lerpn - Linux Engineering RPN calculator

It's a curses RPN calculator written in straight Python. Not actually Linux-dependent, despite the (old) name. Requires Python 3.6 at least.

To install, clone this repository and run python3 setup.py install --user. To test-drive, call python3 -m LerpnApp from the root of the source tree.

Dependencies

Just Python.

Hacking - adding commands

If you want to add your own commands to lerpn, see HACKING.md for a quick intro.

Quick usage notes:

How to use lerpn should be obvious for those familiar with RPN. The basic commands:

+ - * /

_         negative prefix (avoids collision with - for minus)
enter     dup
x         exchange/swap
bksp      drop
?         display help
ctrl-C    exit

For a quick online command reference, see COMMANDS.md.

Single-key commands may be used directly if no number is being entered. During number entry, symbolic commands work directly, but alphabeticals require holding Alt (otherwise they're part of the number; see "Advanced input").

Less-used commands begin with a single-quote, then the command name, then 'enter'. For example, the sequence of keys used to compute the sine of 3.14 would be:

  3 . 1 4 ' s i n <enter>

Advanced input

Numbers can be input in many formats:

	Format	Example	Corresponding display mode
Tagged value	<value>"<tag>	3.14"pi
Duration	day:h:m:s	4:20 (4m20s)	'dur
		1::: (1day)
Complex	<re> j <im>	2 j 3, 2j3
Fraction	num,den	1,2 (½)
	whole,num,den	1,1,4 (1¼)	'wfrac
Scientific notation	<coef>e<mag>	2.3e-24	'sci
Engineering notation	<num><prefix>	2.3k (2300)	'eng
		4u7 (4.7e-6)
Radix notation	0b	0b10010101	'bin
	0o	0o4755	'oct
	0d	0d12345678	'dec
	0x	0xCAFEFACE	'hex

In all modes, use _ instead of - for negatives (to avoid conflict with the subtraction operator). - is permitted after e.

Display modes

The following display modes are available:

Mode
'auto	Chooses between 'dur, 'sci, 'eng, 'bin, 'oct, 'dec and 'hex based on the last used input style.
'eng	Displays in engineering notation, using metrix prefixes like 8.2 k. Use 'eng N to specify precision.
'sci	Displays in scientific notation, like 8.2e3. Use 'sci N to specify precision.
'fix	Displays in fixed-point, like 8200.000000. Use 'fix N for N places.
'nat	Displays in natural notation, like 8200.
'all	Displays numbers to full internal precision.
'hex	Displays integers in hexadecimal.
'dec	Displays integers in decimal.
'oct	Displays integers in octal.
'bin	Displays integers in binary.
'dur	Displays numbers (of seconds) as durations in days, hours, minutes, and seconds.
'wfrac	Render fractions as whole+fraction

The 'auto command can also accept a submode to preset the display to, allowing e.g. the display mode to be reset after accidentally changing it without forcing it to stay at the selected mode. For example:

3k                 Sets the display mode to: 'auto 'eng 7
0x42               Sets the display mode to: 'auto 'hex
'auto 'eng         Oops, I wanted it to stay on 'eng

The 'wfrac command wraps the current display mode, and can be issued again to unwrap it.

Tags

You can tag numbers to help remember what they are. There are two ways to do this. One is to enter the number with the tag, using a double-quote to delimit:

  3 . 1 4 " t h i s   i s   p i

gives:    3.14 : this is pi

Whitespace is acceptable. Alternatively, you can attach a tag to a number already on the stack by pushing the string itself to the stack, then using the lowe-case t command to attach:

  3 . 1 4 <enter>
  " t h i s   i s   p i <enter>
  t

This is currently the only application for strings on the stack.

To remove a tag, use upper-case T. The tag will disappear - if this was unintentional, just use lower-case u to undo.

Basic math can be done to numbers while tagged, but if you perform a binary operation on two tagged numbers, lerpn won't know which tag to keep.

Shell commands

lerpn can run shell commands for you, with the 'sh command. Here's a completely useless example:

39
'sh dc -e "$N 3+pq"

The bottom number on the stack is available as the environment variable $N, which holds the number formatted as a standard float.

The output of the command is regex-searched for anything resembling a float; the first match is parsed and pushed to the stack. The full output is displayed in a scrollbox.