RegEx in AffPub

[Eisbar]

I need to prepare a Publisher Document with hundrets of Endnotes (IDML-import) for online use....

With find&replace I can find every endnote by searching the textstyle "Endnote". But before and after every endnote I had to enter some textcomand. 

For example:

This is1 a text.-> This is<endnote>1</endnote> a text.

Is it possible to do this with find&replace? Finding is not the problem. But is it possible with RegEx to add text before and after the endnote? I just know that RegEx is a strong tool - but I don´t know to work with it 😞

Thx for help!!

[v_kyr]

Since APub has that Find & Replace panel, see therefor ...

• Find and Replace
• Find and Replace Panel

... try out in the replace text field to add a <before> and <after> text which then surrounds your find term.

[Eisbar]

<before> is not working.... 

any other ideas?

[v_kyr]

25 minutes ago, Eisbar said:

<before>

That was of course just meant as a placeholder for whatever you wanted to write there as a replace text. - So you say you can't add any text before the found endnote in a replace run?

[Eisbar]

17 minutes ago, v_kyr said:

So you say you can't add any text before the found endnote in a replace run?

that´s right 😞

[Pyanepsion]

Hello, @Eisbar,

If your note uses a format or style, simply replace the format or style with your sequence:

 

 

 

[Eisbar]

original text: "This is1 a text."

what I need: ""This is<endnote>1</endnote> a text."

[walt.farrell]

9 minutes ago, Eisbar said:

original text: "This is1 a text."

what I need: ""This is<endnote>1</endnote> a text."

But, do you need every 1 replaced that way? Or every is1 replaced? Or every occurrence of any word immediately followed by a 1 (with no space) replaced?  Or any word immediately followed by a single digit? 

You need a general statement of the input and output text to construct a regular expression, not just one example.

 

 

[Eisbar]

every sign with the textstyle: "endnote" (just numbers, from 1 - n) has to get "<endnote> before the number and behind the number must be the text </endnote>

[laurent32]

You could try :

search = 

your textstyle endnote

replace =

<endnote>$0</endnote>

[Old Bruce]

6 hours ago, Eisbar said:

I need to prepare a Publisher Document with hundrets of Endnotes (IDML-import) for online use....

Not having access to the Adobe document means we can't really troubleshoot this. 

Are the note superscripted numbers actually 1, 2, 3 etc or are they markers? I don't know if IDML import converts the Markers to actual text or not. Are the Markers (1, 2, 3, etc) a separate Character Style or part of the overall Paragraph Style for the notes?

[Eisbar]

Laurent32 - thx for your idea. I will try it later! And report...

Old Bruce: The endnote are just superscripted numbers actually 1, 2, 3 (with a textstyle). They are not are part of a Paragraphstye - I mean they are text in a paragraph which has a style. But they have the textstyle by whom I can find very easy every endnote!

[Eisbar]

4 hours ago, laurent32 said:

You could try :

search = 

your textstyle endnote

replace =

<endnote>$0</endnote>

this is not working 😞

the result is exactly this: "<endnote>$0</endnote>"

[walt.farrell]

13 minutes ago, Eisbar said:

this is not working 😞

the result is exactly this: "<endnote>$0</endnote>"

Please show us a screenshot of your Find and Replace panel in Publisher.

[Eisbar]

Here are 2 screenshots: before and after replacing. And for every case: a sample document....

33_Obadija.afpub

[walt.farrell]

Thanks. You need something in the Find field if you're going to use $0, as otherwise you won't have a regular expression search.

You can try using .* and see if it works. There are, however, some forum discussions about problems using that search term successfully. So it might not work.

Perhaps, given that you're only looking for numbers, it would by better to try \d+ or \d+? instead.

And, as another alternative if those don't work, you could try (\d+) in the find and then use $1 or \1 instead of $0 in the replace field.

[laurent32]

You could try v2 :

search = \d+

+ your textstyle endnote

replace =

<endnote>$0</endnote>

[Eisbar]

Dear Laurent32, I am so grateful to you. It works - and you save me so much work! 1000 thanks!!

[Eisbar]

I have one more challenge!
All words or phrases in italic must be preceded by a <i> and followed by a </i>.
I tried it with the regex "." This works insofar as it finds all characters that have the text attribute italic. But it is found letter by letter and each letter is preceded by an <i> and then a </i>.
Is there an expression to tell publisher: all related words are to be considered as a unit?

[walt.farrell]

2 minutes ago, Eisbar said:

I have one more challenge!
All words or phrases in italic must be preceded by a <i> and followed by a </i>.
I tried it with the regex "." This works insofar as it finds all characters that have the text attribute italic. But it is found letter by letter and each letter is preceded by an <i> and then a </i>.
Is there an expression to tell publisher: all related words are to be considered as a unit?

.*? would be one commmon way.

[laurent32]

You could try :

Search = \w+

+ the style you're looking after

Replace = <i>$0</i>

[Eisbar]

can I avoid this? see screenshot!

[laurent32]

You can try, if you have a text style dedicated to italics :

Search = empty

+ character style italics

Replace = <i>$0</i>

[Old Bruce]

1 hour ago, Eisbar said:

can I avoid this? see screenshot!

You could then search for </i> <I> and replace with just the space.

4 hours ago, Eisbar said:

All words or phrases in italic must be preceded by a <i> and followed by a </i>.

It really depends if the spaces are also styled italic. As you said the text is coming in from an IDML file there is no guarantee that they are.

If the spaces are italic then I would use for find: With the italic option from the silly little cog wheel's Format choice.

(.+?)

and then use the following for the replace:

<I>\1</i>

[walt.farrell]

2 hours ago, Eisbar said:

can I avoid this? see screenshot!

My version using .*?  should not have that problem.