Conclusion

Through this book we have covered many of the directions that the BWT has led people in: new algorithms and data structures for performing the transform, fine-tuning the coding that can be done with the transformed file, new ways to sort the strings for the transform, a better understanding of other compression methods by relating them to the BWT, using the BWT data structures to aid searching and pattern matching, and applying its algorithms and data structures in contexts ranging from image analysis to computational biology. Yet there is a strong sense that we are only just beginning to understand the transform and its potential, as new variants and applications continue to be published regularly. The BWT is a powerful idea, and in the process of decoding generates a collection of useful arrays (R, V, W and so on) which can be used to provide a variety of indexes and views of the original text. The transform process thus provides us with data structures that have opened up novel possibilities, and may yet hold more opportunities for future applications.