Re: Deep QCMagritte tutorial?

Hi Esteban, The idea of a QCQuery is to search in a dataset, arbitrarily how it is organised. It does so by recursively visiting the magritte description of the objects, and has protection against loops. For instance: if you have parties in your model, and each party has addresses (telephone, email, post): you can pass your model to your query and it searches through all parties and addresses to find things that match. The query itself is build up in 3 different parts: The select. This can be a simple string (the string value of any field matches: ’06410’, a conditional string (‘name’ = *’die’*), or a complex query that is an and/or of possible selects. The select could easily be extended with extra things, with the only requirement that they implement the “#matches:” (or “matches:description: if more control is needed). The target and search path. That together determine the part of the model that is searched. The results description. The result set itself generates a magritte description. When nothing specified in the results description it returns the magritte description of the common class of the results. You can add an extra filter here, that you only want results of a certain type, or add columns to give additional calculated fields in the results. My query does not sort and/or group and has no translation into SQL. So I have the feeling there is a mismatch here. As my need was to search the objects that are loaded in the image, and allow the user to search through them. So it does have a string representation the is human readable. Regards, Diego