grabbing Album art [Solved]

[zigbert]

Downloading album art should be easy, but I can't get it work every time - even if I know that the album art is present on the webpage.

Can you see what I am doing wrong? - thank you.

Code: Select all

#!/bin/bash 
STRING="madonna like a virgin" #artist album
URL="`echo "http://www.freecovers.net/api/search/$STRING/Music+CD" | tr ' ' '-'`"
URL2=`wget -q -O - "$@" $URL | grep -o -m 1 -i 'http:\/\/www\.freecovers.net\/preview\/0\/[a-z0-9]*\/big.jpg'`
wget "$URL2" -O $HOME/album_art.jpg

[L18L]

# wget $URL -O $HOME/album_art.jpg
--2011-05-17 18:16:54-- http://www.freecovers.net/api/search/ma ... n/Music+CD
Auflösen des Hostnamen »www.freecovers.net (www.freecovers.net)«.... 204.45.50.211
Verbindungsaufbau zu www.freecovers.net (www.freecovers.net)|204.45.50.211|:80... verbunden.
HTTP Anforderung gesendet, warte auf Antwort... 200 OK
Cookie von »www.freecovers.net« versuchte die Domain auf »www.freecovers.net« zu ändern
Länge: nicht spezifiziert [application/xml]
Saving to: `/root/album_art.jpg'

[ <=> ] 28.609 4,86K/s in 5,7s

2011-05-17 18:17:07 (4,86 KB/s) - `/root/album_art.jpg' saved [28609]

#

$URL not $URL2

[Dougal]

Can you see what I am doing wrong? - thank you.

Code: Select all

STRING="madonna like a virgin" #artist album

Sigmund, have you been sniffing your paints??

Code: Select all

URL2=`wget -q -O - "$@" $URL | grep -o -m 1 -i 'http:\/\/www\.freecovers.net\/preview\/0\/[a-z0-9]*\/big.jpg'`

What's supposed to be in $@?

Anyway, the problem is that the search doesn't return anything sometimes... I just tried your example search and three times out of five it returned nothing, so you might want to do something ugly like

Code: Select all

for i in 1 2 3 4 5 ; do
  URL2=`wget -q -O - "$@" $URL | grep -o -m 1 -i 'http:\/\/www\.freecovers.net\/preview\/0\/[a-z0-9]*\/big.jpg'`
  [ -n "$URL2" ] && break
done

if [ -z "$URL2" ] ; then 
# tell user there's no cover
fi

[seaside]

Probably something on the server because the pure url line failed 3 out of five for me.

Dougal's solution brings it down to 1 failure in the 15 tries I attempted with this -

Code: Select all

for i in 1 2 3 4 5 ; do
  URL2=`wget -q -O - "$@" $URL | grep -o -m 1 -i 'http:\/\/www\.freecovers.net\/preview\/0\/[a-z0-9]*\/big.jpg'`
  wget "$URL2" -O $HOME/album_art.jpg
  [ -n "$URL2" ] && break
done

How can something be ugly if it works

Cheers,
s

[Dougal]

Code: Select all

for i in 1 2 3 4 5 ; do
  URL2=`wget -q -O - "$@" $URL | grep -o -m 1 -i 'http:\/\/www\.freecovers.net\/preview\/0\/[a-z0-9]*\/big.jpg'`
  wget "$URL2" -O $HOME/album_art.jpg
  [ -n "$URL2" ] && break
done

That should be

Code: Select all

for i in 1 2 3 4 5 ; do
  URL2=`wget -q -O - "$@" $URL | grep -o -m 1 -i 'http:\/\/www\.freecovers.net\/preview\/0\/[a-z0-9]*\/big.jpg'`
  [ -n "$URL2" ] && break
done
wget "$URL2" -O $HOME/album_art.jpg

there's no point in trying to download if you found nothing (I copied the wrong line in my previous post...)

Anyway, URL2 is unnecessary:

Code: Select all

for i in 1 2 3 4 5 ; do
  URL=`wget -q -O - "$@" "http://www.freecovers.net/api/search/${STRING// /-}/Music+CD" | grep -o -m 1 -i 'http:\/\/www\.freecovers.net\/preview\/0\/[a-z0-9]*\/big.jpg'`
  [ -n "$URL" ] && break
done
wget "$URL" -O $HOME/album_art.jpg

[zigbert]

Dougal, you're just great.

Code: Select all

for i in 1 2 3 4 5 ; do
  URL=`wget -q -O - "$@" "http://www.freecovers.net/api/search/${STRING// /-}/Music+CD" | grep -o -m 1 -i 'http:\/\/www\.freecovers.net\/preview\/0\/[a-z0-9]*\/big.jpg'`
  [ -n "$URL" ] && break
done
wget "$URL" -O $HOME/album_art.jpg

The only thing I don't understand is the "$@"
I have of course grabbed the code somewhere out there....... I rarely do anything on my own


Thank you
Sigmund

[vovchik]

Dear Dougal,

Works great for me too. I only wish there were some way of getting better search results. If you run it with Beatles "Abbey Road" as parameters, for example, www.freecovers.net offers up a rather weird cover. Also, if you change the search to "DVD+Movie" and type bananas, you won't get Woody Allen's film by that name, but you do get Zelig, Manhattan and Sleeper using those words.

With kind regards,
vovchik

[technosaurus]

I had a similar problem with wget in a simple flash download script - the problem ended up being output to stdout (piped through tar) instead of just doing a straight download and then manipulating the file. The for loop seems like an odd fix, but I'm not sure how to narrow down the problem to verify that it isn't the server - maybe try busybox wget?

[Dougal]

I had a similar problem with wget in a simple flash download script - the problem ended up being output to stdout (piped through tar) instead of just doing a straight download and then manipulating the file.

No, I tried without the -q and redirection to stdout and it still fails but wget reports success, so it's a problem with the server.

Sigmund, I just had about 15 minutes where freecovers.net completely failed, no matter how many times I tried... you might want to try a different server, like for example:

Code: Select all

URL=$(wget -q -O - "http://www.albumart.org/index.php?srchkey=${STRING// /+}&itempage=1&newsearch=1&searchindex=Music" | grep -F 'View larger image' | grep -m1 -o 'http://ecx.images-amazon.com/images/.*.jpg"' | cut -d'"' -f1)

(note that this gives you the smaller image, the one you see in the Amazon entry and is clickable.

[Dougal]

if you change the search to "DVD+Movie" and type bananas, you won't get Woody Allen's film by that name, but you do get Zelig, Manhattan and Sleeper using those words.

I don't know... I tried "woody allen bananas" and it got me crap like "bananas in pajamas"! then it stopped working and I gave up.

[zigbert]

Omair
Great stuff, I have implemented it like this:

Code: Select all

album_art (){
	rm $WORKDIR/album_art.jpg
	for i in 1 2 3 4 5 ; do
		URL=`wget -q -O - "http://www.freecovers.net/api/search/${1// /-}/Music+CD" | grep -o -m 1 -i 'http:\/\/www\.freecovers.net\/preview\/0\/[a-z0-9]*\/big.jpg'`
	  [ -n "$URL" ] && break
	done
	#if no go, use another server (smaller image)
	[ ! "$URL" ] && URL=$(wget -q -O - "http://www.albumart.org/index.php?srchkey=${STRING// /+}&itempage=1&newsearch=1&searchindex=Music" | grep -F 'View larger image' | grep -m1 -o 'http://ecx.images-amazon.com/images/.*.jpg"' | cut -d'"' -f1)
	#download
	wget "$URL" -O $WORKDIR/album_art.jpg
}

The smaller alternative is better than nothing.


Thank you
Sigmund

[technosaurus]

if you change the search to "DVD+Movie" and type bananas, you won't get Woody Allen's film by that name, but you do get Zelig, Manhattan and Sleeper using those words.

I don't know... I tried "woody allen bananas" and it got me crap like "bananas in pajamas"! then it stopped working and I gave up.

damn "filter bubble" must assume that children use your computer with all of the puppy googling ... probably even worse if you use pfix=ram