34results about How to "Improve lookup performance" patented technology

The invention discloses an FPGA-based logical operation search method. The method includes the steps of extracting super-quintuple information from an input network message; setting different fields for the super-quintuple information of the network message; comparing the fields of the super-quintuple information with a comparison range to generate corresponding identifier bits; from the identifier bits of the fields, generating a vector with logic bits and operations among elements; decomposing the vector to generate a plurality of sub-vectors according to the number of logics or operations; dividing a TCAM into the corresponding amount of areas according to vector dimensions required by search; configuring an entry of each area according to a descending order of effective states; correspondingly searching for the number of effective bits in the sub-vectors and the number of effective states in the TCAM, generating search result bits, and subjecting the result bits to logics or operations to generate final identifier bits to complete the search; passing the matched message and abandoning the failed message.

The invention discloses a method for checking a null pointer in a source code, comprising: a computer system construction referencable variable type dictionary, a computer system receiving an instruction for scanning the null pointer, formatting the source code to be checked; scanning the global code zone of the source code to be checked and obtaining an uninitialized global variable list corresponding to the variable statement type in the dictionary; scanning the method zone of the source code to be checked and obtaining the uninitialized local variable list corresponding to the variable statement type in the dictionary; acquiring uninitialized global variable list in method based on the uninitialized global variable list; scanning the method zone of the source code to be checked, searching the called variable name, if the called variable name is in the uninitialized local variable list or uninitialized global variable list in method, the found called variable name is the null pointerin the source code. The invention also discloses a computer system.

A classification management and index method for the extensive package box type warehouse is characterized by that a three-layer structure is used to manage the box-type warehouse and the database technique is used for classification management and index and increasing the additional description information about package boxes. Its advantage is high management and index efficiency.

The embodiment of the invention provides a method and system for searching for a useless file. An older-version application program installation directory meeting an installation directory feature corresponding to the application program in a preset feature library besides the installation directory in the parent directory of the application program installation directory obtained according to a registry, and the useless file is searched for in the searched older-version application program installation directory. Due to the fact that the new-version and older-version application program installation directories generally have a certain installation directory feature, the older-version application program installation directory can be found through the installation directory feature, and therefore the useless file is searched for in the older-version application program installation directory. Therefore, according to the method and system for searching for the useless file, the useless file search effect is improved, the situation that the useless file in the older-version installation directory exists in terminal equipment for a long time and occupies the disk space is avoided, and efficient operating of the terminal equipment is effectively guaranteed.

The invention provides a data packet classification method and system based on a convolutional neural network. The method includes merging each rule set in the training rule set to form multiple merging schemes, determining the optimal merging scheme for each rule set in the training rule set based on performance evaluation; combining each rule set in the training rule set and the target rule set The distribution of the prefix combination is converted into an image, and the convolutional neural network model is trained with the image and the corresponding optimal merging scheme; the target rule set is classified based on the image similarity, and the corresponding hash table is constructed for packet classification. The method of the invention can remarkably improve the data packet search performance, increase the data packet search speed, and increase the update speed of the rules. Through the mutual cooperation between the online system and the offline system, the system of the present invention can ensure that the online system realizes efficient search of data packets and rapid update of the rule set, and can monitor the update of the rule set to always reflect the latest state of the network.