Yes, technically it can be done (but not in a single model, unless another forum member wants a challenge that needs a really creative solution :confused:), but honestly you're probably better off just making two models.
One way that I'd envision would be to have a single record CSV file for the English headings, and another for the French headings. Run an English batch file to copy the English names to a new file, then run a Monarch project set to append to the same file without including any field names. Now open the "completed" CSV file as a database source, and output/filter/sort/summarize as necessary.
Do the same for a French batch file, replacing the English heading record with the previously created French heading record.
Like I said, make two models.
Maybe it would be more worthwhile if you needed more than two languages, but probably not.
I think this could be done (mercifully we don't have to use all three official languages at work so I've not needed to implement this). Let's assume that there's a way to detect the language of the report from some text in the page header of the first page. Along with all the other templates, trap this as an append so it's present on every row of data, use a calculated field to convert this to "en" or "fr" as required.
Create a filter called "en", and one called "fr" based on this. Create a project export based on "all filters" of the table data with export filenames based on filter names. Now you have two datasets, one of which will be empty, and you can use two models, one which maps the field names to French ones, and these can be run against the known filenames from the exports of the first project.
Have you looked at the possiblity of using Linked Objects for creating a second model?
I'm wondering whether you could, in effect, have a second version of the model completely, or almost completely, maintained by changes (when required) being applied only to the 'Master' model.
It's just an idea at the moment. However you have a sample report/models/(spec(s)) that you are using to develop a solution and you can share them with others I would be happy to take on a challenge to see how much could be achieved.
I would rather work with a 'real world' challenge than try to make something up only to find that it is an incomplete solution to an incomplete problem!
Depending on what and how much you are exporting, but could you use summaries. Two identical ones with the field names changed as approrpiate. Then all you need to do it set up a couple of Project Exports and choose which one you need to use.
I have done something similar when different people needed to see specific column headers for the same thing. The limiting factor is what you need to export in the way of data.
Just wondering if you could use summaries. Set up the first for English, duplicate it and change the field titles to French.
Then set up 2 project exports, and then choose which one you need to use.
Depending on what and how much data you have this could work and if you ever needed to export the English original in French you could.
I have done something similar when 2 users needed different column headings for the same data.
I have decided to try Nigel's suggestion as it seemed to fit what I was trying to do. Of coarse, this brings another question:rolleyes:.
I created a hidden calculated field to indicate which language the report is in. I used this field to determine that if it's the appropriate language for the summary (one is English and the other French, yeah really:rolleyes:) by using it in a filter. So one summary has no data and the other has the output I wanted.
I created a project file from my model and created two Project Exports for the two summaries I created. The problem is that when I try to create the PDF from the command prompt, I get two of them, but one is actually blank (Adobe complains that it has no pages) but the file is not empty (not zero bytes).
Here is the command I use:
Since I really don't know the language of the report from the command prompt, I thought that I would be able to tell by seeing which file is created, but as explained above, both are created. How would I be able to get it to not create the output file if there is not data, other then getting fancy with Windows command line or using VBS?
What I did in my batch file was to do a "find" in the report file to see if a certain word was in a certain language (Rapport/Report) so after that I knew which Project Export to use; in fact that is the same region of the report that I use to determine the language used in by filter.
Thanks everyone for their suggestions.:D