With AdScan32, you can effectively search within bulky collections of text units (called ads) organized as the entire files or their parts. Classical three-paned Explorer-like interface allows to easily switch between hierarchically arranged ad collections, view a list of ads, and read individual ads at the same time. Preindexed keyword-based full text search can be used when searching within collections of ads of no regular structure. This can be combined with preindexed search by user-defined data items extractable by regular structured text elements, if there are such ones in the ad texts as in the example of messages with data fields. The custom data items are defined by means of regular expression scripts, new friendly way of composing regular expressions, and assigned textual, numerical, date/time or currency data type being represented as columns in the ad list view. For textual columns search criteria are keyword-based for numerical, date/time and currency columns they are based on values of the respective data types. Searches are stored in the form of filters associated with the given ad collection. Switching between the filters gives an effective way to examine various aspects of an ad collection contents. Building/managing column and full text indexes is extremely fast and fully automated.