download all the commitstrip strips

commitstrip-dl

#!/usr/bin/env bash
# 
#          name : commitstrip-dl
#   description : download all the commitstrip strips
#       version : 1.0.0
#        author : José Luis Cruz © 2017
#   dependencies: bash<=4 curl [sxiv]
#       license : MIT
# 


#-------------------------------------------------------------------------------
# ::main
#
commitstrip-dl::main() {

	local image_dir="./imgs"  # directory where to save the images
	local image_viewer="sxiv" # (default) image viewer binary

	local web_src         # the source coude of the current webpage
	local image_url       # the url of the current image (strip)
	local image_filename  # the target filename for the image
	                      # with the format: $post_date-$post_title
	local -l post_title   # the title of the post

	local next_post_link
	local next_post_selector
	
	local operation="$1"  # [all|update]

	case "$operation" in

	all) # start from the first strip, forwards
		next_post_link="http://www.commitstrip.com/en/2012/02/22/interview/"
		next_post_selector='rel="next'
		;;

	update) # start from the last strip, backwards
		next_post_link=$(commitstrip-dl::get-last-post)
		next_post_selector='rel="prev'
		;;

	random) # display a random image with sxiv by default
		if [[ -n $2 ]]; then
			if [[  $(which "$2" 2> /dev/null) ]]; then
				image_viewer="$2"
			else
				echo "ERROR: you need to provide a valid program."
				exit 1
			fi
		fi
		if [[ $(which "$image_viewer" 2> /dev/null) ]]; then
			$image_viewer "$(find "$image_dir" | sort -R | head -1)" &
			exit 0
		else
			echo "ERROR: $image_viewer doesn't exist." \
				" Please install it or use another program."
			exit 1
		fi
		;;
	*)
		commitstrip-dl::usage
		exit 1
		;;
	esac

	# prepare where to save the images
	mkdir -p -- "$image_dir"
	cd -- "$image_dir" || exit

	while true; do
  
		web_src=$(curl -s "$next_post_link")
    
		# get commit image
		image_url=$(grep '<img .*src=.*uploads' <(printf '%s' "$web_src") \
			| head -1 \
			| sed -n 's/.*<img .*src="\([^"]*\)".*/\1/p')

		# generate the img filename
		#
		# 1. extract the date of the post
		post_date=$(egrep -o '[0-9]{4}/[0-9]{2}/[0-9]{2}' \
			<(printf '%s' "$next_post_link") \
			| tr -d '/' )
		# 2. filter the post title
		#    - extract the text
		#    - convert html entities to spaces
		#    - convert special characters to spaces
		#    - convert multiple spaces into one
		#    - remove leading/trailing spaces
		#    - replace spaces with dashes
		post_title=$(grep '<h1 .*entry-title' <(printf '%s' "$web_src") \
			| sed -n 's/.*<h1 .*>\([^<]*\)<.*/\1/p' \
			| sed -e 's/\&[^\s]*;/ /g' \
			| sed -e 's/[^A-Za-z0-9._-]/ /g' \
			| sed -e 's/[[:space:]]\+/ /g' \
			| sed -e 's/^[[:space:]]*//g;s/[[:space:]]*$//g' \
			| sed -e 's/[[:space:]]/-/g' \
		)
		# 3. generate the local image file name, with the appropriate extension
		image_filename="$post_date-$post_title.${image_url##*.}"
		
		printf '"%s" ... ' "$image_filename"

		# detect an already existing image
		if [[ -f $image_filename ]]; then

			printf 'alredy exists!\n'

			if [[ $operation == update ]]; then
				printf '\nDone!\n'
				exit 0
			fi

		else
			# download the image
			curl -so "$image_filename" "$image_url"
			curl_err="$?"

			if [[ $curl_err -eq 0 ]]; then 
				printf 'Ok.\n'
			else
				printf 'ERROR: %s\n' "$curl_err"
			fi
		fi

		# get next post link
		next_post_link=$(grep "$next_post_selector" <(printf '%s' "$web_src") \
			| head -1 \
		   	| sed -n 's/.*<a href="\([^"]*\)".*/\1/p')

		# Detect the end (no more posts)
		if [[ -z $next_post_link ]]; then
			printf '\nDone!\n'
			exit 0
		fi

		# artificial delay
		sleep 0.5
	done

} # ::main


#-------------------------------------------------------------------------------
# ::get-last-post
#
# 
# < the last post URL
#
commitstrip-dl::get-last-post() {

	local web="http://www.commitstrip.com/en/" # the list of posts
	local web_src; web_src=$(curl -s "$web")
	local last_post

	last_post=$(grep -A 20 '<div class="excerpts">' <(printf '%s' "$web_src") \
		| grep -e '[^-]<a href=' \
		| sed -n 's/.*<a href="\([^"]*\)".*/\1/p')

	printf '%s' "$last_post"

} # ::get-last-post


#-------------------------------------------------------------------------------
# ::usage
#
commitstrip-dl::usage() {
	cat -- <<-ENDUSAGE
	Usage: $(basename "$0") [all|update]

	all       will fetch all commit strips since the first one.

	update    will start fetching the last commit strip, until the first one
	          it finds that's already downloaded

	See the comic strips online at <http://www.commitstrip.com/en/>

	ENDUSAGE

} # ::usage


if [[ $0 == "${BASH_SOURCE[0]}" ]]; then commitstrip-dl::main "$@"; fi