There are two approaches one could use to identify the filename extension of documents in a database.
- Include the extension in the Name of documents as they are captured to a database. That can be done by checking the option in Preferences > Import - Titles to use “Filenames with extension”. If that had been the chosen option, a Name search for the extension string can be done. There could be some false positives if the extension string occurs other than at the end of the Name, but the results list of a query, for example, that includes (xls OR xlsx) will include Excel documents that meet the search criteria.
I find extensions added to document Names distracting, so I don’t choose that option. But if I’ve added the Kind column to database views and to a search window (View > Columns > Kind), I can see the extension for each item in a list.
- Create a smart group (or filter a search in the full Search window by using the Advanced button). In this context, the predicate “Kind” isn’t necessarily used for individual filetype extensions. There are four cases of the use of “Kind” in the smart group editor list of options:
2a) Some common filetype extensions, e.g., HTML, plain text, rich text are available for selection in the list of options for the Kind predicate. For example, a search using the filter ‘Kind is HTML’ will restrict searches to items with that extension.
2b) Some Kind predicates in the list of options in the smart group editor are used for multiple filetype extensions within a category of documents, e.g., “audio”, “video”, “image”. For example, a search using the filter ‘Kind is image’ would include images in the results list that may have different extensions, e.g., TIFF, JPEG, etc. If one is only interested in JPEG images in the search results list, adding the sortable Kind column can allow one to ‘pull together’ the list of JPEG images among the search results. If I need to do further work with that collection of JPEG documents I could replicate them to a new group that contains only the selected JPEF documents.
2c) The Kind predicate can ‘split’ documents that have a single extension into categories based on another property. For example, for documents that share the PDF filename extension, Kind is PDF is used to distinguish documents that do not contain searchable text and Kind is PDF+Text is used to distinguish documents that do contain searchable text. For example, a smart group based on Kind is PDF can be used to identify candidates for OCR.
2d) The ‘Kind is Other’ predicate in the smart group editor is used for all other filetype extensions. Excel documents with the XLS or XLSX extensions fall under this predicate, along with many other possible filetypes in a database. If one adds the sortable Kind column to the search window (View > Columns > Kind), a sort of the search results can be used to pull together all Excel documents.
Using the ‘Kind’ predicates in the smart group editor (including the Advanced button in searches) in this way allows presentation of a small number of options, and avoids forcing the user to scroll through hundreds (possibly thousands) of extensions looking for the desired one. (Even were DEVONthink to detect and automatically list only those extensions present in a collection of open databases, I would have to perform scrolling looking for a particular extension among a significant number of them.)