diff options
Diffstat (limited to 'generic/tclScan.c')
| -rw-r--r-- | generic/tclScan.c | 971 |
1 files changed, 493 insertions, 478 deletions
diff --git a/generic/tclScan.c b/generic/tclScan.c index 2e4bf18..54f9b78 100644 --- a/generic/tclScan.c +++ b/generic/tclScan.c @@ -1,14 +1,14 @@ -/* +/* * tclScan.c -- * * This file contains the implementation of the "scan" command. * * Copyright (c) 1998 by Scriptics Corporation. * - * See the file "license.terms" for information on usage and redistribution - * of this file, and for a DISCLAIMER OF ALL WARRANTIES. + * See the file "license.terms" for information on usage and redistribution of + * this file, and for a DISCLAIMER OF ALL WARRANTIES. * - * RCS: @(#) $Id: tclScan.c,v 1.17 2005/05/10 18:34:49 kennykb Exp $ + * RCS: @(#) $Id: tclScan.c,v 1.18 2005/07/21 14:38:51 dkf Exp $ */ #include "tclInt.h" @@ -17,23 +17,23 @@ * Flag values used by Tcl_ScanObjCmd. */ -#define SCAN_NOSKIP 0x1 /* Don't skip blanks. */ -#define SCAN_SUPPRESS 0x2 /* Suppress assignment. */ -#define SCAN_UNSIGNED 0x4 /* Read an unsigned value. */ -#define SCAN_WIDTH 0x8 /* A width value was supplied. */ +#define SCAN_NOSKIP 0x1 /* Don't skip blanks. */ +#define SCAN_SUPPRESS 0x2 /* Suppress assignment. */ +#define SCAN_UNSIGNED 0x4 /* Read an unsigned value. */ +#define SCAN_WIDTH 0x8 /* A width value was supplied. */ -#define SCAN_SIGNOK 0x10 /* A +/- character is allowed. */ -#define SCAN_NODIGITS 0x20 /* No digits have been scanned. */ -#define SCAN_NOZERO 0x40 /* No zero digits have been scanned. */ -#define SCAN_XOK 0x80 /* An 'x' is allowed. */ -#define SCAN_PTOK 0x100 /* Decimal point is allowed. */ -#define SCAN_EXPOK 0x200 /* An exponent is allowed. */ +#define SCAN_SIGNOK 0x10 /* A +/- character is allowed. */ +#define SCAN_NODIGITS 0x20 /* No digits have been scanned. */ +#define SCAN_NOZERO 0x40 /* No zero digits have been scanned. */ +#define SCAN_XOK 0x80 /* An 'x' is allowed. */ +#define SCAN_PTOK 0x100 /* Decimal point is allowed. */ +#define SCAN_EXPOK 0x200 /* An exponent is allowed. */ -#define SCAN_LONGER 0x400 /* Asked for a wide value. */ +#define SCAN_LONGER 0x400 /* Asked for a wide value. */ /* - * The following structure contains the information associated with - * a character set. + * The following structure contains the information associated with a + * character set. */ typedef struct CharSet { @@ -62,9 +62,9 @@ static int ValidateFormat _ANSI_ARGS_((Tcl_Interp *interp, char *format, * * BuildCharSet -- * - * This function examines a character set format specification - * and builds a CharSet containing the individual characters and - * character ranges specified. + * This function examines a character set format specification and builds + * a CharSet containing the individual characters and character ranges + * specified. * * Results: * Returns the next format position. @@ -85,7 +85,7 @@ BuildCharSet(cset, format) char *end; memset(cset, 0, sizeof(CharSet)); - + offset = Tcl_UtfToUniChar(format, &ch); if (ch == '^') { cset->exclude = 1; @@ -131,8 +131,8 @@ BuildCharSet(cset, format) while (ch != ']') { if (*format == '-') { /* - * This may be the first character of a range, so don't add - * it yet. + * This may be the first character of a range, so don't add it + * yet. */ start = ch; @@ -159,7 +159,7 @@ BuildCharSet(cset, format) } else { cset->ranges[cset->nranges].start = ch; cset->ranges[cset->nranges].end = start; - } + } cset->nranges++; } } else { @@ -189,8 +189,8 @@ BuildCharSet(cset, format) static int CharInSet(cset, c) CharSet *cset; - int c; /* Character to test, passed as int because - * of non-ANSI prototypes. */ + int c; /* Character to test, passed as int because of + * non-ANSI prototypes. */ { Tcl_UniChar ch = (Tcl_UniChar) c; int i, match = 0; @@ -209,7 +209,7 @@ CharInSet(cset, c) } } } - return (cset->exclude ? !match : match); + return (cset->exclude ? !match : match); } /* @@ -243,8 +243,8 @@ ReleaseCharSet(cset) * * ValidateFormat -- * - * Parse the format string and verify that it is properly formed - * and that there are exactly enough variables on the command line. + * Parse the format string and verify that it is properly formed and that + * there are exactly enough variables on the command line. * * Results: * A standard Tcl result. @@ -259,8 +259,8 @@ static int ValidateFormat(interp, format, numVars, totalSubs) Tcl_Interp *interp; /* Current interpreter. */ char *format; /* The format string. */ - int numVars; /* The number of variables passed to the - * scan command. */ + int numVars; /* The number of variables passed to the scan + * command. */ int *totalSubs; /* The number of variables that will be * required. */ { @@ -274,9 +274,9 @@ ValidateFormat(interp, format, numVars, totalSubs) char buf[TCL_UTF_MAX+1]; /* - * Initialize an array that records the number of times a variable - * is assigned to by the format string. We use this to detect if - * a variable is multiply assigned or left unassigned. + * Initialize an array that records the number of times a variable is + * assigned to by the format string. We use this to detect if a variable + * is multiply assigned or left unassigned. */ if (numVars > nspace) { @@ -309,9 +309,9 @@ ValidateFormat(interp, format, numVars, totalSubs) if ((ch < 0x80) && isdigit(UCHAR(ch))) { /* INTL: "C" locale. */ /* - * Check for an XPG3-style %n$ specification. Note: there - * must not be a mixture of XPG3 specs and non-XPG3 specs - * in the same format string. + * Check for an XPG3-style %n$ specification. Note: there must + * not be a mixture of XPG3 specs and non-XPG3 specs in the same + * format string. */ value = strtoul(format-1, &end, 10); /* INTL: "C" locale. */ @@ -331,25 +331,25 @@ ValidateFormat(interp, format, numVars, totalSubs) /* * In the case where no vars are specified, the user can * specify %9999$ legally, so we have to consider special - * rules for growing the assign array. 'value' is - * guaranteed to be > 0. + * rules for growing the assign array. 'value' is guaranteed + * to be > 0. */ xpgSize = (xpgSize > value) ? xpgSize : value; } goto xpgCheckDone; } - notXpg: + notXpg: gotSequential = 1; if (gotXpg) { - mixedXPG: + mixedXPG: Tcl_SetResult(interp, "cannot mix \"%\" and \"%n$\" conversion specifiers", TCL_STATIC); goto error; } - xpgCheckDone: + xpgCheckDone: /* * Parse any width specifier. */ @@ -381,73 +381,73 @@ ValidateFormat(interp, format, numVars, totalSubs) */ switch (ch) { - case 'c': - if (flags & SCAN_WIDTH) { - Tcl_SetResult(interp, - "field width may not be specified in %c conversion", - TCL_STATIC); - goto error; - } - /* - * Fall through! - */ - case 'n': - case 's': - if (flags & SCAN_LONGER) { - invalidLonger: - buf[Tcl_UniCharToUtf(ch, buf)] = '\0'; - Tcl_AppendResult(interp, - "'l' modifier may not be specified in %", buf, - " conversion", NULL); - goto error; - } - /* - * Fall through! - */ - case 'd': - case 'e': - case 'f': - case 'g': - case 'i': - case 'o': - case 'u': - case 'x': - break; - /* - * Bracket terms need special checking - */ - case '[': - if (flags & SCAN_LONGER) { - goto invalidLonger; - } + case 'c': + if (flags & SCAN_WIDTH) { + Tcl_SetResult(interp, + "field width may not be specified in %c conversion", + TCL_STATIC); + goto error; + } + /* + * Fall through! + */ + case 'n': + case 's': + if (flags & SCAN_LONGER) { + invalidLonger: + buf[Tcl_UniCharToUtf(ch, buf)] = '\0'; + Tcl_AppendResult(interp, + "'l' modifier may not be specified in %", buf, + " conversion", NULL); + goto error; + } + /* + * Fall through! + */ + case 'd': + case 'e': + case 'f': + case 'g': + case 'i': + case 'o': + case 'u': + case 'x': + break; + /* + * Bracket terms need special checking + */ + case '[': + if (flags & SCAN_LONGER) { + goto invalidLonger; + } + if (*format == '\0') { + goto badSet; + } + format += Tcl_UtfToUniChar(format, &ch); + if (ch == '^') { if (*format == '\0') { goto badSet; } format += Tcl_UtfToUniChar(format, &ch); - if (ch == '^') { - if (*format == '\0') { - goto badSet; - } - format += Tcl_UtfToUniChar(format, &ch); - } - if (ch == ']') { - if (*format == '\0') { - goto badSet; - } - format += Tcl_UtfToUniChar(format, &ch); + } + if (ch == ']') { + if (*format == '\0') { + goto badSet; } - while (ch != ']') { - if (*format == '\0') { - goto badSet; - } - format += Tcl_UtfToUniChar(format, &ch); + format += Tcl_UtfToUniChar(format, &ch); + } + while (ch != ']') { + if (*format == '\0') { + goto badSet; } - break; - badSet: - Tcl_SetResult(interp, "unmatched [ in format string", - TCL_STATIC); - goto error; - default: + format += Tcl_UtfToUniChar(format, &ch); + } + break; + badSet: + Tcl_SetResult(interp, "unmatched [ in format string", + TCL_STATIC); + goto error; + default: { char buf[TCL_UTF_MAX+1]; @@ -460,10 +460,11 @@ ValidateFormat(interp, format, numVars, totalSubs) if (!(flags & SCAN_SUPPRESS)) { if (objIndex >= nspace) { /* - * Expand the nassign buffer. If we are using XPG specifiers, - * make sure that we grow to a large enough size. xpgSize is + * Expand the nassign buffer. If we are using XPG specifiers, + * make sure that we grow to a large enough size. xpgSize is * guaranteed to be at least one larger than objIndex. */ + value = nspace; if (xpgSize) { nspace = xpgSize; @@ -504,14 +505,19 @@ ValidateFormat(interp, format, numVars, totalSubs) } for (i = 0; i < numVars; i++) { if (nassign[i] > 1) { - Tcl_SetResult(interp, "variable is assigned by multiple \"%n$\" conversion specifiers", TCL_STATIC); + Tcl_SetResult(interp, + "variable is assigned by multiple \"%n$\" conversion specifiers", + TCL_STATIC); goto error; } else if (!xpgSize && (nassign[i] == 0)) { /* - * If the space is empty, and xpgSize is 0 (means XPG wasn't - * used, and/or numVars != 0), then too many vars were given + * If the space is empty, and xpgSize is 0 (means XPG wasn't used, + * and/or numVars != 0), then too many vars were given */ - Tcl_SetResult(interp, "variable is not assigned by any conversion specifiers", TCL_STATIC); + + Tcl_SetResult(interp, + "variable is not assigned by any conversion specifiers", + TCL_STATIC); goto error; } } @@ -521,17 +527,17 @@ ValidateFormat(interp, format, numVars, totalSubs) } return TCL_OK; - badIndex: + badIndex: if (gotXpg) { Tcl_SetResult(interp, "\"%n$\" argument index out of range", TCL_STATIC); } else { - Tcl_SetResult(interp, + Tcl_SetResult(interp, "different numbers of variable names and field specifiers", TCL_STATIC); } - error: + error: if (nassign != staticAssign) { ckfree((char *)nassign); } @@ -544,8 +550,8 @@ ValidateFormat(interp, format, numVars, totalSubs) * * Tcl_ScanObjCmd -- * - * This procedure is invoked to process the "scan" Tcl command. - * See the user documentation for details on what it does. + * This function is invoked to process the "scan" Tcl command. See the + * user documentation for details on what it does. * * Results: * A standard Tcl result. @@ -581,12 +587,12 @@ Tcl_ScanObjCmd(dummy, interp, objc, objv) Tcl_UniChar ch, sch; Tcl_Obj **objs = NULL, *objPtr = NULL; int flags; - char buf[513]; /* Temporary buffer to hold scanned - * number strings before they are - * passed to strtoul. */ + char buf[513]; /* Temporary buffer to hold scanned number + * strings before they are passed to + * strtoul. */ if (objc < 3) { - Tcl_WrongNumArgs(interp, 1, objv, + Tcl_WrongNumArgs(interp, 1, objv, "string format ?varName varName ...?"); return TCL_ERROR; } @@ -597,7 +603,7 @@ Tcl_ScanObjCmd(dummy, interp, objc, objv) /* * Check for errors in the format string. */ - + if (ValidateFormat(interp, format, numVars, &totalVars) == TCL_ERROR) { return TCL_ERROR; } @@ -617,9 +623,9 @@ Tcl_ScanObjCmd(dummy, interp, objc, objv) baseString = string; /* - * Iterate over the format string filling in the result objects until - * we reach the end of input, the end of the format string, or there - * is a mismatch. + * Iterate over the format string filling in the result objects until we + * reach the end of input, the end of the format string, or there is a + * mismatch. */ objIndex = 0; @@ -644,9 +650,9 @@ Tcl_ScanObjCmd(dummy, interp, objc, objv) } continue; } - + if (ch != '%') { - literal: + literal: if (*string == '\0') { underflow = 1; goto done; @@ -664,15 +670,15 @@ Tcl_ScanObjCmd(dummy, interp, objc, objv) } /* - * Check for assignment suppression ('*') or an XPG3-style - * assignment ('%n$'). + * Check for assignment suppression ('*') or an XPG3-style assignment + * ('%n$'). */ if (ch == '*') { flags |= SCAN_SUPPRESS; format += Tcl_UtfToUniChar(format, &ch); - } else if ((ch < 0x80) && isdigit(UCHAR(ch))) { /* INTL: "C" locale. */ - value = strtoul(format-1, &end, 10); /* INTL: "C" locale. */ + } else if ((ch < 0x80) && isdigit(UCHAR(ch))) { /* INTL: "C" locale. */ + value = strtoul(format-1, &end, 10); /* INTL: "C" locale. */ if (*end == '$') { format = end+1; format += Tcl_UtfToUniChar(format, &ch); @@ -684,8 +690,8 @@ Tcl_ScanObjCmd(dummy, interp, objc, objv) * Parse any width specifier. */ - if ((ch < 0x80) && isdigit(UCHAR(ch))) { /* INTL: "C" locale. */ - width = strtoul(format-1, &format, 10); /* INTL: "C" locale. */ + if ((ch < 0x80) && isdigit(UCHAR(ch))) { /* INTL: "C" locale. */ + width = strtoul(format-1, &format, 10); /* INTL: "C" locale. */ format += Tcl_UtfToUniChar(format, &ch); } else { width = 0; @@ -711,90 +717,90 @@ Tcl_ScanObjCmd(dummy, interp, objc, objv) */ switch (ch) { - case 'n': - if (!(flags & SCAN_SUPPRESS)) { - objPtr = Tcl_NewIntObj(string - baseString); - Tcl_IncrRefCount(objPtr); - objs[objIndex++] = objPtr; - } - nconversions++; - continue; + case 'n': + if (!(flags & SCAN_SUPPRESS)) { + objPtr = Tcl_NewIntObj(string - baseString); + Tcl_IncrRefCount(objPtr); + objs[objIndex++] = objPtr; + } + nconversions++; + continue; - case 'd': - op = 'i'; - base = 10; - fn = (long (*) _ANSI_ARGS_((char*,void*,int)))strtol; + case 'd': + op = 'i'; + base = 10; + fn = (long (*) _ANSI_ARGS_((char*,void*,int)))strtol; #ifndef TCL_WIDE_INT_IS_LONG - lfn = (Tcl_WideInt (*)_ANSI_ARGS_((char*,void*,int)))strtoll; + lfn = (Tcl_WideInt (*)_ANSI_ARGS_((char*,void*,int)))strtoll; #endif - break; - case 'i': - op = 'i'; - base = 0; - fn = (long (*)_ANSI_ARGS_((char*,void*,int)))strtol; + break; + case 'i': + op = 'i'; + base = 0; + fn = (long (*)_ANSI_ARGS_((char*,void*,int)))strtol; #ifndef TCL_WIDE_INT_IS_LONG - lfn = (Tcl_WideInt (*)_ANSI_ARGS_((char*,void*,int)))strtoll; + lfn = (Tcl_WideInt (*)_ANSI_ARGS_((char*,void*,int)))strtoll; #endif - break; - case 'o': - op = 'i'; - base = 8; - fn = (long (*)_ANSI_ARGS_((char*,void*,int)))strtoul; + break; + case 'o': + op = 'i'; + base = 8; + fn = (long (*)_ANSI_ARGS_((char*,void*,int)))strtoul; #ifndef TCL_WIDE_INT_IS_LONG - lfn = (Tcl_WideInt (*)_ANSI_ARGS_((char*,void*,int)))strtoull; + lfn = (Tcl_WideInt (*)_ANSI_ARGS_((char*,void*,int)))strtoull; #endif - break; - case 'x': - op = 'i'; - base = 16; - fn = (long (*)_ANSI_ARGS_((char*,void*,int)))strtoul; + break; + case 'x': + op = 'i'; + base = 16; + fn = (long (*)_ANSI_ARGS_((char*,void*,int)))strtoul; #ifndef TCL_WIDE_INT_IS_LONG - lfn = (Tcl_WideInt (*)_ANSI_ARGS_((char*,void*,int)))strtoull; + lfn = (Tcl_WideInt (*)_ANSI_ARGS_((char*,void*,int)))strtoull; #endif - break; - case 'u': - op = 'i'; - base = 10; - flags |= SCAN_UNSIGNED; - fn = (long (*)_ANSI_ARGS_((char*,void*,int)))strtoul; + break; + case 'u': + op = 'i'; + base = 10; + flags |= SCAN_UNSIGNED; + fn = (long (*)_ANSI_ARGS_((char*,void*,int)))strtoul; #ifndef TCL_WIDE_INT_IS_LONG - lfn = (Tcl_WideInt (*)_ANSI_ARGS_((char*,void*,int)))strtoull; + lfn = (Tcl_WideInt (*)_ANSI_ARGS_((char*,void*,int)))strtoull; #endif - break; + break; - case 'f': - case 'e': - case 'g': - op = 'f'; - break; + case 'f': + case 'e': + case 'g': + op = 'f'; + break; - case 's': - op = 's'; - break; + case 's': + op = 's'; + break; - case 'c': - op = 'c'; - flags |= SCAN_NOSKIP; - break; - case '[': - op = '['; - flags |= SCAN_NOSKIP; - break; + case 'c': + op = 'c'; + flags |= SCAN_NOSKIP; + break; + case '[': + op = '['; + flags |= SCAN_NOSKIP; + break; } /* - * At this point, we will need additional characters from the - * string to proceed. + * At this point, we will need additional characters from the string + * to proceed. */ if (*string == '\0') { underflow = 1; goto done; } - + /* - * Skip any leading whitespace at the beginning of a field unless - * the format suppresses this behavior. + * Skip any leading whitespace at the beginning of a field unless the + * format suppresses this behavior. */ if (!(flags & SCAN_NOSKIP)) { @@ -814,370 +820,370 @@ Tcl_ScanObjCmd(dummy, interp, objc, objv) /* * Perform the requested scanning operation. */ - + switch (op) { - case 's': - /* - * Scan a string up to width characters or whitespace. - */ + case 's': + /* + * Scan a string up to width characters or whitespace. + */ - if (width == 0) { - width = (size_t) ~0; - } - end = string; - while (*end != '\0') { - offset = Tcl_UtfToUniChar(end, &sch); - if (Tcl_UniCharIsSpace(sch)) { - break; - } - end += offset; - if (--width == 0) { - break; - } + if (width == 0) { + width = (size_t) ~0; + } + end = string; + while (*end != '\0') { + offset = Tcl_UtfToUniChar(end, &sch); + if (Tcl_UniCharIsSpace(sch)) { + break; } - if (!(flags & SCAN_SUPPRESS)) { - objPtr = Tcl_NewStringObj(string, end-string); - Tcl_IncrRefCount(objPtr); - objs[objIndex++] = objPtr; + end += offset; + if (--width == 0) { + break; } - string = end; - break; - - case '[': { - CharSet cset; + } + if (!(flags & SCAN_SUPPRESS)) { + objPtr = Tcl_NewStringObj(string, end-string); + Tcl_IncrRefCount(objPtr); + objs[objIndex++] = objPtr; + } + string = end; + break; - if (width == 0) { - width = (size_t) ~0; - } - end = string; + case '[': { + CharSet cset; - format = BuildCharSet(&cset, format); - while (*end != '\0') { - offset = Tcl_UtfToUniChar(end, &sch); - if (!CharInSet(&cset, (int)sch)) { - break; - } - end += offset; - if (--width == 0) { - break; - } - } - ReleaseCharSet(&cset); + if (width == 0) { + width = (size_t) ~0; + } + end = string; - if (string == end) { - /* - * Nothing matched the range, stop processing - */ - goto done; + format = BuildCharSet(&cset, format); + while (*end != '\0') { + offset = Tcl_UtfToUniChar(end, &sch); + if (!CharInSet(&cset, (int)sch)) { + break; } - if (!(flags & SCAN_SUPPRESS)) { - objPtr = Tcl_NewStringObj(string, end-string); - Tcl_IncrRefCount(objPtr); - objs[objIndex++] = objPtr; + end += offset; + if (--width == 0) { + break; } - string = end; - - break; } - case 'c': - /* - * Scan a single Unicode character. - */ + ReleaseCharSet(&cset); - string += Tcl_UtfToUniChar(string, &sch); - if (!(flags & SCAN_SUPPRESS)) { - objPtr = Tcl_NewIntObj((int)sch); - Tcl_IncrRefCount(objPtr); - objs[objIndex++] = objPtr; - } - break; - - case 'i': + if (string == end) { /* - * Scan an unsigned or signed integer. + * Nothing matched the range, stop processing. */ + goto done; + } + if (!(flags & SCAN_SUPPRESS)) { + objPtr = Tcl_NewStringObj(string, end-string); + Tcl_IncrRefCount(objPtr); + objs[objIndex++] = objPtr; + } + string = end; - if ((width == 0) || (width > sizeof(buf) - 1)) { - width = sizeof(buf) - 1; - } - flags |= SCAN_SIGNOK | SCAN_NODIGITS | SCAN_NOZERO; - for (end = buf; width > 0; width--) { - switch (*string) { - /* - * The 0 digit has special meaning at the beginning of - * a number. If we are unsure of the base, it - * indicates that we are in base 8 or base 16 (if it is - * followed by an 'x'). - * - * 8.1 - 8.3.4 incorrectly handled 0x... base-16 - * cases for %x by not reading the 0x as the - * auto-prelude for base-16. [Bug #495213] - */ - case '0': - if (base == 0) { - base = 8; - flags |= SCAN_XOK; - } - if (base == 16) { - flags |= SCAN_XOK; - } - if (flags & SCAN_NOZERO) { - flags &= ~(SCAN_SIGNOK | SCAN_NODIGITS - | SCAN_NOZERO); - } else { - flags &= ~(SCAN_SIGNOK | SCAN_XOK - | SCAN_NODIGITS); - } - goto addToInt; - - case '1': case '2': case '3': case '4': - case '5': case '6': case '7': - if (base == 0) { - base = 10; - } - flags &= ~(SCAN_SIGNOK | SCAN_XOK | SCAN_NODIGITS); - goto addToInt; - - case '8': case '9': - if (base == 0) { - base = 10; - } - if (base <= 8) { - break; - } - flags &= ~(SCAN_SIGNOK | SCAN_XOK | SCAN_NODIGITS); - goto addToInt; - - case 'A': case 'B': case 'C': - case 'D': case 'E': case 'F': - case 'a': case 'b': case 'c': - case 'd': case 'e': case 'f': - if (base <= 10) { - break; - } - flags &= ~(SCAN_SIGNOK | SCAN_XOK | SCAN_NODIGITS); - goto addToInt; - - case '+': case '-': - if (flags & SCAN_SIGNOK) { - flags &= ~SCAN_SIGNOK; - goto addToInt; - } - break; - - case 'x': case 'X': - if ((flags & SCAN_XOK) && (end == buf+1)) { - base = 16; - flags &= ~SCAN_XOK; - goto addToInt; - } - break; - } + break; + } + case 'c': + /* + * Scan a single Unicode character. + */ - /* - * We got an illegal character so we are done accumulating. - */ + string += Tcl_UtfToUniChar(string, &sch); + if (!(flags & SCAN_SUPPRESS)) { + objPtr = Tcl_NewIntObj((int)sch); + Tcl_IncrRefCount(objPtr); + objs[objIndex++] = objPtr; + } + break; - break; + case 'i': + /* + * Scan an unsigned or signed integer. + */ - addToInt: + if ((width == 0) || (width > sizeof(buf) - 1)) { + width = sizeof(buf) - 1; + } + flags |= SCAN_SIGNOK | SCAN_NODIGITS | SCAN_NOZERO; + for (end = buf; width > 0; width--) { + switch (*string) { /* - * Add the character to the temporary buffer. + * The 0 digit has special meaning at the beginning of a + * number. If we are unsure of the base, it indicates that + * we are in base 8 or base 16 (if it is followed by an + * 'x'). + * + * 8.1 - 8.3.4 incorrectly handled 0x... base-16 cases for + * %x by not reading the 0x as the auto-prelude for + * base-16. [Bug #495213] */ + case '0': + if (base == 0) { + base = 8; + flags |= SCAN_XOK; + } + if (base == 16) { + flags |= SCAN_XOK; + } + if (flags & SCAN_NOZERO) { + flags &= ~(SCAN_SIGNOK | SCAN_NODIGITS | SCAN_NOZERO); + } else { + flags &= ~(SCAN_SIGNOK | SCAN_XOK | SCAN_NODIGITS); + } + goto addToInt; - *end++ = *string++; - if (*string == '\0') { + case '1': case '2': case '3': case '4': + case '5': case '6': case '7': + if (base == 0) { + base = 10; + } + flags &= ~(SCAN_SIGNOK | SCAN_XOK | SCAN_NODIGITS); + goto addToInt; + + case '8': case '9': + if (base == 0) { + base = 10; + } + if (base <= 8) { break; } - } + flags &= ~(SCAN_SIGNOK | SCAN_XOK | SCAN_NODIGITS); + goto addToInt; + + case 'A': case 'B': case 'C': + case 'D': case 'E': case 'F': + case 'a': case 'b': case 'c': + case 'd': case 'e': case 'f': + if (base <= 10) { + break; + } + flags &= ~(SCAN_SIGNOK | SCAN_XOK | SCAN_NODIGITS); + goto addToInt; - /* - * Check to see if we need to back up because we only got a - * sign or a trailing x after a 0. - */ + case '+': case '-': + if (flags & SCAN_SIGNOK) { + flags &= ~SCAN_SIGNOK; + goto addToInt; + } + break; - if (flags & SCAN_NODIGITS) { - if (*string == '\0') { - underflow = 1; + case 'x': case 'X': + if ((flags & SCAN_XOK) && (end == buf+1)) { + base = 16; + flags &= ~SCAN_XOK; + goto addToInt; } - goto done; - } else if (end[-1] == 'x' || end[-1] == 'X') { - end--; - string--; + break; } + /* + * We got an illegal character so we are done accumulating. + */ + + break; + addToInt: /* - * Scan the value from the temporary buffer. If we are - * returning a large unsigned value, we have to convert it back - * to a string since Tcl only supports signed values. + * Add the character to the temporary buffer. */ - if (!(flags & SCAN_SUPPRESS)) { - *end = '\0'; + *end++ = *string++; + if (*string == '\0') { + break; + } + } + + /* + * Check to see if we need to back up because we only got a sign + * or a trailing x after a 0. + */ + + if (flags & SCAN_NODIGITS) { + if (*string == '\0') { + underflow = 1; + } + goto done; + } else if (end[-1] == 'x' || end[-1] == 'X') { + end--; + string--; + } + + /* + * Scan the value from the temporary buffer. If we are returning a + * large unsigned value, we have to convert it back to a string + * since Tcl only supports signed values. + */ + + if (!(flags & SCAN_SUPPRESS)) { + *end = '\0'; #ifndef TCL_WIDE_INT_IS_LONG - if (flags & SCAN_LONGER) { - wideValue = (Tcl_WideInt) (*lfn)(buf, NULL, base); - if ((flags & SCAN_UNSIGNED) && (wideValue < 0)) { - /* INTL: ISO digit */ - sprintf(buf, "%" TCL_LL_MODIFIER "u", - (Tcl_WideUInt)wideValue); - objPtr = Tcl_NewStringObj(buf, -1); - } else { - objPtr = Tcl_NewWideIntObj(wideValue); - } + if (flags & SCAN_LONGER) { + wideValue = (Tcl_WideInt) (*lfn)(buf, NULL, base); + if ((flags & SCAN_UNSIGNED) && (wideValue < 0)) { + /* INTL: ISO digit */ + sprintf(buf, "%" TCL_LL_MODIFIER "u", + (Tcl_WideUInt)wideValue); + objPtr = Tcl_NewStringObj(buf, -1); } else { + objPtr = Tcl_NewWideIntObj(wideValue); + } + } else { #endif /* !TCL_WIDE_INT_IS_LONG */ - value = (long) (*fn)(buf, NULL, base); - if ((flags & SCAN_UNSIGNED) && (value < 0)) { - sprintf(buf, "%lu", value); /* INTL: ISO digit */ - objPtr = Tcl_NewStringObj(buf, -1); - } else if ((flags & SCAN_LONGER) - || (unsigned long) value > UINT_MAX) { - objPtr = Tcl_NewLongObj(value); - } else { - objPtr = Tcl_NewIntObj(value); - } -#ifndef TCL_WIDE_INT_IS_LONG + value = (long) (*fn)(buf, NULL, base); + if ((flags & SCAN_UNSIGNED) && (value < 0)) { + sprintf(buf, "%lu", value); /* INTL: ISO digit */ + objPtr = Tcl_NewStringObj(buf, -1); + } else if ((flags & SCAN_LONGER) + || (unsigned long) value > UINT_MAX) { + objPtr = Tcl_NewLongObj(value); + } else { + objPtr = Tcl_NewIntObj(value); } -#endif - Tcl_IncrRefCount(objPtr); - objs[objIndex++] = objPtr; +#ifndef TCL_WIDE_INT_IS_LONG } +#endif + Tcl_IncrRefCount(objPtr); + objs[objIndex++] = objPtr; + } - break; + break; - case 'f': - /* - * Scan a floating point number - */ + case 'f': + /* + * Scan a floating point number + */ - if ((width == 0) || (width > sizeof(buf) - 1)) { - width = sizeof(buf) - 1; - } - flags &= ~SCAN_LONGER; - flags |= SCAN_SIGNOK | SCAN_NODIGITS | SCAN_PTOK | SCAN_EXPOK; - for (end = buf; width > 0; width--) { - switch (*string) { - case '0': case '1': case '2': case '3': - case '4': case '5': case '6': case '7': - case '8': case '9': - flags &= ~(SCAN_SIGNOK | SCAN_NODIGITS); - goto addToFloat; - case '+': case '-': - if (flags & SCAN_SIGNOK) { - flags &= ~SCAN_SIGNOK; - goto addToFloat; - } - break; - case '.': - if (flags & SCAN_PTOK) { - flags &= ~(SCAN_SIGNOK | SCAN_PTOK); - goto addToFloat; - } - break; - case 'e': case 'E': - /* - * An exponent is not allowed until there has - * been at least one digit. - */ - - if ((flags & (SCAN_NODIGITS | SCAN_EXPOK)) - == SCAN_EXPOK) { - flags = (flags & ~(SCAN_EXPOK|SCAN_PTOK)) - | SCAN_SIGNOK | SCAN_NODIGITS; - goto addToFloat; - } - break; + if ((width == 0) || (width > sizeof(buf) - 1)) { + width = sizeof(buf) - 1; + } + flags &= ~SCAN_LONGER; + flags |= SCAN_SIGNOK | SCAN_NODIGITS | SCAN_PTOK | SCAN_EXPOK; + for (end = buf; width > 0; width--) { + switch (*string) { + case '0': case '1': case '2': case '3': + case '4': case '5': case '6': case '7': + case '8': case '9': + flags &= ~(SCAN_SIGNOK | SCAN_NODIGITS); + goto addToFloat; + case '+': case '-': + if (flags & SCAN_SIGNOK) { + flags &= ~SCAN_SIGNOK; + goto addToFloat; } - - /* - * We got an illegal character so we are done accumulating. - */ - break; - - addToFloat: + case '.': + if (flags & SCAN_PTOK) { + flags &= ~(SCAN_SIGNOK | SCAN_PTOK); + goto addToFloat; + } + break; + case 'e': case 'E': /* - * Add the character to the temporary buffer. + * An exponent is not allowed until there has been at + * least one digit. */ - *end++ = *string++; - if (*string == '\0') { - break; + if ((flags & (SCAN_NODIGITS | SCAN_EXPOK)) == SCAN_EXPOK) { + flags = (flags & ~(SCAN_EXPOK|SCAN_PTOK)) + | SCAN_SIGNOK | SCAN_NODIGITS; + goto addToFloat; } + break; } /* - * Check to see if we need to back up because we saw a - * trailing 'e' or sign. + * We got an illegal character so we are done accumulating. */ - if (flags & SCAN_NODIGITS) { - if (flags & SCAN_EXPOK) { - /* - * There were no digits at all so scanning has - * failed and we are done. - */ - if (*string == '\0') { - underflow = 1; - } - goto done; - } + break; + + addToFloat: + /* + * Add the character to the temporary buffer. + */ + + *end++ = *string++; + if (*string == '\0') { + break; + } + } + /* + * Check to see if we need to back up because we saw a trailing + * 'e' or sign. + */ + + if (flags & SCAN_NODIGITS) { + if (flags & SCAN_EXPOK) { /* - * We got a bad exponent ('e' and maybe a sign). + * There were no digits at all so scanning has failed and + * we are done. */ - end--; - string--; - if (*end != 'e' && *end != 'E') { - end--; - string--; + if (*string == '\0') { + underflow = 1; } + goto done; } /* - * Scan the value from the temporary buffer. + * We got a bad exponent ('e' and maybe a sign). */ - if (!(flags & SCAN_SUPPRESS)) { - double dvalue; - *end = '\0'; - dvalue = TclStrToD(buf, NULL); - objPtr = Tcl_NewDoubleObj(dvalue); - Tcl_IncrRefCount(objPtr); - objs[objIndex++] = objPtr; + end--; + string--; + if (*end != 'e' && *end != 'E') { + end--; + string--; } - break; + } + + /* + * Scan the value from the temporary buffer. + */ + + if (!(flags & SCAN_SUPPRESS)) { + double dvalue; + + *end = '\0'; + dvalue = TclStrToD(buf, NULL); + objPtr = Tcl_NewDoubleObj(dvalue); + Tcl_IncrRefCount(objPtr); + objs[objIndex++] = objPtr; + } + break; } nconversions++; } - done: + done: result = 0; code = TCL_OK; if (numVars) { /* - * In this case, variables were specified (classic scan) + * In this case, variables were specified (classic scan). */ + for (i = 0; i < totalVars; i++) { - if (objs[i] != NULL) { - result++; - if (Tcl_ObjSetVar2(interp, objv[i+3], NULL, - objs[i], 0) == NULL) { - Tcl_AppendResult(interp, "couldn't set variable \"", - Tcl_GetString(objv[i+3]), "\"", (char *) NULL); - code = TCL_ERROR; - } - Tcl_DecrRefCount(objs[i]); + if (objs[i] == NULL) { + continue; + } + result++; + if (Tcl_ObjSetVar2(interp, objv[i+3], NULL, objs[i], 0) == NULL) { + Tcl_AppendResult(interp, "couldn't set variable \"", + TclGetString(objv[i+3]), "\"", (char *) NULL); + code = TCL_ERROR; } + Tcl_DecrRefCount(objs[i]); } } else { /* * Here no vars were specified, we want a list returned (inline scan) */ + objPtr = Tcl_NewObj(); for (i = 0; i < totalVars; i++) { if (objs[i] != NULL) { @@ -1185,9 +1191,10 @@ Tcl_ScanObjCmd(dummy, interp, objc, objv) Tcl_DecrRefCount(objs[i]); } else { /* - * More %-specifiers than matching chars, so we - * just spit out empty strings for these + * More %-specifiers than matching chars, so we just spit out + * empty strings for these. */ + Tcl_ListObjAppendElement(NULL, objPtr, Tcl_NewObj()); } } @@ -1213,3 +1220,11 @@ Tcl_ScanObjCmd(dummy, interp, objc, objv) } return code; } + +/* + * Local Variables: + * mode: c + * c-basic-offset: 4 + * fill-column: 78 + * End: + */ |