ℙ𝕖𝕡 🙴 ℕ𝕠𝕞

A compiler is a program that translates an input language into an output language. The Dragon Book (Compilers etc) - approximate quote

the translation folder for the ℙ𝕖𝕡/ℕ𝕠𝕞 system

This folder contains a number of nom translation scripts which are also capable of translating themselves and other translation scripts as well.

Translation scripts which are recent (2025) have names like /eg/nom.todart.pss or /eg/nom.tolua.pss and they are much better than the older scripts which have names like /eg/translate.java.pss or /eg/translate.tcl.pss But the older ones should still work for translating many scripts.

There is also a helper nom script called /eg/nom.to.pss which can translate any script to any other available language. This script accepts the command in 'english' such as 'translate <file> to ruby'

translate the palindrome example script to rust

 pep -f nom.to.pss -i 'palindrome.pss to rust' | bash

translate the translator to perl

 pep -f nom.to.pss -i 'nom.to.pss to perl' | bash

use the translated translator to translate a script to lua

 echo "palindrome.pss to lua" | ./nomto.pl

The nom scripts which I am currently writing (August 2025) all include their own help system (usually triggered by the '/help' keyword or some other help-word) and they also include self-translation which is provided by the helper nom script /eg/nom.to.pss

translation examples

Here are some examples of how to translate a script into a target language listed above.

translate an inline script to lua and print to stdout

 pep -f nom.tolua.pss -i "r;[\n]{lines;add ' ';}print;d;"

translate an inline script to lua and run with input


    pep -f nom.tolua.pss -i "r;[\n]{lines;add ' ';}print;d;" > test.lua
    echo -e "leaf \nbranch\n twig" | lua test.lua
    # This will print each line numbered, or run with a file at input
    # (but won't number the first line!)
    cat /usr/share/dict/words | lua test.lua

translate a ℕ𝕠𝕞 script file to lua and run with input


    pep -f tr/nom.tolua.pss eg/nom.syntax.reference.pss > syntax.lua
    echo  "read;add '/';print;clear;" | lua syntax.lua
    # or run with a file at input
    cat eg/palindrome.pss | lua syntax.lua
    # or use the command line switches for input
    # (but this is more for testing purposes)
    lua syntax.lua --file eg/palindrome.pss 
    # or 
    lua syntax.lua --input "read;add '/';print;clear;"

BASH HELPER FUNCTIONS

You can also use the bash functions in bumble.sf.net/books/pars/helpers.pars.sh to translate into a particular language. There are functions for each of the translator languages and they usually start with an abbreviation of the language name. If the function name ends with an 's' then it is for testing an in-line script in that language and if the fn name ends in 'f' then it is for translating/testing a script file. A double 's' ending means that second-generation testing will take place.

“Second-generation ” script translation is a fairly spiffing pep,nom concept that is documented elsewhere but basically means translating the translator with itself and then translating the target script into the same language and then testing it. It's sort of a way of testing the “idem-potency” of the translator. (There is an even more bizarrely amazing ℙ𝕖𝕡 🙵 ℕ𝕠𝕞 concept which is contained in the interpret() method of the new perl translator at /tr/nom.toperl.pss . This concept involves translating the perl translator with itself and then using the translated-translator to interpret nom scripts. But I digress... ).

translate an inline script into the Ruby language and run with input

 pep.rbs 'read;add".";print;clear;' 'abcd'

translate an inline script into rust, compile, and run with input

 pep.rus 'read;add".";print;clear;' 'abcd'

(I goes without saying - but I will say it - that you need the Rust compiler rustc to be installed for this to work)

translate a script file into lua and run with input

 pep.luaf script.pss '[a[b[c]]]'

NOTES

The Dart translator uses a library for grapheme-cluster parsing, and that library needs to be referenced in a pubspec.yaml file and downloaded from the ether with 'pub get' or something similar. But actually the nom translation helper script /eg/nom.to.pss will try to do all that on your behalf (create the yaml file etc), but the bash helper scripts will not.

HISTORY

In this section I annotate work carried out on, and updates to the various translation scripts.

28 Aug 2025

I am in the process of adding the system command to each translation script. Also, I added the accumulator as an exit code for the quit command which seems a simple and elegant solution to allowing scripts to be pure recognisers .

June 2025

Created some new and much better translation scripts for dart /tr/nom.todart.pss rust /tr/nom.torust.pss perl /tr/nom.toperl.pss and lua /tr/nom.tolua.pss These scripts have all parsing code in a parse() method and accept command-line switches for parsing from a file or from <stdin> They also have better debugged escape and unescape commands and are much more mindful and thoughtful of Unicode and UTF8 and code-points and grapheme clusters. At the very least they are designed to acknowledge the problem of unicode code-points versus grapheme clusters. However only the Dart translator currently actually parses grapheme clusters.

For much more information about work carried out on these amazing translation scripts see the work journal at /doc/pepnom.doc.journal.html