A descriptor vector is added to every bucket in the EH* file. Then the substrings descriptor vector is compared with the keys descriptors vector and the buckets descriptor vector. At last, the keyset, where every key contains the substring, are retrieved.

• 通过对子串和关键字计算描述符向量，为eh*文件中的每个桶添加一个桶描述符向量，然后把子串描述符向量分别与关键字和桶的描述符向量进行比较，得到包含子串的关键字集。