Cody Kochmann CodyKochmann

I need scripts to run on github webhooks just as bad as you do. But this one-size-fits-all "lets make scripts but in yaml" shit has got to go.

Yaml acts as the "purely ci details". Script logic goes in... SCRIPTS!

Every CI/CD engine has a million different 'input' layers - environment variables, repo variables, workflows from various branches, maybe the git commit object. Oftentimes your scripts will have the ability to jam more variables into the 'input' layer for later scripts - such as in the case of the github environment variables. UNFORTUNATELY, it's never clear WHICH input layers are available to WHICH parts of the script. "Oh, no, you can't use outputs from this script to as arguments for something in our yaml, because the yaml is calculated first." Well thanks, Mr Nadella, I'll just torch my entire pipeline and start over.

My form of dynamic ci materializes all variables into logged and exam-able ci code. Hidden ci assembly magic was one of the first things I got rid of because life triggered

	#!/bin/bash
	# This script contains the steps needed to shrink a massive table on postgres
	# when disks did not have enough room to run a VACUUM FULL after major deletions.
	# by: Cody Kochmann

	# NOTE - These commands were run manually and will need situation specific
	# options added if they were ever to be reused. This gist was only to
	# capture the gist of the steps needed.

	# set a date to clear logs to

	#!/usr/bin/env python3
	# by: Cody Kochmann
	# created: 2023-10-01
	# license: MIT

	'''
	This script takes a json object from stdin and turns it into a directory tree of files.
	'''

	import sys, json, os, contextlib

	#!/bin/bash
	# by: Cody Kochmann
	# how to run an interactive gui program from another user on debian 12
	# inspired by https://gist.github.com/kasunbg/5502cb630429819d07b5dc0cfa26813c

	export NEW_USER_NAME=cody-facebook

	#---------------------------------
	# step 1 - create a new user
	#---------------------------------

	# file /etc/fstab
	# by: Cody Kochmann
	# license: MIT
	# description: how I keep cache to per boot lifetimes

	# Note - pull the uid and gid from /etc/fstab to ensure correct permissions
	# Note - the "size" can be defined as units like "1G" or percentages like "10%"

	# configuration for per user cache
	tmpfs /home/cody/.cache tmpfs defaults,uid=1000,gid=1000,mode=0700,size=10% 0 0

	# Author: Cody Kochmann
	# Date Created: 2023-09-16
	# Last Modified: 2023-09-16
	# License: MIT
	#
	# This makefile spins up an aria2c process for every .magnet file in the current directory.
	#
	# It assumes it will be only one magnet link per each .magnet file.
	#
	# The benefit of this over just having a single aria2c call pull multiple magnet links from

	➜ ~ ssh [email protected] cat /usr/local/bin/sqlite3-user-db
	[email protected]'s password:
	#!/bin/bash

	/usr/bin/sqlite3 "$HOME/$USER.db"

	➜ ~ ssh [email protected]
	[email protected]'s password:
	Linux kali 6.0.0-kali3-amd64 #1 SMP PREEMPT_DYNAMIC Debian 6.0.7-1kali1 (2022-11-07) x86_64

	# by: Cody Kochmann
	# this script pulls the contents of $TARGET_IMAGE
	# and uses pytesseract against it to output the
	# text contents of the image to stdout

	import os, sys, functools
	import requests
	try:
	from PIL import Image
	except ImportError:

	-- -------------------------------------------
	-- This is an implementation of a key, value
	-- store for function based kv storage for
	-- different sql based algorithms to utilize
	-- in postgres.
	-- -------------------------------------------
	-- Author: Cody Kochmann
	-- license: MIT
	-- -------------------------------------------

	from functools import wraps
	from threading import Timer, Semaphore

	import requests

	def service_request_rate_limit(per_minute: int):
	assert isinstance(per_minute, int), per_minute
	assert per_minute > 0, per_minute
	lock = Semaphore(per_minute)