Section "3.1 Preprocessing" of the online GAS manual has a wrong reference to "Using...
[deliverable/binutils-gdb.git] / gas / atof-generic.c
index 42fcfa97a159ef57f8bf7e9c94f6093eee09bf30..6dfd61bb81df6c9030e3f2d702f5b42eb87684c8 100644 (file)
@@ -1,41 +1,38 @@
 /* atof_generic.c - turn a string of digits into a Flonum
-   Copyright (C) 1987, 1990, 1991 Free Software Foundation, Inc.
-   
+   Copyright (C) 1987-2020 Free Software Foundation, Inc.
+
    This file is part of GAS, the GNU Assembler.
-   
+
    GAS is free software; you can redistribute it and/or modify
    it under the terms of the GNU General Public License as published by
-   the Free Software Foundation; either version 2, or (at your option)
+   the Free Software Foundation; either version 3, or (at your option)
    any later version.
-   
-   GAS is distributed in the hope that it will be useful,
-   but WITHOUT ANY WARRANTY; without even the implied warranty of
-   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
-   GNU General Public License for more details.
-   
-   You should have received a copy of the GNU General Public License
-   along with GAS; see the file COPYING.  If not, write to
-   the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.  */
 
-#include <ctype.h>
-#include <string.h>
+   GAS is distributed in the hope that it will be useful, but WITHOUT
+   ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
+   or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public
+   License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with GAS; see the file COPYING.  If not, write to the Free
+   Software Foundation, 51 Franklin Street - Fifth Floor, Boston, MA
+   02110-1301, USA.  */
 
 #include "as.h"
+#include "safe-ctype.h"
 
-#ifdef __GNUC__
-#define alloca __builtin_alloca
-#else
-#ifdef sparc
-#include <alloca.h>
+#ifndef FALSE
+#define FALSE (0)
 #endif
+#ifndef TRUE
+#define TRUE  (1)
 #endif
 
-#ifdef USG
-#define bzero(s,n) memset(s,0,n)
+#ifdef TRACE
+static void flonum_print (const FLONUM_TYPE *);
 #endif
 
-/* #define     FALSE (0) */
-/* #define TRUE  (1) */
+#define ASSUME_DECIMAL_MARK_IS_DOT
 
 /***********************************************************************\
  *                                                                     *
  *             uses base (radix) 2                                     *
  *             this machine uses 2's complement binary integers        *
  *             target flonums use "      "         "       "           *
- *             target flonums exponents fit in a long          *
+ *             target flonums exponents fit in a long                  *
  *                                                                     *
  \***********************************************************************/
 
 /*
-  
+
   Syntax:
-  
-  <flonum>             ::=     <optional-sign> <decimal-number> <optional-exponent>
-  <optional-sign>              ::=     '+' | '-' | {empty}
-  <decimal-number>     ::=       <integer>
-  | <integer> <radix-character> 
-  | <integer> <radix-character> <integer> 
-  |        <radix-character> <integer>
-  <optional-exponent>  ::=     {empty} | <exponent-character> <optional-sign> <integer> 
-  <integer>            ::=     <digit> | <digit> <integer>
-  <digit>                      ::=     '0' | '1' | '2' | '3' | '4' | '5' | '6' | '7' | '8' | '9'
-  <exponent-character> ::=     {one character from "string_of_decimal_exponent_marks"}
-  <radix-character>    ::=     {one character from "string_of_decimal_marks"}
-  
+
+  <flonum> ::= <optional-sign> <decimal-number> <optional-exponent>
+  <optional-sign> ::= '+' | '-' | {empty}
+  <decimal-number> ::= <integer>
+  | <integer> <radix-character>
+  | <integer> <radix-character> <integer>
+  | <radix-character> <integer>
+
+  <optional-exponent> ::= {empty}
+  | <exponent-character> <optional-sign> <integer>
+
+  <integer> ::= <digit> | <digit> <integer>
+  <digit> ::= '0' | '1' | '2' | '3' | '4' | '5' | '6' | '7' | '8' | '9'
+  <exponent-character> ::= {one character from "string_of_decimal_exponent_marks"}
+  <radix-character> ::= {one character from "string_of_decimal_marks"}
+
   */
-\f
-int                            /* 0 if OK */
-    atof_generic (
-                 address_of_string_pointer, /* return pointer to just AFTER number we read. */
-                 string_of_decimal_marks, /* At most one per number. */
-                 string_of_decimal_exponent_marks,
-                 address_of_generic_floating_point_number)
-
-char * *               address_of_string_pointer;
-const char *   string_of_decimal_marks;
-const char *   string_of_decimal_exponent_marks;
-FLONUM_TYPE *  address_of_generic_floating_point_number;
 
+int
+atof_generic (/* return pointer to just AFTER number we read.  */
+             char **address_of_string_pointer,
+             /* At most one per number.  */
+             const char *string_of_decimal_marks,
+             const char *string_of_decimal_exponent_marks,
+             FLONUM_TYPE *address_of_generic_floating_point_number)
 {
-       
-       int                     return_value; /* 0 means OK. */
-       char *          first_digit;
-       /* char *               last_digit; JF unused */
-       int                     number_of_digits_before_decimal;
-       int                     number_of_digits_after_decimal;
-       long            decimal_exponent;
-       int                     number_of_digits_available;
-       char                    digits_sign_char;
-       \f
+  int return_value;            /* 0 means OK.  */
+  char *first_digit;
+  unsigned int number_of_digits_before_decimal;
+  unsigned int number_of_digits_after_decimal;
+  long decimal_exponent;
+  unsigned int number_of_digits_available;
+  char digits_sign_char;
+
+  /*
+   * Scan the input string, abstracting (1)digits (2)decimal mark (3) exponent.
+   * It would be simpler to modify the string, but we don't; just to be nice
+   * to caller.
+   * We need to know how many digits we have, so we can allocate space for
+   * the digits' value.
+   */
+
+  char *p;
+  char c;
+  int seen_significant_digit;
+
+#ifdef ASSUME_DECIMAL_MARK_IS_DOT
+  gas_assert (string_of_decimal_marks[0] == '.'
+         && string_of_decimal_marks[1] == 0);
+#define IS_DECIMAL_MARK(c)     ((c) == '.')
+#else
+#define IS_DECIMAL_MARK(c)     (0 != strchr (string_of_decimal_marks, (c)))
+#endif
+
+  first_digit = *address_of_string_pointer;
+  c = *first_digit;
+
+  if (c == '-' || c == '+')
+    {
+      digits_sign_char = c;
+      first_digit++;
+    }
+  else
+    digits_sign_char = '+';
+
+  switch (first_digit[0])
+    {
+    case 'n':
+    case 'N':
+      if (!strncasecmp ("nan", first_digit, 3))
        {
-               /*
-                * Scan the input string, abstracting (1)digits (2)decimal mark (3) exponent.
-                * It would be simpler to modify the string, but we don't; just to be nice
-                * to caller.
-                * We need to know how many digits we have, so we can allocate space for
-                * the digits' value.
-                */
-               
-               char *          p;
-               char            c;
-               int                     seen_significant_digit;
-               
-               first_digit = * address_of_string_pointer;
-               c= *first_digit;
-               if (c=='-' || c=='+')
-                   {
-                           digits_sign_char = c;
-                           first_digit ++;
-                   }
-               else
-                   digits_sign_char = '+';
-               
-               if(   (first_digit[0]=='n' || first_digit[0]=='N')
-                  && (first_digit[1]=='a' || first_digit[1]=='A')
-                  && (first_digit[2]=='n' || first_digit[2]=='N')) {
-                       address_of_generic_floating_point_number->sign=0;
-                       address_of_generic_floating_point_number->exponent=0;
-                       address_of_generic_floating_point_number->leader=address_of_generic_floating_point_number->low;
-                       (*address_of_string_pointer)=first_digit+3;
-                       return 0;
-               }
-               if(   (first_digit[0]=='i' || first_digit[0]=='I') 
-                  && (first_digit[1]=='n' || first_digit[1]=='N')
-                  && (first_digit[2]=='f' || first_digit[2]=='F')) {
-                       address_of_generic_floating_point_number->sign= digits_sign_char=='+' ? 'P' : 'N';
-                       address_of_generic_floating_point_number->exponent=0;
-                       address_of_generic_floating_point_number->leader=address_of_generic_floating_point_number->low;
-                       if(   (first_digit[3]=='i' || first_digit[3]=='I')
-                          && (first_digit[4]=='n' || first_digit[4]=='N')
-                          && (first_digit[5]=='i' || first_digit[5]=='I')
-                          && (first_digit[6]=='t' || first_digit[6]=='T')
-                          && (first_digit[7]=='y' || first_digit[7]=='Y'))
-                           (*address_of_string_pointer)=first_digit+8;
-                       else
-                           (*address_of_string_pointer)=first_digit+3;
-                       return 0;
-               }
-               
-               number_of_digits_before_decimal = 0;
-               number_of_digits_after_decimal = 0;
-               decimal_exponent = 0;
-               seen_significant_digit = 0;
-               for (p = first_digit;
-                    ((c = * p) != '\0')
-                    && (!c || ! strchr (string_of_decimal_marks,          c) )
-                    && (!c || ! strchr (string_of_decimal_exponent_marks, c) );
-                    p ++)
-                   {
-                           if (isdigit(c))
-                               {
-                                       if (seen_significant_digit || c > '0')
-                                           {
-                                                   number_of_digits_before_decimal ++;
-                                                   seen_significant_digit = 1;
-                                           }
-                                       else
-                                           {
-                                                   first_digit++;
-                                           }
-                               }
-                           else
-                               {
-                                       break;          /* p -> char after pre-decimal digits. */
-                               }
-                   }                           /* For each digit before decimal mark. */
-               
+         address_of_generic_floating_point_number->sign = 0;
+         address_of_generic_floating_point_number->exponent = 0;
+         address_of_generic_floating_point_number->leader =
+           address_of_generic_floating_point_number->low;
+         *address_of_string_pointer = first_digit + 3;
+         return 0;
+       }
+      break;
+
+    case 'i':
+    case 'I':
+      if (!strncasecmp ("inf", first_digit, 3))
+       {
+         address_of_generic_floating_point_number->sign =
+           digits_sign_char == '+' ? 'P' : 'N';
+         address_of_generic_floating_point_number->exponent = 0;
+         address_of_generic_floating_point_number->leader =
+           address_of_generic_floating_point_number->low;
+
+         first_digit += 3;
+         if (!strncasecmp ("inity", first_digit, 5))
+           first_digit += 5;
+
+         *address_of_string_pointer = first_digit;
+
+         return 0;
+       }
+      break;
+    }
+
+  number_of_digits_before_decimal = 0;
+  number_of_digits_after_decimal = 0;
+  decimal_exponent = 0;
+  seen_significant_digit = 0;
+  for (p = first_digit;
+       (((c = *p) != '\0')
+       && (!c || !IS_DECIMAL_MARK (c))
+       && (!c || !strchr (string_of_decimal_exponent_marks, c)));
+       p++)
+    {
+      if (ISDIGIT (c))
+       {
+         if (seen_significant_digit || c > '0')
+           {
+             ++number_of_digits_before_decimal;
+             seen_significant_digit = 1;
+           }
+         else
+           {
+             first_digit++;
+           }
+       }
+      else
+       {
+         break;                /* p -> char after pre-decimal digits.  */
+       }
+    }                          /* For each digit before decimal mark.  */
+
 #ifndef OLD_FLOAT_READS
-               /* Ignore trailing 0's after the decimal point.  The original code here
-                * (ifdef'd out) does not do this, and numbers like
-                *      4.29496729600000000000e+09      (2**31)
-                * come out inexact for some reason related to length of the digit
-                * string.
-                */
-               if ( c && strchr(string_of_decimal_marks,c) ){
-                       int zeros = 0;  /* Length of current string of zeros */
-                       
-                       for (  p++; (c = *p) && isdigit(c); p++ ){
-                               if ( c == '0'){
-                                       zeros++;
-                               } else {
-                                       number_of_digits_after_decimal += 1 + zeros;
-                                       zeros = 0;
-                               }
-                       }
+  /* Ignore trailing 0's after the decimal point.  The original code here
+     (ifdef'd out) does not do this, and numbers like
+       4.29496729600000000000e+09      (2**31)
+     come out inexact for some reason related to length of the digit
+     string.  */
+
+  /* The case number_of_digits_before_decimal = 0 is handled for
+     deleting zeros after decimal.  In this case the decimal mark and
+     the first zero digits after decimal mark are skipped.  */
+  seen_significant_digit = 0;
+  signed long subtract_decimal_exponent = 0;
+
+  if (c && IS_DECIMAL_MARK (c))
+    {
+      unsigned int zeros = 0;  /* Length of current string of zeros.  */
+
+      if (number_of_digits_before_decimal == 0)
+       /* Skip decimal mark.  */
+       first_digit++;
+
+      for (p++; (c = *p) && ISDIGIT (c); p++)
+       {
+         if (c == '0')
+           {
+             if (number_of_digits_before_decimal == 0
+                 && !seen_significant_digit)
+               {
+                 /* Skip '0' and the decimal mark.  */
+                 first_digit++;
+                 subtract_decimal_exponent--;
                }
+             else
+               zeros++;
+           }
+         else
+           {
+             seen_significant_digit = 1;
+             number_of_digits_after_decimal += 1 + zeros;
+             zeros = 0;
+           }
+       }
+    }
 #else
-               if (c && strchr (string_of_decimal_marks, c))
-                   {
-                           for (p ++;
-                                ((c = * p) != '\0')
-                                && (!c || ! strchr (string_of_decimal_exponent_marks, c) );
-                                p ++)
-                               {
-                                       if (isdigit(c))
-                                           {
-                                                   number_of_digits_after_decimal ++; /* This may be retracted below. */
-                                                   if (/* seen_significant_digit || */ c > '0')
-                                                       {
-                                                               seen_significant_digit = TRUE;
-                                                       }
-                                           }
-                                       else
-                                           {
-                                                   if ( ! seen_significant_digit)
-                                                       {
-                                                               number_of_digits_after_decimal = 0;
-                                                       }
-                                                   break;
-                                           }
-                               }                       /* For each digit after decimal mark. */
-                   }
-               while(number_of_digits_after_decimal && first_digit[number_of_digits_before_decimal+number_of_digits_after_decimal]=='0')
-                   --number_of_digits_after_decimal;
-               /*    last_digit = p; JF unused */
+  if (c && IS_DECIMAL_MARK (c))
+    {
+      for (p++;
+          (((c = *p) != '\0')
+           && (!c || !strchr (string_of_decimal_exponent_marks, c)));
+          p++)
+       {
+         if (ISDIGIT (c))
+           {
+             /* This may be retracted below.  */
+             number_of_digits_after_decimal++;
+
+             if ( /* seen_significant_digit || */ c > '0')
+               {
+                 seen_significant_digit = TRUE;
+               }
+           }
+         else
+           {
+             if (!seen_significant_digit)
+               {
+                 number_of_digits_after_decimal = 0;
+               }
+             break;
+           }
+       }                       /* For each digit after decimal mark.  */
+    }
+
+  while (number_of_digits_after_decimal
+        && first_digit[number_of_digits_before_decimal
+                       + number_of_digits_after_decimal] == '0')
+    --number_of_digits_after_decimal;
 #endif
-               
-               if (c && strchr (string_of_decimal_exponent_marks, c) )
-                   {
-                           char                digits_exponent_sign_char;
-                           
-                           c = * ++ p;
-                           if (c && strchr ("+-",c))
-                               {
-                                       digits_exponent_sign_char = c;
-                                       c = * ++ p;
-                               }
-                           else
-                               {
-                                       digits_exponent_sign_char = '+';
-                               }
-                           for (;
-                                (c);
-                                c = * ++ p)
-                               {
-                                       if (isdigit(c))
-                                           {
-                                                   decimal_exponent = decimal_exponent * 10 + c - '0';
-                                                   /*
-                                                    * BUG! If we overflow here, we lose!
-                                                    */
-                                           }
-                                       else
-                                           {
-                                                   break;
-                                           }
-                               }
-                           if (digits_exponent_sign_char == '-')
-                               {
-                                       decimal_exponent = - decimal_exponent;
-                               }
-                   }
-               * address_of_string_pointer = p;
+
+  if (flag_m68k_mri)
+    {
+      while (c == '_')
+       c = *++p;
+    }
+  if (c && strchr (string_of_decimal_exponent_marks, c))
+    {
+      char digits_exponent_sign_char;
+
+      c = *++p;
+      if (flag_m68k_mri)
+       {
+         while (c == '_')
+           c = *++p;
        }
-       \f
-       number_of_digits_available =
-           number_of_digits_before_decimal
-               + number_of_digits_after_decimal;
-       return_value = 0;
-       if (number_of_digits_available == 0)
+      if (c && strchr ("+-", c))
+       {
+         digits_exponent_sign_char = c;
+         c = *++p;
+       }
+      else
+       {
+         digits_exponent_sign_char = '+';
+       }
+
+      for (; (c); c = *++p)
+       {
+         if (ISDIGIT (c))
            {
-                   address_of_generic_floating_point_number -> exponent = 0;   /* Not strictly necessary */
-                   address_of_generic_floating_point_number -> leader
-                       = -1 + address_of_generic_floating_point_number -> low;
-                   address_of_generic_floating_point_number -> sign = digits_sign_char;
-                   /* We have just concocted (+/-)0.0E0 */
+             decimal_exponent = decimal_exponent * 10 + c - '0';
+             /*
+              * BUG! If we overflow here, we lose!
+              */
            }
-       else
+         else
            {
-                   LITTLENUM_TYPE *    digits_binary_low;
-                   int         precision;
-                   int         maximum_useful_digits;
-                   int         number_of_digits_to_use;
-                   int         more_than_enough_bits_for_digits;
-                   int         more_than_enough_littlenums_for_digits;
-                   int         size_of_digits_in_littlenums;
-                   int         size_of_digits_in_chars;
-                   FLONUM_TYPE power_of_10_flonum;
-                   FLONUM_TYPE digits_flonum;
-                   
-                   
-                   precision = (address_of_generic_floating_point_number -> high
-                                - address_of_generic_floating_point_number -> low
-                                + 1
-                                );             /* Number of destination littlenums. */
-                   /* Includes guard bits (two littlenums worth) */
-                   maximum_useful_digits = (  ((double) (precision - 2))
-                                            * ((double) (LITTLENUM_NUMBER_OF_BITS))
-                                            / (LOG_TO_BASE_2_OF_10)
-                                            )
-                       + 2;                    /* 2 :: guard digits. */
-                   if (number_of_digits_available > maximum_useful_digits)
-                       {
-                               number_of_digits_to_use = maximum_useful_digits;
-                       }
-                   else
-                       {
-                               number_of_digits_to_use = number_of_digits_available;
-                       }
-                   decimal_exponent += number_of_digits_before_decimal - number_of_digits_to_use;
-                   
-                   more_than_enough_bits_for_digits
-                       = ((((double)number_of_digits_to_use) * LOG_TO_BASE_2_OF_10) + 1);
-                   more_than_enough_littlenums_for_digits
-                       = (  more_than_enough_bits_for_digits
-                          / LITTLENUM_NUMBER_OF_BITS
-                          )
-                           + 2;
-                   
-                   /*
-                    * Compute (digits) part. In "12.34E56" this is the "1234" part.
-                    * Arithmetic is exact here. If no digits are supplied then
-                    * this part is a 0 valued binary integer.
-                    * Allocate room to build up the binary number as littlenums.
-                    * We want this memory to disappear when we leave this function.
-                    * Assume no alignment problems => (room for n objects) ==
-                    * n * (room for 1 object).
-                    */
-                   
-                   size_of_digits_in_littlenums = more_than_enough_littlenums_for_digits;
-                   size_of_digits_in_chars = size_of_digits_in_littlenums
-                       * sizeof( LITTLENUM_TYPE );
-                   digits_binary_low = (LITTLENUM_TYPE *)
-                       alloca (size_of_digits_in_chars);
-                   bzero ((char *)digits_binary_low, size_of_digits_in_chars);
-                   
-                   /* Digits_binary_low[] is allocated and zeroed. */
-                   
+             break;
+           }
+       }
+
+      if (digits_exponent_sign_char == '-')
+       {
+         decimal_exponent = -decimal_exponent;
+       }
+    }
+
+#ifndef OLD_FLOAT_READS
+  /* Subtract_decimal_exponent != 0 when number_of_digits_before_decimal = 0
+     and first digit after decimal is '0'.  */
+  decimal_exponent += subtract_decimal_exponent;
+#endif
+
+  *address_of_string_pointer = p;
+
+  number_of_digits_available =
+    number_of_digits_before_decimal + number_of_digits_after_decimal;
+  return_value = 0;
+  if (number_of_digits_available == 0)
+    {
+      address_of_generic_floating_point_number->exponent = 0;  /* Not strictly necessary */
+      address_of_generic_floating_point_number->leader
+       = -1 + address_of_generic_floating_point_number->low;
+      address_of_generic_floating_point_number->sign = digits_sign_char;
+      /* We have just concocted (+/-)0.0E0 */
+
+    }
+  else
+    {
+      int count;               /* Number of useful digits left to scan.  */
+
+      LITTLENUM_TYPE *temporary_binary_low = NULL;
+      LITTLENUM_TYPE *power_binary_low = NULL;
+      LITTLENUM_TYPE *digits_binary_low;
+      unsigned int precision;
+      unsigned int maximum_useful_digits;
+      unsigned int number_of_digits_to_use;
+      unsigned int more_than_enough_bits_for_digits;
+      unsigned int more_than_enough_littlenums_for_digits;
+      unsigned int size_of_digits_in_littlenums;
+      unsigned int size_of_digits_in_chars;
+      FLONUM_TYPE power_of_10_flonum;
+      FLONUM_TYPE digits_flonum;
+
+      precision = (address_of_generic_floating_point_number->high
+                  - address_of_generic_floating_point_number->low
+                  + 1);        /* Number of destination littlenums.  */
+
+      /* precision includes two littlenums worth of guard bits,
+        so this gives us 10 decimal guard digits here.  */
+      maximum_useful_digits = (precision
+                              * LITTLENUM_NUMBER_OF_BITS
+                              * 1000000 / 3321928
+                              + 1);    /* round up.  */
+
+      if (number_of_digits_available > maximum_useful_digits)
+       {
+         number_of_digits_to_use = maximum_useful_digits;
+       }
+      else
+       {
+         number_of_digits_to_use = number_of_digits_available;
+       }
+
+      /* Cast these to SIGNED LONG first, otherwise, on systems with
+        LONG wider than INT (such as Alpha OSF/1), unsignedness may
+        cause unexpected results.  */
+      decimal_exponent += ((long) number_of_digits_before_decimal
+                          - (long) number_of_digits_to_use);
+
+      more_than_enough_bits_for_digits
+       = (number_of_digits_to_use * 3321928 / 1000000 + 1);
+
+      more_than_enough_littlenums_for_digits
+       = (more_than_enough_bits_for_digits
+          / LITTLENUM_NUMBER_OF_BITS)
+       + 2;
+
+      /* Compute (digits) part. In "12.34E56" this is the "1234" part.
+        Arithmetic is exact here. If no digits are supplied then this
+        part is a 0 valued binary integer.  Allocate room to build up
+        the binary number as littlenums.  We want this memory to
+        disappear when we leave this function.  Assume no alignment
+        problems => (room for n objects) == n * (room for 1
+        object).  */
+
+      size_of_digits_in_littlenums = more_than_enough_littlenums_for_digits;
+      size_of_digits_in_chars = size_of_digits_in_littlenums
+       * sizeof (LITTLENUM_TYPE);
+
+      digits_binary_low = (LITTLENUM_TYPE *)
+       xmalloc (size_of_digits_in_chars);
+
+      memset ((char *) digits_binary_low, '\0', size_of_digits_in_chars);
+
+      /* Digits_binary_low[] is allocated and zeroed.  */
+
+      /*
+       * Parse the decimal digits as if * digits_low was in the units position.
+       * Emit a binary number into digits_binary_low[].
+       *
+       * Use a large-precision version of:
+       * (((1st-digit) * 10 + 2nd-digit) * 10 + 3rd-digit ...) * 10 + last-digit
+       */
+
+      for (p = first_digit, count = number_of_digits_to_use; count; p++, --count)
+       {
+         c = *p;
+         if (ISDIGIT (c))
+           {
+             /*
+              * Multiply by 10. Assume can never overflow.
+              * Add this digit to digits_binary_low[].
+              */
+
+             long carry;
+             LITTLENUM_TYPE *littlenum_pointer;
+             LITTLENUM_TYPE *littlenum_limit;
+
+             littlenum_limit = digits_binary_low
+               + more_than_enough_littlenums_for_digits
+               - 1;
+
+             carry = c - '0';  /* char -> binary */
+
+             for (littlenum_pointer = digits_binary_low;
+                  littlenum_pointer <= littlenum_limit;
+                  littlenum_pointer++)
+               {
+                 long work;
+
+                 work = carry + 10 * (long) (*littlenum_pointer);
+                 *littlenum_pointer = work & LITTLENUM_MASK;
+                 carry = work >> LITTLENUM_NUMBER_OF_BITS;
+               }
+
+             if (carry != 0)
+               {
+                 /*
+                  * We have a GROSS internal error.
+                  * This should never happen.
+                  */
+                 as_fatal (_("failed sanity check"));
+               }
+           }
+         else
+           {
+             ++count;          /* '.' doesn't alter digits used count.  */
+           }
+       }
+
+      /*
+       * Digits_binary_low[] properly encodes the value of the digits.
+       * Forget about any high-order littlenums that are 0.
+       */
+      while (digits_binary_low[size_of_digits_in_littlenums - 1] == 0
+            && size_of_digits_in_littlenums >= 2)
+       size_of_digits_in_littlenums--;
+
+      digits_flonum.low = digits_binary_low;
+      digits_flonum.high = digits_binary_low + size_of_digits_in_littlenums - 1;
+      digits_flonum.leader = digits_flonum.high;
+      digits_flonum.exponent = 0;
+      /*
+       * The value of digits_flonum . sign should not be important.
+       * We have already decided the output's sign.
+       * We trust that the sign won't influence the other parts of the number!
+       * So we give it a value for these reasons:
+       * (1) courtesy to humans reading/debugging
+       *     these numbers so they don't get excited about strange values
+       * (2) in future there may be more meaning attached to sign,
+       *     and what was
+       *     harmless noise may become disruptive, ill-conditioned (or worse)
+       *     input.
+       */
+      digits_flonum.sign = '+';
+
+      {
+       /*
+        * Compute the mantissa (& exponent) of the power of 10.
+        * If successful, then multiply the power of 10 by the digits
+        * giving return_binary_mantissa and return_binary_exponent.
+        */
+
+       int decimal_exponent_is_negative;
+       /* This refers to the "-56" in "12.34E-56".  */
+       /* FALSE: decimal_exponent is positive (or 0) */
+       /* TRUE:  decimal_exponent is negative */
+       FLONUM_TYPE temporary_flonum;
+       unsigned int size_of_power_in_littlenums;
+       unsigned int size_of_power_in_chars;
+
+       size_of_power_in_littlenums = precision;
+       /* Precision has a built-in fudge factor so we get a few guard bits.  */
+
+       decimal_exponent_is_negative = decimal_exponent < 0;
+       if (decimal_exponent_is_negative)
+         {
+           decimal_exponent = -decimal_exponent;
+         }
+
+       /* From now on: the decimal exponent is > 0. Its sign is separate.  */
+
+       size_of_power_in_chars = size_of_power_in_littlenums
+         * sizeof (LITTLENUM_TYPE) + 2;
+
+       power_binary_low = (LITTLENUM_TYPE *) xmalloc (size_of_power_in_chars);
+       temporary_binary_low = (LITTLENUM_TYPE *) xmalloc (size_of_power_in_chars);
+
+       memset ((char *) power_binary_low, '\0', size_of_power_in_chars);
+       *power_binary_low = 1;
+       power_of_10_flonum.exponent = 0;
+       power_of_10_flonum.low = power_binary_low;
+       power_of_10_flonum.leader = power_binary_low;
+       power_of_10_flonum.high = power_binary_low + size_of_power_in_littlenums - 1;
+       power_of_10_flonum.sign = '+';
+       temporary_flonum.low = temporary_binary_low;
+       temporary_flonum.high = temporary_binary_low + size_of_power_in_littlenums - 1;
+       /*
+        * (power) == 1.
+        * Space for temporary_flonum allocated.
+        */
+
+       /*
+        * ...
+        *
+        * WHILE        more bits
+        * DO   find next bit (with place value)
+        *      multiply into power mantissa
+        * OD
+        */
+       {
+         int place_number_limit;
+         /* Any 10^(2^n) whose "n" exceeds this */
+         /* value will fall off the end of */
+         /* flonum_XXXX_powers_of_ten[].  */
+         int place_number;
+         const FLONUM_TYPE *multiplicand;      /* -> 10^(2^n) */
+
+         place_number_limit = table_size_of_flonum_powers_of_ten;
+
+         multiplicand = (decimal_exponent_is_negative
+                         ? flonum_negative_powers_of_ten
+                         : flonum_positive_powers_of_ten);
+
+         for (place_number = 1;/* Place value of this bit of exponent.  */
+              decimal_exponent;/* Quit when no more 1 bits in exponent.  */
+              decimal_exponent >>= 1, place_number++)
+           {
+             if (decimal_exponent & 1)
+               {
+                 if (place_number > place_number_limit)
                    {
-                           /*
-                            * Parse the decimal digits as if * digits_low was in the units position.
-                            * Emit a binary number into digits_binary_low[].
-                            *
-                            * Use a large-precision version of:
-                            * (((1st-digit) * 10 + 2nd-digit) * 10 + 3rd-digit ...) * 10 + last-digit
-                            */
-                           
-                           char *              p;
-                           char                c;
-                           int         count;  /* Number of useful digits left to scan. */
-                           
-                           for (p = first_digit, count = number_of_digits_to_use;
-                                count;
-                                p ++,  -- count)
-                               {
-                                       c = * p;
-                                       if (isdigit(c))
-                                           {
-                                                   /*
-                                                    * Multiply by 10. Assume can never overflow.
-                                                    * Add this digit to digits_binary_low[].
-                                                    */
-                                                   
-                                                   long        carry;
-                                                   LITTLENUM_TYPE *    littlenum_pointer;
-                                                   LITTLENUM_TYPE *    littlenum_limit;
-                                                   
-                                                   littlenum_limit
-                                                       =     digits_binary_low
-                                                           +   more_than_enough_littlenums_for_digits
-                                                               - 1;
-                                                   carry = c - '0';    /* char -> binary */
-                                                   for (littlenum_pointer = digits_binary_low;
-                                                        littlenum_pointer <= littlenum_limit;
-                                                        littlenum_pointer ++)
-                                                       {
-                                                               long    work;
-                                                               
-                                                               work = carry + 10 * (long)(*littlenum_pointer);
-                                                               * littlenum_pointer = work & LITTLENUM_MASK;
-                                                               carry = work >> LITTLENUM_NUMBER_OF_BITS;
-                                                       }
-                                                   if (carry != 0)
-                                                       {
-                                                               /*
-                                                                * We have a GROSS internal error.
-                                                                * This should never happen.
-                                                                */
-                                                               as_fatal("failed sanity check.");       /* RMS prefers abort() to any message. */
-                                                       }
-                                           }
-                                       else
-                                           {
-                                                   ++ count;   /* '.' doesn't alter digits used count. */
-                                           }           /* if valid digit */
-                               }                       /* for each digit */
+                     /* The decimal exponent has a magnitude so great
+                        that our tables can't help us fragment it.
+                        Although this routine is in error because it
+                        can't imagine a number that big, signal an
+                        error as if it is the user's fault for
+                        presenting such a big number.  */
+                     return_value = ERROR_EXPONENT_OVERFLOW;
+                     /* quit out of loop gracefully */
+                     decimal_exponent = 0;
                    }
-                   
-                   /*
-                    * Digits_binary_low[] properly encodes the value of the digits.
-                    * Forget about any high-order littlenums that are 0.
-                    */
-                   while (digits_binary_low [size_of_digits_in_littlenums - 1] == 0
-                          && size_of_digits_in_littlenums >= 2)
-                       size_of_digits_in_littlenums --;
-                   
-                   digits_flonum . low = digits_binary_low;
-                   digits_flonum . high        = digits_binary_low + size_of_digits_in_littlenums - 1;
-                   digits_flonum . leader      = digits_flonum . high;
-                   digits_flonum . exponent    = 0;
-                   /*
-                    * The value of digits_flonum . sign should not be important.
-                    * We have already decided the output's sign.
-                    * We trust that the sign won't influence the other parts of the number!
-                    * So we give it a value for these reasons:
-                    * (1) courtesy to humans reading/debugging
-                    *     these numbers so they don't get excited about strange values
-                    * (2) in future there may be more meaning attached to sign,
-                    *     and what was
-                    *     harmless noise may become disruptive, ill-conditioned (or worse)
-                    *     input.
-                    */
-                   digits_flonum . sign        = '+';
-                   
+                 else
                    {
-                           /*
-                            * Compute the mantssa (& exponent) of the power of 10.
-                            * If sucessful, then multiply the power of 10 by the digits
-                            * giving return_binary_mantissa and return_binary_exponent.
-                            */
-                           
-                           LITTLENUM_TYPE *power_binary_low;
-                           int         decimal_exponent_is_negative;
-                           /* This refers to the "-56" in "12.34E-56". */
-                           /* FALSE: decimal_exponent is positive (or 0) */
-                           /* TRUE:  decimal_exponent is negative */
-                           FLONUM_TYPE temporary_flonum;
-                           LITTLENUM_TYPE *temporary_binary_low;
-                           int         size_of_power_in_littlenums;
-                           int         size_of_power_in_chars;
-                           
-                           size_of_power_in_littlenums = precision;
-                           /* Precision has a built-in fudge factor so we get a few guard bits. */
-                           
-                           
-                           decimal_exponent_is_negative = decimal_exponent < 0;
-                           if (decimal_exponent_is_negative)
-                               {
-                                       decimal_exponent = - decimal_exponent;
-                               }
-                           /* From now on: the decimal exponent is > 0. Its sign is seperate. */
-                           
-                           size_of_power_in_chars
-                               =   size_of_power_in_littlenums
-                                   * sizeof( LITTLENUM_TYPE ) + 2;
-                           power_binary_low = (LITTLENUM_TYPE *) alloca ( size_of_power_in_chars );
-                           temporary_binary_low = (LITTLENUM_TYPE *) alloca ( size_of_power_in_chars );
-                           bzero ((char *)power_binary_low, size_of_power_in_chars);
-                           * power_binary_low = 1;
-                           power_of_10_flonum . exponent       = 0;
-                           power_of_10_flonum . low    = power_binary_low;
-                           power_of_10_flonum . leader = power_binary_low;
-                           power_of_10_flonum . high   = power_binary_low      + size_of_power_in_littlenums - 1;
-                           power_of_10_flonum . sign   = '+';
-                           temporary_flonum . low      = temporary_binary_low;
-                           temporary_flonum . high     = temporary_binary_low          + size_of_power_in_littlenums - 1;
-                           /*
-                            * (power) == 1.
-                            * Space for temporary_flonum allocated.
-                            */
-                           
-                           /*
-                            * ...
-                            *
-                            * WHILE    more bits
-                            * DO       find next bit (with place value)
-                            *  multiply into power mantissa
-                            * OD
-                            */
-                           {
-                                   int         place_number_limit;
-                                   /* Any 10^(2^n) whose "n" exceeds this */
-                                   /* value will fall off the end of */
-                                   /* flonum_XXXX_powers_of_ten[]. */
-                                   int         place_number;
-                                   const FLONUM_TYPE * multiplicand; /* -> 10^(2^n) */
-                                   
-                                   place_number_limit = table_size_of_flonum_powers_of_ten;
-                                   multiplicand
-                                       = (  decimal_exponent_is_negative
-                                          ? flonum_negative_powers_of_ten
-                                          : flonum_positive_powers_of_ten);
-                                   for (place_number = 1;      /* Place value of this bit of exponent. */
-                                        decimal_exponent;      /* Quit when no more 1 bits in exponent. */
-                                        decimal_exponent >>= 1
-                                        , place_number ++)
-                                       {
-                                               if (decimal_exponent & 1)
-                                                   {
-                                                           if (place_number > place_number_limit)
-                                                               {
-                                                                       /*
-                                                                        * The decimal exponent has a magnitude so great that
-                                                                        * our tables can't help us fragment it.  Although this
-                                                                        * routine is in error because it can't imagine a
-                                                                        * number that big, signal an error as if it is the
-                                                                        * user's fault for presenting such a big number.
-                                                                        */
-                                                                       return_value = ERROR_EXPONENT_OVERFLOW;
-                                                                       /*
-                                                                        * quit out of loop gracefully
-                                                                        */
-                                                                       decimal_exponent = 0;
-                                                               }
-                                                           else
-                                                               {
 #ifdef TRACE
-                                                                       printf("before multiply, place_number = %d., power_of_10_flonum:\n", place_number);
-                                                                       flonum_print( & power_of_10_flonum );
-                                                                       (void)putchar('\n');
+                     printf ("before multiply, place_number = %d., power_of_10_flonum:\n",
+                             place_number);
+
+                     flonum_print (&power_of_10_flonum);
+                     (void) putchar ('\n');
 #endif
-                                                                       flonum_multip(multiplicand + place_number, &power_of_10_flonum, &temporary_flonum);
-                                                                       flonum_copy (& temporary_flonum, & power_of_10_flonum);
-                                                               }               /* If this bit of decimal_exponent was computable.*/
-                                                   }                   /* If this bit of decimal_exponent was set. */
-                                       }                       /* For each bit of binary representation of exponent */
 #ifdef TRACE
-                                   printf( " after computing power_of_10_flonum: " );
-                                   flonum_print( & power_of_10_flonum );
-                                   (void)putchar('\n');
+                     printf ("multiplier:\n");
+                     flonum_print (multiplicand + place_number);
+                     (void) putchar ('\n');
+#endif
+                     flonum_multip (multiplicand + place_number,
+                                    &power_of_10_flonum, &temporary_flonum);
+#ifdef TRACE
+                     printf ("after multiply:\n");
+                     flonum_print (&temporary_flonum);
+                     (void) putchar ('\n');
+#endif
+                     flonum_copy (&temporary_flonum, &power_of_10_flonum);
+#ifdef TRACE
+                     printf ("after copy:\n");
+                     flonum_print (&power_of_10_flonum);
+                     (void) putchar ('\n');
+#endif
+                   } /* If this bit of decimal_exponent was computable.*/
+               } /* If this bit of decimal_exponent was set.  */
+           } /* For each bit of binary representation of exponent */
+#ifdef TRACE
+         printf ("after computing power_of_10_flonum:\n");
+         flonum_print (&power_of_10_flonum);
+         (void) putchar ('\n');
+#endif
+       }
+      }
+
+      /*
+       * power_of_10_flonum is power of ten in binary (mantissa) , (exponent).
+       * It may be the number 1, in which case we don't NEED to multiply.
+       *
+       * Multiply (decimal digits) by power_of_10_flonum.
+       */
+
+      flonum_multip (&power_of_10_flonum, &digits_flonum, address_of_generic_floating_point_number);
+      /* Assert sign of the number we made is '+'.  */
+      address_of_generic_floating_point_number->sign = digits_sign_char;
+
+      if (temporary_binary_low)
+       free (temporary_binary_low);
+      if (power_binary_low)
+       free (power_binary_low);
+      free (digits_binary_low);
+    }
+  return return_value;
+}
+
+#ifdef TRACE
+static void
+flonum_print (f)
+     const FLONUM_TYPE *f;
+{
+  LITTLENUM_TYPE *lp;
+  char littlenum_format[10];
+  sprintf (littlenum_format, " %%0%dx", sizeof (LITTLENUM_TYPE) * 2);
+#define print_littlenum(LP)    (printf (littlenum_format, LP))
+  printf ("flonum @%p %c e%ld", f, f->sign, f->exponent);
+  if (f->low < f->high)
+    for (lp = f->high; lp >= f->low; lp--)
+      print_littlenum (*lp);
+  else
+    for (lp = f->low; lp <= f->high; lp++)
+      print_littlenum (*lp);
+  printf ("\n");
+  fflush (stdout);
+}
 #endif
-                           }
-                           
-                   }
-                   
-                   /*
-                    * power_of_10_flonum is power of ten in binary (mantissa) , (exponent).
-                    * It may be the number 1, in which case we don't NEED to multiply.
-                    *
-                    * Multiply (decimal digits) by power_of_10_flonum.
-                    */
-                   
-                   flonum_multip (& power_of_10_flonum, & digits_flonum, address_of_generic_floating_point_number);
-                   /* Assert sign of the number we made is '+'. */
-                   address_of_generic_floating_point_number -> sign = digits_sign_char;
-                   
-           }                           /* If we had any significant digits. */
-       return (return_value);
-} /* atof_generic () */
 
 /* end of atof_generic.c */
This page took 0.03525 seconds and 4 git commands to generate.