Digital Library of India: Download all that you can...

[:http://dli.iiit.ac.in/|Digital Library of India] has been unveiled, but with a shocker of an interface. But not much can be expected out of a "Government of India" project, as they always manage to find just the right technologies (or people?) for their job (Why, e-governance in India is all ready to go Microsoft's way. When M$ boasts riches, we can all show our kids its logo and say that our government of poor people is one of their key customers). They use TIFF format for image scans of thousands of books probably from libraries all over India. There are two petty interfaces which need you to download a software to be able to view. The software, in turn, needs to be registered to be able to use. Last month, it was the [:http://ildc.gov.in/Kannada/kdownload.htm|disappointing set of tools] released by TDIL, and this month, the DLI. Not to mention, the projects are obviously worth a lot, but shabbily done. Idea seems to be right, but implementation has been terribly bad.

A Quick Script

To overcome the toil between interest for books on DLI (which otherwise are not easily available) and the irritation of shabby interface, I wrote a (shabby) script that batch downloads the TIFFs. It is a quickly written script with pieces from here and there that has many stupid parts (which undoubtedly would be mine). But mainly it works, just like the projects I've been mentioning here. Serves right, in a way. It needs you to paste URLs pointing to TIFF of starting page for each book, with the filename removed. Pretty clumsy, yes. But that was convenient for me, since I removed the frames from their web page while viewing, browsed through the list, and clicked on each to check the quality. Saves the irritation for the next many pages. Ah, and you'll need to paste the URLs onto a file. For use on just a single book, it is easy to modify, anyway. Try it, modify it and let me know if you improve it. #!/bin/sh #Get your favourite book from DLI: Specify the start page and end page, and this script takes care of the rest. #Caveat: you'll need to specify the base URL, though. PATH=/bin:/usr/bin:/usr/local/bin progname=`basename $0` case $# in 0) 1>&2 echo $progname: usage $progname start end; exit 1 ;; esac start=$1 end=$2 prefix=0000 echo "Enter path for the file to read:" read file #exec > $HOME/log_dli.txt x=1 lns=`wc -l $file` echo "LNS: $lns" index=`expr $start` while [ $x -lt $(wc -l <$file) ] do url=`head -n $x $file | tail -n 1` index=`expr $start` mkdir $x cd $x while [ $index -le $end ] do if [ $index -lt 10 ] then digits=000 elif [ $index -ge 10 ] then if [ $index -lt 100 ] then digits=00 else digits=0 fi else echo "bah. \n" fi WGET_OUTPUT=$(2>&1 wget --timestamping --progress=dot:mega \ "$url$prefix$digits$index.tif") # wget $url$prefix$digits$index.tif if [ $? -ne 0 ] then # wget had problems. echo 1>&2 $0: "$WGET_OUTPUT" Exiting. fi if (echo "$WGET_OUTPUT" | fgrep 'Not Found') > /dev/null then break else echo "~~~~ Page found. Downloaded. ~~~~ \n" fi index=`expr $index + 1` done cd .. x=`expr $x + 1` done


Note: Make sure the URL is in the form of http://dli.iiit.ac.in//server12/disk3a/TO%20SUBMISSION/KANNADA/Bharatiya%20Tatva%20Shastra%20Samgraha//PTIFF and each URL is to be placed on the file in new line (the script doesn't detect empty lines).

See also:

• [:http://hpnadig.net/notes/converting-and-merging-tiff-to-pdf|Converting Tiffs to PDF].
• [:http://sampada.net/Kannada-ebooks-torrent-1-and-2-Index|Torrents to several books] I prepared using this script.

Enjaaay!