Hi,
I would like to access the Wiki dumps for Wikipedia and Wikitravel. Essentially, I am looking to get dumps for some US cities from both these sources for product research. However, it is not very clear at http://dumps.wikimedia.org/backup-index.html, which are the relevant files to pick up.
Which Wikipedia do you want? What data do you want? If you want the dumps of the English Wikipedia, find “enwiki” in the index page, which will link to the most recent dump files (currently http://dumps.wikimedia.org/enwiki/20120211/). If you want current text of articles, you want the pages-articles dump (currently http://dumps.wikimedia.org/enwiki/20120211/enwiki-20120211-pages-articles.xm..., 7.6 GB).
Wikitravel is not run by Wikimedia, so you won't find their dump at the same page. You can find information about their dump at http://wikitravel.org/en/Wikitravel:Database_dump.
The other question I had was at what level of atomicity data is available in dumps. The Web Service allows us to retrieve a Wiki entry but it's not easily parsed out into different sections or in more granular form. I was wondering if the dumps solve this problem.
No, the dumps contain the article text is the same form as what you can edit as a user or what you can get though the API. So, if you want to get any useful information out of it, you have to parse it by yourself.
Petr Onderka [[en:User:Svick]]