Option to open prc/mobi files with either FBReader or CoolReader

[Louvre]

Right now prc files are automatically opened with FBReader. But even the latest desktop version 0.10.7 does not open highly compressed prc files. Examples of files that does not work in OpenInkpot (worse: a reset is necessary because Hanlin/BeBook/Lbook is totally blocked) but works with CoolReader (in the April version of the Lbook firmware):

• German Wikipedia of March 2005, the largest one file version (versions that come in several files do not work in CoolReader either):  http://www.axelschaefermdb.de/service/wikipedia.html
• Webster English dictionary version 2.1 (high compression and searchable with virtual keyboard):  at the end of this message

[dottedmag]

long-press to open menu.

If you want the permanent switch - there's already open ticket, search for it.

[lunohod]

CoolReader doesn't support prc/mobi files. There is a proprietary plugin in the stock firmware which isn't ported to the OpenInkpot for obvious reasons.

[Louvre]

Replying to dottedmag:

long-press to open menu.

If you want the permanent switch - there's already open ticket, search for it.

Thank you very much. I browsed the active tickets but did not find the one you told me about. Permanent switch...if you are talking of replacing FBReader by CoolReader for prc/mobi files in the next version/snapshot of OpenInkpot, yes, that sounds right since CoolReader is superior right now for this kind of file (the highly compressed ones I tested at least). A semi-permanent switch adustable in a MadShelf menu (a preference to use FBreader or CoolReader depending on the kind of file when the user does a short press of the file number) would be interesting too.

[Louvre]

lunohod I answered before reading your message.

Too bad about that proprietary code.

I will register at fbreader to inquire about high compression and hyperlinks between several files. What about adding a keyboard? Does it depend on them or on you? Is the keyboard code from CoolReader free and can it be ported?

The Webster dictionary for example also has a standard compression version ( here), and it works fine in FBReader desktop version 0.10.7. Searches work well.

[Louvre]

I tried this Webster dictionary with standard compression in OpenInkpot. It does not work, locks up everything, and I have to do a reset to reboot. The version of FBReader in OpenInkpot is not 0.10.7 ?

[lunohod]

Our FBReader can open both files. HUFF/CDIC decompression is already implemented. The only problem is the lack of RAM, because FBReader loads the entire file into RAM. This approach isn't that good if you have big files and little RAM.

There is a virtual keyboard in our FBReader. Check the User Guide.

[Louvre]

Could FBReader be adapted to cache some memory in a swap on the internal storage? I already thought before that it might be a good thing to tell OpenInkpot users to leave 100MB free on the internal storage. This way the limits of OpenInkpot could be pushed: more fonts,...

[Louvre]

I managed to load a 1.8 MB .prc dictionary (a simple one with small definitions), but it took a long time, and then the index was not available. Could you add the possibility to use the index like in CoolReader, or must I file a bug report at the FBReader site?

I managed to do a search with the virtual keyboard, and by pressing 0 (next result) I got to the word quickly enough. With words used in the definitions of too many other words, it could take a very long time.

Big (good quality) dictionaries are unusable because of the memory problem.

Did you file a bug at the FBReader site to stop it from loading the whole file in RAM, or must I do it? Or can you yourself adapt FBReader to the limited memory of the Hanlin V3?