diff options
Diffstat (limited to 'generic/tclParse.c')
| -rw-r--r-- | generic/tclParse.c | 70 |
1 files changed, 41 insertions, 29 deletions
diff --git a/generic/tclParse.c b/generic/tclParse.c index 164905a..6143cb7 100644 --- a/generic/tclParse.c +++ b/generic/tclParse.c @@ -195,19 +195,19 @@ Tcl_ParseCommand( * NULL, then no error message is provided. */ const char *start, /* First character of string containing one or * more Tcl commands. */ - register int numBytes, /* Total number of bytes in string. If < 0, + int numBytes, /* Total number of bytes in string. If < 0, * the script consists of all bytes up to the * first null character. */ int nested, /* Non-zero means this is a nested command: * close bracket should be considered a * command terminator. If zero, then close * bracket has no special meaning. */ - register Tcl_Parse *parsePtr) + Tcl_Parse *parsePtr) /* Structure to fill in with information about * the parsed command; any previous * information in the structure is ignored. */ { - register const char *src; /* Points to current character in the + const char *src; /* Points to current character in the * command. */ char type; /* Result returned by CHAR_TYPE(*src). */ Tcl_Token *tokenPtr; /* Pointer to token being filled in. */ @@ -620,14 +620,14 @@ TclIsBareword( static int ParseWhiteSpace( const char *src, /* First character to parse. */ - register int numBytes, /* Max number of bytes to scan. */ + int numBytes, /* Max number of bytes to scan. */ int *incompletePtr, /* Set this boolean memory to true if parsing * indicates an incomplete command. */ char *typePtr) /* Points to location to store character type * of character that ends run of whitespace */ { - register char type = TYPE_NORMAL; - register const char *p = src; + char type = TYPE_NORMAL; + const char *p = src; while (1) { while (numBytes && ((type = CHAR_TYPE(*p)) & TYPE_SPACE)) { @@ -729,7 +729,7 @@ TclParseHex( * conversion is to be written. */ { int result = 0; - register const char *p = src; + const char *p = src; while (numBytes--) { unsigned char digit = UCHAR(*p); @@ -784,10 +784,9 @@ TclParseBackslash( * of bytes scanned should be written. */ char *dst) /* NULL, or points to buffer where the UTF-8 * encoding of the backslash sequence is to be - * written. At most TCL_UTF_MAX bytes will be - * written there. */ + * written. At most 4 bytes will be written there. */ { - register const char *p = src+1; + const char *p = src+1; Tcl_UniChar unichar = 0; int result; int count; @@ -817,7 +816,7 @@ TclParseBackslash( count = 2; switch (*p) { /* - * Note: in the conversions below, use absolute values (e.g., 0xa) + * Note: in the conversions below, use absolute values (e.g., 0xA) * rather than symbolic values (e.g. \n) that get converted by the * compiler. It's possible that compilers on some platforms will do * the symbolic conversions differently, which could result in @@ -831,19 +830,19 @@ TclParseBackslash( result = 0x8; break; case 'f': - result = 0xc; + result = 0xC; break; case 'n': - result = 0xa; + result = 0xA; break; case 'r': - result = 0xd; + result = 0xD; break; case 't': result = 0x9; break; case 'v': - result = 0xb; + result = 0xB; break; case 'x': count += TclParseHex(p+1, (numBytes > 3) ? 2 : numBytes-2, &result); @@ -867,6 +866,16 @@ TclParseBackslash( * No hexdigits -> This is just "u". */ result = 'u'; + } else if (((result & 0xDC00) == 0xD800) && (count == 6) + && (p[5] == '\\') && (p[6] == 'u') && (numBytes >= 10)) { + /* If high surrogate is immediately followed by a low surrogate + * escape, combine them into one character. */ + int low; + int count2 = TclParseHex(p+7, 4, &low); + if ((count2 == 4) && ((low & 0xDC00) == 0xDC00)) { + result = ((result & 0x3FF)<<10 | (low & 0x3FF)) + 0x10000; + count += count2 + 2; + } } break; case 'U': @@ -876,6 +885,9 @@ TclParseBackslash( * No hexdigits -> This is just "U". */ result = 'U'; + } else if ((result | 0x7FF) == 0xDFFF) { + /* Upper or lower surrogate, not allowed in this syntax. */ + result = 0xFFFD; } break; case '\n': @@ -967,12 +979,12 @@ TclParseBackslash( static int ParseComment( const char *src, /* First character to parse. */ - register int numBytes, /* Max number of bytes to scan. */ + int numBytes, /* Max number of bytes to scan. */ Tcl_Parse *parsePtr) /* Information about parse in progress. * Updated if parsing indicates an incomplete * command. */ { - register const char *p = src; + const char *p = src; int incomplete = parsePtr->incomplete; while (numBytes) { @@ -1039,8 +1051,8 @@ ParseComment( static int ParseTokens( - register const char *src, /* First character to parse. */ - register int numBytes, /* Max number of bytes to scan. */ + const char *src, /* First character to parse. */ + int numBytes, /* Max number of bytes to scan. */ int mask, /* Specifies when to stop parsing. The parse * stops at the first unquoted character whose * CHAR_TYPE contains any of the bits in @@ -1318,7 +1330,7 @@ Tcl_ParseVarName( * NULL, then no error message is provided. */ const char *start, /* Start of variable substitution string. * First character must be "$". */ - register int numBytes, /* Total number of bytes in string. If < 0, + int numBytes, /* Total number of bytes in string. If < 0, * the string consists of all bytes up to the * first null character. */ Tcl_Parse *parsePtr, /* Structure to fill in with information about @@ -1329,7 +1341,7 @@ Tcl_ParseVarName( * reinitialize it. */ { Tcl_Token *tokenPtr; - register const char *src; + const char *src; int varIndex; unsigned array; @@ -1511,13 +1523,13 @@ Tcl_ParseVarName( const char * Tcl_ParseVar( Tcl_Interp *interp, /* Context for looking up variable. */ - register const char *start, /* Start of variable substitution. First + const char *start, /* Start of variable substitution. First * character must be "$". */ const char **termPtr) /* If non-NULL, points to word to fill in with * character just after last one in the * variable specifier. */ { - register Tcl_Obj *objPtr; + Tcl_Obj *objPtr; int code; Tcl_Parse *parsePtr = TclStackAlloc(interp, sizeof(Tcl_Parse)); @@ -1596,10 +1608,10 @@ Tcl_ParseBraces( * NULL, then no error message is provided. */ const char *start, /* Start of string enclosed in braces. The * first character must be {'. */ - register int numBytes, /* Total number of bytes in string. If < 0, + int numBytes, /* Total number of bytes in string. If < 0, * the string consists of all bytes up to the * first null character. */ - register Tcl_Parse *parsePtr, + Tcl_Parse *parsePtr, /* Structure to fill in with information about * the string. */ int append, /* Non-zero means append tokens to existing @@ -1612,7 +1624,7 @@ Tcl_ParseBraces( * successful. */ { Tcl_Token *tokenPtr; - register const char *src; + const char *src; int startIndex, level, length; if ((numBytes == 0) || (start == NULL)) { @@ -1738,7 +1750,7 @@ Tcl_ParseBraces( */ { - register int openBrace = 0; + int openBrace = 0; while (--src > start) { switch (*src) { @@ -1798,10 +1810,10 @@ Tcl_ParseQuotedString( * NULL, then no error message is provided. */ const char *start, /* Start of the quoted string. The first * character must be '"'. */ - register int numBytes, /* Total number of bytes in string. If < 0, + int numBytes, /* Total number of bytes in string. If < 0, * the string consists of all bytes up to the * first null character. */ - register Tcl_Parse *parsePtr, + Tcl_Parse *parsePtr, /* Structure to fill in with information about * the string. */ int append, /* Non-zero means append tokens to existing |