Importing and converting html fles

[Starson17]

Originally Posted by citac

Unfortunately it doesn't [S&R the head tag in html files]

I never was really sure why you wanted calibre to correct the head tag for you. If you read the html file, you don't see the head tag. If you convert to some other format, it gets stripped and you still don't see it.

[citac]

Because it shows up in the cover, and when I open my file in CoolReader, there's the incorrect author and it drives me bonkers.

When I convert the file to epub, the first line of the cover generated by calibre is the title, the second line is the author. That second line contains what's written in the head tag, so calibre must read it in order to create the cover. What it doesn't do is search and replace it with the correct author.

So I have to correct the head tag; at this point it doesn't matter whether it's via calibre or another program. I cannot leave it as is just because when I read the html file I don't see it. It causes something to happen, and that something then results in incorrect information. Am I making myself clear?

(Plus it just plan bothers me and my OCD tendencies when something is not correct. I am working around it at the moment by making calibre read metadata from file names instead of html tags, which is why I'm trying to find a naming scheme which would work both for fics that belong to a series and those that don't.)

[DoctorOhh]

Originally Posted by citac

When I convert the file to epub, the first line of the cover generated by calibre is the title, the second line is the author. That second line contains what's written in the head tag, so calibre must read it in order to create the cover.

The title and author are generated from the Title and Author fields of calibre's library view.

You might want to check out the Generate Cover Plugin.

[citac]

Originally Posted by dwanthny

The title and author are generated from the Title and Author fields of calibre's library view.

You might want to check out the Generate Cover Plugin.

I will, thank you.

Look, while I still had calibre read metadata from file, the cover would be generated with the incorrect author even though I had entered the correct author in the proper field in Edit metadata. It is really late here now, and tomorrow I will switch it back to reading metadata from files, repeat the conversion and check the results.

[citac]

Here are the screenshots I made after I switched to collect metadata from file instead of title.


The first one is the screenshot of the head tags of the original file before import.



The second one is after import before conversion



The third one is bulk convert search and replace settings



The fourth one is after conversion to epub



The fifth is the cover of the epub



The sixth is the head tag of the html I extracted from the zip



The seventh is epub viewed in Sigil,



which shows that the head tag was rearranged and the correct name of the author inserted, so why does the cover still show the incorrect author? Where is that information stored and how can I get rid of it/correct it? Do you understand what I want to do?

[citac]

And can someone please explain to me why is each and every paragraph styled? What is it with all these calibre classes? Isn't it enough to just set one style for <p> in the css and have it apply to all the paragraphs in a text? I am far from an expert on css, but isn't that a bit redundant??

[DoctorOhh]

Originally Posted by citac

which shows that the head tag was rearranged and the correct name of the author inserted, so why does the cover still show the incorrect author? Where is that information stored and how can I get rid of it/correct it? Do you understand what I want to do?

This is all as expected. The metadata on any book inside the library is not updated unless a conversion happens, so when you converted your html to epub the new epub used the metadata you had updated in the library view. The only other time metadata is updated (at least for mobi and epub) is upon export from the library, using Save to disk, Send to device, Connect to folder or the Content server.

The new epub also uses whatever cover you have set at the time of conversion. On my system the kind of cover you had is not auto-generated when you add the html to the library, you actually have to click on the generate cover button in the edit metadata window. When converting the html to epub make you you correct the cover first. Cover management is up to the user, calibre just applies the cover you give it.

Originally Posted by citac

And can someone please explain to me why is each and every paragraph styled? What is it with all these calibre classes? Isn't it enough to just set one style for <p> in the css and have it apply to all the paragraphs in a text? I am far from an expert on css, but isn't that a bit redundant??

If the original document had one style for <p> tags then calibre will assign the same class to every paragraph. If the original document had varying styles for different paragraphs then that too is reflected in the outputted calibre classes.

Not a technical answer, but I'm sure someone with specific knowledge will give you a more thorough answer shortly.

[citac]

But I didn't click on the Generate cover button, even though I have the plugin installed. I converted the files to epub, and when I opened the epub, it showed me that cover. Look:



I did not apply cover after importing the original file to the library.

Also, I just checked the original file: there's absolutely no paragraph styles. Just <p>Some text</p> and that's it.

I just sent that one particular book to disk, and when I open the epub in the internal vievwer, it still showed me the same cover.

A question: if I chose Send only EPUB to disk, why is there an opf file with it? What is its purpose? Can I delete it if I don't want it on my reader?

[citac]

Another question: should I just forget about having calibre read metadata from file and keep it reading it from the title? Because if I'm having these problems now, what's gonna happen when I start importing my old txt files from 2004? They have no metadata inside the file, no html tags for calibre to use.

[theducks]

Originally Posted by citac

But I didn't click on the Generate cover button, even though I have the plugin installed. I converted the files to epub, and when I opened the epub, it showed me that cover. Look:



I did not apply cover after importing the original file to the library.

Also, I just checked the original file: there's absolutely no paragraph styles. Just <p>Some text</p> and that's it.

I just sent that one particular book to disk, and when I open the epub in the internal vievwer, it still showed me the same cover.

A question: if I chose Send only EPUB to disk, why is there an opf file with it? What is its purpose? Can I delete it if I don't want it on my reader?

Generate cover makes one for the book browser (to be used instead in your PIC). Calibre will make a cover IN THE BOOK if one is not available. At least that is my observation.
Since the cover is now there, Calibre will not update it.
You might try converting again, with Ignore first image ticked.

[citac]

Originally Posted by theducks

Generate cover makes one for the book browser (to be used instead in your PIC). Calibre will make a cover IN THE BOOK if one is not available. At least that is my observation.
Since the cover is now there, Calibre will not update it.
You might try converting again, with Ignore first image ticked.

(Is there a "pulls hair" emoticon?) It did not work. The cover is still there, with programmer's name instead of author's.

This is really annoying. I don't understand why it won't correct the author. I just went and extracted the contents of the epub, and looked inside the html files where the meta information in the head tag is corrected. Yet even though that's true, calibre still creates a cover containing wrong data. I don't understand what I need to do to make it stop.

[DoctorOhh]

I see your problem and I should have paid closer attention to your images in sequence previously. You are doing a search and replace to update the author name in the header, then you open the epub and look at the header to prove it is there. The problem is what you show in the ePub has nothing to do with metadata in a ePub. Metadata in a ePub is contained in the content.opf file.

The whole search and replace step your doing is a waste of time. You need to update the author's name in the library view. This field is used when creating the metadata and your cover in your ePub.

Originally Posted by citac

I just sent that one particular book to disk, and when I open the epub in the internal vievwer, it still showed me the same cover.

This cover is only generated during a conversion, see epub output during conversion. Save to disk etc... will update the metadata and only update the cover (IE. replace a cover in the book with a new one you add/show to calibre's library view) when a properly set cover already exists in the epub.

It has literally been years since I had calibre generate a cover during conversion to epub so I forgot about this feature.

Originally Posted by citac

A question: if I chose Send only EPUB to disk, why is there an opf file with it? What is its purpose? Can I delete it if I don't want it on my reader?

You can delete that file and avoid sending it by deselecting the option to Save metadata in opf file under Preferences - Saving books to disk. While in the area uncheck save cover separately.

Is there a reason you didn't use Send to device or Connect to folder to send books to your reader?

Originally Posted by citac

Another question: should I just forget about having calibre read metadata from file and keep it reading it from the title? Because if I'm having these problems now, what's gonna happen when I start importing my old txt files from 2004? They have no metadata inside the file, no html tags for calibre to use.

If you are adding text then you should not read metadata from file on adding books read from filename instead.

Originally Posted by citac

This is really annoying. I don't understand why it won't correct the author. I just went and extracted the contents of the epub, and looked inside the html files where the meta information in the head tag is corrected. Yet even though that's true, calibre still creates a cover containing wrong data. I don't understand what I need to do to make it stop.

Again I apologize for not catching this sooner but you are jumping through hoops that aren't accomplishing anything. The metadata is created in the epub via the fields in the calibre library and are stored in the content.opf file in the epub. In Sigil use F8 to open the metadata editor to view the metadata in the epub.

[theducks]

Originally Posted by citac

(Is there a "pulls hair" emoticon?)

There is

[speakingtohe]

There is a checkbox for no default cover under preferences output options epub.

To remove cover from epub I use tweak epub as the remove first image hasn't worked for me for a long time unless you are putting in a new image. Used to work in version 6xx.

select book
hit t
explode epub
delete cover.jpg or cover1.jpg
delete titlepage.xhtml
preview epub
If satisfied with result rebuild epub

Takes under a minute

Helen

[speakingtohe]

A question: if I chose Send only EPUB to disk, why is there an opf file with it? What is its purpose? Can I delete it if I don't want it on my reader?

There is a checkbox in the saving books to disk preferences that decides whether to send the cover and opf files when saving to disk. Just uncheck it and you won't have to delete.

Helen