How to use regular expressions in XML file types to mark placeholders?

Question

I created a custom XML file type for an XML file with embedded HTML content and would like to mark formatting strings enclosed in rectangular brackets as placeholders. Here's an excerpt: This is [B] an [/B] example.

]]> What kind of XPATH query will I need to use to select all strings enclosed by rectangular brackets?

Jerzy Czopik · Accepted Answer

Unfortunately you cannot combine more than two parser settings, so you need to define both "normal" tags and []-tags in one parser, using regex.

Daniel Hug · Answer

Hans Kröger I think you are expecting too much from the Studio parser. There is an XML parser to handle the XML content and an HTML parser to handle the embedded content. If you have custom requirements, you might have to find a custom solution. One simple way might be to turn the [B]stuff[/B] into tags. If you pre-process your files to arrive at something like this: This is example.

]]> (I used the regex ($$B$$.*?$$/B$$) for matching and for replacing.) How easy that is depends a bit on the content: Is it always [B] or can it be [All kind of things]? What is enclosed? If it's characters like <, ", ', & the content might need escaping. If it is as simple as it is in your sample, then the above will show in the editor a like that: Daniel EDIT: I should add that if you go this route, you will have to post-process the files accordingly. Just to state the obvious.

Paul · Answer

Hans Kröger said: This is [B] an [/B] example.

]]> Hans Kröger Perhaps the easiest solution is to simply use the Data Protection Suite after you have created your project. This will allow you to tag up the placeholders in your project and remove them before saving the target files. Simple enough.

Jerzy Czopik · Answer

I'm afraid xpath won't work here. Even the typical regex rule for tag pairs will not work. So what remains is creating placeholders for such expressions, preferably $$[^]]*$$

Paul · Answer

Hans Kröger Hans Kröger said: Can you please provide an example of the text() syntax that SDL Studio supports? It supports the correct way to use it in XPath: //*[contains(text(), 'toast')] In this case any segments at all containing the word toast. //VariableAssignment[answer/text()='42']/Value A way to extract the content of the element ONLY if the value of the element is 42 - using a conditional xpath expression;

Trados Studio > 5. Regex and XPath

How to use regular expressions in XML file types to mark placeholders?

Top Replies