JDBC/ODBC bridge - JDBC driver for ODBC databases, comes as part of the JDK (up to Java 7, it's been removed from Java 8 without a replacement); on other OSes, you'll have to get ODBC up and running first: http://www.unixodbc.org/
ImageJ - Java image processing application and library that has plugins for lots of image file formats
JIMI - library to read and write BMP, CUR, GIF, ICO, JPEG, PICT, PNG, PSD, Sun Raster, TGA, TIFF, XBM and XPM. There's a plugin for using JIMI with ImageJ, which also includes a couple of JIMI patches.
PDFBox - library that can merge, split and print PDFs, extract text, create images from PDFs, encrypt/decrypt PDFs, fill in PDF forms and more. See PDFBoxExample for example code of how to use it to create a PDF.
JPedal - library for viewing and printing PDFs, can also extract text (how to print PDFs); commercial (the LGPL version provides PDF viewing only)
PDFxStream - commercial library to extract text from PDFs
POI - library to read and write DOC and DOCX files. It can also be used for extracting the text of a document.
WordApi.exe is native Windows component with a Java interface, which lets you create Word documents, and alter word templates. Some impressions about it can be found here.
Java2Word - library to create Word documents, especially reports, on the fly.
If you encounter an obscure format for which no library is available, it may be feasible to create a reader for it if you have a file format description (which may be available on Wotsit, see link above). Several libraries, so-called lexers and parsers, are available that help in creating a reader, especially if the file format is ASCII, and not binary. You will need knowledge of regular expressions, though. Some file formats that have been tackled using this approach include RTF, CSV, HPGL and PBM/PGM/PPM. Lexers are easier to start with, but parsers can do more of the work for you. All these have ready-to-use examples on their web sites.