rca is a convenient, non-programmable, command-line calculator, for typical one-off professional and everyday problems. It uses the mpdecimal math library for more reliable accuracy than native floating point. rca is similar to both the bc and dc Unix commands, but unlike them, it can seamlessly evaluate both RPN and infix. It's less technical, and perhaps less rigorous, than either, but hopefully it's also more user-accessible. rca is written in C, and is maintained in a single source file.

rca is very versatile. Examples follow, but the man page and rca's help text have all the details.

rca is hosted at github.

• At its heart, rca is a reverse polish notation calculator in the HP tradition.

      34 11 *
                                   374
      sto
      52 13 /
                                     4
      rcl
                                   374
      *
                                 1,496
  

• It seamlessly evaluates infix-style expressions, either mixed with RPN input, or in a full-time infix mode (not shown here).

      ((lastx + pi) * 2)
                                 2,998.28
      (8 - 5) /
                                   999.428
  

• It runs in either floating point or integer mode.

      D
       Mode is signed decimal (D).  Integer math with 64 bits.
                                   374
                                 1,496
                                   999     # warning: accuracy lost, was 999.4277284357265288
  

• It has the trig and math functions of a scientific calculator.

      30 sin 2 ^ 30 cos 2 ^ + sqrt
                                     1.000
  

  or, using infix:

      (sqrt (sin(30) ** 2 + cos(30) ** 2))
                                     1.000
  

  and with lots of digits:

      30 digits 0 degrees
       trig functions will now use radians
      pi
                                     3.14159265358979323846264338328
      4 /
                                     0.78539816339744830961566084582
      sin
                                     0.707106781186547524400844362105
      2 ^
                                     0.5
  

• It has the word width selection, bitwise operators, and multiple bases of a programmer's calculator.

      32 bits  H
       Mode is hex (H).  Integer math with 32 bits.
      1 zf
       Zero fill of hex/octal/binary output is now on
      5
                           0x0000,0005
      8 rol
                           0x0000,0500
      0x80010000 |
                           0x8001,0500
      b
                           0b10000000,00000001,00000101,00000000
  

• It supports variables.

      (_a = 7 + 10; _b = 11 + 43)
                                    54
      (_sum = _a + _b)
                                    71
      variables
                         _a  17
                         _b  54
                       _sum  71
  

• It also has a selection of convenient unit conversions.

      3.5 i2mm
                                    88.9
      mm2i
                                     3.5
      1 oz2g
                                    28.3495
      2 m2ft
                                     6.56168
      30 mpg2l100km
                                     7.84049
  

• And it can be trivially incorporated into shell scripts.

      $ source rca_float
      $ foo=$(fe "22 / 7")
      $ echo $foo
      3.143
      $ fc "$foo > 3.1" && echo yes
      yes
  

---

rca's BSD license can be found in the source, at https://github.com/foxharp/rca, or viewed with the "license" command.