RNG Query

CLI to use pseudorandomness the easy way and help me with decision paralysis

"a, b, c"                 => choose a, b or c at random
"a, b, c / 2"             => chose 2 values
"a, b, c / shuffle"       => shuffle the list
"2d20"                    => roll 2 20 sided dice and sum the result
"[1..10]" or "1..=10"     => random number between 1 and 10, including both
"[1..10)" or "1..10"      => number between 1 and 10 NOT including 1
"(1..5.5]"                => decimal number between 1 and 5.5 not including 5.5
"coin"                    => toss a coin
"a, coin, d6, -5..5 / 2e" => 2 values from the list and evaluate the result

CLI

cargo install rng-query

rq "your query"   # run a quick query
rq                # will get you into a REPL
rq --help         # see all options

There are also precompiled binaries in the github releases.

With the CLI you can read from stdin, files, treat files as only data and combine it with an inline query.

For example, choose 2 lines from a file keeping the input order:

rq -F input.txt "/ 2o"

Input files will be stored in memory with a little overhead. Therefore, very large files may use a lot of memory. It is possible to avoid this, but it's currently not in the scope of this project.

Syntax

Each query works like a stack where you push entries separated by , or a newline. Then, everything AFTER /, UNTIL ; or the end of the line are options. If no / is given, default options are used, which will just select one random entry.

In a single line, inside balanced [], () or {}, input is escaped so ,, / and ; can be used freely. Also, [] and () balance together, so [) and (] are balanced.[^1] If it's not balanced, it's an error.[^2] Finally, you can also wrap anything between " to treat it as a string.

[^1]: This may be triggering to some people, I'm sorry, but I like the interval syntax. [^2]: For input from a file, this can be bypassed.

Entries can be expressions, however, by default, if there are more than 1 entry, then entries are treated as text, not expressions.

When the query get's executed, it consumes the stack and leaves it empty for future queries.

Options

If the input ends without options, / 1 is the default.

The options have the format / [n] [flags], where flags are just chars. Spaces are ignored. [n] is a non negative integer or all, so you don't have to count. If not given, it's 1.

• r: repeating. Allow options to repeat.
• e: (default if only 1 entry) each entry is a expression like an interval, coin or a dice roll.
• E: (default if more than 1 entry) each entry is just text.
• o: keep the order when choosing multiple.
• p: push the selected entries back to the stack instead of returning them.

There are some shorthands:

• / shuffle same as / all E
• / list same as / all Eo
• / eval same as / all eo

Choose number from an interval

[1..4]   # 1 to 4 including both 1 and 4
[1..4)   # 1 to 4 NOT including 4
(1..4]   # 1 to 4 NOT including 1
(1..4)   # 1 to 4 NOT including 1 nor 4
(0, 1)   # 0 to 1 float
(0..1.0) # 0 to 1 float
(.5, 1]  # 0.5 to 1 float including 1

[ and ] to include endpoint, ( and ) to exclude endpoint. If the middle is ,, values are floats. If the middle is .. the values are int by default except if any of them is a float. A float can be of the form 1.5, 1. or .5.

Also <start>..<end> or <start>..=<end> to include end number is an alternative notation only for intergers.

1..=4  # 1 to 4 including both
1..4   # 1 to 4 NOT including 4

Negatives number are supported both in integers and floats.

Open/half-open intervals are not supported because I don't know a good way to handle max/min values.

Toss coin

coin

When evaluated you will get heads or tails.

Roll dice

[amount]d<sides>[!][select][modifier*]

d6        => roll a 6 sided die
2d6       => 2 x 6s dice and sum
2d20k     => 2 x 20s dice and keep the highest

Sides can also be % which equals to 100.

! is exploding. If rolled the maximum value, roll another die.

For select you can add <k|d>[h|l][n]. If n is not given, it's 1. You can have.

• k or kh to keep the highest n dice.
• kl to keep the lowest n dice.
• d or dl to drop the lowest n dice.
• dh to drop the n highest dice.

The modifer is <+|->[m] to add or subtract a value to the total result. You can specify more than one.

When evaluated you will get the sum of all the dice rolls.

There are many more ways to expand this dice notation, but please don't use this tool for your D&D game, roll real dice! If you really really really think more modifiers can be useful, submit an issue.

Notes on pseudorandomness

Currently, randomness should be statistically valid, but NOT cryptographically secure. If you need this to change, submit an issue.