|
#!/bin/bash |
|
# dependencies: wget, libxml-xpath-perl, xmlstarlet, imagemagick, and dropbox_uploader.sh from https://github.com/andreafabrizi/Dropbox-Uploader |
|
|
|
# check for dependencies |
|
command -v wget >/dev/null 2>&1 || { echo "$(tput setaf 1)Please install $(tput bold)wget$(tput sgr0)$(tput setaf 1) and make sure that the $(tput bold)wget$(tput sgr0)$(tput setaf 1) command is working$(tput sgr0)" >&2; exit 1; } |
|
command -v xpath >/dev/null 2>&1 || { echo "$(tput setaf 1)Please install $(tput bold)libxml-xpath-perl$(tput sgr0)$(tput setaf 1) and make sure that the $(tput bold)xpath$(tput sgr0)$(tput setaf 1) command is working$(tput sgr0)" >&2; exit 1; } |
|
command -v xmlstarlet >/dev/null 2>&1 || { echo "$(tput setaf 1)Please install $(tput bold)xmlstarlet$(tput sgr0)$(tput setaf 1) and make sure that the $(tput bold)xmlstarlet$(tput sgr0)$(tput setaf 1) command is working$(tput sgr0)" >&2; exit 1; } |
|
command -v mogrify >/dev/null 2>&1 || { echo "$(tput setaf 1)Please install $(tput bold)imagemagick$(tput sgr0)$(tput setaf 1) and make sure that the $(tput bold)mogrify$(tput sgr0)$(tput setaf 1) command is working$(tput sgr0)" >&2; exit 1; } |
|
|
|
# clean up from any old broken jobs |
|
rm -rf tmp |
|
|
|
# let's go |
|
|
|
feedurl='https://what-if.xkcd.com/feed.atom' |
|
|
|
mkdir tmp |
|
cd tmp |
|
|
|
wget -q -O feed.tmp "$feedurl" |
|
|
|
# edit these if you want to use a different feed |
|
# for example, engadget's feed at http://www.engadget.com/rss-full.xml would be |
|
# feedname=$(cat feed.tmp | xpath -q -e "//rss/channel/title/text()") |
|
# entrytitle=$(cat feed.tmp | xpath -q -e "//rss/channel/item/title/text()") |
|
# entryurl=$(cat feed.tmp | xpath -q -e "//rss/channel/item/link/text()") |
|
# etc... |
|
feedname=$(cat feed.tmp | xpath -q -e "//feed/title/text()") |
|
entrytitle=$(cat feed.tmp | xpath -q -e "//feed/entry[1]/title/text()") |
|
entryurl=$(cat feed.tmp | xpath -q -e "//feed/entry[1]/id/text()") |
|
entrycontent=$(cat feed.tmp | xpath -q -e "//feed/entry[1]/content/text()") |
|
entrydate=$(cat feed.tmp | xpath -q -e "//feed/entry[1]/published/text()") |
|
# i used these to make sure i was extracting properly |
|
#echo "$feedname" |
|
#echo "$entrytitle" |
|
#echo "$entryurl" |
|
#echo "$entrycontent" |
|
#echo "$entrydate" |
|
#echo "$entrynumber" |
|
|
|
# check to make sure the feed we downloaded is legit |
|
if [ -s feed.tmp ]; then |
|
echo "$(tput setaf 2)Fetched feed$(tput sgr0)" |
|
else |
|
echo "$(tput setaf 1)Problem fetching the feed. Quitting...$(tput sgr0)" |
|
cd .. |
|
rm -rf tmp |
|
exit 1 |
|
fi |
|
|
|
# time to build a new html doc |
|
echo "<!DOCTYPE html><html><head><meta charset=\"utf-8\"><meta http-equiv=\"Content-Type\" content=\"text/html; charset=utf-8\" /></head><body>" > tmp.html |
|
# grab the feed's content, feed it through the xmlstarlet fixer, replace all src="/blah/foo/pic.png" with src="pic.png", and change all png references to jpg :( |
|
echo "$entrycontent" | xmlstarlet unesc | sed -e 's#src="[^"]*/#src="#g' | sed -e 's#.png#.jpg#g' >> tmp.html |
|
echo "<p><i>Published in <a href=\"$entryurl\" target=\"_blank\">$feedname</a> at $entrydate</i></p>" >> tmp.html |
|
echo "</body></html>" >> tmp.html |
|
if [ -s tmp.html ]; then |
|
echo "$(tput setaf 2)New temporary HTML file built$(tput sgr0)" |
|
else |
|
echo "$(tput setaf 1)Problem building temporary HTML file. Quitting...$(tput sgr0)" |
|
cd .. |
|
rm -rf tmp |
|
exit 1 |
|
fi |
|
|
|
# if there's no difference from last time to this time, clean up and exit |
|
if cmp -s "tmp.html" "../lastfetch.html"; then |
|
echo "$(tput setaf 3)The fetched/built html file is the same as last time. Quitting...$(tput sgr0)" |
|
cd .. |
|
rm -rf tmp |
|
exit 1 |
|
fi |
|
|
|
# if you've reached this point in the script, then the downloaded and built file is new |
|
# copy this new one we've built to be the "last fetched" html... |
|
cp tmp.html ../lastfetch.html |
|
# ...and start the mobi building process |
|
mv tmp.html "$feedname: $entrytitle.html" |
|
|
|
# wget all of the images from the entry page |
|
wget --quiet --page-requisites --span-hosts --convert-links --no-directories "$entryurl" |
|
|
|
# kindles suck at turning transparent pngs into jpgs |
|
mogrify -colorspace sRGB -background white -format jpg -alpha remove ./*.png |
|
rm -rf ./*.png |
|
|
|
# move back up, build the book with kindlegen, move the book out of tmp, and delete the tmp directory |
|
# https://smile.amazon.com/gp/feature.html?docId=1000765211 |
|
|
|
cd .. |
|
|
|
# download the kindlegen binary if you don't have it in the working directory |
|
if [ ! -s kindlegen ]; then |
|
echo "$(tput setaf 1)Kindlegen binary not found. Downloading now...$(tput sgr0)" |
|
wget http://kindlegen.s3.amazonaws.com/kindlegen_linux_2.6_i386_v2_9.tar.gz -O kindlegen.tar.gz |
|
tar -xzf kindlegen.tar.gz -C tmp |
|
rm kindlegen.tar.gz |
|
mv tmp/kindlegen . |
|
fi |
|
|
|
./kindlegen -c0 -o "$entrytitle".mobi tmp/"$feedname: $entrytitle.html" |
|
mv tmp/"$entrytitle".mobi . |
|
rm -rf tmp |
|
|
|
if [ -s "$entrytitle".mobi ]; then |
|
echo "$(tput setaf 2)Book built!$(tput sgr0)" |
|
else |
|
echo "$(tput setaf 1)Something went wrong!$(tput sgr0)" |
|
exit 1 |
|
fi |
|
|
|
# send a pushbullet notification that it worked, if you want |
|
#pbtitle="New $feedname Article Generated" |
|
#pbbody="Article Name: $entrytitle at $entryurl" |
|
|
|
# you'll need a pushbullet api key from https://www.pushbullet.com/#settings/account |
|
# curl --silent \ |
|
# --header 'Access-Token: YOUR_PUSHBULLET_API_KEY' \ |
|
# --header 'Content-Type: application/json' \ |
|
# --data-binary '{"title":"'"$pbtitle"'","body":"'"$pbbody"'","type":"note"}' \ |
|
# --request POST \ |
|
# https://api.pushbullet.com/v2/pushes > /dev/null |
|
|
|
# upload the new file to dropbox folder |
|
./dropbox_uploader.sh upload "$entrytitle".mobi "/To Kindle" |
|
echo "$(tput setaf 2)Book File Uploaded!$(tput sgr0)" |
|
echo "$(tput setaf 2)All done!$(tput sgr0)" |
