CN100410933C - XML data based highly effective path indexing method - Google Patents
XML data based highly effective path indexing method Download PDFInfo
- Publication number
- CN100410933C CN100410933C CNB2004100992728A CN200410099272A CN100410933C CN 100410933 C CN100410933 C CN 100410933C CN B2004100992728 A CNB2004100992728 A CN B2004100992728A CN 200410099272 A CN200410099272 A CN 200410099272A CN 100410933 C CN100410933 C CN 100410933C
- Authority
- CN
- China
- Prior art keywords
- node
- index
- path
- similarity
- state
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Images
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The present invention belongs to the technical field of a database, which particularly to provides a novel XML path index-UD (k, l) index. The present invention is an effective approximate index structure, and the classification of data nodes are according to an upward path of a K length and a downward path of an l length. The present invention roundly uses information of upward local approximate degree and downward local approximate degree of XML data nodes, and thus, the present invention can be used for effectively executing path expressing modes and is especially used for executing sub-path expressing modes.
Description
Technical field
The invention belongs to database technical field, be specifically related to a kind of novel and efficient is carried out index to the XML data method.
Background technology
In recent years, XML (extend markup language) has become the main standard of data distribution and exchange on the Internet.Along with a large amount of appearance of XML document, also obtained people at the inquiry of XML data and more and more paid close attention to.Various query languages are suggested one after another.Wherein, the execution route expression formula is the importance of XML inquiry.The manner of execution of simple path expression formula is directly to carry out detailed searching on whole XML document, and this obviously is a poor efficiency very.Path indexing is by being limited in the execution efficient that improves path expression on only relevant with the inquiry XML partial document with searching.Therefore, how from semi-structured data, to extract the path indexing structure recently by extensive concern.
The path indexing structure that has proposed has DataGuide, 1-index, Fabric index and A (k) index etc.The method that DataGuide and follow-up work thereof propose is to set up the structural summary of the digraph form of a tape label.Be intended that with the least possible node and Bian Lai and preserve all routing informations in the data plot.Yet, such index size or even raw data several times.This is because they are accurate index: from the root node of XML data, all paths comprise that those very long paths of seldom using all have been recorded, and in DataGuide, the chance that each back end occurs may be more than once.A (k) index reduces the index size by the local degree of approximation of considering back end, so can effectively support simple (no branch) path query.The downward similarity of having ignored them but it has only considered the upwards similarity of node.So this node is a poor efficiency when handling the inquiry of band individual path.
Summary of the invention
The objective of the invention is to propose a kind of indexing means efficiently based on the XML data, this method has been introduced and has been called UD (k, the efficient execution route expression formula of novel approximate index structure l) is especially with the path expression of branch.
Full use of the present invention the upwards local similar degree and the information of local similar degree downwards of XML back end, though it is the storage space of index is more bigger than A (k) index, all more faster than A (k) index and 1-index when handling the path expression of band branch.
1. some notion and the definition relevant with the present invention.
[1] XML document:
Among the present invention, XML document is turned to an oriented label figure G=(V by pattern
G, E
G, ∑
G, lab, oid, val root), is called the XML data plot.Here, V
GBe the set of node, E
GBe the set on limit, wherein the set membership or the element-numerical relation of an element represented on every limit.∑
GIt is the set of all labels in the XML document.Lab, oid and val are three kinds of mapping functions, and lab is mapping V
GIn a node to ∑
GIn the function of a label, oid is mapping V
GIn the function of node to a unique indications, val then is that mapping (enum) data value to is not exported on the leaf node on limit.Last root is V
GIn be labeled as unique root node of ROOT.
Fig. 1 is the partial information that has shown the top football league in England with the mode of XML data plot.Digitized representation oid in the node.Set membership between the solid wire representative element.Leaf elements among the XML figure has data value.The generality link between element is represented on the dotted line limit.
[2] tag path and node path:
The a series of label l that one strip label path representation is separated by decollator "/"
1... l
p(p 〉=1).One data paths is then represented a series of node n that separated by decollator "/"
1... n
p(p 〉=1), and for 1≤i＜p, n
iBe n
I+1Father node.
[3] regular path expression:
By separator "/", two fens symbol " | " circulation symbols " * " and substitute accord with "? " define basic regular path expression R, as follows: R::=ε | l|_|R/R|R|R| (R) | R? | R
*Here, l ∈ E
G, ' _ ' is one can mate E
GIn the special symbol of any one label.
Further, can define the path expression of band branch, as follows: BR::=R[R] | R/BR|BR/BR|BR/R.Here R is a basic regular path expression.Path expression R for band branch
1[R
2], individual path appears at path R
1Last label on.This label is called as take-off point.The path expression of a band branch is made of a main path and these two kinds of paths, some condition paths.Main path is meant the path expression that obtains after the part (comprising bracket itself) that brackets of all brackets in the path expression that removes band branch, and a condition path then is meant that part of path that a bracket brackets.The path expression with any take-off point is not called the simple path expression formula.
[4] similarity relation and similarity:
We are V
GOn a kind of binary relation ≈ of symmetry
uBe called upwards similarity relation,, concern u ≈ if they exist for any two data node u and v
uV has then a) that u has identical label with v, and b is arranged) for any child node u ' of u, there is the child node v ' of v, make u ' ≈
uV ', vice versa.If exist between two node u among the G and the v and concern u ≈
uV, we just say that these two nodes make progress similar.
We are V
GOn a kind of binary relation ≈ of symmetry
dBe called downward similarity relation,, concern u ≈ if they exist for any two data node u and v
dV has then a) that u has identical label with v, and b is arranged) for any child node u ' of u, there is the child node v ' of v, make u ' ≈
dV ', vice versa.If exist between two node u among the G and the v and concern u ≈
dV, we just say that these two nodes are similar downwards.
If symmetrical binary relation be upwards similarity relation also be downward similarity relation, we just claim that it is a similarity relation up and down, is designated as ≈
d uIf two node u among the G and v make progress similarly, also are similar, claim that then they are similar up and down downwards.
≈
l k(k-l-similarity) must meet following three conditions a) for any two node u and v,
And if only if, and u has identical label with v; B)
And if only if
And, exist the father node v ' of node v to make for any father node u ' of node u
Vice versa; C)
And if only if
And, exist the child node v ' of node v to make for any child node u ' of node u
Vice versa.
[5] UD (k, l) index structure and structure thereof
(k, l) index is that the input path of k and outgoing route that length is l come the node in the data plot is classified according to the length of node to UD.The k-l similarity has defined a relation of equivalence on the data plot node, and we are referred to as the k-l similarity relation.We set up an index node for each such equivalence class earlier, regard the node in the equivalence class expansion of this index node as, if index node is designated as X, corresponding equivalence class just is designated as ext[X].Then, if figure G in ext[A] in certain node and ext[B] in certain node between have the limit, just index node A is linked to each other with the limit with index node B.Like this, we have just constructed a key map based on the k-l similarity, are called UD (k, l) key map.
(k, l) index has the following character to UD.
Character 1: for node u and v, if
Then their all lengths are that (the input label path of m≤k+1) all is identical to m, and all length is that (outgoing route of n≤l) also all is identical to n.
(k, l) index is that (the simple path expression formula of m≤k+1) and length are that (Query Result of the condition path expression of n≤l) all is accurate to n to m for any length to character 2:UD.
Character 3: for length is that (simple path of m 〉=k+1) and length are that (the condition path of n 〉=l), (k, Query Result l) is similar to UD n to m.The result set that it returns is a possible result set, might not be real result set, need arrive the back end in the checking result set on the source data figure.
Character 4:UD (k, l) index is safe, that is to say, for any paths expression formula, the result set that it returns has always comprised carries out the result set that this inquiry should be returned in source data figure.Like this, not the result set that really needs even just guaranteed the result set that returns, also can remedy by on source data figure, verifying.
Character 5:UD (k+1, l) index or and UD (k, l) index is consistent, or is UD (k, l) index segmentation.Similarly, UD (k, l+1) index or and UD (k, l) index is consistent, or is UD (k, l) index segmentation.
(k, above-mentioned character l) can (k, l) index structure figure be used for the execution route expression formula, especially for carrying out the individual path expression formula, inquires about with UD according to UD.Concrete steps are: (with similarity l is as input downwards, ((k, l) document is as output for UD to set up good UD with the upwards similarity k of expection for k, l) index at first the source XML document to be made up UD.The first step is classified by its label to all back end, obtains a set of node tabulation, and each set of node all has a unique label; Calculate the upwards similarity of each set of node then, division has the minimum upwards set of node of similarity repeatedly, and upwards similarity ub minimum in all sets of node is not less than given parameter k; Then, calculate the downward similarity of each set of node, divide node minimum downward similarity dp in all sets of node equally repeatedly and be not less than given parameter l; Second step was the limit of setting up index node and connecting index node: at first each set of node is set up an index node; For any two summit u and vertex v that have the limit to link to each other among the figure G, if do not have the limit between index node A and the index node B, just add a last limit, the set of node of node A correspondence comprises u here, and the corresponding node collection of Node B comprises v.
A regular path expression can be converted into a DFA (determining finte-state machine).Can be of the input of XML data as this DFA, move this DFA and find those can Driven DF A to arrive the node of end-state.These node matching this regular path expression.The path expression of a band branch can convert a series of DFA to.Can carry out the path expression of band branch by constantly select corresponding D FA for data plot.If there is key map, key map surrogate data method figure can be inquired about as the input of DFA, to reduce the inquiry cost.A UD (k, it is as follows l) to carry out the path expression query steps of band branch on the key map:
Main path structure DFA A for band individual path expression formula
0, and be each condition path i structure DFA A
iThen on key map, move A
0, index is carried out depth-first traversal, automat carries out state exchange automatically according to the node of coupling, and sets up a table for branch node and write down their intermediate result; Work as A
0When arriving a branch node, suspend its execution, start corresponding A according to the condition path
iWork as A
iWhen arriving end-state, stop A
i, checking automat A
0The set of node of that index node correspondence that stops, the correct node of storage empirical tests in the intermediate result table continues to carry out automat A afterwards then
0If A
iCan not arrive end-state, automat A
0Just can not continue to have carried out forward, but will return its preceding state earlier, down carry out again; Work as A
0When arriving end-state, this moment, key map arrived certain node, and the set of node of this index is confirmed, will confirm that correct result adds net result to and concentrates.
Description of drawings
Fig. 1 is the XML document of the partial information of the top football league in description England.
Fig. 2 is the building process of a UD (2,2) index.
Fig. 3 represents the insertion process of node.
Embodiment
[1] UD (k, l) structure of key map:
UD (k, l) construct in two steps by key map.The first step is classified by its label to all back end, and the node that just will have same label is classified as a class, can obtain a set of node tabulation like this, and each set of node all has a unique label.Calculate the upwards similarity of each set of node then, division has the minimum upwards set of node of similarity repeatedly, and upwards similarity ub minimum in all sets of node is not less than given parameter k.Calculate the downward similarity of each set of node then, divide node minimum downward similarity dp in all sets of node equally repeatedly and be not less than given parameter l.The value of k and l need be determined that for different set of source data and query path collection, it may be different making performance reach optimum k and l value by experiment.Second goes on foot, and sets up the limit of index node and connection index node.At first each set of node is set up an index node.For any two summit u and v that have the limit to link to each other among the figure G, if (its corresponding set of node does not have the limit to link to each other between comprising v), then adds a limit for index node A (its corresponding set of node comprises u) and index node B.
When l gets 0, and UD (k, l) key map deteriorates to A (k) key map, and different is that each set of node has all kept its upwards similarity ub and downward similarity dp.Therefore when needs divide structure repeatedly, can directly find the minimum index node of ub or dp value to divide, and when structure A (k) index, need each set of node of inspection to observe it and whether need further division.So (k, l) structure time ratio A (k) index of index needs will reduce a lot UD.
Fig. 2 is to be the building process that example has been showed index with UD (2,2) index.Figure G is the structural drawing of source data, need set up UD (2,2) key map for it.At first, with node-classification, obtain UD (0,0) key map according to node label; Difference according to father node divides same category node then, and for example node 6 and 7 label all are E, but because their one of father node is B, another is C, so come by division, has so just obtained UD (1,0) key map; When requiring all with the label of grandfather's node of category node when also identical, promptly requiring the upwards similarity with category node is 2 o'clock, has just obtained UD (2,0) key map.Now, the upwards similarity of node has met the requirements, and in order to improve downward similarity, needs to continue division partial index node.According to the difference of child node separately, distinguish node 2 and 3, obtain UD (2,1) key map.At last, when the downward similarity that requires all with category node is 2, just obtained UD (2,2) key map.
[2] UD (k, l) index is used to the inquiry of the path expression of branch:
Be the main path structure DFAA of band individual path expression formula, and be each condition path configuration DFAA
iThen move A on key map, index is carried out depth-first traversal, automat carries out state exchange automatically according to the node of coupling, and sets up a table for branch node and write down their intermediate result; When A arrives a branch node, suspend its execution, start corresponding A according to the condition path
iWork as A
iWhen arriving end-state, stop A
i, the set of node of that index node correspondence that checking automat A stops, the correct node of storage empirical tests in the intermediate result table continues to carry out automat A afterwards then; If A
iCan not arrive end-state, automat A just can not continue to have carried out forward, but will return its preceding state earlier, down carries out again; When A arrived end-state, this moment, key map arrived certain node, and the set of node of this index is confirmed, will confirm that correct result adds net result to and concentrates.
(k, l) character of index self has determined UD, no matter be that the intermediate result of a branch node or the net result that key map returns all are a possible result set.So, for the node that need on data plot, verify than long query path in the result set.
The path expression of band branch is branched node and is divided into several parts.If index node N is at branch node A
I1Intermediate result in, ext[N so] in back end just might be intermediate result, need carry out the checking of direction up and down to them.If N.db 〉=m
1, (k, l) character 2 of index does not need to verify downwards as can be known by UD.If N.ub+1 〉=n
1, do not need the checking that makes progress.Otherwise, need on data plot, verify.
For example, on UD shown in Figure 2 (2,2) key map, carry out inquiry A/B[E/F], at first start from the DFA of main path structure, promptly A → B imports key map in the mode of depth-first, and automat suspends at node 2.Start the DFA of individual path, i.e. E → F, it is the child node of E that node 2 does not have label, so automat can not arrive end-state.Continue to carry out the DFA of main path, automat suspends at node 3, then starts the DFA of individual path, and there is the downward path with the E/F coupling in node 3, so automatic function arrives end-state, therefore obtains intermediate result node 3.Because in this inquiry, the upwards path of B label is 1, (its downward path is 2 for k, l) the k value 2 of index, is no more than this UD (k, l) the l value 2 of index to be no more than this UD.So need not on raw-data map, to verify.
[3] UD (k, l) increment type of index is safeguarded:
When source data changes, (k, l) corresponding adjustment made in index, could guarantee the validity of index to query expression like this to UD.Because rebulid the time cost of index is very high, so be necessary to set up an Incremental Updating Algorithm.
UD (k, l) to have defined the upwards similarity of back end be k to index, similarity is l downwards, is no more than l so be limited in upwards for the coverage of the renewal of any index node apart from it, is no more than in the node of k apart from it downwards.When upgrading the operation generation for one, adopt following steps that key map is carried out incremental maintenance:
If upgrade operation is to insert a subtree T on node υ, and depth-first traversal T inserts each node successively.When node θ inserts a node δ, at first divide and be binned in the θ top and the θ distance is the node of l, find suitable index node position insertion node δ then.If upgrading operation is the subtree T that deletion is positioned at node γ, at first from the index node of correspondence, remove all index nodes among the T; Then to node γ top and γ apart from dividing for the node of l and recombinating.
As shown in Figure 3, if will insert node 6 below node 4, its label is D, then earlier with the node in node 4 and the key map 5 separately, and then inserts node 6.Because the downward similarity that key map requires is 2, behind the insertion new node, may influence the ancestor node of new insertion node, so must advanced line splitting.
Claims (4)
1. highly effective path indexing method based on the XML data, it is characterized in that by making up UD (k, l) key map, be used for the execution route expression formula or carry out the individual path expression formula and inquire about, concrete steps are: the source XML document is made up UD (k, l) index, with the upwards similarity k of expection and downwards similarity l as input, set up good UD (k, l) document is as output, and the first step is classified by its label to all back end, obtain a set of node tabulation, each set of node all has a unique label; Calculate the upwards similarity of each set of node then, division has the minimum upwards set of node of similarity repeatedly, and upwards similarity ub minimum in all sets of node is not less than given parameter k; Then, calculate the downward similarity of each set of node, divide node minimum downward similarity db in all sets of node equally repeatedly and be not less than given parameter l; Second step was the limit of setting up index node and connecting index node: at first each set of node is set up an index node; For any two summit u and vertex v that have the limit to link to each other in the XML data plot, if do not have the limit between index node A and the index node B, just add a last limit, the set of node of node A correspondence comprises u here, and the corresponding node collection of Node B comprises v.
2. indexing means according to claim 1, it is characterized in that utilizing the UD (k that builds, l) key map is carried out band individual path expression formula, be with determining limited automata representation with path expression, as the input of determining finte-state machine, operation determines that finte-state machine finds those to drive and determines that finte-state machine arrives the node of end-state with index data.
3. indexing means according to claim 1 is characterized in that (k, l) the path i expression formula query steps of execution band branch is as follows on the key map: the main path structure that is band individual path expression formula is determined finte-state machine A at UD
0, and be that each condition path i structure is determined finte-state machine A
iThen on key map, move A
0, index is carried out depth-first traversal, A
0Node according to coupling carries out state exchange automatically, and sets up a table for branch node and write down their intermediate result; Work as A
0When arriving a branch node, suspend its execution, i starts corresponding A according to the condition path
iWork as A
iWhen arriving end-state, stop A
i, checking A
0The set of node of that index node correspondence that stops, the correct node of storage empirical tests in the intermediate result table continues to carry out A afterwards then
0If A
iCan not arrive end-state, A
0Just can not continue to carry out forward, but will return its preceding state earlier, down carry out again; Work as A
0When arriving end-state, key map arrives certain node, and the set of node of this index is confirmed, will confirm that correct result adds net result to and concentrates.
4. indexing means according to claim 1, it is characterized in that when the data of former document change, to UD (k, l) the increment type maintenance made in index, concrete operation method is as follows: if upgrade node δ of position insertion that operation is node θ in former document, then at first divide and be binned in θ top and the θ distance is the node of l, find suitable index node position insertion node δ then; If upgrade operation is to insert a subtree T on node γ, then with the method traversal subtree T of depth-first, inserts each node successively; If upgrading operation is the subtree T that deletion is positioned at node γ, then at first from the index node of correspondence, remove all index nodes among the subtree T, then to node γ top and γ apart from dividing for the node of l and recombinating.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2004100992728A CN100410933C (en) | 2004-12-29 | 2004-12-29 | XML data based highly effective path indexing method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2004100992728A CN100410933C (en) | 2004-12-29 | 2004-12-29 | XML data based highly effective path indexing method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1632792A CN1632792A (en) | 2005-06-29 |
| CN100410933C true CN100410933C (en) | 2008-08-13 |
Family
ID=34848013
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2004100992728A Expired - Fee Related CN100410933C (en) | 2004-12-29 | 2004-12-29 | XML data based highly effective path indexing method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100410933C (en) |
Families Citing this family (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101216824B (en) * | 2007-01-05 | 2010-04-21 | 冯卫国 | Method for publishing tree -type structure database as distributed XML database |
| JP4445509B2 (en) * | 2007-03-20 | 2010-04-07 | 株式会社東芝 | Structured document retrieval system and program |
| EP1986108B1 (en) * | 2007-04-27 | 2009-09-16 | Software Ag | Method and database system for executing an XML database query |
| CN101887458A (en) * | 2010-07-06 | 2010-11-17 | 江苏大学 | Path coding-based XML document index method |
| EP2492824B1 (en) * | 2011-02-23 | 2020-04-01 | Harman Becker Automotive Systems GmbH | Method of searching a data base, navigation device and method of generating an index structure |
| CN102682028A (en) * | 2011-03-17 | 2012-09-19 | 新奥特（北京）视频技术有限公司 | Cataloguing description indexing method and device based on dynamic field storage |
| CN102193988A (en) * | 2011-03-25 | 2011-09-21 | 北京世纪互联工程技术服务有限公司 | Method and system for retrieving node data in graphic database |
| CN104598474B (en) * | 2013-10-30 | 2018-04-27 | 同济大学 | Information recommendation method based on data semantic under cloud environment |
| CN104376084B (en) * | 2014-11-18 | 2018-01-30 | 百度在线网络技术（北京）有限公司 | Similarity of paths computational methods and device |
| CN105488162B (en) * | 2015-11-30 | 2019-04-16 | 用友优普信息技术有限公司 | Data processing method and device |
| CN105912877A (en) * | 2016-05-12 | 2016-08-31 | 成都鼎智汇科技有限公司 | Data processing method of medicine product |
| CN107678953A (en) * | 2017-09-22 | 2018-02-09 | 深圳航天科技创新研究院 | Path generating method, system and storage medium based on uml diagram shape |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030212662A1 (en) * | 2002-05-08 | 2003-11-13 | Samsung Electronics Co., Ltd. | Extended markup language (XML) indexing method for processing regular path expression queries in a relational database and a data structure thereof |
-
2004
- 2004-12-29 CN CNB2004100992728A patent/CN100410933C/en not_active Expired - Fee Related
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030212662A1 (en) * | 2002-05-08 | 2003-11-13 | Samsung Electronics Co., Ltd. | Extended markup language (XML) indexing method for processing regular path expression queries in a relational database and a data structure thereof |
Non-Patent Citations (4)
| Title |
|---|
| XML数据的路径表达式查询优化技术. 吕建华，王国仁，于戈.软件学报，第14卷第9期. 2003 |
| XML数据的路径表达式查询优化技术. 吕建华，王国仁，于戈.软件学报，第14卷第9期. 2003 * |
| 基于相容关系的XML索引机制. 徐海渊，吴泉源，王怀民，贾焰.电子学报，第31卷第8期. 2003 |
| 基于相容关系的XML索引机制. 徐海渊，吴泉源，王怀民，贾焰.电子学报，第31卷第8期. 2003 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1632792A (en) | 2005-06-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100410933C (en) | XML data based highly effective path indexing method | |
| US7831608B2 (en) | Service identification in legacy source code using structured and unstructured analyses | |
| CN1606004B (en) | Method and apparatus for identifying semantic structures from text | |
| US5929858A (en) | Device for aiding analysis of infeasible solution and unbounded solution | |
| CN104679867B (en) | Address method of knowledge processing and device based on figure | |
| CN102110109B (en) | Digital report topic making method and system | |
| CN103473171A (en) | Coverage rate dynamic tracking method and device based on function call paths | |
| CN105138601A (en) | Graph pattern matching method for supporting fuzzy constraint relation | |
| US8732116B1 (en) | Harvesting relational tables from lists on the web | |
| CN103198149A (en) | Method and system for query error correction | |
| CN107145584A (en) | A kind of resume analytic method based on n gram models | |
| JP2008191833A (en) | Logical structure recognition processing program, logical structure recognition processing method and logical structure recognition processor | |
| Schimm | Generic linear business process modeling | |
| CN103106294A (en) | Territory programming method for static random access memory compiler | |
| CN102968431B (en) | A kind of control device that the Chinese entity relationship based on dependency tree is excavated | |
| CN104239290A (en) | Statistical machine translation method and system based on dependency tree | |
| CN104125300B (en) | The synchronous method of separation between machine and card type home gateway service configuration data | |
| CN111369294B (en) | Software cost estimation method and device | |
| CN102857949A (en) | Method and device for planning data consistency guarantees | |
| CN104536877A (en) | Mixed strategy based test data generation method | |
| CN102591931B (en) | Recognition and extraction method for webpage data records based on tree weight | |
| CN102629244A (en) | Multi-language work card generating system and method | |
| CN106649219B (en) | A kind of telecommunication satellite design document automatic generation method | |
| CN115858865A (en) | MBSE-oriented demand model rapid query and visualization method | |
| CN101635711A (en) | Programmable character communication method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20080813 Termination date: 20101229 |