SQL Reloaded: Pulling the Strings of SQL Server VII (List of Values)

Introduction

    Lists are one of the basic structures in Mathematics, the term referring to an (ordered) set of elements separated by comma, space or any other delimiter (e.g. “:”, “;”). The elements of a list can be numbers, words, functions, or any other type of objects. In the world of databases, a list is typically formed out of the values of a given column or a given record, however it could span also a combination of rows and records, is such cases two delimiters being needed – one for column and one for row. From here comes probably the denomination of list of values. In a more general accept a list of values could be regarded as a delimited/concatenated subset. Such lists are formed when needed to send the data between the layers of an application or applications, this type of encoding being quite natural. In fact, also the data in a database are stored in similar tabular delimited structure, more complex though.  

    An useful example in which the list of values are quite handy is the passing of multiple values within the parameter of stored procedure or function (see example). This supposes first building the list and then use the values in a dynamic build query (like in the before mentioned example) or by building a table on the fly. We can call the two operations composition, respectively decomposition of list of values.

Composition 

Composition, whether on vertical or horizontal is nothing but a concatenation in which the values alternate with one or more delimiters. Let’s reconsider the concatenation based on the values of a Person.AddressType AdventureWorks table. As the logic for concatenating for one or more attributes is the same, the below example concatenates a list based on a single attribute, namely AddressTypeID in SingleList, respectively two attributes, AddressTypeID and Name.

-- concatenation of values across a table 
;WITH CTE (AddressTypeID, Name, Ranking) 
AS (--preparing the data       
     SELECT AddressTypeID  
     , Name 
     , ROW_NUMBER () OVER(ORDER BY Name) Ranking 
     FROM Person.AddressType 
     -- WHERE ... 
) 
, DAT (SingleList, DoubleList, Ranking) 
AS ( -- concatenating the values 
     SELECT Cast(AddressTypeID as varchar(max)) SingleList 
     , Cast('('+ Cast(AddressTypeID as varchar(10)) + ',''' + Name + ''')' as varchar(max)) DoubleList 
     , Ranking 
     FROM CTE 
     WHERE Ranking = 1 
     UNION ALL 
     SELECT DAT.SingleList + ',' + Cast(CTE.AddressTypeID as varchar(20)) SingleList 
    , Cast(DAT.DoubleList + ', ('+ Cast(CTE.AddressTypeID as varchar(10)) + ',''' + CTE.Name + ''')' as varchar(max)) DoubleList 
    , CTE.Ranking  
     FROM CTE          
       JOIN DAT           
          ON CTE.Ranking = DAT.Ranking + 1       
)  

-- the lists 
SELECT SingleList 
, DoubleList 
FROM DAT 
WHERE Ranking = (SELECT MAX(Ranking) FROM DAT) 

 
 
   The second example is based on atypical delimiters, resembling to the structure built for a batch insert or table value constructor-based statement, and as we’ll see later, ideal to be used in a dynamically-built query

Decomposition

Decomposition follows the inverse path, though it’s much easier to exemplify. In fact it’s used the same technique introduced in the last example from the previous post belonging to the same cycle, Subparts of a String, in which a space was used as delimiter. Another example is the dbo.SplitList function which decomposes a string using a loop.

-- decomposition of a string to a table using CTE 
CREATE FUNCTION dbo.StringToTable( 
 @str varchar(500) 
,@Delimiter char(1)) 
RETURNS @Temp TABLE ( 
Id int NOT NULL 
,Value varchar(50)) 
AS 
BEGIN  
     ;WITH CTE (PrevString, Position, Word)  
     AS (  
     SELECT LTrim(RTrim( CASE  
           WHEN CharIndex(@Delimiter, @str)>;0 THEN Right(@str, Len(@str)-CharIndex(@Delimiter, @str))  
           ELSE ''  
      END)) PrevString  
     , 1 Position  
     , LTrim(RTrim(CASE  
           WHEN CharIndex(@Delimiter, @str)>0 THEN LEFT(@str, CharIndex(@Delimiter, @str)-1)  
           ELSE @str  
       END)) Word  
      UNION ALL  
      SELECT LTrim(RTrim(CASE  
            WHEN CharIndex(@Delimiter, PrevString)>0 THEN Right(PrevString, Len(PrevString)-CharIndex(@Delimiter, PrevString))  
             ELSE ''  
       END)) PrevString  
      , Position + 1 Position  
      , LTrim(RTrim(CASE  
           WHEN CharIndex(@Delimiter, PrevString)>0 THEN LEFT(PrevString, CharIndex(@Delimiter, PrevString)-1)  
          ELSE PrevString  
      END)) Word      FROM CTE  
     WHERE Len(PrevString)>0  
    )  
     INSERT @Temp(Id, Value) 
     SELECT Position  
     , Word      FROM CTE  
     OPTION (maxrecursion 100)  
     RETURN 
END    

Here are two examples based on the single list created above and another one based on alphabet:

-- decomposing a list
SELECT Id 
, value 
FROM dbo.StringToTable('6,1,2,3,4,5', ',')     


-- decomposing the "alphabet" 
SELECT Id 
, value 
FROM dbo.StringToTable('a,b,c,d,e,f,g,h,i,j,k,l,m,n,o,p,q,r,s,t,u,v,w,x,y,z', ',') 

    


   
Even if the function deals only with a delimiter, it could be used to decompose lists involving multiple delimiters, as long the list is adequately built:

-- decomposing double list 
SELECT Id 
, value 
, Left(value, CHARINDEX(',', value)-1) LeftValue 
, Right(value, len(value)-CHARINDEX(',', value)) RightValue 
FROM dbo.StringToTable('6,Archive;1,Billing;2,Home;3,Main Office;4,Primary;5,Shipping', ';')     

 
 The tables built thus from list of values can be further used in queries when needed to create a table on the fly. It would be interesting maybe to show that the composition and decomposition are inverse functions, however that’s out of scope, at least for current set of posts. 

SQL Reloaded: Pulling the Strings of SQL Server VIII (Insertions, Deletions and Replacements)

Until now, the operations with strings resumed to concatenation and its reverse operation(s) - extracting a substring or splitting a string into substrings. It was just the warm up! There are several other important operations that involve the internal manipulation of strings – insertion, deletion and replacement of a substring in a given string, operations performed using the Replace and Stuff functions.

Replace function, as its name denotes, replaces all occurrences of a specified string value with another. Several scenarios in which the function is quite useful: the replacement of delimiters, special characters, correcting misspelled words or any other chunks of text. Here are some basic simple examples, following to consider the before mentioned applications in other posts:

-- examples with replace 
DECLARE @str varchar(30) 
SET @str = 'this is a test string' 
SELECT replace(@str, ' ', ',') Example1 
, replace(@str, ' ', ' ') Example2 
, replace(@str, ' ', '') Example3 
, replace(@str, 'is', 'as') Example4  

Output:

Example1	Example2	Example3	Example4
this,is,a,test,string	this is a test string	thisisateststring	thas as a test string

When there are good chances that the searched string won’t appear in the “searched” string, and especially when additional logic is depending on the replacement, logic that could be included in the same expression with the replacement, then maybe it makes sense to check first if the searched character is present:

-- replacement with check 
DECLARE @str varchar(30) 
DECLARE @search varchar(30) 
DECLARE @replacememt varchar(30) 
SET @str = 'this is a test string' 
SET @search = 'this string' 
SET @replacememt = 'other string' 
SELECT CASE            
    WHEN CharIndex(@search, @str)>0 THEN Replace(@str, @search, @replacememt)             
    ELSE @str        
END result 

Unfortunately the function doesn’t have the flexibility of the homonym functions provided by the languages from the family of VB (VBScript, VB.NET), which allow to do the replacement starting with a given position, and/or for a given number of occurrences. This type of behavior could be obtained with a simple trick – splitting the string into two other strings, performing the replacement on the second string, and then concatenating the first string and the result of the replacement:

-- replacement starting with a given position 
DECLARE @str varchar(30) 
DECLARE @search varchar(30) 
DECLARE @replacememt varchar(30) 
DECLARE @start int 
SET @str = 'this is a test string' 
SET @search = 's' 
SET @replacememt = 'x' 
SET @start = 7 
SELECT Left(@str, @start-1) FirstPart 
, RIGHT(@str, Len(@str)-@start+1) SecondPart 
, CASE        
    WHEN @start <= LEN(@str) THEN Left(@str, @start-1) + Replace(RIGHT(@str, Len(@str)-@start+1), @search, @replacememt)        
     ELSE @str  
END Replacement 

Output:

FirstPart	SecondPart	Replacement
this i	s a test string	this ix a text xtring

The logic can be encapsulated in a function together with additional validation logic.

Stuff function inserts a string into another string starting with a given position and deleting a specified number of characters. Even if seldom used, the function it’s quite powerful allowing to insert a string in another, to remove a part of a string or more general, to replace a single occurrence of a string with another string, as can be seen from the below examples:
 

-- Stuff-based examples DECLARE @str varchar(30) 
SET @str = 'this is a test string' 
SELECT STUFF(@str, 6, 2, 'was ') Example1 
, STUFF(@str, 1, 0, 'and ') Example2 
, STUFF(@str, 1, 0, 'that') Example3 
, STUFF(@str, LEN(@str) + 1, 0, '!') Example4 

Output:

Example1	Example2	Example3	Example4
this was a test string	and this is a test string	thatthis is a test string	NULL

If in the first example is done a replacement of a text from a fix position, in the next examples are attempted insert on a first, middle respectively end position. As can be seen, the last example doesn’t work as expected, this because the insert position can’t go over the length of the target string. Actually, if the insert needs to be done at the beginning, respectively the end of a string, a concatenation can be much easier to use. A such example is the padding of strings with leading or trailing characters, typically in order to arrive to a given length. SQL Server doesn’t provide such a function, however the function is quite easy to build.
 

-- left/right padding DECLARE @str varchar(30) 
DECLARE @length int  
DECLARE @padchar varchar(1) 
SET @str = '12345'  
SET @length = 10 
SET @padchar = '0' 
SELECT @str StringToPad  
, CASE  
     WHEN LEN(@str)<@length THEN Replicate(@padchar, @length-LEN(@str)) + @str       
     ELSE @str  
END LeftPadding  
, CASE  
     WHEN LEN(@str)<@length THEN @str + Replicate(@padchar, @length-LEN(@str))      
     ELSE @str  
END RightPadding 

 
Output:

StringToPad	LeftPadding	RightPadding
12345	0000012345	1234500000

Happy Coding!

SQL Reloaded: Pulling the Strings of SQL Server III: (Concatenation)

    Typically a database in general, and a table in particular, that follows the normalization rules, is designed to have the columns contain the smallest semantic chunks of data, it could be a Street, a Zip Code, City, a person’s First or Last Name, but also a large chunk of text like a Description or a Comment. No matter how well designed is a database, there will always be the need to do various operations with strings, typically concatenation, extraction of subpart of a string, insertion or deletion of characters, rearangement of string’s characters, trimming, splitting it in substrings, or of getting various numeric values: length, position of a given text, number of not empty characters, on whether the text represents a valid numeric or date values, etc. In the following posts I will attempt to address the respective operations in the context of select statements, and let’s start with concatenation.

    Concatenation is the operation of joining two or more string values within an expression. In SQL Server the “+” operator is used for concatenation, and it could be used to concatenate two or more members. In order to concatenate two members from which one of them is a string, the second term needs to be explicitly converted to a string data type, otherwise an error will occur. For readability or post-processing purposes, the strings are concatenated using a delimiter in order to delimit the boundaries of the initial value, it could be used a space, a comma, a semicolon, a pipe, a tab or any other character that could be used typically in delimiting columns.  

-- concatenation of strings 
SELECT 'a string' + ' and ' + 'a second string' Example1  
, 'a string' + ',' + 'a second string' Example2  
, '1245' + '67890' Example3  
, '1245' + '.' + '67890' Example4 

   The concatenation of string variables or columns functions based on the same principles: 

-- concatenating string variables 
DECLARE @string1 varchar(50) 
DECLARE @string2 varchar(50) 
DECLARE @string3 varchar(50) 

SET @string1 = 'this is a string'  
SET @string2 = 'this is another string'  

SELECT @string1 + ' and ' + @string2 Example1 
, @string1 + char(31) + @string2 Example2 
, @string1 + ', ' + @string2 Example3 
, @string1 + ' ' + @string3 Example4 
, @string1 + IsNull(@string3, '!') Example5 

 
 

    Here’s another example based on the concatenation of columns coming from two joined tables from AdventureWorks database:
 

-- concatenating columns of joined tables 
SELECT PAD.AddressID 
, IsNull(PAD.AddressLine1, '')  
+ IsNull(', ' + PAD.AddressLine2, '') 
+ IsNull(', ' + PAD.City, '') 
+ IsNull(', ' + PAD.PostalCode, '') 
+ IsNull(', ' + PSP.Name, '') Address 
FROM Person.Address PAD 
      JOIN Person.StateProvince PSP 
         ON PAD.StateProvinceID = PSP.StateProvinceID 

 


   As stressed in the previous post, the NULL values need to be adequately handled either by initializing values or by using the IsNull or COALESCE functions. The concatenation of strings combined with IsNull function could be used creatively in order to add a comma only when a value is not null, as in the above example.

   There are scenarios in which is needed to concatenate the values belonging to the same column but from different records, for example concatenating the email values in order to send a single email to all the users in one single action. Before the introduction of common table expressions (CTE), wasn’t possible to concatenate the string values belonging to different records, at least not in a query, this functionality being achieved by using cursors or loops, or simply performed on client or intermediate layers. As I already gave an example on how to use cursor in order to loop through the values of a table and concatenate them (see “Cursors and Lists” post), I will focus on the use of loops and simple CTEs.

    Loops are one of the basic functionality in programming languages, no matter of their complexity or type. Either if are talking about WHILE, FOR, foreach or do … until loops, the principles are the same: perform a list of actions until one or more conditions are met. In this case the actions performed is reduced to a set of concatenations based on the letters of the (English) alphabet:

-- concatenation within a loop 
DECLARE @list varchar(max) 
DECLARE @index int  
SET @list = '' 
SET @index = ASCII('a') WHILE (@index<ASCII('z')) 
BEGIN 
     SET @list = @list + ', ' + Char(@index) 
     SET @index = @index + 1 
END 
SELECT @list Result 

    There is more work that needs to be performed in order to remove the leading comma from the output, but that’s a topic for the next post, when discussing about decomposition of strings.

    CTEs are a little more complex to use than the loops, though the concatenation could be achieved across records and this in one query and not in procedural language as in the above example. In order delimit the two components of a CTE, I made use of a second CTE which simulates the existence of a given table:
 

-- concatenation across records 
;WITH Data(Column1, Ranking)  
AS  
( -- preparing test data 
       SELECT 'A' Column1, 0 Ranking  
       UNION ALL  
       SELECT 'B' Column1, 1 Ranking  
       UNION ALL  
       SELECT 'C' Column1, 2 Ranking 
) 
, Result(Column1, Ranking)  
AS 
(  -- performing the actual concatenation 
      SELECT Cast(Column1 as nvarchar(max)) Column1 , Ranking 

      FROM Data 

      WHERE Ranking = 0  
      UNION ALL 
      SELECT Cast(B.Column1 + ',' + A.Column1 as nvarchar(max)) Column1 , A.Ranking  
      FROM Data A  
         JOIN Result  B  
             ON A.Ranking - 1 = B.Ranking  
) SELECT Column1  
FROM Result  
WHERE Ranking IN (SELECT MAX(Ranking) FROM Result) 

  
    The logic for doing a simple concatenation seems maybe complicated, though the volume of work is not so big if we ignore the first CTE. On the other side I introduced an “index” within the Ranking column, fact that allows processing easier the records. When dealing with the records coming from a table it’s probably much easier to use one of the ranking functions that suits best.

DBMS: Dynamic SQL (Definitions)

"Dynamic SQL is a method of binding between an application program, in one of a number of standard programming languages, and an SQL-implementation." (ISO/IEC 9075-1, 1999)

"In Embedded SQL for C, a SQL statement built and executed at run time." (Microsoft Corporation, "SQL Server 7.0 System Administration Training Kit", 1999)

"SQL that is interpreted during execution of the statement." (Ajay Gupta et al, "Informix Dynamic Server 11", 2007)

"A SQL statement built and executed at runtime by concatenating different parts of the statement based on provided variables." (Darril Gibson, "MCITP SQL Server 2005 Database Developer All-in-One Exam Guide", 2008)

"[...] dynamic SQL is a means of constructing SQL statements as strings in your server-side (or even client-side) applications and executing them dynamically, on the fly. When used properly, dynamic SQL can be used to generate complex queries at runtime, in some cases improve performance, and do tasks that just aren't possible (or are extremely difficult) in standard, nondynamic T-SQL." (Michael Coles, "Pro T-SQL 2008 Programmer's Guide", 2008)

"SQL statements that are constructed at runtime; for example, the application may allow users to enter their own queries. These types of SQL statements are not hard-coded into the application. See also static SQL." (John Goodson & Robert A Steward, "The Data Access Handbook", 2009)

[Dynamic embedded SQL:] "Embedded SQL in which the entire SQL statement cannot be assembled prior to running the program. The SQL statement is therefore completed and processed during the program run." (Jan L Harrington, "SQL Clearly Explained” 3rd Ed., 2010)

"A term used to describe an environment in which the SQL statement is not known in advance, but instead is generated at run time. In a dynamic SQL environment, a program can generate the SQL statements at run time that are required to respond to ad hoc queries." (Carlos Coronel et al, "Database Systems: Design, Implementation, and Management” 9th Ed., 2011)

"Dynamically constructed SQL queries that are not pre-processed, and whose access paths are determined at run time prior to execution." (DAMA International, "The DAMA Dictionary of Data Management", 2011)

"SQL statements that are prepared and executed within an application program while the program is executing. In dynamic SQL, the SQL source is contained in host language variables rather than being coded into the application program. The SQL statement can change several times during the application program's execution." (Kirsten A Larsen et al, "Extremely pureXML in DB2 10 for z/OS", 2011)

"SQL queries that are not preprocessed and whose access paths are determined at run time right before execution." (Craig S Mullins, "Database Administration", 2012)

"Allows an Embedded SQL or Client-Library application to execute SQL statements containing variables whose values are determined at runtime." (Sybase, "Open Server Server-Library/C Reference Manual", 2019)

[dynamic SQL statements:] "An SQL statement that is prepared and executed at run time. In dynamic SQL, the SQL statement is contained as a character string in a host variable and is not precompiled." (Sybase, "Open Server Server-Library/C Reference Manual", 2019)

"A feature that lets you create and execute prepared statements using more robust, secure, and efficient methods to substitute parameter values than the naive technique of concatenating the parts of the statement into a string variable." (MySQL)

"A program in which user input determines how the program interacts with the database. Put another way, the SQL statements are not part of the program when it is launched; instead they are built as the program runs." (Microfocus)

"A type of embedded SQL in which SQL statements are created and compiled at run time." (Microsoft) 

"Dynamic SQL is a programming technique that enables you to build SQL statements dynamically at runtime. You can create more general purpose, flexible applications by using dynamic SQL because the full text of a SQL statement may be unknown at compilation." (Oracle) [source]

"Dynamic SQL is an application that uses dynamic code to increase flexibility in querying a database. Dynamic code is generated at execution of a SQL application by inserting statements into variables." (Idera) [source]