fnk

python from the command line

what it is

A command line interface with functional tendencies that allows you to parse information in the shell using Python.

examples

note on defining closures/lambdas Functions (-f and --function) can be defined using a variety of function formats, one of them being the default Rust format for closures and another explicitly showing the mapping. The below are all valid ways to write your functions (each example does the same thing, add one to a list of input numbers):

1. |x: int| x + 1
2. x: int -> x + 1
3. (x: int) -> x + 1
4. |x: int| -> x + 1 I prefer the top format shown in 1), but you do you boo boo. Examples below are written in the mixed form show in 4) above just to make it absolutely clear what the variable and function declarations are.

Note on the note: if you want to treat the input data as strings, you don't need to specify type. For example: |s| -> s + "concat to end of s" Additionally, if you're using a uniform type for all variables, you can also just specify the type using the -t/--type flag: fnk map -f '|n| -> n + 1' -t int The above function is functionally equivalent to the examples shown in 1 - 4).

add one to each input number and print

echo -e "1\n2\n3" | fnk map -f '|n: int| -> n + 1'
>>> 2
>>> 3
>>> 4

add one to each input number and sum

echo -e "1\n2\n3" | fnk map -f '|n: int| -> n + 1' | fnk agg -f sum
>>> 11

take list of full paths and filter out just the parent dir and current file name

ls -d $PWD/*
>>> ~/scripts/fnpy/fnk/fnk.py
>>> ~/scripts/fnpy/fnk/parser.py

ls -d $PWD/* | fnk map -f '|f: str| -> "/".join(f.split(".")[-1].split("/")[-2:])'
>>> fnk/fnk
>>> fnk/parser

take input and split each word into it's own line

echo -e "words on a line\nand between lines" | fnk map --split_whitespace
>>> words
>>> on
>>> a
>>> line
>>> and
>>> between
>>> lines

filter out filenames that are .py files

ls
>>> fnk.py
>>> parser.py
>>> README.md
>>> random_text.txt

ls | fnk filter -f '|s: str| ".py" in s'
>>> fnk.py
>>> parser.py

removed duplicate lines from input data

echo -e "same\nsame\nsame\ndifferent" | fnk map -container_type set
>>> same
>>> different

read input as json, filter name name key, filter out entries without a name key, and prepend with "Name: "

echo '[{"name": "shane", "age": 35}, {"name": "dan", "age": 31}, {"age": 22}]' | fnk filtermap -f '|entry| -> "Name: " + entry.get("name", "")' --json_in
>>> shane
>>> dan

read current list of files, import os module, and use to delete everything with "tempnvim" in it

ls
>>> c:usersstephlocaltempvnim.alk1231l
>>> c:userstephlocaltempnvim.asdl213k1
>>> fnk.py
>>> parser.py

ls | fnk filter -f '|f| -> "tempnvim" in f' | fnk apply -f '|f| os.remove(f)' --modules os
>>> fnk.py
>>> parser.py

evaluate a simple expression from the command line

If you want to manually input your input data instead of piping it in, you can do that as follows:

# note that both functions below are functionally equivalent, eval just evaluates each line of input.
fnk eval "1 + 1"
>>> 2

fnk filtermap "1 + 1" -f "|e| eval(e)"
>>> 2

calculate stats on a string of numbers from the command line

fnk agg "1,2,3" -f stats --input_separator ,
>>> Mean   -> 2.0
>>> Median -> 2.0
>>> Mode   -> 1.0
>>> Stdev  -> 1.0
>>> Var    -> 1.0

echo "1 2 3" | fnk agg -f stats --split_whitespace
>>> Mean   -> 2.0
>>> Median -> 2.0
>>> Mode   -> 1.0
>>> Stdev  -> 1.0
>>> Var    -> 1.0

echo -e "1\n2\n3" | fnk agg -f stats
>>> Mean   -> 2.0
>>> Median -> 2.0
>>> Mode   -> 1.0
>>> Stdev  -> 1.0
>>> Var    -> 1.0

future

defo needs unit tests written and defo needs multithreading support bcuz python is slow slow slow.