Digital Workshop

hello dw,

to me this sounds easy to implement but i'm sure its not, duncan...

would it be possible to have a utility in opus which grabs all the text being used in a publication and outputs it as a text file. this is before it is published, in the opus editor.

so it would go through the publication and make note of all the text in text boxes and compile a nice file which can be printed.

why you ask, well for ease of translation into different languages of course. in a perfect world i'd love to be able to edit the text once compiled in the file, this would save a lot of copying and pasting and going through the publication.

and yes i know you could store all the text in a database or text file and do it yourself with more variables than god but as usual i'd prefer opus to do that for me.

cheers lads,

luke

ps any word on a 2nd service pack this side of xmas?

i agree, because it could be usefull too, for spelling and grammar correction when you use another langage than american or english...

benoit

I can see that this would be a useful addition, but let me suggest a workaround by describing what I do.

I have several pubs which exceed 1000 A4 pages in length. I prepare the complete text for these in WordPerfect.

By doing this I can manipulate the layout, use shortcuts to include boilerplate text, apply various text attributes, spell check, grammar check, etc. etc.

When I've finished, I save the file as RFT. From here, I can either cut and paste into Opus, or preferably, break the main file into text object sized chunks of text which I then save as separate files and import into Opus.

I prefer the second, though longer method because it gives me an independent backup of all the text in the way in which I used it in Opus. This means that if (when ) I have a disaster, I can go straight to the block of text I need and re-use it.

For this method to be effective, you must develop a way of readily linking the Opus text object to its RTF file. I just give them the same name: Opus page name + text object ID.

This method may not work for others, but I've been successfuly using it for 5 years.

so would this be a possible future feature DW?

being able to have the text file external would be even better, so the user could select which language they wanted.


luke

It might be possible to do something like this in the editor; though there are complications such as hypertext and variables that would either get lost or take some effort to work around). I'll certainly add it to the list of requested future enhancements.

Doing it at runtime would really be in the realm of the database solution you suggested. It's something so few people would use that it would not be worth the growth in the player to support it - and then you would have problems with web publications, Flash support, etc. So this is probably a "no".

cool thanks duncan,

yeah db would be the best for that alright. just one note, say you have the following text.

Built with Opus <version>, the best multimedia developer.

where <version> is a variable.

if that's stored in a db, and pulled in as text into an Opus publication, the variable is left as text. ideally if possible it would be great if anything pulled from a db or txt file with text enclosed within <> would be treated as a variable, (or some other token on each side).

maybe an option which could be turned on in the text object property could be added so it would check for variables.

this would make creating multiple language versions much easier.

I am now searching the forum for a solution... lots of text sections I expect will be pulled from a DB. [ I will 'ask' that question separately. ]

For this thread, I will offer this idea for the "front end" during creation of the content. I researched what tools will help me organize all the text sections I will put into an Opus pub.

Besides other CMSs, these two KBs stand out for me. IdeaMason and Baltsoft's General Knowledge Base. The first takes many separate sections of text and enables creating a single composition (can tag or categorize sections... so very useful)

Baltsoft likes to output a collection of topics not to one doc, but to multiple files. So again, useful for manipulating collections. (I don't like their long file names though)

I think having organized collections would help with the language translation, if not done in Opus arena.

Can I highlight this feature request for future consideration.


Thanks.

Both the extraction of the text and parsing of variables from a DB / INI could be achieved using OpusScript and a little forethought.

Mack

I think Mack's right, especially about the 'forethought'. Well... maybe more right about OpusScript! Wonderful tool.

Currently Opus enables you to Export current variables. All you have to do is go to Publication and Page Properties >>> Variables Tab... then select All or some subset. That will enable the Export button.

For the text 'round up' (i.e., gather and harness), if anyone wants to tackle this here are a couple starter pieces of the puzzle.
First, Brenden Knifton of DW did a neat implementation/starter_pub for an export-to-disk (.html file) and print of same output.
Here: viewtopic.php?t=442&highlight=print+html2
A couple notes on that.
- it is currently set up in the download .imp to handle TextInput Objects on 'a page' where those objects have variables associated with them.
- -- to do this more universally, you will have to identify more variables... and if you want to include standard TextObjects that do not have a variable to populate the text, then you need to work with SetSelection() and GetSelection() for each of those objects.

That may sound like a discouraging uphill trek. But continue.

- I posted a script you can run on any Pub during development of the Pub, to document all of its objects in a tree structure. ('all' being ~8 levels anyway). The key point here is that script runs through the Pub's full object hierarchy quite well.
- -- I think all you'd have to do is add a FILTER that would GetType() == "Text"
- - -- -- and then work on getting at the text 'selection' or the applicable variable for those objects -- accumulate these in Brenden Knifton's approach.
- - -- -- BTW, don't forget Button Objects. So in addition to the above Filter, add type == "Button", and get the Child of the Button (it holds label text)

I managed to play around with the Structure/Tree output when I developed it -- it was very easy to Copy the output and Paste it into Excel. Pic attached. (don't know if I tweaked it after posting the script, but just let me know if needed).
Here: viewtopic.php?t=3619&highlight=structure

BTW -- I made this Tree structure script partly for documenting and viewing and partly in anticipation of wanting to save Object positions and states ... so if you have anything along the lines of Saving / Restoring states, I'd sure like to hear/get some help there.

Just to highlight this again.

The idea of just being able to have an option in Opus Editor, something like Tools > Extract All Text > Saves it as a text file.


I understand that variables would get lost here, but for anyone using blocks of embedded text it would be very useful indeed.