In many applications, it makes sense from a workflow and production point of view to create the content of repository items outside of ATG, on a file system, and subsequently load the items into the repository. The ATG Repository Loader provides a flexible way to take files that are stored in a file system, convert them into repository items, and load the items into the repository.
In the ATG platform, the Repository Loader has been implemented to handle the loading of HTML files, XML files, and binary media items (like image or audio files) into a SQL repository. The Repository Loader uses ATG’s xml2repository system to transform files into XML files that use the xml2repository schema, which are then transformed into repository items. The xml2repository system is described in full in the Repository to XML Data Binding chapter of the Web Services and Integration Framework Guide. Make sure you are familiar with that chapter, as it is essential to understanding how to use the Repository Loader.
The Repository Loader has been designed to be flexible. When you configure the Repository Loader, you can choose to let the Repository Loader scan the file system, identify the files that need to be loaded, and load the files, either on command or on a specified schedule. Alternatively, you can submit to the Repository Loader a manifest file that specifies the files to be loaded. The first alternative requires less effort on the part of the Web site administrators; the second alternative loads faster and puts less of a burden on your system. Which to choose depends in part on the size of your repository. The larger the number of files that need to be scanned, the longer it will take to load them, and the greater the burden that file system scanning will place on your system.
This chapter includes the following sections:
