title	Terminal RPN Calculator
subtitle	an RPN calculator shell for the terminal
section	1
author	Simon Widmer
date	31.05.2026
lang	en
license	GNU GPL3
cover-image	rpn-calc.svg
description	an RPN calculator for the terminal
header	RPN
footer	RPN

Terminal RPN Calculator

an RPN calculator shell for the terminal

Screenshots

rpn calc icon

rpn calc v1.9 running in xfce4-terminal

rpn calc v1.9 with custom colorscheme in xterm

Preface and history 📜

First of all: This calculator is not another exactly emulated Hewlett Packard ™ calculator nor is it using ROM files to provide the full functionality. At the current state, it is not able to store and execute RPL-programs. Even so, if you are used to those calculators, you might remember and recognize the most useful commands. Lets say, the calculator mimics its great predecessors. In addition to the usual abbreviated commands, these can also be written as full words. Note that some commands are not implemented in the famous calculators but I found them to be useful anyway.

I am using RPN calculators since I was a teenager and still appreciate it. Hence, it is my intention to program a calculator which supports RPN (reverse polish notation) on the one hand and is executable in a terminal on the other hand. It is written in pure python and designed to be a simple and distraction free terminal shell tool with no dependencies. Despite the simple design, it has become a quite powerful calculator with a lot of functions. (Actually, this calculator has come a long way: I originally planned to create a pure Bash implementation using bc in the background, but ultimately decided against it during the implementation process. The current Python implementation is much more powerful and easier to maintain than the original Bash version would have been.)

Again, keep in mind: This is not the real and famous HP42S™ or HP48SX™ that is supporting programming, unit conversions, graphs, equation solvers, libraries and lots of things I do not even know of. Those real calculators are an industrial and technical work of art I admire and may never be compared with.

Installation 📦

Linux installation 🐧

• Run sudo apt install python3 for Debian. Installation for other distributions may vary. There is no other dependencies.
• run ./rpn.py to start the calculator 🚀
• to install it permanently including man page, run
  • apt install pandoc
  • make
  • sudo make install
• to uninstall it, run sudo make uninstall

Windows installation 🪟️

• install python (Tested with Windows 11 only)
• use run.bat to start the calculator or open a cmd-shell and execute py ./rpn.py 🚀
• Occasionally, I create an executable file, which can also be used and makes installing Python unnecessary. It is stored in the bin/ directory.

Usage 💡

Disclaimer ⚠️

Don't use this calculator if you plan to fly to the moon. I'm not kidding — expect that some results might be wrong. Testing all functions and edge cases is a huge task.

General Usage 👇

• If you are not familiar with rpn calculators and with reverse polish notation, you might not want to learn this notation unless you know the advantage of this concept and are willing to rethink.
• Typing in either small or capital letters is valid.
• Errors are shown in red and cleared after 2 seconds.
• Decimal numbers are separated by a period.
• the size of the stack is unlimited (in theory).

User Interface 🖥️

CTRL-N for help, CTRL-X to exit
┌──────────────────────────────────────────────┐
│ Terminal RPN Calculator vX.x.x               │
├──────────────────────────────────────────────┤
│ 4:                                           │
│ 3:                                         64│
│ 2:                                        100│
│ 1:                                      45054│
└──────────────────────────────────────────────┘
input > ▒

Example Text User interface above with some numbers entered in the visible stack and cursor after input.

Hotkeys ⌨️

• CTRL-X → exit
• CTRL-L → clears entire stack
• BACKSPACE →drop stack 1
• DEL → drop stack 1
• CTRL-K → drop actual input >
• TAB → swap stack 1 and stack 2
• ENTER → duplicate stack 1 (if input is empty) else perform the operation or function
• CTRL-E → edit stack 1
• CTRL-N or ? → help

Hint (ℹ): If a hotkey is not working, this is most likely due to the terminal shell in use and it's predefined hotkeys. Workaround: enter the according command at the input prompt and hit ENTER.

Commands ⚡

Info about commands 📝️

commands are executed by typing in and hit ENTER. This might feel rather like a shell therefore.

• Advantages:
  • no polling for keys but just waiting for input.
    • No CPU usage at this point.
    • Simple Code.
    • shell feeling
    • commands should be mostly the sames as the commands of RPN calculators
    • easy to memorize
  • Disadvantages:
    • typing commands for a calculator might feel unusual at first

General commands 📢️

• exit or quit or just off followed by hitting the ENTER-key to leave the rpn calculator.
• help or ? or hlp to get a quite detailed help.
• about or info to get a little bit info about myself, the calculator and the license
• refresh to clear a destroyed UI.
• getkey to check the key code (used for debugging purposes only)

Stack commands 📑

• clr or clear will clear the entire stack.
• swap or swp or x<>y will swap stack 1 and 2.
• drop will delete stack 1.
• drop2 will delete stack 1 and stack 2.
• dup or duplicate or enter will duplicate stack 1
• dup2 will duplicate stack1 and stack 2
• edit to edit stack 1
• depth → number of elements in stack.
• avg or average → average of entire stack.
• save → save entire stack to ~/stack.txt in decimal values. Windows: C:\Users\YourUserName\stack.txt
• save <filename> → save entire stack to ~/<filename> in decimal values
• sort → sort entire stack in ascending order and if already, in descending order.
• sum → sum of entire stack.
• rollup → Moves the very last element of the stack to Level 1 (Bottom to Top).
• rolldown → Moves the element from Level 1 to the very bottom of the stack (Top to Bottom).
• over → copies stack 2 to stack 1 (without deleting stack 1).
• pick → copy stack[n] to top (n from stack 1, 1-based)

Basic Arithmetic Operations 🧮

As with the real rpn calculators, hitting the operators does execute an immediate operation. You do NOT have to hit the ENTER-key to do so.

• + or plus → addition
• - or minus → subtraction
• * or mul → multiplication
• / or : or div → division
• % → percentage
• ^ or pow → exponentiation (yˣ)
• _ → changes the sign of a number (+/-)

Math functions ƒ(𝑥)

The following functions can be used followed by hitting ENTER:

• Sign and basic operation ⊕/⊖
  • neg or chs → negation
  • abs→ absolute value
  • comb or combination(s) → returns the number of combinations of two numbers in the stack (nCr)
  • ip or int or integer → integer of a given value (by cutting, not rounding)
  • mant or mantissa → mantissa of a number
  • xpon → exponent of argument (floor of log10 of abs value)
  • fact or factorial → factorial (x! = Γ(x + 1)). Note that the returned value can be very high, as the factorial grows very fast.
  • frac or frac or fractional → fractional part of a number (x - ip(x))
  • ceil → returns next greater integer
  • floor → returns next smaller integer
  • mod or modulo → returns modulo of two numbers in the stack
  • perm or permutation(s) → returns the number of permutations of two numbers in the stack (nPr)
  • ran rand or random → returns random number 0 < x < 1
  • rnd or round → Rounds number as specified in level 1.
  • sign → sign of x → -1, 0, or 1
  • gcd → greatest common divisor
  • lcm → least common multiple
• Powers and roots 📈
  • reci or reciprocal or inv → reciproke (1/𝑥)
  • powor power → exponentiation (yˣ)
  • sq or square → square (𝑥²)
  • sqrt or squareroot → square root (√𝑥)
  • root or rt or xroot → ⁿ√𝑥
  • exp → eˣ (inverse of ln)
  • exp10 or alog → 10ˣ (inverse of log10)
  • expm → eˣ - 1 (more accurate for small x)
• Logarithms 📏
  • ln → natural (base e) logarithm
  • log10 or log → common logarithm (base 10)
  • log2 → binary logarithm (base 2)
• Trigonometric functions (in degrees) 📐
  • sin or sine → sine
  • asin or arcsine → arc sine
  • sinh or sinushyperbolicus → hyperbolic sine
  • cos or cosine → cosine
  • acos or arccosine→ arc cosine
  • cosh or cosinushyperbolicus → hyperbolic cosine
  • tan or tangent → tangent
  • atan arctangent → arc tangent
  • tanh tangenthyperbolicus → hyperbolic tangent
• Logic operations ✅❌
  • setwordsize or stws → sets word size for logic ops (the default word size is 64 bits)
  • recallwordsize or rcws → recalls word size to stack
  • and → Logical or binary AND.
  • or → Logical or binary OR
  • xor → Logical or binary exclusive OR.
  • not → Logical or binary NOT.
  • sl or lsl → logical shift left 1 bit
  • sr or lsr → logical shift right 1 bit
  • slb → shift left 1 byte (8 bits)
  • srb → shift right 1 byte (8 bits)
  • asr → arithmetic shift right (sign bit replicated)
  • rl → rotate left 1 bit
  • rr → rotate right 1 bit
  • rlb → rotate left 1 byte (8 bits)
  • rrb → rotate right 1 byte (8 bits)

Constants ◆

The following constants are implemented for now and can be entered followed by hitting ENTER:

• pi : circle constant (π = 3.1415927…)
• tau: "true" circle constant (𝜏 = 2π = 6.2831853…)
• c or lightspeed: speed of light in a vacuum (𝑐) 299792458 m/s
• euler : 2.7182818… (ℯ)
• gravity or g: 9.80665m/s² (𝑔)
• phyor goldenratio: 1.6180339887… (φ = (1 + √5) / 2)

Memory commands 🧠

• mem or memory → shows all available memory of the OS, if supported.
• sto or store will store value of stack 1 into memory and delete the value in the stack.
• rcl or recall will recall value from memory and append it to the stack. Memory contains 0 by default.
• mc or memclr or mclear will clear the memory and set it to 0.

Modes 🔧

Display Modes 👁️

• fix → Fixed-Decimal (default), will set and show n display digits by the stack 1 value. E.g. 3 *ENTER* fix *ENTER* will limit to 3 digits. Allowed value is 0-16
• eng → Engineering notation, will set and show n display digits by the stack 1 value
• sci → Scientific notation, will set and show n display digits by the stack 1 value

Base Modes ⚙️

• dec will show the stack in decimal values
• hex will show the stack in hexadecimal values
• bin will show the stack in binary values
• oct will show the stack in octal values
• the current mode is indicated below the stack at the input prompt if not in decimal-mode

Notation of dec, hex or bin value(s) ✒️

The notation of a hexadecimal or binary number can be typed in as follows:

• decimal notation of exponential numbers: (e.g. 3e6, 1.5E4, 5.2e-3, 1E+6, 3E-5 …)
• binary: e.g.: 0b1111, 0b10101010, …
• hexadecimal : e.g. like 0xabcd, 0xAFAEBD, … or directly without leading 0b or 0x.

Further notice 💬:

• Mixing decimal, hexadecimal and binary values in the stack is not possible.
• if in decimal mode, all entered values with leading 0x, 0b or 0o are directly converted into the decimal value.
• same behavior and conversion for the other modes

Conversion Modes 🪄

angle conversions

• deg2rad or `d>r or deg>rad → Degrees-to-radians conversion.
• rad2deg or `r>d or rad>deg → radians → degrees

temperature conversions

• c2f or c>f or °c2°f or °c>°f or celsius2fahrenheit : Celsius → Fahrenheit
• f2c or f>c or °f2°c or °f>°c or fahrenheit2celsius : Fahrenheit → Celsius
• c2k or c>k or °c2°k or °c>°k or celsius2kelvin : Celsius → Kelvin
• k2c or k>c or °k2°c or °k>°c or kelvin2celsius : Kelvin → Celsius
• f2k or f>k or °f2°k or °f>°k or fahrenheit2kelvin : Fahrenheit → Kelvin
• k2f or k>f or °k2°f or °k>°f or kelvin2fahrenheit : Kelvin → Fahrenheit

Note that values below absolute zero (-273.15°C, -459.67°F or 0 K) are returned as error.

time conversions

• >hms or 2hms → decimal hours → H.MMSSss
• >h or 2hours → H.MMSSss → decimal hours

length conversions

english conversions

• inch2cm or inch>cm : inches → centimeters
• cm2inch or cm>inch : centimeters → inches
• inch2mm or inch>mm : inches → millimeters
• mm2inch or mm>inch : millimeters → inches
• inch2m or inch>m : inches → meters
• m2inch or m>inch : meters → inches
• foot2m or foot>m : feet → meters
• m2foot or m>foot : meters → feet{RESET}
• mile2km or mile>km : miles → kilometers
• km2mile or km>mile : kilometers → miles
• mile2m or mile>m : miles → meters
• m2mile or m>mile : meters → miles
• seamile2km or seamile>km : miles → kilometers
• km2seamile or km>seamile : kilometers → miles

metric conversions

• km2m or km>m : kilometers → meters
• m2km or m>km : meters → kilometers
• m2cm or m>cm : meters → centimeters
• cm2m or cm>m : centimeters → meters
• m2mm or m>mm : meters → millimeters
• mm2m or mm>m : millimeters → meters

japanese conversions

• sun2m or sun>m : sun → meters
• m2sun or m>sun : meters → sun
• ken2m or ken>m : ken → meters
• m2ken or m>ken : meters → ken
• shaku2m or shaku>m : shaku → meters
• m2shaku or m>shaku : meters → shaku
• shaku2cm or shaku>cm : shaku → centimeters
• cm2shaku or cm>shaku : centimeters → shaku
• shaku2mm or shaku>mm : shaku → millimeters
• mm2shaku or mm>shaku : millimeters → shaku
• ri2m or ri>m : ri → meters
• m2ri or m>ri : meters → ri

Limitations ▲ ▼

• the digits are limited and tested to the value 16.

I'm developing this pocket calculator in my spare time, so you'll notice that many features and functions — such as the following — are missing:

• programming
• matrix
• plotting
• variables
• Equation solving
• arrays
• complex number handling
• Unit conversion (some conversions implemented)
• IO and many other fancy features
• small subset of hotkeys — Esc-Key and F-Keys are non-functional

For the moment, I do not have intentions to implement these features. There might be other apropriate tools to use.

To Do's and Niceties 🚧

The following features are on my to do list:

• missing commands 🕵️
  • incr, decr (increment and decrement stack 1 by 1)
  • sto stack as array elements
  • beep
  • date and time functions (date, date+, ddays)
  • key (would be nice for debugging new hotkeys)
  • max, min, maxr, minr
  • sto+-/* (are those useful commands?)
• missing but cool features 🌟
  • getting track of command-history to implement hotkey "↑" (up) to recall the commands for the "shell-feeling"
  • autocompletion and suggestions
  • status of stored value

Error reporting 🐛

Errors and bugs are possibly included. Error reports or corrections are welcome: ✉️ sery@solnet.ch

Special Thanks 🤝

• Jan Łukasiewicz for inventing RPN.
• Bill Joy for his eternal editor vi and Bram Moolenaar et al for Vim, the improved Version of vi for Linux and other platforms.
• Chris Kempson for the wonderful and calm colorscheme Tomorrow-Night-Blue.vim.
• John Gruber in collaboration with Aaron Swartz for Markdown
• John MacFarlane for the document conversion tool pandoc that seems to be from outer space with all its functionality.
• Hewlett Packard for the great calculators (that can be programmed with RPL) over all the years.
• Bob Bemer for the invention of ANSI Escape Code that is still supported on many platforms including Windows 11. Unbelievable after all those years!
• Guido van Rossum for inventing the Python programming language.
• Ken Thompson, Dennis Ritchie, Brian Kernighan, Douglas McIlroy and Joe Ossanna et all for inventing the Unix operating system and all the developers who contributed to the Unix ecosystem. Without Unix, Linux would possibly have never been invented.
• Richard Stallman for inventing the GNU operating system and all the developers who contributed to the GNU ecosystem. Without GNU, Linux would be just a kernel and not a full operating system.
• Linus Torvalds for inventing the Linux operating system kernel and all the developers who contributed to the Linux ecosystem.

see also

bc(1), dc(1)