Imagine that you are running a news site that has a collection of articles from various authors.
Your directory structure is as follows:

AuthorName
DocumentName. txt - actual content of a published article
DocumentName. meta - a companion metadata document (pick CSV, XML or JSON format) containing the title, date published, and the subject tags

Create an application that, when given the author's name and a date range, will generate a text document (CSV, XML or JSON) containing the titles, dates of publication and relative paths to the corresponding content of the matching articles. Demo application or unit tests are accepted.