normalization - Should a SQL VIEW always be in 1NF? -


A SQL view is a global, logical table that can be continuous or not. But it is still a table, therefore, should a sequence always follow the usual form (1 NF)? I.e. no duplicate rows, scalar type only, no top-to-bottom or left to right sequence, etc. What about high normal forms?

For me, the results of my application 'Consumption' stored procs, my thoughts are 'consumed' by SQL queries, and these two experiments are mutually exclusive (ie I stored using SQL procs Resultsets do not query and my applications do not include SQL code). I have noticed that others use a view for 'adding' multiple values ​​in one column in a row, usually a comma-separated format. A SQL query against such a column says that such a clause is required:

  ',' concat_col + ',' likes '%' + ',' + Search_value + ',' + '%'  

So I think the expectation of all tables is justified, which are only scaly types. Am I even 'pure' by thinking?

It is absolutely perfect to make sure that your thoughts are normal to at least 1 NF example To allow duplicates, it is a loss that the meaning of the scene is ambiguous and information may be misidentified by users. Incorrect data may arise if tables are updated based on such ambiguities.

E.F. The codes were not necessarily agreed. In his RM version 2 book, he offered to allow visualization of no-keys - I think that's a big mistake, the idea of ​​coded does not really allow duplicates, but they allow every column to be empty and So they have no keys and are not in 1 NF.

String value with a comma-delimited list is not itself 1 violation of NF A string value is a scalar like any other value, which is also in it. Most SQL DBMS does not allow multi-valuable features.


Comments

Popular posts from this blog

c++ - Linux and clipboard -

What is expire header and how to achive them in ASP.NET and PHP? -

sql server - How can I determine which of my SQL 2005 statistics are unused? -