txt/epub - Guillemets conversion problem

julo · 10-16-2011, 01:04 PM

Whenever I have text with guillemets ( » « ) and convert to epub I am getting small t & capital T with caron [ ť Ť ] instead. Is there a setting I am missing?

theducks · 10-16-2011, 01:10 PM

Quote:

Originally Posted by julo

Whenever I have text with guillemets ( » « ) and convert to epub I am getting small t & capital T with caron [ ť Ť ] instead. Is there a setting I am missing?

TXT files don't always contain which code page (character set) was used

In Look and Feel options (on the convert screen) be sure the Code page is set to one that includes them. Usually it is blank and Calibre guesses. This time it got it wrong

julo · 10-16-2011, 03:03 PM

Quote:

Originally Posted by theducks

...Usually it is blank and Calibre guesses. This time it got it wrong

Thanks for the answer. Is there any way to find out what Calibre guessed?

theducks · 10-16-2011, 03:57 PM

Quote:

Originally Posted by julo

Thanks for the answer. Is there any way to find out what Calibre guessed?

Only if you had the Debug mode (must be turned on) log from when you converted. Then you would have to wade through (and know what you are looking at/for

)

jackie_w · 10-17-2011, 12:41 PM

Try to make sure your TXT file is saved in utf-8 format. Notepad++ (free) should be able to do that.

If Calibre doesn't guess correctly when you convert txt-to-epub then make sure you set Convert - Look&Feel - Input character encoding to utf-8 so Calibre won't have to guess.

julo · 10-17-2011, 09:01 PM

I will have to do a better job of keeping track of what I am actually doing. There are several steps involved before I even get to the Calibre conversion.

Quote:

Originally Posted by jackie_w

Try to make sure your TXT file is saved in utf-8 format....

Actually I would much rather keep it Latin1/ANSI. But what I don't understand, since the guillemets are part of ansi/latin1, why I have to say anything at all. All the other characters make it fine.

jackie_w · 10-17-2011, 09:27 PM

Quote:

Originally Posted by julo

Actually I would much rather keep it Latin1/ANSI. But what I don't understand, since the guillemets are part of ansi/latin1, why I have to say anything at all. All the other characters make it fine.

If you're sure that the TXT is in latin1 encoding then try setting Convert - Look&Feel - Input character encoding to latin1, again so Calibre won't have to guess. It will convert your source to utf-8 during the conversion to epub.

julo · 11-07-2011, 05:25 PM

Sorry it took so long to get back. Let me try to explain it different:

If I make a short test txt file (ANSI encoding using Notepad) and copy the following text into it:

Quote:

This is a test text using »guillemets«.

I get the following output in ePub:

Quote:

This is a test text using ťguillemetsŤ.

I tried convert settings of blank, ascii, cp1252, latin1 and utf-8.

Do you get the same?

theducks · 11-07-2011, 06:48 PM

Try HTML entities « and »

julo · 11-07-2011, 08:08 PM

Quote:

Originally Posted by theducks

Try HTML entities « and »

I need to maintain the txt file and I consider it worse having:

Quote:

This is a test text using &raquoguillemets&laquo.

which is what it will then also look in the ePub file. Now you really can't read it.

I do not see a way to edit ePub within Calibre, and if I understand you right I should make an html file first and from that the ePub. That is what I have been doing all along, not using Calibre though, but I thought Calibre could go direct from a txt file to ePub.

Am I wrong?

theducks · 11-07-2011, 08:50 PM

Quote:

Originally Posted by julo

I need to maintain the txt file and I consider it worse having:which is what it will then also look in the ePub file. Now you really can't read it.

I do not see a way to edit ePub within Calibre, and if I understand you right I should make an html file first and from that the ePub. That is what I have been doing all along, not using Calibre though, but I thought Calibre could go direct from a txt file to ePub.

Am I wrong?

No you are not wrong.

But there are exceptions (especially when you step outside ASCII).You may have found one

ldolse · 11-07-2011, 09:19 PM

I'm not sure why you have a preference for ANSI - ANSI means it will only work on your computer and people who change Windows to say it's in the same country as your computer. UTF8 will guarantee it will work on any computer on the planet.

Sometimes I want to kill/throttle whatever idiots at Microsoft insist on neverending support and propagation their asinine default ANSI settings - it's 2011 and those idiots are the number one reason this question comes up over and over again for well over a decade now.

ldolse · 11-07-2011, 09:24 PM

btw, if you save as ANSI the encoding in Calibre will definitely not be UTF-8. And I'm guessing Latin 1 may be iso-8859-1 - the problem with using ANSI is unless you're familiar with what ANSI is for your local country you need to play this guessing game. You also need to know what Python's name for the encoding is (and since we don't know what encoding you're using it's difficult for us to tell you.)

kovidgoyal · 11-07-2011, 09:27 PM

Generally speaking, windows' determination to not break backward compatibility is to be admired, but in this instance, I agree with ldolse, the crazy mix of encodings that windows systems typically operate under causes endless headaches.

ldolse · 11-07-2011, 09:45 PM

And in this day and age I can't actually imagine that 'anything' would break... It's not like ANSI support would go away, they'd just change their save dialogs to default to UTF8 and all newly saved text files going forward would suddenly have a BOM and work on every PC on the planet - files without a BOM will still open as ANSI....

The occasional Windows 95 era program that actually didn't understand a UTF-8 BOM would still render 98% of the text correctly for a western language.. I imagine in Asia there may be some workflows that assume gb2312/shift-JIS/big5 etc as the default, but you could even keep a registry setting around for people who insisted on staying in the dark ages.

10-16-2011, 01:04 PM	#1
julo Connoisseur Posts: 70 Karma: 12 Join Date: Apr 2010 Location: Pittsburgh area Device: prs-505,900,T2	txt/epub - Guillemets conversion problem Whenever I have text with guillemets ( » « ) and convert to epub I am getting small t & capital T with caron [ ť Ť ] instead. Is there a setting I am missing?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Txt->EPUB Conversion Issue	nrich61	Conversion	7	07-07-2011 06:03 PM
TXT to EPUB conversion issue	rockeh	Conversion	2	03-18-2011 10:50 AM
Preserving <br /> on epub -> txt conversion	billingd	Calibre	1	08-11-2010 06:24 AM
Conversion: EPUB to TXT	Starson17	Calibre	11	05-29-2010 12:31 PM
TXT conversion to ePub or LRF - paragraph formatting	Zapped	Calibre	6	10-23-2009 05:06 PM

10-17-2011, 12:41 PM	#5
jackie_w Grand Sorcerer Posts: 6,251 Karma: 16539642 Join Date: Sep 2009 Location: UK Device: ClaraHD, Forma, Libra2, Clara2E, LibraCol, PBTouchHD3	Try to make sure your TXT file is saved in utf-8 format. Notepad++ (free) should be able to do that. If Calibre doesn't guess correctly when you convert txt-to-epub then make sure you set Convert - Look&Feel - Input character encoding to utf-8 so Calibre won't have to guess.

11-07-2011, 06:48 PM	#9
theducks Well trained by Cats Posts: 30,990 Karma: 60358908 Join Date: Aug 2009 Location: The Central Coast of California Device: Kobo Libra2,Kobo Aura2v1, K4NT(Fixed: New Bat.), Galaxy Tab A	Try HTML entities « and »

11-07-2011, 09:19 PM	#12
ldolse Wizard Posts: 1,337 Karma: 123455 Join Date: Apr 2009 Location: Malaysia Device: PRS-650, iPhone	I'm not sure why you have a preference for ANSI - ANSI means it will only work on your computer and people who change Windows to say it's in the same country as your computer. UTF8 will guarantee it will work on any computer on the planet. Sometimes I want to kill/throttle whatever idiots at Microsoft insist on neverending support and propagation their asinine default ANSI settings - it's 2011 and those idiots are the number one reason this question comes up over and over again for well over a decade now.

11-07-2011, 09:24 PM	#13
ldolse Wizard Posts: 1,337 Karma: 123455 Join Date: Apr 2009 Location: Malaysia Device: PRS-650, iPhone	btw, if you save as ANSI the encoding in Calibre will definitely not be UTF-8. And I'm guessing Latin 1 may be iso-8859-1 - the problem with using ANSI is unless you're familiar with what ANSI is for your local country you need to play this guessing game. You also need to know what Python's name for the encoding is (and since we don't know what encoding you're using it's difficult for us to tell you.)

11-07-2011, 09:27 PM	#14
kovidgoyal creator of calibre Posts: 45,289 Karma: 27111240 Join Date: Oct 2006 Location: Mumbai, India Device: Various	Generally speaking, windows' determination to not break backward compatibility is to be admired, but in this instance, I agree with ldolse, the crazy mix of encodings that windows systems typically operate under causes endless headaches.

11-07-2011, 09:45 PM	#15
ldolse Wizard Posts: 1,337 Karma: 123455 Join Date: Apr 2009 Location: Malaysia Device: PRS-650, iPhone	And in this day and age I can't actually imagine that 'anything' would break... It's not like ANSI support would go away, they'd just change their save dialogs to default to UTF8 and all newly saved text files going forward would suddenly have a BOM and work on every PC on the planet - files without a BOM will still open as ANSI.... The occasional Windows 95 era program that actually didn't understand a UTF-8 BOM would still render 98% of the text correctly for a western language.. I imagine in Asia there may be some workflows that assume gb2312/shift-JIS/big5 etc as the default, but you could even keep a registry setting around for people who insisted on staying in the dark ages.

Advert

Advert