| 摘要 |
The invention relates to a file management method that comprises a first step of organising the data file base that comprises creating a tree structure of directories with M levels of N directories each, M being an integer higher than 1, and the steps of storing data files including: applying a hash function on the identifier of a data file F<SUB>i</SUB> to be stored, determining the path of the directory R<SUB>di</SUB> at several destination level in the tree structure based on the result of the preceding step, storing the data file is said R<SUB>di</SUB> directory determined by the hashing function at a location depending on the identifier of the data file; and data file reading steps comprising: applying the same hashing function on the identifier of the data file F<SUB>j</SUB> to be read, determining the path of the target directory R<SUB>cj</SUB> in the tree structure based on the result of the preceding step, reading the data file in said directory R<SUB>cj</SUB> determined by the hashing function at a location depending on the identifier of the data file. |
