UniConversion.h b/src/UniConversion.h index aeb13f0c2..c5867a664 100644 --- a/src/UniConversion.h +++ b/src/UniConversion.h @@ -23,6 +23,7 @@ size_t UTF16Length(const char *s, size_t len); size_t UTF16FromUTF8(const char *s, size_t len, wchar_t *tbuf, size_t tlen); unsigned int UTF32FromUTF8(const char *s, unsigned int len, unsigned int *tbuf, unsigned int tlen); unsigned int UTF16FromUTF32Character(unsigned int val, wchar_t *tbuf); +std::string FixInvalidUTF8(const std::string &text); extern int UTF8BytesOfLead[256]; void UTF8BytesOfLeadInitialise(); |