VMS Help RTL Routines, OTS$, OTS$CVT T x  *Conan The Librarian (sorry for the slow response - running on an old VAX)

    The Convert Numeric Text to D-, F-, G-, or H-Floating routines
    convert an ASCII text string representation of a numeric value to
    a D-floating, F-floating, G-floating, or H-floating value.

    Format

      OTS$CVT_T_D  fixed-or-dynamic-input-string

                   ,floating-point-value [,digits-in-fraction]

                   [,scale-factor] [,flags-value] [,extension-bits]

      OTS$CVT_T_F  fixed-or-dynamic-input-string

                   ,floating-point-value [,digits-in-fraction]

                   [,scale-factor] [,flags-value] [,extension-bits]

      OTS$CVT_T_G  fixed-or-dynamic-input-string

                   ,floating-point-value [,digits-in-fraction]

                   [,scale-factor] [,flags-value] [,extension-bits]

      OTS$CVT_T_H  fixed-or-dynamic-input-string

                   ,floating-point-value [,digits-in-fraction]

                   [,scale-factor] [,flags-value] [,extension-bits]

1 - Returns

    OpenVMS usage:cond_value
    type:         longword (unsigned)
    access:       write only
    mechanism:    by value

2 - Arguments

 fixed-or-dynamic-input-string

    OpenVMS usage:char_string
    type:         character string
    access:       read only
    mechanism:    by descriptor, fixed-length or dynamic string

    Input string containing an ASCII text string representation of
    a numeric value that OTS$CVT_T_x converts to a D-floating, F-
    floating, G-floating, or H-floating value. The fixed-or-dynamic-
    input-string argument is the address of a descriptor pointing to
    the input string.

    The syntax of a valid input string is as follows:

    [blanks][+ or -][digits][.][digits][+ or -][digits]

       or

    [blanks][+ or -][digits][.][digits][ X ][digits]

       where X = {E e D d Q q}[blanks][+ or -]

    E, e, D, d, Q, and q are the possible exponent letters. They are
    semantically equivalent. Other elements in the preceding syntax
    are defined as follows:

    Term           Description

    blanks         One or more blanks
    digits         One or more decimal digits

 floating-point-value

    OpenVMS usage:floating_point
    type:         D_floating, F_floating, G_floating, H_floating
    access:       write only
    mechanism:    by reference

    Floating-point value that OTS$CVT_T_x creates when it converts
    the input string. The floating-point-value argument is the
    address of the floating-point value. The data type of floating-
    point-value depends on the called routine as shown in the
    following table:

    Routine          floating-point-value Data Type

    OTS$CVT_T_D      D-floating
    OTS$CVT_T_F      F-floating
    OTS$CVT_T_G      G-floating
    OTS$CVT_T_H      H-floating

 digits-in-fraction

    OpenVMS usage:longword_unsigned
    type:         longword (unsigned)
    access:       read only
    mechanism:    by value

    Number of digits in the fraction if no decimal point is included
    in the input string. The digits-in-fraction argument contains the
    number of digits. If the number of digits is omitted, the default
    is zero.

 scale-factor

    OpenVMS usage:longword_signed
    type:         longword (signed)
    access:       read only
    mechanism:    by value

    Scale factor. The scale-factor argument contains the value of
    the scale factor. If bit 6 of the flags-value argument is clear,
    the resultant value is divided by 10**scale-factor unless the
    exponent is present. If bit 6 of flags-value is set, the scale
    factor is always applied. If the scale factor is omitted, the
    default is zero.

 flags-value

    OpenVMS usage:mask_longword
    type:         longword (unsigned)
    access:       read only
    mechanism:    by value

    User-supplied flags. The flags-value argument contains the user-
    supplied flags described in the following table:

    Bit  Action if Set         Action if Clear

    0    Ignore blanks.        Interpret blanks as zeros.
    1    Allow only E or e     Allow E, e, D, d, Q and q exponents.
         exponents. (This      (This is consistent with BASIC
         is consistent with    semantics.)
         Fortran semantics.)
    2    Interpret an          Do not interpret an underflow as an
         underflow as an       error.
         error.
    3    Truncate the value.   Round the value.
    4    Ignore tabs.          Interpret tabs as invalid characters.
    5    An exponent must      The exponent letter can be omitted.
         begin with a valid
         exponent letter.
    6    Always apply the      Apply the scale factor only if there
         scale factor.         is no exponent present in the string.

    If you omit the flags-value argument, OTS$CVT_T_x defaults all
    flags to clear.

 extension-bits (D-, F-floating)

    OpenVMS usage:byte_unsigned
    type:         byte (unsigned)
    access:       write only
    mechanism:    by reference

 extension-bits (G-, H-floating)

    OpenVMS usage:word_unsigned
    type:         word (unsigned)
    access:       write only
    mechanism:    by reference

    Extra precision bits. The extension-bits argument is the address
    of a word containing the extra precision bits. If extension-
    bits is present, floating-point-value is not rounded, and the
    first n bits after truncation are returned left-justified in this
    argument, as follows:

                     Number
                     of Bits
    Routine          Returned  Data Type

    OTS$CVT_T_D      8         Byte (unsigned)
    OTS$CVT_T_F      8         Byte (unsigned)
    OTS$CVT_T_G      11        Word (unsigned)
    OTS$CVT_T_H      15        Word (unsigned)

    A value represented by extension bits is suitable for use as the
    extension operand in an EMOD instruction.

    The extra precision bits returned for H-floating may not be
    precise because OTS$CVT_T_H carries its calculations to only
    128 bits. However the error should be small.