Regex examples

[BetterRed]

Originally Posted by Leonatus

I'm at a book where quotations are marked by right- (opening) and left-pointing double angle quotes (» and «).

<snip>

Any help appreciated!

Try doing a search for guillemets (« ») in the Workshop forum. IIRC the Second problem has been discussed in a couple threads, and I recall mention of some older German texts using a convention much like MLA/AP Style Guides for multi-paragraph quotes in English:

Quotations that extend over more than one paragraph must have an opening quotation mark at the beginning of each paragraph and a closing quotation mark at the end of the final paragraph.

BR

[Tex2002ans]

Originally Posted by Leonatus

I'm at a book where quotations are marked by right- (opening) and left-pointing double angle quotes (» and «).

First problem: Frequently, in the middle of direct speech marked as reported, there are citations that are as well marked by right- and left-pointing double angle quotes. I would like to replace those citations by single angle quotes. Is there a good way to find them (only in the middle of direct speech, not outside)?

Which language are you working in? Slovak?

There was a similar thread in the Calibre Editor subforum discussing guillemets:

Regex Function about «» and “”

where I posted a very basic Regex I use. senhal posted his in Post #9.

You may just have to flip around some of the inner/outer »« directions, and substitute in some ›‹, but the logic should all be the same.

Note: And I still stand by Toxaris's EPUBTools Dialogue Checker being the best tool for this job. This problem really requires something a bit smarter than just Regex.

Originally Posted by Leonatus

Second problem: It appears, that at direct speech passages, there is an opening double angle, but the closing one is missing (by error of OCR, perhaps). How can I find (and replace) such items, please?
Edit: Sometimes it's vice-versa: the closing mark is there, but the opening one is missing.

Like BetterRed said, English has an opening quote across multiple paragraphs if it's the same character talking:

Sue continued to drone on, “This is a very long example. [...] And she keeps talking.
“And talking.
“And talking.
“And talking until the roosters crow.”

I am not too sure if other languages follow a similar "no closing quote" rule.

[BetterRed]

Originally Posted by Tex2002ans

Note: And I still stand by Toxaris's EPUBTools Dialogue Checker being the best tool for this job. This problem really requires something a bit smarter than just Regex.

↑ ↑ ↑ ✔️

BR

[Leonatus]

Thank you both, BetterRed and Tex2002ans! The text is in german. And yes, I also find that Toxaris' tool is excellent for this task, but the book has come as epub already; I have to do some modifications (transform in modern spelling and so on). Only that there are quite a lot of mistakes of the reported sort that I wished to mend.
I'll have a look to the recommended threads.
Fyi: In older german books, you'll frequently find the same way of formatting direct speech across multiple paragraghs, i. e. beginning with an opening quote at each paragraph, but ending with the closing one only at the end of the direct speech. This is no longer the case in modern books. the book I'm working on shows this as well, and I changed this. This is not the problem. There are some erroneously missing quotes.

[Leonatus]

Your "simple" check:

(»[^«]*)</p>

found me missing closing quotes within a paragraph, whereas

»([^«]*)»

found me as well missing closing quotes as the "citation problem", also across multiple paragraphs.
That's indeed very helpful!

[cereburn]

I've got a PDF that I used Calibre to convert to ePUB but at the top of each page in the PDF was a piece of page bling next to the page number that is now mixed in with the text of the doc.

file:///K|/eMule/Incoming/88%20sci-fi%20aWizard.html (77 of 309)16-8-2007 23:50:31

When I try to search for this just using:

\A file

it doesn't result at all

if I use:

\A file:///K|/eMule/Incoming/88

then it finds and highlights eMule/Incoming/88

I've tried adding \ to each of the escape required characters above, but that breaks the search back to where I was when I started.

My goal is to setup a search and replace for everything starting with file up to and including the first following <p>

[famfam]

Better text structuring with regex:
Better text structuring allows better orientation in reading and thus better understanding of the texts. Of course this applies in particular to extensive and complicated texts.
That's why I came up with the idea. To emphasize sentence beginnings by increasing the first letter of each sentence two degrees and bold.
This can be easily implemented in Word with regex. But since I also want to do that with Sigil in finished epubs, I'm looking for the right regex solution for Sigil.
Here is my approach:
Search: [.?!] [A-Z]
Replace: ^ &

At this point I do not know, because I do not know if and how you can achieve the character formatting on fat when replacing.

With Word, as I said, that's no problem. But how can I do that in Sigil with the finished epub?

Does anyone have a suggestion?

In german:
Bessere Textstrukturierung mit Regex:
Bessere Textstrukturierung ermöglicht bessere Orientierung beim Lesen und dadurch besseres Verstehen der Texte. Das gilt natürlich insbesondere für umfangreiche und komplizierte Texte.
Daher bin ich auf die Idee gekommen. Satzanfänge herforzuheben, indem ich den ersten Buchstaben jedes Satzes zwei Grad vergrößere und fett markieren.
Dies läßt sich in Word problemlos mit regex realisieren. Da ich aber das auch mit Sigil in fertigen epubs machen möchte, suche ich nach der passenden regex-Lösung für Sigil.
Hier mein Ansatz:
Suchen: [.?!] [A-Z]
Ersetzen: ^&

An dieser Stelle weiß ich nicht weiter, weil ich nicht weiß, ob bzw. wie man beim Ersetzen die Zeichenformatierung auf Fett erreichen kann.

Mit Word, wie gesagt, ist das kein Problem. Aber wie kann ich das in Sigil mit dem fertigen epub hinkriegen?

Hat jemand einen Vorschlag?

[Doitsu]

Originally Posted by famfam

To emphasize sentence beginnings by increasing the first letter of each sentence two degrees and bold.

If your reading app supports pseudo elements, you don't need to use any regular expressions at all. Simply add the following code to your main style-sheet:

p::first-letter {font-size: 2em; font-weight: bold; }

[DiapDealer]

I'm curious what the replace expression is that makes this "easily implemented in Word with regex"?

[BetterRed]

Originally Posted by DiapDealer

i'm curious what the replace expression is that makes this "easily implemented in Word with regex"?

I would guess the OP means using the Formats->Fonts modifiers on the Replace, something like this:



That Find will not find all sentences, e.g. those following speech where writer's style is to punctuate inside the quotes. For that reason alone I would use a VBA macro.

BR

[DiapDealer]

Originally Posted by BetterRed

I would guess the OP means using the Formats->Fonts modifiers on the Replace, something like this:

Ahhh... that makes sense. Thanks. So not actually regex, then. An additional F&R feature. I thought maybe Word had some sort proprietary regular expression syntax that allowed text formatting to be specified in the replace expression.

[BetterRed]

Originally Posted by DiapDealer

Ahhh... that makes sense. Thanks. So not actually regex, then. An additional F&R feature. I thought maybe Word had some sort proprietary regular expression syntax that allowed text formatting to be specified in the replace expression.

If you're looking for something to do you could reverse engineer it into Sigil

[DiapDealer]

Originally Posted by BetterRed

If you're looking for something to do you could reverse engineer it into Sigil

Zero interest on my part. But since the regex module has always been a part of Sigil's bundled Python, there's nothing to stop someone from creating a plugin that could combine regex find-and-replace with css styling.

Though regardless of how it was handled (plugin or inherent Sigil feature), it would very hard to ensure that no pre-existing styling would ever get stomped on in the process.

[Klecks]

Originally Posted by famfam

... To emphasize sentence beginnings by increasing the first letter of each sentence two degrees and bold.

You can try the following expression:

search for:

(?<!St|Mr|Mrs|<|\d)(<p[^>]*>|\!|\?|\.|…|:)( ?)(»|«|“|”|„)?( ?)([A-ZÖÄÜ])

replace with:

\1\2\3\4<b>\5</b>

and in the CSS:

b
{font-weight: bold;
 font-size:1.2em;
 line-height:0.7em;}

Greatings
Klecks.

[famfam]

@Doitsu
Yes, the code works, but only for the first character od paragraphe. Thanks a lot.
@Klecks
Yes, that seems to work very good. I just tried it and will prove it more and mor the next day. My reading of long and complicated ebooks will be better with this new orientationsystem, and better understandig will be the result of that.
Thank you so much.