Librerator - multi-format e-reader, fork of KPV

[Kai771]

@dider

I'm sure this is not what you'd like to hear, but - sorry, I can't help you.

I don't have K4NT, only K3, and I was unable to reproduce this problem. Admitedly, I don't have DuoKan installed, and don't want to fiddle with it, but I do have both KUAL and Cool Reader 3, and it works just as it's supposed to.

Launchpad.ini file has nothing to do with either KUAL or Librerator, especially on K4NT - it's used by launchpad application starter, which is unavailable on K4NT, and therefore irrelevant.

As far as I know, "invalid capture index" is not one of the messages that Librerator will report - it's one of the messages of luajit. I can't even begin to imagine what caused it.

Some ideas that might or might not help:

1. did you use librerator_kual.zip, or you made your own KUAL config file? If you used librerator_kual.zip, did you use it as is or you tried to edit it? If you edited it, did you use editor that supports unix style EOL?

2. did you change your kindle locale? Maybe DuoKan changed it for you?

It might be helpful to supply your kindle model (black or silver), your firmware version, and full list of hacks and modifications that you did. Some other K4NT user might've encountered the similar problem and might be able to help you more then I can.

Regards,
Kai

[twobob]

I have a k4 but no duokan. All the other stuff and cannot reproduce the error.

I'm sorry but I too refuse to run doukan so my usefulness also ends here.

Apologies.

[dider]

Originally Posted by Kai771

@dider
1. did you use librerator_kual.zip, or you made your own KUAL config file? If you used librerator_kual.zip, did you use it as is or you tried to edit it? If you edited it, did you use editor that supports unix style EOL?

Thank you for your replay!
I have not modified any files, everything was copied as it is from the archives linked in the first post. (I'm aware of the bugs caused by editing files with windows notepad, but that's not the case here.)

The only thing which I added to KUAL besides Librerator is "switch to Duokan" extension. It was added and used before I added the Librerator extension (might be important?).

2. did you change your kindle locale? Maybe DuoKan changed it for you?

I didn't change it myself, manually, no. I don't know if Duokan changed that though, how can I check that?

It might be helpful to supply your kindle model (black or silver), your firmware version, and full list of hacks and modifications that you did. Some other K4NT user might've encountered the similar problem and might be able to help you more then I can.

Kindle 4NT Black, Firmware 4.1.1
Modifications:

• Jailbreak
• devcerts+kindlet_jailbreak (KUAL prerequisites)
• Kindle Launcher.azw2 (KUAL)
• update_cr3runner_2.7_k4x_install.bin (cool reader)
• cr3-kindle-2012-10-04.zip (cool reader folders: cr3, launchpad, qtKindle, usrapp) link
• newest Duokan for K4NT
• DuokanLauncher for KUAL link
• appended "DefaultSystem=3" to DK_System\Lite\config.ini

This is by no means a very serious issue, I'm just curious of what is happening. Thank you for your interest Kai, great program!

Originally Posted by twobob

I have a k4 but no duokan. All the other stuff and cannot reproduce the error.
I'm sorry but I too refuse to run doukan so my usefulness also ends here.
Apologies.

Thank you very much for checking that! I'm probably going to uninstall Duokan to see if anything changes. I'm going to report when I'm done with that.

[dider]

OK, It finally works! I still don't know what causes the "invalid capture index" error, but I'm closer.

I uninstalled every modification on my Kindle (except for KAUL and Librerator of course) but the problem persisted.

The only thing that could contain the root of the problem was the "documents" folder itself.

So, first I suspected the collections to cause the problem, as they were created with Calibre's plugin, which used to be very buggy when I was using it back in the day. That didn't change a thing, Librerator still reported "invalid capture index" error.

OK, so I decided to delete all the files from the "documents" folder (except for KUAL of course). It works - no more "invalid capture index" error.

So now I'm wondering which file exactly causes the problem. Is it the name of a file that is problematic or something else? Polish diacritics? Too many files?

I have my buggy "documents" folder copied on my HDD (178MB of files), I'm attaching the list of files from this folder. (Those are mainly books, articles and webpages sent via Calibre and Readability, in English and Polish language). I can upload the files if it helps.

[knc1]

Originally Posted by dider

OK, It finally works! I still don't know what causes the "invalid capture index" error, but I'm closer.

I uninstalled every modification on my Kindle (except for KAUL and Librerator of course) but the problem persisted.

The only thing that could contain the root of the problem was the "documents" folder itself.

So, first I suspected the collections to cause the problem, as they were created with Calibre's plugin, which used to be very buggy when I was using it back in the day. That didn't change a thing, Librerator still reported "invalid capture index" error.

OK, so I decided to delete all the files from the "documents" folder (except for KUAL of course). It works - no more "invalid capture index" error.

So now I'm wondering which file exactly causes the problem. Is it the name of a file that is problematic or something else? Polish diacritics? Too many files?

I have my buggy "documents" folder copied on my HDD (178MB of files), I'm attaching the list of files in this folder. (Those are mainly books, articles and webpages sent via Calibre and Readability, in English and Polish language). I can upload the files if it helps.

start with the simplest assumption - only one of the books and/or collections causes the problem . . . .

then do a binary search for the problem -
put 1/2 the books and/or collections back on -
if problem reoccurs, then problem is in that half -
if problem does not reoccur, then the problem is in that half you did not put on -
eleminate the 1/2 which does not cause the problem -
repeat with the half that does cause the problem by choosing only half of it.
repeat ^^^

For 256 books - will take at most 8 tries to find it.

If more than one book and/or collection causes the problem - then when doing the above, be prepared to find more than one "1/2" sample to cause the problem - adapt your search testing in that case.

[dider]

Originally Posted by knc1

start with the simplest assumption - only one of the books and/or collections causes the problem . . . .

then do a binary search for the problem -
put 1/2 the books and/or collections back on -
if problem reoccurs, then problem is in that half -
if problem does not reoccur, then the problem is in that half you did not put on -
eleminate the 1/2 which does not cause the problem -
repeat with the half that does cause the problem by choosing only half of it.
repeat ^^^

For 256 books - will take at most 8 tries to find it.

If more than one book and/or collection causes the problem - then when doing the above, be prepared to find more than one "1/2" sample to cause the problem - adapt your search testing in that case.

Using binary search I have found the file that was bugging Librerator!
It's name is:
Wisdom from Psychopaths-asin_TWUKDKFYQKO6PVBAVH2EWS5KFB7YI7PG-type_PDOC-v_0.azw%22;

So, it appears that it was the bugged extension of the file that caused Librerator to report the "invalid capture index" message. The default Kindle's software simply doesn't see the bugged file, while Librerator can't see the whole directory containing this bugged file.

I have no idea how that extension came to be. When I change the extension to azw the file reappears in Kindle's software and the directory becomes accessible in Librerator. Don't remeber how I created the file unfortunately, it contains only the first page of this article.

thank you all for assistance

[Kai771]

@dider

I'm glad you managed to sort it out. Thank you very much for reporting back.

Regards,
Kai

[kacir]

Originally Posted by dider

It's name is:
Wisdom from Psychopaths-asin_TWUKDKFYQKO6PVBAVH2EWS5KFB7YI7PG-type_PDOC-v_0.azw%22;

I have no idea how that extension came to be.

You have probably copied a quote at the end of the filename and quote " is [sometimes] encoded as %22 when it is a part of URL, so instead of http://example.com/"foobar" you use http://example.com/%22foobar%22

http://www.w3schools.com/tags/ref_urlencode.asp
http://en.wikipedia.org/wiki/Help:URL

[dider]

Originally Posted by kacir

You have probably copied a quote at the end of the filename and quote " is [sometimes] encoded as %22 when it is a part of URL, so instead of http://example.com/"foobar" you use http://example.com/%22foobar%22

http://www.w3schools.com/tags/ref_urlencode.asp
http://en.wikipedia.org/wiki/Help:URL

The file wasn't created or edited manually though (at least not intentionally). I've never fiddled manually with the files on my Kindle, especially not with their names or extensions.

Neither was it sent via "send to kindle" function, as I get "The following document could not be delivered to the Kindle" message when I try to do it now.

And what about that semicolon? You may not have noticed that, but there's a semicolon in the extension - it seems like some badly written script could do that...

[twobob]

it's a tough one.

Perhaps the lesson learned here might be greater corner-case name detection.

.azw"; is a pretty nasty corner case though!

Either way, lesson learned by all perhaps : )

[sk8terek]

I am really keen on using Librerator on my K4NT but recently I have noticed a little bug. Librerator's table of contents does not display letters other than latin (a-z and 0-9). Instead, it shows a little rectangle. For example "Lech Wałęsa" is displayed as "Lech Wa__sa" (_ means rectangle). This bug occures in every e-book format (pdf, mobi, epub).

[twobob]

Font lacking unicode support perhaps?

That is odd because just a few posts up https://www.mobileread.com/forums/sho...&postcount=124 that guy says his displays polish diacritcs?

I am going to butt out now until I have used this a bit more myself (don't have diacritics my side) but thought I should point out the OP post.

[Kai771]

@sk8terek

It would be helpful if you could provide a test document.

Regards,
Kai

[PoP]

Originally Posted by twobob

Font lacking unicode support perhaps?
...
I am going to butt out now until I have used this a bit more myself (don't have diacritics my side) ...

This site might come handy for cutting&pasting diacritics when testing press "toggle reference sheet".

[dider]

Originally Posted by twobob

Font lacking unicode support perhaps?

That is odd because just a few posts up https://www.mobileread.com/forums/sho...&postcount=124 that guy says his displays polish diacritcs?

I am going to butt out now until I have used this a bit more myself (don't have diacritics my side) but thought I should point out the OP post.

I've never got to actually test if Librerator properly lists files with polish diactricts in their names/titles as I couldn't get a list of files at all! (due to the azw"; bug).

Now, that I uploaded some Polish books with proper extensions I can confirm what sk8terek wrote - I see squares too.

edit: It looks good when I changed the default font to Helvetica!