SQL Server 2012 :: Getting String Out Of Data From A Column?
May 7, 2015
I have system id information in table system_ids and productids and systemidinsformation has lot of data but I am looking two strings in tire data to pull into two separate columns. details below
Database versions :ms sql 2008/2012
tablename:system_id's
column:system id information
sample data from system_id_information column
########################################
<obj xmlns:xsd="http://www.w3.org/2001/XMLSchema" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns="urn:vim25" versionId="5.5" xsi:type="ArrayOfHostSystemIdentificationInfo"><HostSystemIdentificationInfo xsi:type="HostSystemIdentificationInfo"><identifierValue> unknown</identifierValue><identifierType><label>Asset Tag</label><summary>Asset tag of the system</summary><key>AssetTag</key></identifierType>
[Code] .....
I am looking output of two columns, which are bolded
product_id snumber
654081-B21 MXQ43905SW
for serial number this is common
before string :HostSystemIdentificationInfo"><identifierValue>
and after string </identifierValue><identifierType><label>Service tag
and snumber is always between the before and after string and number of characters of snumber varies and entire data for a row also varies
I am trying to find a solution to get the result set to fetch a particular string format from a table in my database, which has a column of NVARCHAR data type
INSERT INTO #ActivityComments VALUES('This is the study code for Field Phase S14-04932-01') INSERT INTO #ActivityComments VALUES('Phase reporting has the study S15-04931-01 which is obselete') INSERT INTO #ActivityComments VALUES('Phase running study code S14-04932-02 is not valid')
All the column names in upper case are actually symptom names, and in those columns are values {NULL, 1, 2, 3, 4, 5} and they belong in a column, so the normalized structure should be like this:
CREATE TABLE Symptom ( PatientID INT NOT NULL, Cycle TINYINT NOT NULL, SymptomName VARCHAR(20) NOT NULL, -- from the source column *name* Grade TINYINT NOT NULL -- from the value in the column with the name in uppercase PRIMARY KEY (PatientID, Cycle, SymptomName));
I can untwist the repeating groups with the code I borrowed from Kenneth Fisher's article [ here ], but the part I'm having a harder time with is grabbing the information that's still left in the column name and integrating it into the solution...
I can retrieve all the column names that are in uppercase using this:
DECLARE @db_id int; DECLARE @object_id int; SET @db_id = DB_ID(N'SCRIDB'); SET @object_id = OBJECT_ID(N'SCRIDB.dbo.BadTox'); SELECT name AS column_name , column_id AS col_order FROM sys.all_columns WHERE name = UPPER(name) COLLATE SQL_Latin1_General_CP1_CS_AS AND object_id = @object_id;
but I can't figure out how to work it into this (that I built by mimicking Kenneth Fisher's article...):
ALTER PROC [dbo].[UnpivotMaxGradeUsingCrossApply] AS SELECT PatientID , Toxicity , MAX(Grade) AS MaxGrade
[code]....
The problem is that I need to extract the column names (where ToxicityName[n] would be). I can do that by querying the sys.all_columns view, but I can't figure out how to integrate the two pieces. About the only thing I have even dreamed up is to build the VALUES(...) statements dynamically from the values returned by the system view.
So how do I get both the value from the ToxicityName[n] column and the column name into my final data query?
I have created some dynamic sql to check a temporary table that is created on the fly for any columns that do contain data. If they do the column name is added to a dynamic sql, if not they are excluded. This looks like:
If (select sum(Case when [Sat] is null then 0 else 1 end) from #TABLE) >= 1 begin set @OIL_BULK = @OIL_BULK + '[Sat]' +',' END
However, I am currently running this on over 230 columns and large tables 1.3 mil rows and it is quite slow. How I can dynamically create a sql script that only selects the columns in the table where there is data in a speedier manner. Unfortunately it has to be on the fly because the temporary table is created on the fly.
We are storing changed data of tables into XML format for auditing purpose. The functionality is already achieved. We are using FOR XML Path clause to convert relational data of tables into XML format.
Now, a table is having column name with '(' . For example name of the column is, ColumnName(). In this case we can not convert into XML using For XML clause. Showing error as,
Column name 'columnName()' contains an invalid XML identifier as required by FOR XML; '(' (0x0028) is the first character at fault.
Need to change the datatype of existing column which has huge data.
I'm performing below steps
1. Create new column with correct datatype in the same table 2. copy data into new column 3. drop indexes on column 4. <<<>>> now the existing column also has many SP dependent and I do not wish to drop them. 5. rename existing column to xxx 6. rename new column to correct column 7. drop old column 8. make required indexes
Currently I have a column with multiple postcodes in one value which are split with the “/” character along with the corresponding location data. What I need to do is split these postcode values into separate rows while keeping their corresponding location data.
We have some URLs within a bulk block of text some of which are very long. I need to identify rows where such urls exceed say 100 characters in length in amongst other text.So the rule would be return a record if within the string there is a string (without spaces) longer than 100 characters.
I'm moving data from one database to another (INSERT INTO ... SELECT ... FROM ....) and am encountering this error:
Msg 8114, Level 16, State 5, Line 6 Error converting data type varchar to numeric.
My problem is that Line 6 is:
set @brn_pk = '0D4BDE66347C440F'
so that is obviously not the problem and my query has almost 200 columns. I can go through one by one and compare what column is int in my destination table and what is varchar in my source tables, but that could take quite a while. How I can work out what column is causing the problem?
I am studying indexes and keys. I have a table that has a fixed width of data to be loaded in the first column which is parsed in a view based on data types within the fixed width specifications.
Example column A: (name phone house cost of house,zipcodecountystatecountry) -a view will later split this large varchar string based column b: is the source filename of the data load (varchar 256) ....
a. would there be a benefit of adding a clustered or nonclustered index (if so which/point in direction on why)
b. is there benefit of making one of these two columns a primary key (millions of records) or for adding a 3rd new column as a pk?
c. view: this parses the data in column a so it ends up looking more like "name phone house cost of house zipcode county state country" each having their own column.
-any pros/cons of adding indexes (if so which) to the view instead of the tables or both for once the data is parsed?
I have a large poorly designed table (inherited) With a Name field that contains comma delimited text containing address information. I need to do several things with it but unfortunately there doesn't appear to be any true consistency in it. When it displays in its own text box it works by placing each section on a new Line and looks ok.But I need to pull it apart and place things like unit number, Building Name in its own column etc. In the data it could be in either the 2nd,3rd, 4th, dependent on what came 1st. the data looks some thing like the following
unitNumber/StreetNumber Space StreetName (Building Name), Subub,City,Country
Some addresses won't have unit number or Suburb or country so when splitting you could have Suburbs and Citys in multiple columns even if you try and stagger the split process.Has any body go a good tool or reference site for dealing for this sort of problem. I have a table that I have made up that has some of the street names that could be used for comparing against existing records but it is by no means fool proof due to spelling inconsistencies . I also have another list of Common building names that could be used to compare, remove and place in the new building column.
I'm trying to parse out a line of data that is separated by the text "atc1.", "atc2." etc.
For example,
[atc1.123/atc2.456/atc3.789/atc4.xyz/]
If I only want the data after atc2., then I could search the string for "atc2." and collect all the characters afterwards. But how can I make sure to trim off all the data after "atc3." to make sure I'm only collecting "456" from the example above?
select notes, jobid, caller from contact where status in (6) and jobid = 173 I am getting this:
This job will be posted to Monster for 2 weeks. 173 906 Waiting for full budget approval 173 906 TUrns out we're uppin 173 906
What should I do so that these three columns for the same jobid from the same caller appears in only one column, either separated by a comma or semicolon?
I got a sales cost and cost amount table for my budget. the sales cost table is getting updated with FOBB items which makes the total incorrect . the FOBB values needs to be moved from the sales cost column to the cost amount column. how can i do it with an SQL script.
I wrote this query to pull some information involving 3 tables dbo.NewFamNbrs$, dbo.CGIItemMaster and dbo.CGIFamilyMaster. In addition I want to perform an update statement to update column dbo.CGIItemMaster.FamilyIDX with the information of dbo.CGIFamilyMaster.IDX. I also know that truncation is indicating that I'm trying to insert data into a field not large enough. So I performed a len() function on the both columns and order by desc and they both came up with 4 as being the max.
Query:
SELECT a.PART, a.Family, b.FamilyIDX, c.FamilyID, c.IDX FROM dbo.NewFamNbrs$ a LEFT OUTER JOIN dbo.CGIFamilyMaster c ON a.Family = c.FamilyID LEFT OUTER JOIN dbo.CGIItemMaster b ON a.PART = b.PART
Results of query:
PART Family FamilyIDX FamilyID IDX 000127233TF011468 TF01 506 000129880TF011468 TF01 506 003110 MET061468 MET06 1915
UPDATE STATEMENT:
begin tran update dbo.CGIItemMaster set FamilyIDX=c.IDX FROM dbo.NewFamNbrs$ a LEFT OUTER JOIN CGIFamilyMaster c ON a.Family = c.FamilyID LEFT OUTER JOIN dbo.CGIItemMaster b ON a.PART = b.PART
I have a SQL text column from SP_who2 in table #SqlStatement:
like 1row shown below :
"update Panel set PanelValue=7286 where PanelFirmwareID=4 and PanelSettingID=9004000"
I want to find what table and column names are in the text ..
I tried like below ..
Select B.Statement from #sp_who2 A LEFT JOIN #SqlStatement B ON A.spid = B.spid where B.Statement IN ( SELECT T.name, C.name FROM sys.tables T JOIN sys.columns C ON T.object_id=C.object_id WHERE T.type='U' )
Something like this : find the column names and tables name
When I set a column to have a default definition that uses a UDF, I am receiving the "String or binary data would be truncated" error.
The UDF:
Code BlockALTER FUNCTION GetDefaultClientTier ( @ClientAssets decimal(15,2) ) RETURNS char(1) AS BEGIN DECLARE @Result char(1) -- Get the first result in case of overlaps. SET @Result = CAST((SELECT TOP 1 ClientTier FROM ClientTiers WHERE @ClientAssets BETWEEN ClientAssetsFloor AND ClientAssetsCeiling) AS char(1)) RETURN @Result END
ClientTier is defined as char(1) in the table. I simply have (isnull([dbo].[GetDefaultClientTier]([ClientAssets])),(null))) as the definition. I can't use a computed column because I the values need to be editable. When inserting with SSIS, the insert works fine but the column has a value of null for each row.
When putting a character as the default (like 'A') the insert works fine.
The cast is there only because I have tried everything I can think of to get around this.
I have a student table like this studentid, schoolID, previousschoolid, gradelevel.
I would like to load this table every day from student system.
During the year, the student could change schoolid, whenever there is a change, I would put current records schoolid to the previous schoolid column, and set the schoolid as the newschoolid from student system.
My question in my merge statement something like below
Merge into student st using (select * from InputStudent ins) on st.id=ins.studentid
When matched then update
set st.schoolid=ins.schoolid , st.previouschoolid= case when (st.schoolid<>ins.schoolid) then st.schoolid else st.previouschoolid end , st.grade_level=ins.grade_level ;
My question is since schoolid is et at the first line of set statement, will the second line still catch what is the previous schoolid?
I have a string and i want to get only the numbers from right.
For example if I have the string Like '123756zxfggr123456' then it will show me only 123456 or if i have the string like '4vbz67xfggr123dfd' then it will show me only 123 or if i have the string like '123756zxfgg43r5' then it will show me only 5.
I got a function where it gives me all the numbers in a string but I don't need that
CREATE FUNCTION dbo.udf_GetNumeric (@strAlphaNumeric VARCHAR(256)) RETURNS VARCHAR(256) AS BEGIN DECLARE @intAlpha INT
[Code] ....
If I ran the select statement it gives me the result 111123456 but i want only 123456 or if i select
SELECT dbo.udf_GetNumeric('111zxfggr6587fhhfkwee') AS 'Num' it will show me 6587.
I have log files that I am loading into a table with duration data in the format "xx hrs xx min xx sec". Only the parts that are required will be there so if duration is only 2 seconds , the column will show "2 sec".
I am trying to get the duration into in to do some analysis on it and I have come up with this query so far which returns the correct data but i am wondering if there is a way to do what I am trying in a more readable format.
CREATE TABLE #tmp(duration VARCHAR(20))
INSERT INTO #tmp VALUES ('1 hrs 20 min 12 sec'), --4812 sec ('48 sec'), --48 sec ('39 min 1 sec'), --2341 sec ('11 hrs 1 min 1 sec'), --39661 sec ('59 min 0 sec'), --3540 sec ('2 min 50 sec') --170 sec
I need extracting string that is between certain characters that are in certain position.
Here is the DDL:
DROP TABLE [dbo].[StoreNumberTest] CREATE TABLE [dbo].[StoreNumberTest]( [StoreNumber] [varchar](50) NULL, [StoreNumberParsed] [varchar](50) NULL) INSERT INTO [dbo].[StoreNumberTest]
[Code] ....
What I need to accomplish is to extract the string that is between the third and fifth '-' (dash) and insert it into the StoreNumberParsed while eliminating the fourth dash.
Sample output would be:
KY117 CA132 OH174 MD163 FL191
I know that parse, charindex, patindex all might come in play, but not sure how to construct the statement.
I am trying to make a query that will group my errors messages together - my problem is that each of the error messages is unique, due to them having an unique id in them.
