数据库中候选码求法.(比拟全的哦)（国外英文资料）.docVIP

数据库中候选码求法.(比较全的哦)（国外英文资料）First, the concrete step of solving the basic algorithm of candidate code.Step 1. Find the minimum function of R U, F , which depends on the set FStep 2, according to the definition above, calculate the UL respectively, UR, UB (UL said the function only depend on the concentration of each depending on the left of the set of attributes; only in UR said function depend on the concentration of each collection of attribute dependence relation between the right side; the UB = U - UL - remember UR)Step 3, if UL indicates Φ, calculation of UL closures, if UL + = U, UL as the only candidate code, R end algorithm. If the UL + indicates a U, step 4. If the UL = Φ, turn to step 5.Step 4, combine UL with the attributes in UB, and use the definition 4 to determine whether the combination attribute is a candidate; After finding all the candidates, the algorithm ends.Step 5, use the above definition 4 to determine the properties and properties of the UB. After finding all the candidates, the algorithm ends.In short: take minimum rely on sets, calculate the UL closures, if UL closures containing all attributes, it is the only candidate codewords UL, if does not contain, in turn and UB attributes combination again after asking whether closures containing all attributes.When UL is empty, take the UB in order to find the closure.Two, multiple attribute dependent set candidate code solution methodInput: relational schema R and its function dependency set F.Output: all candidate code for R.Specific steps:1) all the properties of R are divided into L, R, N, and LR, and X stands for L, N, and Y for the LR class.X +, if X + contains all the attributes of R, then X is the only candidate code for R, turn (5); Otherwise, turn (3).3) take an attribute of A in Y, (XA) +, if it contains all the attributes of R, turn (4); Otherwise, switch an attribute over and over again until youve tried all the properties in Y.4) if all candidate codes have been found, then (5) Otherw