diff options
| author | jan.nijtmans <nijtmans@users.sourceforge.net> | 2023-03-20 11:31:14 (GMT) |
|---|---|---|
| committer | jan.nijtmans <nijtmans@users.sourceforge.net> | 2023-03-20 11:31:14 (GMT) |
| commit | 2475fa8cb8310271c733ae65497746174c93a889 (patch) | |
| tree | cf743c26fa7e0cdb7e4cd958bd0978f1fdf4222d /generic | |
| parent | 921b0dfdeab784bd9d8f8071100db8156c48a1ae (diff) | |
| parent | 5d63902332fb63571e32e9124190aa3f9c98526b (diff) | |
| download | tcl-2475fa8cb8310271c733ae65497746174c93a889.zip tcl-2475fa8cb8310271c733ae65497746174c93a889.tar.gz tcl-2475fa8cb8310271c733ae65497746174c93a889.tar.bz2 | |
Merge 8.7
Diffstat (limited to 'generic')
| -rw-r--r-- | generic/tclEncoding.c | 31 |
1 files changed, 18 insertions, 13 deletions
diff --git a/generic/tclEncoding.c b/generic/tclEncoding.c index ec9e130..27e7278 100644 --- a/generic/tclEncoding.c +++ b/generic/tclEncoding.c @@ -3444,18 +3444,23 @@ TableToUtfProc( if (prefixBytes[byte]) { src++; if (src >= srcEnd) { - /* - * TODO - this is broken. For consistency with other - * decoders, an error should be raised only if strict. - * However, doing that check cause a whole bunch of test - * failures. Need to verify if those tests are in fact - * correct. - */ - src--; - result = TCL_CONVERT_MULTIBYTE; - break; + if (!(flags & TCL_ENCODING_END)) { + src--; + result = TCL_CONVERT_MULTIBYTE; + break; + } else if (PROFILE_STRICT(flags)) { + src--; + result = TCL_CONVERT_SYNTAX; + break; + } else if (PROFILE_REPLACE(flags)) { + ch = UNICODE_REPLACE_CHAR; + } else { + numChars++; /* Silently consume */ + break; + } + } else { + ch = toUnicode[byte][*((unsigned char *)src)]; } - ch = toUnicode[byte][*((unsigned char *)src)]; ch = toUnicode[byte][*((unsigned char *) src)]; } else { ch = pageZero[byte]; @@ -3479,7 +3484,7 @@ TableToUtfProc( * Special case for 1-byte utf chars for speed. */ - if (ch && ch < 0x80) { + if ((unsigned)ch - 1 < 0x7F) { *dst++ = (char) ch; } else { dst += Tcl_UniCharToUtf(ch, dst); @@ -3680,7 +3685,7 @@ Iso88591ToUtfProc( * Special case for 1-byte utf chars for speed. */ - if (ch && ch < 0x80) { + if ((unsigned)ch - 1 < 0x7F) { *dst++ = (char) ch; } else { dst += Tcl_UniCharToUtf(ch, dst); |