Convert comments field to HTML in bulk

[d3m0sth3n3s]

Hi!

I have read in these forums that while comments are displayed/edited as HTML they are not always stored internally as HTML code, depending on where they originally come from. Sometimes they are just stored as plain text. After you edit a comment using the calibre metadata editor it is always stored as HTML internally.

That said, is there any way to get all the comments in a bunch of books converted to HTML the same way the metadata editor would do? Some template that performs this operation in bulk?

[theducks]

While you may paste plain text, there is always HTML on that tab.
It is just simple. <p>words</p>
Same as you would get by 'using the remove formatting (of selected text)

[d3m0sth3n3s]

Originally Posted by theducks

While you may paste plain text, there is always HTML on that tab.
It is just simple. <p>words</p>

I know that. The thing is I’m not talking about manually pasting text (for an individual e-book) into the Comments field using the calibre GUI metadata editor.

I have lots of e-books in ePub format whose Comments field was filled during the import process and it is stored internally in plain text (in the Comments field in the calibre database). I want to convert the Comments format for those e-books to HTML prior to some regex Search & Replace tweak. To that end, I need that all my e-books have their Comments field stored internally in HTML.

[JSWolf]

Originally Posted by d3m0sth3n3s

I know that. The thing is I’m not talking about manually pasting text (for an individual e-book) into the Comments field using the calibre GUI metadata editor.

I have lots of e-books in ePub format whose Comments field was filled during the import process and it is stored internally in plain text (in the Comments field in the calibre database). I want to convert the Comments format for those e-books to HTML prior to some regex Search & Replace tweak. To that end, I need that all my e-books have their Comments field stored internally in HTML.

Why do you need all of your comments to be HTML?

All of your comments are converted to HTML if they are not HTML. So you don't need to convert them. Select a book you think has the comments not in HTML. Press the E key and then under the blurb, click the HTML Source tab. You'll see that it is HTML.

[d3m0sth3n3s]

Originally Posted by JSWolf

All of your comments are converted to HTML if they are not HTML. So you don't need to convert them. Select a book you think has the comments not in HTML. Press the E key and then under the blurb, click the HTML Source tab. You'll see that it is HTML.

No, you are mistaken. That is what the Qt GUI component implementing the Comments metadata editor shows you, not what it is really stored in the calibre database. Check this:

Originally Posted by kovidgoyal

That's probably because it is actually just a <p><br> (the markup you see via the edit metadata dialog is not an accurate representation of what is present internally as it is roundtripped through Qt's text edit widget.

[JSWolf]

Originally Posted by d3m0sth3n3s

No, you are mistaken. That is what the Qt GUI component implementing the Comments metadata editor shows you, not what it is really stored in the calibre database. Check this:

If it doesn't store the blurb as HTML, then how is text being made bold? I've some blurbs from the eBook that have some text bold.

Why is it you need the blurb to be HTML?

[d3m0sth3n3s]

Originally Posted by JSWolf

If it doesn't store the blurb as HTML, then how is text being made bold?

It’s bold because calibre stores it as HTML But that is AFTER you edited the Comments field with the GUI metadata editor. If it was imported as plain text and you left the Comments field untouched in the GUI it is kept as plain text internally in calibre database. The HTML code you see in the Comments metadata editor is not an accurate representation of the real information stored in the database. That can be plain text!

[DNSB]

Odd. When I download metadata, it shows bold, italic, underlines, etc. and those seem to survive closing and re-opening calibre. Perhaps what is stored in the database is not HTML code but the HTML code seems to survive that storage.

[BetterRed]

I have books with plain text comments



I'll try to find a way to HTML-ise them in bulk.

@d3m0sth3n3s - I suggest you ask un_pogaz if the Comments Cleaner plugin can convert plain text to simple HTML.

BR

[kovidgoyal]

THere isnt any such tool, though I dont really understand why you need it. All such a tool would do is mostly just wrap blocks in <p> or <div> tags ( dont recall the details).

[d3m0sth3n3s]

Originally Posted by kovidgoyal

THere isnt any such tool, though I dont really understand why you need it. All such a tool would do is mostly just wrap blocks in <p> or <div> tags ( dont recall the details).

I want to use a Search & Replace regex that relies on the comments to be in HTML to add some structure and styling to them in bulk. It doesn’t work well for plain text comments so I intended to HTML-ize them as a prior step.

I’m thinking that I would be better coding a more complex template that takes plain text comments into account. Just more work, I’m afraid.

PS: I know it’s not a big problem, it just seems inconsistent to me that the calibre database may store imported comments without styles both in HTML code and in plain text just because you edited them in the GUI metadata editor or not.

[BetterRed]

Originally Posted by d3m0sth3n3s

I want to use a Search & Replace regex that relies on the comments to be in HTML to add some structure and styling to them in bulk. It doesn’t work well for plain text comments so I intended to HTML-ize them as a prior step.

I’m thinking that I would be better coding a more complex template that takes plain text comments into account. Just more work, I’m afraid.

PS: I know it’s not a big problem, it just seems inconsistent to me that the calibre database may store imported comments without styles both in HTML code and in plain text just because you edited them in the GUI metadata editor or not.

It has nothing to do with whether they were edited - mainly tto do with what was embedded in the format file that was used to create the book in the calibre library.

The books in my Test library that have plain text comments would have acquired them from the format files I added… e.g. Book 801 came as a PDF from The Lancet website - calibre extracts the Description property from the PDF metadata and puts it in the Comments column, viz:

The Lancet, Corrected proof. doi:10.1016/S0140-6736(22)01585-9

Whereas Book 671 came as an EPUB from Verso, the Comments were extracted from the OPF file in the EPUB, viz:

<description xmlns="http://purl.org/dc/elements/1.1/">&lt;i&gt;&amp;amp;ldquo;It would be an endless task to trace the variety of meannesses, cares, and sorrows, into which women are plunged by the prevailing opinion that they were created rather to feel than reason, and that all the power they obtain, must be obtained by their charms and weakness.&amp;amp;rdquo;&lt;/i&gt;&amp;#8212;Mary Wollstonecraft &lt;br&gt;Composed in 1790, Mary Wollstonecraft&amp;amp;rsquo;s seminal feminist tract &lt;i&gt;A Vindication of the Rights of Woman&lt;/i&gt; broke new ground in its demand for women&amp;amp;rsquo;s education. A Vindication remains one of history&amp;amp;rsquo;s most important and elegant broadsides against sexual oppression. In her introduction, renowned socialist feminist Sheila Rowbotham casts Wollstonecraft&amp;amp;rsquo;s life and work in a new light.</description>

BR

[d3m0sth3n3s]

Originally Posted by BetterRed

It has nothing to do with whether they were edited - mainly to do with what was embedded in the format file that was used to create the book in the calibre library.

Yes, you are right. My explanation was incomplete/misleading.

I just think that if calibre turns a simple comment into HTML code just by, for example, simply adding a comma with no styling involved, whenever you edit comments with the GUI metadata editor, then it would be better for consistency to just store them all in HTML from the very beginning, no matter how comments were defined in the description embedded within the source e-book file you import into your library. That’s just my opinion on the subject.

[un_pogaz]

The Plugin Comments Cleaner can convert plain text Comments to HTML comment.
(caution: default settings do many other things)

To search the Not-HTML comment use the query:

comments:true and not comments:"^</div>"

and to quickly know at a the first glance if a book has Not-HTML comment, go to "Preference > Look & Feel > Column colors" and sélectionné les options

• If the Column 'Comments' is defined
• If the Column 'Comments' do not contain the value "</div>"

Note: A html comment has at least one </div>

[d3m0sth3n3s]

Originally Posted by un_pogaz

The Plugin Comments Cleaner can convert plain text Comments to HTML comment.
(caution: default settings do many other things)

Thank you! It works for me.

Also, the tricks to highlight books without HTML comments are quite useful, too.