Der Bundesbeauftragte für die Unterlagen des Staatssicherheitsdienstes der ehemaligen Deutschen Demokratischen Republik
Efficient migration with 3pc: Fast and secure migration from Government Site Builder to TYPO3
The Federal Commissioner for the Records of the State Security Service of the former German Democratic Republic (BStU) wanted a new website that would reliably preserve the extensive archive and at the same time simplify access to the documents. The basis was to be a stable, reliable CMS that offers a large and constantly updated range of modern functions in a barrier-free and scalable manner. Parallel to the user-centred redesign of the entire BStU website, 3pc set up a TYPO3 backend into which the old content was transferred in a time-saving manner using a specially programmed migration tool.
Creative approach to the new system
The starting point on the way to the new TYPO3 site was the Government Site Builder, which was in use as a CMS before the relaunch. The obvious problem: it is not possible to transfer the data by means of database export and import, as the website data for such websites is stored on protected servers within the federal IT structure. At the same time, the BStU wanted to transfer a lot of content (text, images, PDF documents and videos) to the new system and gradually revise it there.
Address-specific migration via JSON
What content should appear where on the new page? In a first step, we recorded the data situation within the HTML structure of the previous site. By comparing this with the future page structure, a migration concept was created that formed the basis for programming a new application: After collecting the old content, it was delivered to TYPO3 via JSON - even non-programmers were able to use the corresponding tool easily using the input interface. After storing the old URL with the new TYPO3 page ID, the import started; a log contained the essential import information for all content that was inserted into the future page structure.
And go! - Automation ensures speed
During the transfer, headings were semantically structured and images that were too small were transferred as image galleries - a newly developed page template without header images compensated for the missing images in such places. PDF documents with descriptive texts, images with metadata (captions, copyright information) or video iFrames were also extracted and transferred alongside the texts. And in line with the new information architecture, we also created some content - such as the news - as easily manageable data records.
Almost 1,000 URLs with around 3,600 objects were automatically incorporated into the new site - this would have taken around two months of manual work. The editors only had to finalise the articles in accordance with the current accessibility requirements (BITV) in the TYPO3 backend and were able to publish them directly.
