Browse Prior Art Database

Method of Merging Huge Models with Full-Context Using Sparse Model Sections

IP.com Disclosure Number: IPCOM000199646D
Publication Date: 2010-Sep-13
Document File: 4 page(s) / 96K

Publishing Venue

The IP.com Prior Art Database

Abstract

Disclosed is a process for the merging of logical model and sparse models. The disclosed process enables merging parallel changes to multiple files as a single full-context merge, and writing of all final versions of the artifacts into a repository to maintain synchronization of the logical artifact.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 32% of the total text.

Page 1 of 4

Method of Merging Huge Models with Full -Context Using Sparse Model Sections

Disclosed is a process for the merging of logical model and sparse models. The disclosed process enables merging parallel changes to multiple files as a single full-context merge, and writing of all final versions of the artifacts into a repository to maintain synchronization of the logical artifact.

A logical model (or document) can be defined as a complete group of files outside of which there is no further referenced information required to complete information being captured. A logical model can grow in size, outstripping an ability of a client computer to hold the entire model in memory. Further, a fully-featured compare and merge subsystem typically loads four models comprising a local version of the model being submitted back to a software configuration management system (SCM), a remote version containing all changes merged so far descending from a common ancestor version already in the SCM system, the common ancestor version from which all parallel changes descend, and a merged model enabling previewing of changes. These four models typically require large amounts of memory as the models grow and a method capable of managing the growth.

Figure 1 depicts physical structure of a logical model using the terms of logical and physical resource to mean the logical model and constituent files.

Figure 1

The constituent files must remain consistent over time. When a logical model is modified, all changed files need to be processed together to ensure stored files are never out of synchronization. Typical tools consider desynchronized files a form of data corruption and fail to open such a model.

For example, Eclipse (trademark of Eclipse Foundation, Inc.) open-source tooling contains a mechanism enabling grouping such files using a "model provider", which informs the tooling that a specific group of files must be treated as a logical unit. Eclipse further provides a capability of "full-context" compare and merge to compare logical artifacts as a group as an alternative to comparing or merging individual fragments in a low-context fashion. In another example, an implementation of IBM® Rational® Software Architect (IBM and Rational are registered trademarks of IBM in the United States) (RSA) logical models relies on a root file (model or package) and a hierarchy of files, called fragments, comprising the package structure with additional classifiers and diagrams.

The example RSA implementation displays each fragment as the fragment is loaded, maintaining the smallest possible footprint while editing the model. Figure 2 depicts an example in which the user has opened a root model file, opened diagram 3 of package 4, which further references class 4 in package 3. The memory footprint is a small subset of the overall model.

1

[This page contains 1 picture or other non-text object]

Page 2 of 4

Figure 2

When the model contains 5000 fragments and the user perform...