Need RegEx help for making the latest version of User Manual

[narayan]

Hi all,

I have finished writing the latest manual. Now I need to provide different character styles to the various types of text (e.g. user input, menu paths, keyboard shortcuts, etc.).

For this, I am going to use LibreOffice with an Addon called AltSearch.
This addon is capable of running find-and-replace operations with extended RegEx (for example, it can search within frames, footnotes, etc.).

I need RegEx patterns to search for the following:

1. Menu paths

These are typically enclosed by two words: "the" on the left and "menu" on the right. Check for other patterns in the User Manual v 1800 (pdf).

2. Keyboard shortcuts

These involve terms such as CTRL, SHFT, ALT, UpArrow, DownArrow, LeftArrow, RightArrow, ENTER, ESC, the + sign and a letter. This could be a single key (e.g. ESC or ENTER), or a combination (e.g. "CTRL+SHFT+D", "CTRL+LeftArrow").

Those who are good in composing RegEx patterns may please come forward and post the "find" patterns.

*****
I also need help from those of you who have experience in  OpenOffice/LibreOffice in general and AltSearch in particular.

Please drop a line here, and I will post further topics of discussion.

Thanks in advance!

[Tuxman]

I need RegEx patterns to search for the following

"RegEx", oh dear. There are a plenty of RegEx syntaxes around. (Is "syntaxes" a word at all?)

You'd need PCRE, I presume?

1. Menu paths

These are typically enclosed by two words: "the" on the left and "menu" on the right.

Code: Select all

the (.*?) menu

2. Keyboard shortcuts

These involve terms such as CTRL, SHFT, ALT, UpArrow, DownArrow, LeftArrow, RightArrow, ENTER, ESC, the + sign and a letter. This could be a single key (e.g. ESC or ENTER), or a combination (e.g. "CTRL+SHFT+D", "CTRL+LeftArrow").

Code: Select all

((CTRL|SHFT|ALT|UpArrow|DownArrow|LeftArrow|RightArrow|ENTER|ESC)\+)+\w+

No warranties, I'd recommend testing them in something like the Regex Tester ("Preg" setting) first.

[kunkel321]

The hotkey regex seems to work.  
Here's a shot of expresso:


Also regarding Expresso...  
It's free http://www.ultrapico.com/ExpressoDownload.htm  I'm pretty sure it used to be quite expensive.

[kunkel321]

I've tinkered with other ways to retain the font styles of the Usber Manual, while also converting the text to raw TXT.

One is to use the LibreWriter ability of "finding styles." For example if the RTF text is "To  jump  to  the quick bookmark,  press  CTRL+F1  (or use  the  Go  to Quick Bookmark menu option)."  

You could have LO search for
Font = Bold AND Text = any text string

Then replace with the HTML code
Text = < b > any text string < / b >

resulting in "To  jump  to  the quick bookmark,  press  CTRL+F1  (or use  the  [ b ] Go  to Quick Bookmark [ / b ] menu option)."

That when you strip all formatting, the HTML markers are still there, and you use the same process to convert it back.  

I am able to get LO to look for Font = Bold, but I can only get it to find one word at a time.  I can't get it to find multiple words separated by spaces.  Also, I can't get it to remember the 'any text string' and put it back in the replace box.  (I can do this with MS Word).

[Tuxman]

Also regarding Expresso...  
It's free http://www.ultrapico.com/ExpressoDownload.htm  I'm pretty sure it used to be quite expensive.

Not sure, I've been using it for years now without "pls buy me" - however, it seems to be dead.

[narayan]

Thanks!

I need these RegEx patterns to use with AltSearch, a macro addon for LibreOffice (the successor for OpenOffice, in case you haven't heard of it).

This macro can do some pretty neat tricks with find-and-replace operations.
It uses some extended notations too.

I will be using it to find the matching text and apply a different character style to make it stand out.

In other words, the idea is to find keyboard shortcuts and menu paths and make them bold, sans serif, etc.

[Tuxman]

I know (and use) both AltSearch and LibreOffice.

[narayan]

That's good, because Steve (Kunkel) originally volunteered but is going to be unavailable for some months for personal reasons...

I contacted Tom, the Czech author of AltSearch, about our needs.

He has modified AltSearch to extract bookmark links from a document. (AFAIK, he has not implemented extraction of headings so far.)

He sent a beta to me for testing, with instructions. But there may be some communication gaps, because I am not getting the results we expected.

We are supposed to give proper feedback to Tom so that he can make further changes to AltSearch, to make this work.

Can I share the mails from Tom with you? (If so, please send your email ID through a pm). Thanks!

[Tuxman]

If you feel like it...

[narayan]

Steve,

Our problem is reverse now: To avoid any crashes, I have already stripped all character attributes and now I am going to rely on styles alone.

(LO literature insists on that approach, but I had the old MSO habit of highlighting any text and hitting a shortcut like CTRL+U/I/B.)

So now the document does not have any character attributes like bold or underlined. On the contrary, I have to find the shortcuts and menu paths (using RegEx patterns), and then apply a distinctive character style.

[Tuxman]

I'll look at your mail later, I'm at work now. But I received it.