Mihai Cazac wrote
Also I'm thinking that the information presented by the server can be aggregated from different sources, so in this case how can the browser know what physical file is parsed?
"Knowing" what files have to be fetched happens in OSI Layer 7. The
DOM is where the textarea (the editable) part of the web page is identified. I'm using ItsAllText to write this - there is no communication with the DokuWiki webserver during editing, only when the post is saved (you can check that by viewing the live HTTP headers in the Chrome Web Developer tools).
But that is really off-topic for this forum, as you've noted. If you want to know more you'd be better off asking somewhere like StackExchange.
Mihai Cazac wroteOh, and I think that my example with notepad (or other text editor) misleads; in fact I have to do a lot of other kind of processing on the file representing a DokuWiki page. (convert to another format, out and in, save parts of it in a database, aggregate it with other documents, and so on) so I'm not seeking a text editor.
A "good" text editor will allow you to do many of those thing, especially in
combination with other tools. For everything else there is scripting (and ssh). If you are doing variations of the same thing more than a couple of times you
should be automating the process. It's a subject I'm interested in, and I've attached an image from XKCD you may find useful in determining how much time is profitable to invest in automating tasks (I do).