diff options
| -rw-r--r-- | generic/tclEncoding.c | 37 | ||||
| -rw-r--r-- | tests/encoding.test | 21 |
2 files changed, 35 insertions, 23 deletions
diff --git a/generic/tclEncoding.c b/generic/tclEncoding.c index 470f8f3..a11e696 100644 --- a/generic/tclEncoding.c +++ b/generic/tclEncoding.c @@ -2412,11 +2412,11 @@ UtfToUtfProc( *dst++ = *src++; } else if ((UCHAR(*src) == 0xC0) && (src + 1 < srcEnd) && - (UCHAR(src[1]) == 0x80) && (flags & ENCODING_UTF) && + (UCHAR(src[1]) == 0x80) && !(flags & TCL_ENCODING_MODIFIED) && (!(flags & ENCODING_INPUT) || PROFILE_STRICT(profile) || PROFILE_REPLACE(profile))) { /* Special sequence \xC0\x80 */ - if (flags & ENCODING_INPUT) { + if ((PROFILE_STRICT(profile) || PROFILE_REPLACE(profile)) && (flags & ENCODING_INPUT)) { if (PROFILE_REPLACE(profile)) { dst += Tcl_UniCharToUtf(UNICODE_REPLACE_CHAR, dst); src += 2; @@ -2485,26 +2485,28 @@ UtfToUtfProc( } else { int low; - const char *saveSrc = src; + int isInvalid = 0; size_t len = TclUtfToUCS4(src, &ch); - - /* - * Valid single char encodings were already handled earlier. - * So len==1 means an invalid byte that is magically transformed - * to a code point unless it resulted from the special - * \xC0\x80 sequence. Tests io-75.* - */ - if ((len < 2) && (ch != 0) && (flags & ENCODING_INPUT)) { - if (PROFILE_STRICT(profile)) { - result = TCL_CONVERT_SYNTAX; - break; - } else if (PROFILE_REPLACE(profile)) { - ch = UNICODE_REPLACE_CHAR; + if (flags & ENCODING_INPUT) { + if ((len < 2) && (ch != 0)) { + isInvalid = 1; + } else if ((ch > 0xFFFF) && !(flags & ENCODING_UTF)) { + isInvalid = 1; + } + if (isInvalid) { + if (PROFILE_STRICT(profile)) { + result = TCL_CONVERT_SYNTAX; + break; + } + else if (PROFILE_REPLACE(profile)) { + ch = UNICODE_REPLACE_CHAR; + } } } + const char *saveSrc = src; src += len; - if (!(flags & ENCODING_UTF) && (ch > 0x3FF)) { + if (!(flags & ENCODING_UTF) && !(flags & ENCODING_INPUT) && (ch > 0x3FF)) { if (ch > 0xFFFF) { /* CESU-8 6-byte sequence for chars > U+FFFF */ ch -= 0x10000; @@ -2670,6 +2672,7 @@ Utf32ToUtfProc( if ((unsigned)ch > 0x10FFFF || SURROGATE(ch)) { if (PROFILE_STRICT(flags)) { result = TCL_CONVERT_SYNTAX; + ch = 0; break; } if (PROFILE_REPLACE(flags)) { diff --git a/tests/encoding.test b/tests/encoding.test index 7b3304b..de6b87b 100644 --- a/tests/encoding.test +++ b/tests/encoding.test @@ -459,17 +459,20 @@ test encoding-15.26 {UtfToUtfProc CESU-8} { encoding convertfrom cesu-8 \xC0\x80 } \x00 test encoding-15.27 {UtfToUtfProc -profile strict CESU-8} { - encoding convertfrom -profile strict cesu-8 \xC0\x80 + encoding convertfrom -profile strict cesu-8 \x00 } \x00 -test encoding-15.28 {UtfToUtfProc -profile strict CESU-8} { +test encoding-15.28 {UtfToUtfProc -profile strict CESU-8} -body { encoding convertfrom -profile strict cesu-8 \xC0\x80 -} \x00 +} -returnCodes 1 -result {unexpected byte sequence starting at index 0: '\xC0'} test encoding-15.29 {UtfToUtfProc CESU-8} { encoding convertto cesu-8 \x00 -} \xC0\x80 +} \x00 test encoding-15.30 {UtfToUtfProc -profile strict CESU-8} { encoding convertto -profile strict cesu-8 \x00 -} \xC0\x80 +} \x00 +test encoding-15.31 {UtfToUtfProc -profile strict CESU-8 (bytes F0-F4 are invalid)} -body { + encoding convertfrom -profile strict cesu-8 \xF1\x86\x83\x9C +} -returnCodes 1 -result {unexpected byte sequence starting at index 0: '\xF1'} test encoding-16.1 {Utf16ToUtfProc} -body { set val [encoding convertfrom utf-16 NN] @@ -527,6 +530,9 @@ test encoding-16.15 {Utf16ToUtfProc} -body { test encoding-16.16 {Utf16ToUtfProc} -body { encoding convertfrom utf-16le \x00\xDC\x00\xD8 } -result \uDC00\uD800 +test encoding-16.17 {Utf32ToUtfProc} -body { + list [encoding convertfrom -profile strict -failindex idx utf-32le \x41\x00\x00\x00\x00\xD8\x00\x00\x42\x00\x00\x00] [set idx] +} -result {A 4} test encoding-16.9 { Utf16ToUtfProc, Tcl_UniCharToUtf, surrogate pairs in utf-16 @@ -614,9 +620,12 @@ test encoding-19.3 {TableFromUtfProc} -body { test encoding-19.4 {TableFromUtfProc} -body { list [encoding convertfrom -failindex idx ascii AÁ] [set idx] } -result [list A\xC1 -1] -test encoding-19.4 {TableFromUtfProc} -body { +test encoding-19.5 {TableFromUtfProc} -body { list [encoding convertfrom -failindex idx -profile strict ascii A\xC1] [set idx] } -result {A 1} +test encoding-19.6 {TableFromUtfProc} -body { + list [encoding convertfrom -failindex idx -profile strict ascii AÁB] [set idx] +} -result {A 1} test encoding-20.1 {TableFreefProc} { } {} |