davidar

var jsdom = require("jsdom").jsdom;
var Readability = require("./index").Readability;
var express = require("express");

var app = express();

function removeCommentNodesRecursively(node) {
  for (var i = node.childNodes.length - 1; i >= 0; i--) {
    var child = node.childNodes[i];
    if (child.nodeType === child.COMMENT_NODE) {

davidar

var jsdom = require("jsdom").jsdom;
var Readability = require("./index").Readability;

function removeCommentNodesRecursively(node) {
  for (var i = node.childNodes.length - 1; i >= 0; i--) {
    var child = node.childNodes[i];
    if (child.nodeType === child.COMMENT_NODE) {
      node.removeChild(child);
    } else if (child.nodeType === child.ELEMENT_NODE) {
      removeCommentNodesRecursively(child);

davidar

#!/usr/bin/env python
import requests

d = requests.get('https://www.reddit.com/r/EarthPorn/top/.json', headers={'User-agent':'IOTD bot'}).json()
for c in d['data']['children']:
        src = c['data']['preview']['images'][0]['source']
        if src['width'] >= 4000:
                print 'Location: ' + src['url'] + '\n'
                break

davidar

#!/bin/bash
# http://superuser.com/a/1068709/55113
for file in ~/.local/share/applications/*.desktop; do
    CLASS="`grep '^StartupWMClass=' "$file" | cut -d= -f2`"
    if [ -n "$CLASS" ] && ! grep -q xdotool "$file"; then
        EXEC="`grep '^Exec=' "$file"`"
        EXEC="$EXEC \\&\\& xdotool search --sync --classname $CLASS set_window --class $CLASS"
        sed -i~ "s!^Exec=.*!$EXEC!" "$file"
    fi
done

davidar

/*  passable motion blur effect using frame blending
 *  basically move your 'draw()' into 'sample()', time runs from 0 to 1
 *  by dave
 *  http://beesandbombs.tumblr.com
 */

int samplesPerFrame = 32;  // more is better but slower. 32 is enough probably
int numFrames = 48;        
float shutterAngle = 2.0;  // this should be between 0 and 1 realistically. exaggerated for effect here
int[][] result;

davidar

[
  "abstract.coffee",
  "animal.coffee",
  "applause.coffee",
  "aww.coffee",
  "bang-bang.coffee",
  "base58.coffee",
  "base64.coffee",
  "botsnack.coffee",
  "coin.coffee",

davidar

#!/bin/sh
# https://nixos.org/wiki/How_to_install_nix_in_home_%28on_another_distribution%29#PRoot_Installation

OPT=$HOME/opt
ARCH=x86_64
VER_NIX=nix-1.10-$ARCH-linux
URL_NIX=http://hydra.nixos.org/build/25489771/download/1/$VER_NIX.tar.bz2

mkdir $OPT
mkdir $OPT/bin && cd $OPT/bin && wget http://static.proot.me/proot-$ARCH && chmod u+x proot-$ARCH

davidar

package main

// This program takes a .zim-file and dumps all contained articles below the
// current directory. I tested it with a Wikipedia snapshot; that did not
// contain deleted articles or LinkTargetEntrys, so I'm unsure how to handle
// those, for now I'm ignoring them.
//
// Redirects are handled by simply writing out the page pointed to by the
// redirect. IPFS deduplication should take care of it, so I think this is the
// most economical solution, even better than writing out small HTML files with

davidar

Keybase proof

I hereby claim:

• I am davidar on github.
• I am davidar (https://keybase.io/davidar) on keybase.
• I have a public key whose fingerprint is 0B4C 4F82 6A9E AD7C A1B4 9D4B 520B 476E 1344 C707

To claim this, I am signing this object:

davidar

NLP stuff

##Bookshelf

• AIMA chapters 22-23
• Bishop chapter 13 (sequential data)
• Intro to IR
• CGEL optional but good for background

Bibliography