* is because extraction backends are free to implement an
* extraction algorithm that might be more efficient than
* processing every file in the "extract file structure" and
- * "extract file metadata" phases. For example, the current
- * implementation of the UNIX extraction backend will create
- * files on-demand during the "extract file data" phase.
- * Therefore, when using that particular extraction backend, @p
- * end_file_count will only include directories and empty files.
- */
+ * "extract file metadata" phases. */
uint64_t end_file_count;
} extract;