Classes play a central role in Main//Pyrus. The documents became different information and different retention periods through the classes. Classes consist of several columns. Columns define the features of the attributes for the documents of the class. The features of these attributes could be format templates, e.g. for a limited number of ciphers or letters, for a date or for a value from a lookup table. Thereby it is guaranteed that every attribute of the document have a valid format. If for example a company has a 5-digit number for their customers, the system could guarantee that just 5-digit numbers could be entered in the field for this attribute. So the stored data will be constant.

While e.g. the class “taxes” needs information like account number or depositor, the class “ordering” needs different information like the number of the customer. If just one class existed, there would be available the same field information for every document, the important ones and these who are not needed, or there would exist just general information which would disallow a specified search.

Because of these differentiation of the classes, you can limit the needed field information for the class. This assures storage place make sure a high abridgement.  But the biggest advantage of the classes is, that is ensures a precise search which finds the values just for its context.

A document can belong to different classes. If a document will be added to a additional class, it will be created a reference. So it could be possible, that document “xyz” is bill and delivery note to the same time. If you searched for “xyz”, you would find two apparently different documents with the same name in different classes. I you used the class search, you would find the document through boot classes. Additional to this, it is possible to specify different retention periods for the classes, where the belonging document would be deleted after the expiration. The native file of these documents will be deleted after every reference of it has been deleted.

To point these fact out, there will be differenced between logical and physical documents. The physical document is basically the file of itself. The logical document contains the several attributes like the number of the customers. That means, that the users just see the logical documents.

For the example stated above it means the following: there would exist one physical document “delivery note/bill xyz” and two logical documents: “delivery note xyz” and “bill xyz”.

Because of the creation of logical documents and the fact, that permissions are linked to them and not to the physical, it is possible, that different people can see different information about a document. So it is for example possible to add a document to a class where general information are shown and to add it to a class with private information fields. The persons who have just the permissions to read the classes with the general information could not see the private ones therefore. So they are two different views of one document.

The class has to be specified for every document by the import, because users can just read the logical documents and every physical document must have at least one logical.