The code is the basic concept in the data system. The so-called code is to uniquely identify the properties of the entity, which is the nature of the entire entity set, not the nature of a single entity. It includes supercodes, candidate codes, master code. Optical is a collection of one or more attributes that allow us to uniquely identify an entity in a entity. If k is a super code, then k 's superchard is also supercoding, that is, if k is a super code, all the collection of K is also supercoded. Candidate code is selected from the ultra code, natural candidate code is also a collection of one or more attributes. Because the range of supercodes is too wide, many are we are not interested in useless. So the candidate code is the smallest super code, and all of their true subsets cannot be supercoded. For example, if k is a supercode, all the collections containing K cannot be a candidate code; if k, j is not overca, then the set of K and J (K, J) may be a candidate code. It is arbitrarily selected from multiple candidate codes, if the candidate code is only one, then the candidate code is the master code. Although the choice of the principal code is relatively casual, it is still necessary to rely on certain experience in actual development, and there will be many problems in the system. Generally speaking, the main code should choose this from the attributes from not or very varied.
