Output driven generation of a combined schema from a plurality of input data schemas

• US 9,892,135 B2
• Filed: 09/30/2014
• Issued: 02/13/2018
• Est. Priority Date: 03/13/2013
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Li, Jeff J., Nusbickel, Wendi L.
• Primary Examiner(s): Trujillo, James
• Assistant Examiner(s): Khoshnoodi, Fariborz
• Application Number:
  US14/501,955
• Publication Number:
  US 20150019477A1
• Time in Patent Office: 1,232 Days
• Field of Search: 707602, 707803
• CPC Class Codes: G06F 16/211 Schema design and management
• CPC Class Codes: G06F 16/254 Extract, transform and load [ETL] procedures, e.g. ETL data flows in data warehouses

Abstract

A computational device receives a plurality of versions of an input data schema. At least one element is selected from the plurality of versions of the input data schema based on an expected result. A combined schema is generated based on the at least one selected element. The input data is processed according to the combined schema.

Claims

1. A method, comprising:

receiving, via a computational device, a first version of an input data schema comprising a first plurality of elements and a second version of the input data schema comprising a second plurality of elements;

in response to determining that a user is interested in an output corresponding to selected elements of elements that are present in both the first plurality of elements and the second plurality of elements, and is not interested in outputs corresponding to other elements that are present in the first plurality of elements or the second plurality of elements, performing a selection of the selected elements, wherein the selected elements are common to the first version and the second version of the input data schema but are fewer in number than at least one of the first plurality of elements and the second plurality of elements, based on an expected result corresponding to the selected elements;

generating, by an Extract Transform Load (ETL) component, a combined schema from the first version of the input data schema and the second version of the input data schema, wherein the combined schema includes the selected elements and no other elements;

in response to generating the combined schema, processing input data according to the combined schema to generate an output by generating and sending streaming events based on the combined schema from a parser to a state machine to validate incoming data and generate the output, wherein the streaming events correspond to tokens generated by the parser from the input data, and the combined schema corresponds to the state machine; and

during the processing of the input data according to the combined schema, performing;

defining another version of the input data schema, wherein the another version does not affect a structure of elements of interest from the output; and

in response to defining the another version of the input data schema, the ETL component continues to execute and still generates the output in accordance with the combined schema.

2. The method of claim 1, wherein unused data is discarded at the parser.

3. The method of claim 1, wherein the ETL component extracts information from the input data, transforms information in conformance with operational needs, and generates the output.

4. The method of claim 1, wherein the generating of the combined schema further comprises:

configuring the combined schema to have all elements corresponding to a schema of the expected result, wherein all the elements corresponding to the schema of the expected result are a subset of an union of each of the first version and the second version of the input data schema.

5. The method of claim 1, wherein the selected elements include a first element that is of interest to the user, wherein the first element has an identical name in each of the first version and the second version of the input data schema, wherein the user desires an output based on the first element, wherein at least one of the first version and the second version of the input data schema include a second element, and wherein the user does not want an output based on the second element.

6. The method of claim 5, wherein the second element is omitted from the combined schema.

7. The method of claim 1, wherein the parser and the state machine disregards elements that are not needed in the output, and wherein the user maintains only one ETL component.