About Atril Solutions
Atril Solutions has a unique reputation for innovating translation technology from the standpoint of the user, not the technologist. We have never considered new functions in isolation or introduced them solely for the sake of the technology, but we have always worked towards a bigger picture.
Right from our very beginnings in 1993, we have been committed to much more than simply automating the translation process to enhance productivity, aiming above all to provide a tightly-integrated tool that helps everyone in the translation supply chain work. The ultimate goal has been to make it as effective and efficient, as possible. This all shouldn’t be surprising, as these are our roots.
During our system design, we never involved any trade-offs between quality and productivity. Just as they have always been conceived, so they can be easily and flexibly customized to individual needs.
This thinking has always enabled us to constant technological innovation. Just as it has involved us in close dialogue with our users, this approach itself has been innovative while ensuring our product development makes a real impact on quality, productivity, and reputation.
Over the years, we have added a large number of features. As we continued research into the techniques that drive Déjà Vu, we have steadily added new translation features and refined existing techniques to increase both speed and output. And we have also added a large number of filters, both for documents and for database importing and exporting.
Problem
Our translation tool, Déjà Vu, provides an interface that allows users to translate documents in a wide variety of formats. It includes a number of filters that extract the translatable content from the documents, and store any inline markup or formatting information in placeholders. The user, therefore, does not see the real formatting of the documents; while in some cases, this can be an advantage. Because having to rearrange numerous placeholders in the translated text is a time-consuming task.
The increase in complexity of the Microsoft Office applications over the years means that Office documents, and Word documents, in particular, are littered with markup that is not visible to the user when working in Office. But often it results in a large number of unnecessary placeholders when those documents are imported into Déjà Vu.
Until now, processing the older pre-Office 2007 formats required the users to have Office installed and to partly or completely process the files through Office automation. Having to rely on automation also limits the stability of the filters, since they cannot be safely used in server-side applications. Having separate filters for the Open Office XML formats creates unnecessary complexity for the end-user, who is not interested in knowing whether a DOC file has to be converted into RTF using Word automation before being imported, or a DOCX file can be processed by extracting the XML content inside it.
When started to plan the new version of our product, where we decided to change all of this. We required components that would allow us to process different types of Office documents consistently without relying on OLE automation. It shall allow us to easily extract formatting information so that we could display this to the user in a pseudo-WYSIWYG interface, and easily convert document content to HTML format. This could provide users with a more realistic real-time preview of the formatted translation.
For all our users – over 10,000 across the world – this would be an incredibly welcoming improvement, as it will improve the ergonomics and usability of our product. It will also reduce the amount of additional formatting they need to do outside our tool, thus considerably increasing their productivity.
Solution
We chose Aspose because it provided nearly all of the functionality of the Office Open XML SDK, and it also provided consistent solutions across different file formats (for example RTF, DOC, and DOCX) and allowed us to easily extract and manipulate basic formatting information.
We completely re-wrote our existing Office document filters using Aspose.Total for .NET, which resulted in incredibly clean code, compared to our previous codebase which had extensive classes to convert, clean and parse RTF, or to extract and parse XML files from Open XML files. In one of the Aspose.Total for .NET Case Study regarding Excel versus the way Aspose.Cells for .NET handles the document processing, the difference has been obvious where thousands of lines of codes in our old codebase were reduced to a few hundred lines.
With our new solution based on Aspose.Words for .NET, Aspose.Cells for .NET and Aspose.Slides for .NET, users simply instruct our translation tool to import a document, and the user is presented with all of the translatable content, in a visual representation of basic formatting attributes such as bold, italics, superscript, etc. In addition, users have an HTML-based real-time preview of the translated content, based on HTML generated from the source documents using Aspose components. Once the user has finished translating the document, our tool modifies the original document, by replacing the original text with the translation. Furthermore, while using Aspose components we can modify any formatting information that is required which includes changing language identifiers, text directionality, etc.
Experience
Finding a solution
Apart from the Office Open XML SDK, we did not find any other tool that provided the functionality that we required. We weren’t interested in simply generating Office documents, but needed to open existing documents and extract information from them. We saw recommendations for Aspose on various web sites, so we downloaded a trial version and began testing it and before the trial had ended, we had a fully working version of the two of three new filters. We did not require to contact support because the examples and the documentation were extremely clear, and the only issues we came across was already addressed in the Aspose forums.
Implementation
It only took a single programmer 3 weeks to implement the first phase of the solution, which included complete import and export including inline formatting. The object model exposed by the various Aspose components made the newly written code very clean, concise and made it easy to maintain as well as extend.
Outcome
We are about to begin external beta testing of the new functionality, so at the moment, we don’t have feedback from our users.
Next Steps
The object model exposed by the Aspose components means we need to spend less time worrying about the internal structure of the document formats and can easily access the document content in a variety of ways. This enabled us to easily extend the filters to add new functionality in future releases. For our future release versions, we have also considered using Aspose.Tasks for.NET and Aspose.Diagram for .NET to process Project and Visio files, respectively.
Summary
We were very pleased that, as we began working with Aspose products, and they lived up to our expectations. They made the design and coding of the new document filters simpler, cleaner and quicker, helping to greatly reduce complexity (and hence they increased the quality) in one of the more complex parts of our codebase. We would readily recommend Aspose components to anybody who needs to process or generate Office documents, either in a desktop tool or in a server application.
Blandine Loze, Managing Director,
Daniel Benito, CTO,
Atril Solutions
