How Astoria Handles Reuse in Translations

Reuse is handled differently in translations for DITA and non-DITA structured documents.

Non-DITA XML documents encode reuse by means of entity references (sometimes called "classic reuse"). When Astoria exports a non-DITA document containing an entity reference, the referenced element is included in the document. For example, a document with an entity reference to a <div1> element from another document is exported as a single document; the <div1> is exported inline; that is, as part of the referencing document.

DITA XML documents encode reuse without the use of entity references. When Astoria exports a DITA document containing these kinds of content references, each referenced document is exported as its own file. For example, a map containing a <topicref> element pointing to a <topic> element in another file is exported as two files: one file containing the map and the other file containing the topic.

When building translation packages, Astoria matches the way the file-export operation works. When packaging non-DITA XML documents, any reused content is included directly within the referencing documents. In most cases, the packaging process creates one big document to be sent for translation. When this one big document is translated and then imported back into Astoria, Astoria breaks apart the document and stores the translated content in the appropriate subdocuments, thereby matching the source-language document structure. In the example above, the inline <div> element in the translated document is separated and returned to the document that contains it.

When packaging DITA documents, any reused content is retained within its containing document. The individual documents are bundled into a single package sent for translation. When the package is translated and then imported back into Astoria, Astoria stores each translated file in the appropriate location. All content references are updated appropriately.