This invention relates generally to digital information processing. More particularly, this invention relates to techniques for forming and using a tree structured database with top-down trees and bottom-up indices.
A variety of markup languages are known in the art. For example, Extensible Markup Language (XML) is a restricted form of SGML, the Standard Generalized Markup Language defined in ISO 8879 and XML is one form of structuring data. XML is more fully described in “Extensible Markup Language (XML) 1.0 (Second Edition)”, W3C Recommendation (6 Oct. 2000), which is incorporated by reference herein for all purposes [and available at http://www.w3.org/TR/2000/REC-xml-20001006] (hereinafter, “XML Recommendation”). XML is a useful form of structuring data because it is an open format that is human-readable and machine-interpretable. Other structured languages without these features or with similar features might be used instead of XML, but XML is currently a popular structured language used to encapsulate (obtain, store, process, etc.) data in a structured manner.
An XML document has two parts: 1) a markup document and 2) a document schema. The markup document and the schema are made up of storage units called “elements”, which can be nested to form a hierarchical structure. The following is an example of an XML markup document:
The Cerisent XQE patent application describes a high-performance.
XML search and database system.
This document contains data for one “citation” element. The “citation” element has within it a “title” element, and “author” element and an “abstract” element. In turn, the “author” element has within it a “last” element (last name of the author) and a “first” element (first name of the author). Thus, an XML document comprises text organized in freely-structured outline form with tags indicating the beginning and end of each outline element. Generally, an XML document comprises text organized in freely-structured outline form with tags indicating the beginning and end of each outline element. In XML, a tag is delimited with angle brackets followed by the tag's name, with the opening and closing tags distinguished by having the closing tag beginning with a forward slash after the initial angle bracket.
Elements can contain either parsed or unparsed data. Only parsed data is shown for the example document above. Unparsed data is made up of arbitrary character sequences. Parsed data is made up of characters, some of which form character data and some of which form markup. The markup encodes a description of the document's storage layout and logical structure. XML elements can have associated attributes in the form of name-value pairs, such as the publication date attribute of the “citation” element. The name-value pairs appear within the angle brackets of an XML tag, following the tag name.
XML schemas specify constraints on the structures and types of elements and attribute values in an XML document. The basic schema for XML is the XML Schema, which is described in “XML Schema Part 1: Structures”, W3C Working Draft (24 Sep. 1999), which is incorporated by reference herein for all purposes [and available at http://www.w3.org/TR/1999/WD-xmlschema-1-19990924]. A previous and very widely used schema format is the DTD (Document Type Definition), which is described in the XML Recommendation.
Since XML documents are typically in text format, they can be searched using conventional text search tools. However, such tools might ignore the information content provided by the structure of the document, one of the key benefits of XML. Several query languages have been proposed for searching and reformatting XML documents that do consider the XML documents as structured documents. One such language is XQuery, which is described in “XQuery 1.0: An XML Query Language”, W3C Working Draft (20 Dec. 2001), which is incorporated by reference herein for all purposes [and available at http://www.w3.org/TR/XQuery].
XQuery is derived from an XML query language called Quilt [described at http://www.almaden.ibm.com/cs/people/chamberlin/quilt.html], which in turn borrowed features from several other languages, including XPath 1.0 [described at http://www.w3.org/TR/XPath.html], XQL [described at Http://www.w3.org/TandS/QL/QL98/pp/xql.html], XML-QL [described at http://www.research.att.com/.about.mfflfiles/final.html] and OQL.
Query languages predated the development of XML and many relational databases use a standardized query language called SQL, as described in ISO/IEC 9075-1:1999. The SQL language has established itself as the linquafranca for relational database management and provides the basis for systems interoperability, application portability, client/server operation, and distributed databases. XQuery is proposed to fulfill a similar role with respect to XML database systems. As XML becomes the standard for information exchange between peer data stores, and between client visualization tools and data servers, XQuery may become the standard method for storing and retrieving data from XML databases.
With SQL query systems, much work has been done on the issue of efficiency, such as how to process a query, retrieve matching data and present that to the human or computer query issuer with efficient use of computing resources to allow responses to be quickly made to queries. As XQuery and other tools are relied on more and more for querying XML documents, efficiency will be more essential.
One problem with data analysis is that qualities of data often need to be determined for classification, comparison or other analytical purposes. A simple quality is whether or not the data contains a specified element. With text documents, an inquiry can be made as to whether a text document contains a string of interest. A search system, for example, can find all files in a corpus that contain a particular string, set of strings, regular expression, etc. It is desirable to avoid the inefficiency of searching an entire corpus for a particular string or path expression. Therefore, there is a need for improved indexing schemes.
A method for loading information into a tree structured database includes receiving a document and forming a top-down tree characterizing the document. Leaf nodes in the top-down tree are identified. Bottom-up indices are formed for the leaf nodes, where the bottom-up indices characterizes paths from selected leaf nodes to a root node of the top-down tree. The top-down tree and bottom-up indices are stored as separately searchable entities in the tree structured database.
A method of processing a query to a tree structured database includes resolving a query to path constraints. The path constraints are matched to separately searchable entities of the tree structured database to form matched paths. The tree structured database includes top-down trees characterizing path structures for documents and bottom-up indices for nodes of the path structures for the documents. The bottom-up indices characterize paths from selected leaf nodes to root nodes of the top-down trees.
The invention is more fully appreciated in connection with the following detailed description taken in conjunction with the accompanying drawings, in which:
Like reference numerals refer to corresponding parts throughout the several views of the drawings.
A memory 110 is also connected to the bus 106. The memory 110 includes data and executable instructions to implement operations of the invention. A data loader 112 includes executable instructions to process documents and form top-down trees and bottom-up indices, as described herein. These trees and indices are then stored in a tree structured database 114. A query processor 116 includes executable instructions to decompose a query and apply it against the database 114, as discussed below. A user interface 118 includes executable instructions to define an interface to coordinate operations of the invention. A database manager 120 includes executable instructions to perform various database management operations.
The modules in memory 110 are exemplary. These modules may be combined or be reduced into additional modules. The modules may be implemented on any number of machines in a networked environment. It is the operations of the invention that are significant, not the particular architecture by which the operations are implemented.
While top-down trees have been used in the prior art to support various search mechanisms, the disclosed technology supplements such top-down trees with the bottom-up indices, which may be conveniently formulated while producing the top-down trees. As demonstrated below, the bottom-up indices allow one to reduce the amount of searched content during query processing. In particular, a match at a leaf node allows one to follow a path to a root node. That is, the path from a leaf node to a root node is a deterministic path up a tree structure. Utilization of the bottom-up indices allows one to limit a search to the relevant segments of a tree structure.
The path parameters may be expressed relative to the root node or absolute to the root node. Parameter expressions support descendent operators at any depth within a tree. Similarly, wildcards may be in paths at any depth within a tree. Paths may also contain predicates at any depth. Paths may also contain union operators at any depth.
The database manager 120 is responsive to inputs from the user interface 118. The database manager 120 includes executable instructions to coordinate operations associated with the database 114.
Embodiments of the invention support the expression of various path constraints including equal-to, greater-than, greater-than-or-equal-to, less-than, less-than-or-equal-to and not-equal to. Path constraints may be specified for each data type. In the case of a string data type, a user specified collation may be defined.
The operations of the invention are more fully appreciated with some specific examples.
Various path expressions may be used to query the structure of
The indices used in accordance with embodiments of the invention provide summaries of data stored in the database. The indices are used to quickly locate information requested in a query. Typically, indices store keys (e.g., a summary of some part of data) and the location of the corresponding data. When a user queries a database for information, the system initially performs index look-ups based on keys and then accesses the data using locations specified in the index. If there is no suitable index to perform look-ups, then the database system scans the entire data set to find a match. The invention provides bottom-up indices to facilitate search operations. Advantageously, the bottom-up indices are formed while forming top-down tree structures. Consequently, an incremental amount of additional processing provides bottom-up indices that may be effectively used to enhance search operations.
User queries typically have two types of patterns including point searches and range searches. In a point search a user is looking for a particular value, for example, give me last names of people with first-name=“John”. In a range search, a user is searching for a range of values, for example, give me last names of people with first-name>“John” AND first-name<“Pamela”.
Observe that the type of indices required for these two types of queries are different. Point search does not need keys in the index to be stored in a sorted order, but the range index must store sorted values. Database systems usually exploit this subtle difference for efficiently implementing the two types of indices. Range indices contain the entire range of values in a sorted order stored in a data structure that is more suitable for extracting ranges. On the other hand, value indices are stored in structures that are efficient for insertion and retrieval of point value, such as hash tables. A path range index is a collection of sorted values, for example found in an XML document using a user specified path expression. It is useful for queries that search a range of values on a particular path in the database.
The structure 500 of
A path expression is a branch of a tree. A path expression can be traversed both top-down and bottom-up. Document trees may be traversed at various times, such as when the document gets inserted into the database and after an index look-up has identified the document for filtering. Paths are traversed at various times: (1) when a document is inserted into a database, (2) during index resolution to identify matching indices, (3) during index look-up to identify all the values matching the user specified path range and (4) during filtering. The bottom-up indices of the invention may be utilized during these different path traversal operations.
Top-down traversal can be viewed as forward traversal and bottom-up traversal can be viewed as backward or reverse path traversal. The advantage of top-down traversal is that it is natural and starts with the first node in the document tree or path expression. The database system has to keep track of all the nodes traversed subsequently until the traversal hits a leaf. If there are multiple path indices defined in a system, the system has to traverse all the paths starting at the root to the leaf. This can be very inefficient when there are many paths with large depths. The state of the art implementations of path indices use top-down traversals. They are not only inefficient, but also have a limitation that each path must start from the root of a document. In contrast, the invention uses a combination of top-down document traversal and bottom-up path traversal for efficient document processing. The reverse path traversals are used for index resolution and look-up. This provides high flexibility in path expression syntax and further provides higher performance than top-down path traversal techniques.
Thus, the invention provides bottom-up path traversals at all phases of query processing. This is accomplished through the disclosed data structures used for bottom-up processing. Advantageously, these data structures may be efficiently generated while traversing top-down trees. The structures and techniques of the invention allow for a variety of path operators in queries that use reverse path validation. The path operators may include absolute, relative and descendant paths. Wildcards, union and predicates (relational and existential) may also be used. Further, the disclosed techniques support element and attribute paths.
Next, path expression keys and path leaf keys are computed 602. That is, namespaces are resolved into elements and attributes. Keys are computed for individual operators in the parsed tree. Then, an overall path expression key is computed. A leaf key for a path expression may be based on criteria, such as an element, an attribute or a wildcard.
A path expression key may be based on criteria, such as different types of nodes and operators in the path and values of elements and attributess in the path predicates. A range index configuration table is then loaded 604.
Returning to
The leaf path type is determined in block 608. If an element type is identified, then it is determined at block 610 if the element is a wildcard. If so (610—Yes), the key is loaded into an element leaf wildcard path vector 612.
Returning to
Thus,
Block 1402 of
Returning to
If a leaf is not an element (1706—No), it is determined whether the attribute is a wildcard 1714. If not (1714—No), a lookup is made 1716 to the attribute leaf path table 1200 and attribute leaf wildcard path vector 1300. If so (1714—Yes), a lookup is made 1718 to the attribute leaf wildcard path vector 1300. Processing from blocks 1710, 1712, 1716 and 1718 proceeds to block 1720. If more paths need processing (1720—No), a path expression in the query path is matched to a path expression in the path range index 1722. The matched path is then added to the query plan 1724. This is repeated until the last path is processed (1720—Yes), which terminates processing. Thus, paths in a query are matched to bottom-up path expressions, which may be used to reduce the number evaluated documents.
The disclosed indices allow a query to be performed which only considers documents 1918. As shown in the figure, this is a small sub-set of all of the documents 1900. Thus, the indices of the invention allow a focused search on a small number of documents. Consequently, data filtering is minimized, if not eliminated. The disclosed query resolution process identifies the set of indices that can produce the smallest set of documents to inspect. The query plan includes the steps to compute the query results. The query plan includes the indices identified through the disclosed index resolution.
The disclosed indexing techniques enable a high-performance query evaluation engine. The query evaluator is capable of using multiple indices in evaluating a single complex query. While each individual index can improve query performance by reducing the amount of data fetched off disk, the query evaluator can aggregate the gains of all indices by composing the use of the indices during a single query evaluation. Therefore, index and query evaluation designs allow the evaluator to use multiple indices at the same time.
The disclosed techniques may be used in connection with geospatial constraints. For example, the system can find all data items that meet a geospatial constraint quickly by using an index to identify and fetch only matching items off disk. For example, a query request all data items that contain the phrase “hello world” and contain a coordinate within 500 miles of latitude 10 degrees and longitude 24 degrees. The full-text index is used in conjunction with a geospatial index.
The indices allow for query evaluation of complex queries. A simple query is a restriction that can be efficiently resolved with a single index. A complex query is a composition of multiple simple queries using Boolean operators, such as AND, OR, AND-NOT. Thus, multiple indices support queries, such as element-value, element-word and geospatial queries.
An embodiment of the present invention relates to a computer storage product with a computer readable storage medium having computer code thereon for performing various computer-implemented operations. The media and computer code may be those specially designed and constructed for the purposes of the present invention, or they may be of the kind well known and available to those having skill in the computer software arts. Examples of computer-readable media include, but are not limited to: magnetic media such as hard disks, floppy disks, and magnetic tape; optical media such as CD-ROMs, DVDs and holographic devices; magneto-optical media; and hardware devices that are specially configured to store and execute program code, such as application-specific integrated circuits (“ASICs”), programmable logic devices (“PLDs”) and ROM and RAM devices. Examples of computer code include machine code, such as produced by a compiler, and files containing higher-level code that are executed by a computer using an interpreter. For example, an embodiment of the invention may be implemented using JAVA®, C++, or other object-oriented programming language and development tools. Another embodiment of the invention may be implemented in hardwired circuitry in place of, or in combination with, machine-executable software instructions.
The foregoing description, for purposes of explanation, used specific nomenclature to provide a thorough understanding of the invention. However, it will be apparent to one skilled in the art that specific details are not required in order to practice the invention. Thus, the foregoing descriptions of specific embodiments of the invention are presented for purposes of illustration and description. They are not intended to be exhaustive or to limit the invention to the precise forms disclosed; obviously, many modifications and variations are possible in view of the above teachings. The embodiments were chosen and described in order to best explain the principles of the invention and its practical applications, they thereby enable others skilled in the art to best utilize the invention and various embodiments with various modifications as are suited to the particular use contemplated. It is intended that the following claims and their equivalents define the scope of the invention.