Extract clean text from webpage w/ Mercury API and save to PDF

convert.sh

#!/bin/bash -e

APIURL="https://mercury.postlight.com/parser"
APIKEY="ErUeufK8WIFsi6769SZuQEEn1b6CS4D5JAfrvBhJ"
URL="$1"
TEMPDIR=$(mktemp -d)

# change working dir
cd $TEMPDIR

# download article with Postlight's Mercury API
curl -G --data-urlencode "url=$URL" -H "x-api-key: $APIKEY" $APIURL > result.json

# extract data from downloaded json
title=$(jq -r .title result.json)
content=$(jq -r .content result.json)
author=$(sed 's/null/?/g' <<<"$(jq -r .author result.json)")
date=$(date -jf "%Y-%m-%d" $(jq -r .date_published result.json | cut -c 1-10) +"%-d %b %Y" 2>/dev/null || echo '?')
url=$(jq -r .url result.json)
domain=$(jq -r .domain result.json)
wordcount=$(jq -r .word_count result.json | awk '{ print ($1 > 5) ? $1 : "?" }')

# write html
cat <<EOF > result.html
<style>
html {
	font-family: Georgia;
}
img {
  width: auto;
  height: auto;
  max-width: 100%;
  max-height: 300px;
}
code {
  background-color: #eee;
}
.aligncenter {
  text-align: center;
}
.wp-caption-text {
  font-style: italic;
}
</style>
<h1>$title</h1>
<div><b>$wordcount</b> words by <b>$author</b> on <a href="$url"><b>$domain</b></a> at <b>$date</b></div>
<hr>
$content
EOF

# filter html
xidel --html result.html --xquery 'transform(/, function($e) { 
  if (string($e) = "Subscribe to Praxis, our members-only blog exploring the future of productivity, for just $10/month. Or follow us for free content via Twitter, Facebook, LinkedIn, or YouTube.") then <i></i> else $e
})' | sponge result.html

# convert html to pdf
fname="$title".pdf
wkhtmltopdf --lowquality --image-dpi 300 --image-quality 80 --no-outline --encoding 'utf-8' --disable-smart-shrinking result.html "$fname"

# import into Notability
open -a Notability "$fname"

# clean up
sleep 5
rm -rf $TEMPDIR