334 results about "Trie" patented technology

A translation memory adapted to sentence-based translation includes a first retrieval dictionary of a first language having a first dictionary TRIE structure and a first index TRIE structure, and a second retrieval dictionary of a second language having a second dictionary TRIE structure and a second index TRIE structure. The first and second dictionary TRIE structures respectively store sub character strings on respective paths from roots to multiple nodes thereof. The first and second index TRIE structure respectively store index information to specify the sub character strings stored in the first and second dictionary TRIE structures. The index information stored in the first retrieval dictionary includes identification information that identifies the index information to specify the sub character strings in the second retrieval dictionary that are translations of those in the first retrieval dictionary.

Methods and apparatus are disclosed for generating and using an enhanced tree bitmap data structure in determining a longest prefix match, such as in a router, packet switching system. One implementation organizes the tree bitmap to minimize the number of internal nodes that must be accessed during a lookup operation. A pointer is included in each of the trie or search nodes to the best match so far entry in the leaf or results array which allows direct access to this result without having to parse a corresponding internal node. Moreover, one implementation stores the internal node for a particular level as a first element in its child array. Additionally, one implementation uses a general purpose lookup engine that can traverse multiple tree bitmaps or other data structures simultaneously, and perform complete searches, partial searches, and resume partial searches such as after receiving additional data on which to search.

The present invention provides a data-structure to store a search database and provides techniques to build this datastructure given a list of prefixes (P) and to search this database efficiently for a best matching prefix for an address D. The data-structure can be stored in standard memory (14), where values are stored associated with memory address locations. The data structure includes representations of addressable linked tables (FIG. 3b). The representations are related to a binary search trie (FIG. 1) and each linked table (T) has at least one entry. Entries in a table span more than one level of the binary search trie. The spanning feature relates to compression of a binary search trie into a finite number of levels (and hence tables). The finite number is less than the number of levels in the binary search trie. Hence the search algorithm is restricted to a finite, and predetermined number of search accesses to the tables to obtain a best-match result.

A method and system are provided for approximate string matching of a target string to a trie data structure. The trie data structure has a root node and generations of child nodes each node representing at least one character in an alphabet to provide a lexicon of words and word fragments. The method involves traversing the trie data structure starting from the root node by comparing each node of a branch of the trie data structure to characters in the target string and adding characters traversed in a branch of the trie data structure to a gathered string to provide suggestions of approximate matches. If the method reaches a node flagged as a node for a word or a word fragment and, if the target string is longer than the gathered string, the method loops back to the root node, and continues the traverse from the root node. This enables the trie data structure to use word fragments for compound words and to split non-delimited words where appropriate. The method also includes, at each node, determining if there is a correction rule for one or more characters in the remainder of the target string from the current node, and if so, applying the correction rule to the target string to obtain a modified target string.

In a database file management system for accessing data records that correspond to items in a directory. The directory items are linked to a trie index that is arranged in blocks and being stored in a storage medium. The trie index enables accessing or updating the directory items data records by key or keys and being susceptible to an unbalanced structure of blocks. There is provided a method for constructing a layered index arranged in blocks, which includes the steps of providing the trie index and constructing a representative index over the representative keys of the trie index. The layered index enables accessing or updating the directory items by key or keys and it constitutes a balanced structure of blocks.

Fast, scalable methods and devices are provided for layer four switching in a router as might be found in the Internet. In a first method, a grid of tries, which are binary branching trees, is constructed from the set of routing filters. The grid includes a dest-trie and a number of source tries. To avoid memory blowup, each filter is stored in exactly one trie. The tries are traversed to find the lowest cost routing. Switch pointers are used to improve the search cost. In an extension of this method, hash tables may be constructed that point to grid-of-tries structures. The hash tables may be used to handle combinations of port fields and protocol fields. Another method is based on hashing, in which searches for lowest cost matching filters take place in bit length tuple space. Rectangle searching with precomputation and markers are used to eliminate a whole column of tuple space when a match occurs, and to eliminate the rest of a row when no match is found. Various optimizations of these methods are also provided. A router incorporating memory and processors implementing these methods is capable of rapid, selective switching of data packets on various types of networks, and is particularly suited to switching on Internet Protocol networks.

Techniques for representing nodes of tries. Associated with the nodes are keys and rules. A node of a trie having a stride n>1 is represented by a trie having a stride of 1 and the stride 1 trie is represented by a bit string termed a structural enumeration. The structural enumeration has a bit for each node of the trie of stride 1. If the node has a key and rule associated with it, the bit is set; otherwise it is not. The representation of a node of stride n>1 includes the node's structural enumeration and an array of rule pointers. The array has an entry for each rule associated with the node and the entries have the same order in the array as the set bits for their keys in the structural enumeration. Nodes having large strides may be represented by subdividing them into subtries.

An improved method for creating an index based on a path-compressed binary trie in a database system comprising database tables and indexes on those tables is described. For a given index to be created, a path-compressed binary trie for the given index is determined. The path-compressed binary trie comprises internal nodes and leaf nodes. Based on a traversal of the path-compressed binary trie, an index is created comprising a first array of internal nodes encountered during the traversal, and a second array of leaf nodes encountered during the traversal. The database system employs said first and second arrays for providing index-based access for a given key value.

A method, computer program product, and data processing system for efficiently storing a set of hierarchically-specified names in a modular hardware design are disclosed. In accordance with a preferred embodiment of the present invention, a data structure for storing the names is built from a master trie. The master trie is used to store names of instances of modules contained within the design. The node in the master trie corresponding to a particular instance name is associated with an additional trie (“class trie”) corresponding to the class of module to which that instance belongs. In this additional trie are stored the names of the individual signals associated with that class of module. Where there are multiple instances of the same class of module within a design, each instance name may be associated with a single class trie storing each of the individual signal names associated with that class of module.

In a database file management system for accessing data records and being executed on data processing system. The data records are linked to a trie index that is arranged in blocks and being stored in a storage medium. The trie index enables accessing or updating the data records by key or keys and being susceptible to an unbalanced structure of blocks. There is provided a method for constructing a layered index arranged in blocks, which includes the steps of providing the trie index and constructing a representative index over the representative keys of the trie index. The layered index enables accessing or updating the data records by key or keys and it constitutes a balanced structure of blocks.

Improved method, data structure and computer readable medium for searching for digital information files. Files referenced by URLs may be quickly located by finding a minimum unique prefix for the desired URL, breaking the prefix into substrings, and traversing a trie data structure to find indices to another trie data structure that will yield the physical location of the stored digital information file. A node data structure may be used to construct the trie data structures, and may be compressed to allow the tries to occupy less memory, thus allowing the tries to be maintained in memory and less access to storage devices. The result is faster retrieval times for digital information files.

An improved method for creating an index based on a path-compressed binary trie in a database system comprising database tables and indexes on those tables is described. For a given index to be created, a path-compressed binary trie for the given index is determined. The path-compressed binary trie comprises internal nodes and leaf nodes. Based on a traversal of the path-compressed binary trie, an index is created comprising a first array of internal nodes encountered during the traversal, and a second array of leaf nodes encountered during the traversal. The database system employs said first and second arrays for providing index-based access for a given key value.

A block data compression system comprising a Compression unit and a Decompression unit, and an Algorithm for fast block data compression using multi-byte search. Objective of the invention is to develop a block data compression system and algorithm for fast block data compression with multi-byte search for optimal encoding during the learning phase of substitutional methods, allowing length-limited and relative small blocks of input data symbols to be compressed independently, as required by random-access storage or telecommunication devices; and reaching high-performance characteristics by employed accelerating architectures and highly pipelines data-flow principles. According to the present invention these objectives are accomplished by a Compression Unit comprising an Input-FIFO (8), connected to a Modeling Unit (6), where said Modeling Unit (6) is connected to a multitude of memory locations represented by Trie-Dictionary (4) memory, to a Zero-Finder Look-Up Table (3), to search means in the form of a Comparison Unit (5), to memory means in the form of a Literal-Dictionary (2) and also to an Encoder Unit (7); said Encoder Unit (7) also connected through an Aligning Unit (7) to an Output-FIFO (12). The invention comprises a block data compression system, composed of a Compression Unit and a Decompression Unit, and an algorithm for fast block data compression using multi-byte search; and is related to the field of data compression, specifically to the implementation of a lossless, adaptive, reversible and fast block data compression for storage and telecommunication devices.

An input lookup key is hashed and the hashed key divided into stride bits into a multi-level Trie structure. A compression function logically combines the stride bits to generate the compressed index bits into the stride tables. The bucket in the last stride table found by the hashed key may have several keys that collide at the same hash value. Discriminant bits are read from the key and select a stored key in the bucket table for verification of its result. Since the hashed key is a compression of the longer input key, more information is contained per bit of the hashed key than in the long key. The multi-stride lookup is performed first on the hashed key, allowing a faster convergence to the lookup result. The first stride can index a single hash table, with the remaining hash bits and discriminant bits used to select from among colliding keys.

Proprietary suite of underlying document image analysis capabilities, including a novel forms enhancement, segmentation and modeling component, forms recognition and optical character recognition. Future version of the system will include form reasoning to detect and classify fields on forms with varying layout. Product provides acquisition, modeling, recognition and processing components, and has the ability to verify recognized data on the image with a line by line comparison. The key enabling technologies center around the recognition and processing of the scanned forms. The system learns the positions of lines and the location of text on the pre-printed form, and associates various regions of the form with specific required fields in the electronic version. Once the form is recognized, the preprinted material is removed and individual regions are passed to an optical character recognition component. The current proprietary OCR engine is trained with a variety of Roman text fonts and has a back end dictionary that can be customized to account for the fact that the system knows which field it is recognizing. The engine performs segmentation to obtain isolated characters and computes a structure based feature vector. The characters are normalized and classified using a cluster centric classifier, which responds well to variations in the symbols contour. An efficient dictionary lookup scheme provides exact and edit distance lookup using a TRIE structure. An edit distance is computed and a collection of near misses can be output in a lattice to enhance the final recognition result. The current classification rate can exceed 99% with context. The ultimate goal of this system is to enable the processing of all tax forms including forms with handwritten material.

Online spelling correction / phrase completion is described herein. A computer-executable application receives a phrase prefix from a user, wherein the phrase prefix includes a first character sequence. A transformation probability is retrieved responsive to receipt of the phrase prefix, wherein the transformation probability indicates a probability that a second character sequence has been transformed into a first character sequence. A search is then executed over a trie to locate a most probable phrase completion based at least in part upon the transformation probability.

A method for searching for an input symbol string, includes receiving (B) an input symbol string, proceeding (C) in a trie data structure to a calculation point indicated by the next symbol, calculating (D) distances at the calculation point, selecting (E) repeatedly the next branch to follow (C) to the next calculation point to repeat the calculation (D). After the calculation (G), selecting the symbol string having the shortest distance to the input symbol string on the basis of the performed calculations. To minimize the number of calculations, not only the distances are calculated (D) at the calculation points, but also the smallest possible length difference corresponding to each distance, and on the basis of each distance and corresponding length difference a reference value is calculated, and the branch is selected (E) in such a manner that next the routine proceeds from the calculation point producing the lowest reference value.

In an embodiment, different aspects of a packet header and data included in the packet are singled out for attention, rather that just the four byte IP destination address. Different information is included in nodes of the trie that enables matching and branching on different header fields. In an embodiment, the ACL of a configuration file in a router or switch is compiled into a trie data structure located in the memory of the router or switch. In an embodiment, a trie data structure is used to map a multicast packet header by a sequence of nodes that match on destination address or source address.

A multi-stage (e.g., two-stage) packet-based lookup process using a Ternary Content Addressable Memory (TCAM) divided into partitions. The result of a first stage lookup is used to selectively search one of a plurality of TCAM partitions during the second stage. A subset of destination address bits may be used in the first stage to hash to a selected partition for the second stage. Alternatively, a partitioning algorithm segments a routing trie into partitions, and then, either a separate, small TCAM or one of the partitions is used in the first stage to map a prefix of the destination address to one of the TCAM partitions for use in the second stage. The “trie-based” algorithms may advantageously partition the trie such that each second stage partition comprises a substantially contiguous sequence of routing prefixes in a post-order traversal of the routing trie, together with one or more covering prefixes thereof.

The invention discloses an implementation method for multi-field rule matching, and relates to the technical field of high-speed stream classification. The implementation method comprises the steps that firstly, a rule list is established for each tuple according to quintuple definition of IP message data streams, then a Hash rapid stream list is established, and all information matched with rules is added in the rule lists; secondly, IP message data stream matching is conducted, corresponding actions are executed if matched data streams are found, otherwise the rule lists are searched, and mask matching inquiries are conducted. According to the implementation method, the mode that a prefix extension Trie tree high in inquiring speed is improved, part of memory space is sacrificed, and data compression and the Hash list are combined is adopted, and thus under the premise that the inquiring speed is ensured, memory explosion caused by a large-capacity rule base can be avoided.

A method and apparatus for updating stored data structures representing network forwarding information used for network route lookup is described. By making sure there is only one level of dependency between data structures storing forwarding information, these data structures may be updated quickly and with minimal overhead

A search engine improves search speed and reduces required memory for a longest prefix matching (LPM) router that routes packets using IP addresses. The search engine includes a first bit vector with set bits corresponding to address ranges. A set bit counter counts the set bits in the bit vector based on a first portion of the address of the a first packet. A first next hop table contains first pointers for each of the set bits. One of the first pointers is selected based on a number of set bits counted by the set bit counter. For longer addresses, the addresses are split into address portions. The search engine includes a trie data structure that has n levels. The n levels of the trie data structure include nodes representing non-overlapping address space.

A method for bursting a hash table of a key-value database comprises receiving a key and a value, traversing trie nodes of the key-value database from a root node to a leaf node by recursively dividing the key into a prefix and a suffix, reaching the leaf node, the leaf node being the hash table, determining that the key is not stored in the hash table, determining that the hash table is not able to store the key and the value, removing the hash table, associating a new trie node with a parent trie node of the hash table, associating two or more new hash tables with the new trie node, moving all keys and associated values from the hash table into one of the two or more new hash tables, and inserting the key and the associated value into one of the two or more new hash tables.

A router has a shape graph that is a compressed form of a trie that represents routing information for routing data packets in a network, and an update data structure that includes plural entries corresponding to nodes of the shape graph, the plural entries containing count values indicating respective numbers of nodes of the tie represented by the corresponding nodes of the shape graph. The router incrementally updates the shape graph as a portion of the routing information changes, where the incremental updating uses information in the update data structure.

A block data compression system comprising a Compression unit and a Decompression unit, and an Algorithm for fast block data compression using multi-byte search.Objective of the invention is to develop a block data compression system and algorithm for fast block data compression with multi-byte search for optimal encoding during the learning phase of substitutional methods, allowing length-limited and relative small blocks of input data symbols to be compressed independently, as required by random-access storage or telecommunication devices; and reaching high-performance characteristics by employed accelerating architectures and highly pipelines data-flow principles.According to the present invention these objectives are accomplished by a Compression Unit comprising an Input-FIFO (8), connected to a Modeling Unit (6), where said Modeling Unit (6) is connected to a multitude of memory locations represented by Trie-Dictionary (4) memory, to a Zero-Finder Look-Up Table (3), to search means in the form of a Comparison Unit (5), to memory means in the form of a Literal-Dictionary (2) and also to an Encoder Unit (7); said Encoder Unit (7) also connected through an Aligning Unit (7) to an Output-FIFO (12).The invention comprises a block data compression system, composed of a Compression Unit and a Decompression Unit, and an algorithm for fast block data compression using multi-byte search; and is related to the field of data compression, specifically to the implementation of a lossless, adaptive, reversible and fast block data compression for storage and telecommunication devices.

The invention relates to a method for implementing a functional memory and to a memory arrangement. The memory is implemented as a directory structure comprising a tree-shaped hierarchy having nodes at several different hierarchy levels, wherein an individual node can be (i) a trie node associated with a logical table wherein an individual element may contain a pointer pointing to a lower node in the hierarchy, or (ii) a bucket containing at least one element so that the type of an individual element in the bucket is selected from a group including of e.g. a data unit or a pointer to a stored data unit. To optimize the performance of the functional trie structure, the trie nodes are implemented as quad nodes of four elements, and in at least part of the directory structure groups of successive quad nodes are replaced by compressed nodes in such a way that (a) an individual group comprising a given quad node and its child nodes is replaced by a node whose logical table has 16 elements, and (b) a compressed node known per se is formed from said node of 16 elements by physically storing in the node only non-nil pointers and in addition a bit pattern on the basis of which the physical storage location in the node, corresponding to the search word, can be determined. The invention also relates to a structure in which no buckets are used.

A method for clustering together network IP addresses is disclosed. A number of IP addresses are received and processed to determine which IP addresses share a longest prefix matching. The longest prefix matching process is performed according to radix encoded trie which facilitates on-line clustering of the IP addresses. Client and / or server IP addresses may be clustered in accordance with the teachings herein.

The present invention is a method for improving delivery of content to a client communicating with a server on the Web. Groups or clusters of clients are formed by processing the IP addresses of the clients according to a network-aware, radix-encoded trie classification process. The groups of clients are categorized based on information about one or more clients in each group that can be determined by the server. That information is used to help drive tailored actions on the part of Web servers. Users with poor connectivity may choose not to spend much time at a Web site if it takes a long time to receive a page, even if the Web server at the site is not the bottleneck. Retaining such clients may be of interest to a Web site. Better-connected clients may be able to receive enhanced representations of Web pages such as with higher quality images.

The present invention relates to a searching technology for the routing forwarding table, in particular to a searching technology and a device which can enhance a multidigit Trie tree. Through optimizing and improving the searching method to the multidigit Trie, and the data structure of a tree node, particularly the relation between an expansion bitmap and an inferior tree node, and though adopting an FPGA and a DDRII SDRAM, the searching of the high-capacity routing forwarding table is realized, as well as the wire-speed searching under the 20 Gbps flow, even the 40 Gbps flow, and the present invention has obvious advantages in the aspects of table capacity and the cost etc. compared with the commercial three-state content addressable memorizer.