Sejmou

# TODO: remove project-specific wording, turn into more generic stuff
# A reusable Dockerfile for building worker images.
# Includes only the worker package, its workspace deps (e.g. flow-utils), and their
# transitive dependencies. Final image contains only the installed venv (no full repo).
#
# Build args:
#   WORKER_PACKAGE_NAME - uv package name (e.g. aa_torrents), must match name of a directory in workers/

ARG WORKER_PACKAGE_NAME

Sejmou

import nbformat as nbf
from typing import TypedDict


class CodeCellMeta(TypedDict):
    cell_num: int
    output_size_bytes: int
    first_lines: list[str]

Sejmou

// ==UserScript==
// @name         Genius Lyrics Copier
// @namespace    genius-lyrics-copier
// @version      1
// @description  Adds a button to copy lyrics from Genius website to clipboard
// @match        https://genius.com/*
// @grant        GM_setClipboard
// @noframes
// ==/UserScript==

Sejmou

jupyter nbconvert --stdout --no-prompt --to script notebook.ipynb | grep -Ev '^\s*\(*[a-zA-Z0-9_\.,\s]+\)*\s*$'

Sejmou

// https://stackoverflow.com/a/23013726/13727176
function swap(obj) {
  const ret = {};
  for (const key in obj) {
    ret[json[key]] = key;
  }
  return ret;
}

Sejmou

snippet imfn
	```{r, echo=F}
	# This is a custom function I wrote for showing images
	# It is more robust than embedding images in R markdown directly
	# With imgs embedded in R markdown, compilation of whole document fails if img missing
	# This function also displays the image if possible and just logs an error otherwise

	# Under the hood, knitr::include_graphics() is used, which also accepts multiple image paths as parameter
	# Therefore, with this function, side-by-side images are also supported
	# We then need to pass a vector of paths in the function call and add fig.show='hold' to the code cell

Sejmou

# Search for pattern in PDFs inside current folder (and subfolders, recursively)
pdfgrep pattern $(find -name '*.pdf' | sort)

Sejmou

snippet imfn
	```{r, echo=F}
	# This is a custom function I wrote for showing images
	# It is more robust than embedding images in R markdown directly
	# With imgs embedded in R markdown, compilation of whole document fails if img missing
	# This function also displays the image if possible and just logs an error otherwise

	# Under the hood, knitr::include_graphics() is used, which also accepts multiple image paths as parameter
	# Therefore, with this function, side-by-side images are also supported
	# We then need to pass a vector of paths in the function call and add fig.show='hold' to the code cell

Sejmou

{"lastUpload":"2021-03-24T16:57:43.413Z","extensionVersion":"v3.4.3"}