KOReader: a document reader for PDF, DJVU, EPUB, FB2, HTML, ... (GPLv3)

[Doitsu]

Originally Posted by barobach_so

but koreader how has a RTL and LTR option in reflow engine ? when I read a pdf with koreader I can change this option.

AFAIK, KOReader uses K2pdfopt as the reflow engine, which treats PDF files like images and doesn't require character shaping.

Originally Posted by barobach_so

there is a glossary.BGL that say it is graphical and no need font ( I don't know really waht is it ), can it be somehow helpful ?

Since sdcv is a command line program and the KOReader code only returns text, you can't use dictionaries with pictures in them.

Originally Posted by barobach_so

I'm thinking of change bgl to html (babylontohtml program ) and have a html glossary ( already have ) then revers the characters somehow and maybe connect them to each other by test mechanics....)

You could convert the .bgl file to a .babylon (GLS) file with StarDict Editor, edit the .babylon file and finally compile it to a StarDict dictionary, however, character shaping and RTL handling isn't exactly a trivial matter.

[Ebri]

Originally Posted by guerrilla

I'm still unsure about how I got stuck in this though. At a closer look it seems that that message is printed only when sdcv creates the cache file (.oft extension) and this should be the first time you search for a term. That's why this issue got unnoticed so far: you install your dictionary and search for a term, sdcv gives the "save to cache" message, json.decode fails to decode, koreader returns the dummy table, you think: "uh, word not present in the dictionary" and make another search which, from now on, goes smoothlessly.
For some reason I was getting that warning again and again.

Anyway, I will gladly do a report to github, I'm very happy to help

I am unsure about it, but in some cases on Kobo Aura HD dictionary gave up after some time, and haven't found anything, then after restarting koreader, it worked again. I am unsure, but it may be the same case (symptoms are the same, and it starts eg. after a wakeup.) I think I closed the related issue on github though.

[guerrilla]

@Ebri You can find out if it's the same issue if you change in koreader.sh the line that goes:

./reader.lua "$@" 2> crash.log

in

./reader.lua -d "$@" 2> crash.log 1> koreader-debug.txt

You can use a convenient path for koreader-debug.txt, on kindle I use /mnt/us/documents/koreader-debug.txt so that I can easily access the file in the stock reader.
Then, next time you encounter a problem with a dictionary search, close koreader and open the log file. If there's a line in it that goes:

# JSON data cannot be decoded frontend/JSON.lua:525: frontend/JSON.lua:240: can't parse JSON at char 1 of: save to cache

that's this issue.

[barobach_so]

Originally Posted by Doitsu

AFAIK, KOReader uses K2pdfopt as the reflow engine, which treats PDF files like images and doesn't require character shaping.


Since sdcv is a command line program and the KOReader code only returns text, you can't use dictionaries with pictures in them.


You could convert the .bgl file to a .babylon (GLS) file with StarDict Editor, edit the .babylon file and finally compile it to a StarDict dictionary, however, character shaping and RTL handling isn't exactly a trivial matter.

hi
I have a txt glossary of en 2 persian dic that i reversed with arabshaping ( convert to srt then use arabshaping convert to txt )
but when I try to convert it to stardict format ( ifo ) in PyGlossary in ubuntu
it shows this error

usr/share/pyglossary/src/interface.py:88: DeprecationWarning: os.popen3 is deprecated. Use the subprocess module.
p3 = os.popen3('dictzip')
/usr/share/pyglossary/src/interface.py:98: DeprecationWarning: os.popen3 is deprecated. Use the subprocess module.
p3 = os.popen3("%s '%s.dict'"%(dictzipCmd,filename))

and the output file ( have a size e.g. 26 mb ) doesnt show in koreader

any idea ?

[Doitsu]

Originally Posted by barobach_so

I have a txt glossary of en 2 persian dic that i reversed with arabshaping ( convert to srt then use arabshaping convert to txt ) but when I try to convert it to stardict format ( ifo ) in PyGlossary in ubuntu it shows this error [...]

According to the error message, Python complained about a line in interface.py. However, the latest pyGlossary version, 2010.05.29, doesn't contain interface.py.
It looks like you're using an older version of pyGlossary and/or maybe Python 3.4 instead of Python 2.7.

If you've installed pyGlossary with a package manager, download the latest version from the developer.

Originally Posted by barobach_so

[...] the output file ( have a size e.g. 26 mb ) doesnt show in koreader

To exclude conversion errors you may want to test your dictionary with GoldenDict first, which is also available for Linux.
If your dictionary works fine with GoldenDict, but doesn't work with KOReader, you might have to extract the .dict file from the .dz file, if pyGlossary has created one.

[barobach_so]

Originally Posted by Doitsu

According to the error message, Python complained about a line in interface.py. However, the latest pyGlossary version, 2010.05.29, doesn't contain interface.py.
It looks like you're using an older version of pyGlossary and/or maybe Python 3.4 instead of Python 2.7.

If you've installed pyGlossary with a package manager, download the latest version from the developer.


To exclude conversion errors you may want to test your dictionary with GoldenDict first, which is also available for Linux.
If your dictionary works fine with GoldenDict, but doesn't work with KOReader, you might have to extract the .dict file from the .dz file, if pyGlossary has created one.

hi
thanks for reply
i downloaded pyglossary from ubuntu software center
and output format is alreadey dz
.ifo .idx .idz

but i will test with other versions too

I should add
the owner of this link you sent to me is Saeed Rasuli also an Iranian and he send me the glossary
and second thing is when I run pyglossary in windows ( both GUI and cmd form ) the output file doesn't have any size only an icon

[barobach_so]

Hi
with Doitsu & hawhill Guide and my friend help we finally have RTL Persian Dictionary in koreader
here is the picture : ( You can compare with post #745 )


here is the process:
convert babylon glossay to HTML file with BabylontoHtml
reverse the character order by Arabshaping
convert reveresed HTML to stardict ( ifo ) by stardict editor

add Persian unicode font to koreader font folder and change the defaults in font.lua in koreader

[Belenos]

Does it support AZW3 files?

[hawhill]

@barobach_so: Great! And thanks for sharing the processing instructions!
@Belenos: No. You'd have to convert those.

[guerrilla]

@barobach_so You should be able to right align the text in the quick lookup window too, maybe playing a little with frontend/ui/widget/dictquicklookup.lua?

[barobach_so]

Originally Posted by guerrilla

@barobach_so You should be able to right align the text in the quick lookup window too, maybe playing a little with frontend/ui/widget/dictquicklookup.lua?

Hi
on a trip now, I will test later
I think its because arabshaping it work on srt files in subtitles
and it happens there too
but I will test it

Thank you

[hawhill]

I don't think right alignment is currently implemented. That would have to get implemented here: https://github.com/koreader/koreader...tboxwidget.lua - so don't bother to fiddle with the - higher level - dictquicklookup.

[Belenos]

Originally Posted by hawhill

@Belenos: No. You'd have to convert those.

Thanks, any chances for a future support?

[hawhill]

You'd have to ask this the CREngine developer(s), but I think it is highly unlikely. CREngine started as a FB2 renderer, had EPUB2 (at least more than enough of it to be useful) added later. Switching that to EPUB3 compatibility is a major feat alone, adding support for a proprietary format on top of that is an immense amount of work. So no, I don't think that will happen soon.

Maybe if some new reader engine comes up, but where from?

[NiLuJe]

The Readium SDK? Granted, AFAICT, there's no proper Linux port (yet?), and it may or may not rely on a full webkit setup, but that's still potentially the best contender. (And what Kobo's using for their kepub ACCESS renderer, if I understood everything properly...).