For each section of a buffer, a string of characters is retrieved and reordered to generate a reordered string of characters for the section. Thereafter, for each section of the buffer, the string of characters within the section is copied over with the reordered string of characters for a different section of the buffer. The buffer is then output to a file. If a desired length of the file has not yet been reached, then this process is repeated, or the buffer is again output to the file, based on a desired amount of duplicate data to be included within the file. The resulting file can be used to test a single-instance storage algorithm. The file is thus adapted to mimic a non-artificially generated file in relation to testing a single-instance storage algorithm.