### xml

#### Targetted #INCLUDE-like directive in markup languages for documentation purposes

The goal
I'm looking for a markup language for writing documentation that would support a very flexible #INCLUDE-like directive that can target specific elements of an included file or resource. The markup language should also have some LaTeX (and ideally also HTML) exporter.
However, the elements should be included from the file as they are without any additional tagging of the elements.
One additional requirement is to be able to use LaTeX math fragments as this is to be used in academia.
Example usage
For example, I may have several documents about specific components of a device and one day I decide I want to create a report about certain aspects of certain components. Using this #INCLUDE-like directive I would include the paragraphs/resources/nodes from the various documents and then export it to PDF through LaTeX.
Research so far
I've had a look at most of the common markup (and even non-markup) languages and most do have something like and #INCLUDE directive, but usually it's only capable of including whole files or at best a range of lines from a file. However, line ranges are very brittle and may become quickly obsolete if the included document is e.g. extended.
Here I list some of the most promising candidates:
EMACS Org-mode
I've been playing around with the newly extended #INCLUDE directive functionality in Org-mode version 8.3 for extracting parts of files, documented in the manual and an example is in the answer of this question.
I'm aware of DocBook, but editing XML files is certainly not user-friendly. Or is there some good frontend these days? It seems LyX might be even able to use MathML in the result.