PFE 0.33.70

generated
(C) Guido U. Draheim
guidod@gmx.de

FLOATING-Misc

Compatibility words

---

FLIT  => "FORTH"

(no description)

primitive code = [p4_f_literal_execution]

---

FP@( -- addr )  => "FORTH"

returns the floating point stack pointer

primitive code = [p4_f_p_fetch]

---

FP!( addr -- )  => "FORTH"

sets the floating point stack pointer - this is the inverse of FP@

primitive code = [p4_f_p_store]

---

F=  => "FORTH"

(no description)

primitive code = [p4_f_equal]

---

F<>( f: a b -- s: a!=b )  => "FORTH"

primitive code = [p4_f_not_equal]

---

F>  => "FORTH"

(no description)

primitive code = [p4_f_greater_than]

---

F<=  => "FORTH"

(no description)

primitive code = [p4_f_less_than_or_equal]

---

F>=  => "FORTH"

(no description)

primitive code = [p4_f_greater_than_or_equal]

---

S>F( n -- f: x )  => "FORTH"

it's inverse is F>S - convert a cell parameter to floating-point.

primitive code = [p4_s_to_f]

---

FTRUNC>S(f: x -- s: n )  => "FORTH"

The word F>S was sometimes defined with a different behavior than FTRUNC>S which is the type-cast behaviour of C according to C99 section 6.3.1.4 - truncation would also match the ANS-Forth specification for F>D.

Some systems used F>S defined to FROUND>S instead. The pfe provides explicit words for both conversions, the word FROUND>S and FTRUNC>S which return single-cell parameters for a floating point number with the conversion method of FTRUNC or FROUND.

In PFE, F>S is a synonym pointing to FTRUNC>S in analogy of the behavior of F>D where no explicit word exists. The inverse of F>S is the cast conversion of S>F.

primitive code = [p4_f_trunc_to_s]

---

FROUND>S(f: x -- s: n)  => "FORTH"

complements FTRUNC>S for applications that expect F>S to be defined with a rounding behavior like

  : FROUND>S FROUND FTRUNC>S ;
  

primitive code = [p4_f_round_to_s]

---

FTRUNC(f: x -- x' )  => "FORTH"

truncate towards zero, discard a fractional part. See also FTRUNC>S conversion and the FROUND and FLOOR adaptors.

  : FTRUNC FDUP F0< IF FCEIL ELSE FLOOR THEN ;

(When available, uses a single call to C99 trunc() internally)

primitive code = [p4_f_trunc]

---

-FROT(f: x1 x2 x3 -- x3 x1 x2 )  => "FORTH"

F-stack equivalent of -ROT

note, some systems call this work F-ROT, here it is the inverse of FROT

primitive code = [p4_minus_f_rot]

---

FNIP(f: x1 x2 -- x2 )  => "FORTH"

F-stack equivalent of NIP

primitive code = [p4_f_nip]

---

FTUCK(f: x1 x2 -- x2 x1 x2 )  => "FORTH"

F-stack equivalent of TUCK

primitive code = [p4_f_tuck]

---

1/F(f: x -- 1/x )  => "FORTH"

primitive code = [p4_one_over_f]

---

F^2(f: x -- x^2 )  => "FORTH"

primitive code = [p4_f_square]

---

F^N( u f: x -- x^u )  => "FORTH"

For large exponents, use F** instead. Of course u=-1 is large.

primitive code = [p4_f_power_n]

---

F2/(f: x -- x/2 )  => "FORTH"

primitive code = [p4_f_two_slash]

---

F2*(f: x -- x*2 )  => "FORTH"

primitive code = [p4_f_two_star]

---

F0>(f: x -- s: flag )  => "FORTH"

primitive code = [p4_f_zero_greater]

---

F0<>(f: x -- s: flag )  => "FORTH"

primitive code = [p4_f_zero_not_equal]