[Kindle Touch] Instafetch

[seird]

InstaFetcher 1.1

This little Kindle Touch hack adds the option "Fetch Instapaper Articles" to the GUI launcher provided by YiFanLu. (https://wiki.mobileread.com/wiki/Kind...g#GUI_Launcher)
It will download a .mobi (or .epub) ebook from instapaper containing your Read Later articles.

Change the username and password in bin/instafetch.sh and download your Instapaper to your Kindle whenever you want!

Edit:
Statically linked curl binary included as sugested below (Thanks NiLuJe!). This should work on K4 and is tested on K5 Touch.

Edit2:
Added credits to README. Changed the name of the script because Insafetch as a name is already used by the great android tool.

[knc1]

Originally Posted by seird

InstaFetch 1.0

This little Kindle Touch hack adds the option "Fetch Instapaper Articles" to the GUI launcher provided by YiFanLu. (https://wiki.mobileread.com/wiki/Kind...g#GUI_Launcher)
It will download a .mobi (or .epub) ebook from instapaper containing your Read Later articles.

The script will only work with the libcurl optware package installed provided by the 'extend' Kindle hack: https://www.mobileread.com/forums/sho...d.php?t=161704
$ ipkg install libcurl

Change the username and password in bin/instafetch.sh and download your Instapaper to your Kindle whenever you want!

Why don't you include a libcurl binary with your script?
We have similar, worked examples, indexed.
That would save noobie users the need to install/learn optware.

[knc1]

Originally Posted by knc1

Why don't you include a libcurl binary with your script?
We have similar, worked examples, indexed.
That would save noobie users the need to install/learn optware.

Correction, it needs the curl application, not just libcurl.

[geekmaster]

Originally Posted by knc1

Correction, it needs the curl application, not just libcurl.

Preferably static linked, so we do not have to support multiple versions or curl for multiple kindle versions. Static linking is not bad for a kindle, being mostly a "single task" device, and having a bunch of static-linked apps only consumes a tiny portion of the mmc.

[NiLuJe]

Haven't looked at the code, but I'm pretty sure there's also at least a basic wget binary in the base Kindle system, wouldn't that be sufficient?

EDIT: Okay, looked at the code, and, yeah, no, pretty sure busybox's wget doesn't handle cookies either .

So, yep, building a curl binary shouldn't be much trouble, there's no real dependencies to speak of if you don't need SSL support. (And if you do, it supports both gnutls and openssl, and while openssl is a bit of a nightmare to cross-compile, gnutls, being a GNU project, should be using a sane buildsystem ^^).

EDIT²: Here you go (targeted at a K4/K5, it won't run on older devices). (OpenSSL static linked to avoid the OpenSSl 0.9.8/1.0.x mess, which explains the size). [I didn't test it, but it looks okay].

[seird]

Originally Posted by NiLuJe

Haven't looked at the code, but I'm pretty sure there's also at least a basic wget binary in the base Kindle system, wouldn't that be sufficient?

EDIT: Okay, looked at the code, and, yeah, no, pretty sure busybox's wget doesn't handle cookies either .

So, yep, building a curl binary shouldn't be much trouble, there's no real dependencies to speak of if you don't need SSL support. (And if you do, it supports both gnutls and openssl, and while openssl is a bit of a nightmare to cross-compile, gnutls, being a GNU project, should be using a sane buildsystem ^^).

EDIT²: Here you go (targeted at a K4/K5, it won't run on older devices). (OpenSSL static linked to avoid the OpenSSl 0.9.8/1.0.x mess, which explains the size). [I didn't test it, but it looks okay].

Good point! Makes installing a lot easier.

I have edited the initial post and uncluded the statically linked binary. Thanks NiLuJe!

[NiLuJe]

@seird: Does that mean it actually runs correctly? (Remember I said I hadn't actually tested it ;D).

Sidenote: You might want to include the contents of the CREDITS file in your readme, though .

[seird]

Originally Posted by NiLuJe

@seird: Does that mean it actually runs correctly? (Remember I said I hadn't actually tested it ;D).

Sidenote: You might want to include the contents of the CREDITS file in your readme, though .

It actually works!

I added the CREDITS, and changed the name of the script.

Thanks!

[NiLuJe]

@seird: Good to know, thanks .