... but how to do it?

http://www.softlab.ntua.gr/~ttsiod/buildWikipediaOffline.html

is this possibly, oh mighty iLiad developers? 2.7 GB is a lot, but with modern 4 or 8 GB flash card I'd gladly reserve that space on my cf card. As long as you stay away from media files, storage is plentiful...

... but how to do it?

http://www.softlab.ntua.gr/~ttsiod/buildWikipediaOffline.html

is this possibly, oh mighty iLiad developers? 2.7 GB is a lot, but with modern 4 or 8 GB flash card I'd gladly reserve that space on my cf card. As long as you stay away from media files, storage is plentiful...


There already is one in Mobipocket format in German, and Tome Raider in English.


http://www.mobileread.com/forums/showthread.php?t=11778&highlight=mobipocket+offline+wikipedia

... but how to do it?

http://www.softlab.ntua.gr/~ttsiod/buildWikipediaOffline.html

is this possibly, oh mighty iLiad developers? 2.7 GB is a lot, but with modern 4 or 8 GB flash card I'd gladly reserve that space on my cf card. As long as you stay away from media files, storage is plentiful...

I'm playing with that script now. I'll see what I can come up with for the iLiad.

The problem is distributing it... If I do make an iLiad specific version, it's way too big to host anywhere...

I would be willing to set up a host.

A reader version would be very nice too, although I am wondering how easily one would be able to navigate such a thing.

A reader version would be very nice too, although I am wondering how easily one would be able to navigate such a thing.

I'd say that without text or stylus entry, it'd be darn near impossible to even search for anything with it...

I doubt that anyone has the resources to do the english Wikipedia as Mobipocket ebook.
The german version at 500 MB took several days on several computers to generate.
It would also be over 2 GB which is the practical limit for SD cards.

I doubt that anyone has the resources to do the english Wikipedia as Mobipocket ebook.
The german version at 500 MB took several days on several computers to generate.
It would also be over 2 GB which is the practical limit for SD cards.

The iLiad supports 4GB CF cards, but I'm not at all sure that the MobiPocket Reader (or any other) would be happy to load a 4GB file :).

It would be several files anyway. The real problem is that not all devices being able to show Mobipocket files can handle 4 GB SD cards. That means you have a device specific Mobipocket ebook. I am sure Mobipocket would frown upon this.

Well, I've got the posted solution working from the terminal. It required a port of php, perl, bzip2, and xapian. With the database, it uses almost my entire 4gb CF card.

Extracting from the bzip archive and generating the html is a bit slow (~12 seconds), but it's doable. If someone could write a gtk gui for it....we would have the wikipedia in the palm of our hands! :2thumbsup :grin2: :book2:

root@ereader:/mnt/cf/offline.wikipedia# ./mywiki.pl iliad
0: (abort)
1: (100%) Iliad
2: (72%) The Iliad
3: (72%) Little Iliad
4: (72%) Ambrosian Iliad
5: (63%) Category:Characters in the Iliad
6: (63%) Wikipedia:Articles for deletion/Ambrosian Iliad
7: (63%) Wikipedia:Votes for deletion/Ambrosian Iliad

Select a number: 1

This is the search program. It's case insensitive, so I think the article about the iRex iLiad is overwritten by the book...

Selected:
Iliad
rec00153enwiki-20070802-pages-articles.xml.bz2p

And now that the result html has been generated, I open it from the contentlister.

Wow - that's impressive, Adam!

I'd be willing to set up a host too, even if it was just for for the initial seeding of the file. Afterwards we could generate a torrent and I'm sure people would help seed the file.

And for text entry:
Isn't it possible to develop a hook to provide "selection search" as an option while reading another book?
Hardcore users could use a usb keyboard anyway...

That said, I'm not sure how many people would be willing to dedicate a 4GB card just for wikipedia.

The problem is that right now, the only way to search is from the terminal... Perhaps I can get python running with that mini webserver posted in the original topic.

As always adam ports faster than other people can follow the forum discussions :)

I can host it on my company's server. Well, unless lots of people keeps downloading it, but I don't think so - it is not _that_ useful for common users. Besides, there are plenty of filesharing portals, most of them are free.

But a quick-and-dirty gui would be needed, I guess... otherwise it is too cumbersome to use :S

And now that the result html has been generated, I open it from the contentlister.

Adam, you never cease to amaze me!

But a quick-and-dirty gui would be needed, I guess... otherwise it is too cumbersome to use :S

I was having a bit of sucess yesterday using the version of python compiled for the zaurus. Setting up a mini webserver, and using dillo to search and browse would probably be a pretty good soution...

But, since I messed up a system file and had to reflash, thus loosing some progress... Not to mention the fact that I have a new toy (http://www.mobileread.com/forums/showthread.php?t=12786) to play with, I may not be able to work on it further this week.