Related
I have a column in a SQL Server table that has strings of varying lengths. I need to find the position of the first occurrence of the string , -- that's not enclosed in single quotes or square brackets.
For example, in the following two strings, I've bolded the portion I would like to get the position of. Notice in the first string, the first time , -- appears on its own (without being between single quote or square bracket delimiters) is at position 13 and in the second string, it's at position 16.
'a, --'[, --]**, --**[, --]
[a, --b]aaaaaaa_ **, --**', --'
Also I should mention that , -- itself could appear multiple times in the string.
Here's a simple query that shows the strings and my desired output.
SELECT
t.string, t.desired_pos
FROM
(VALUES (N'''a, --''[, --], --[, --]', 14),
(N'[a, —-b]aaaaaaa_ , --'', --''', 18)) t(string, desired_pos)
Is there any way to accomplish this using a SELECT query (or multiple) without using a function?
Thank you in advance!
I've tried variations of SUBSTRING, CHARINDEX, and even some CROSS APPLYs but I can't seem to get the result I'm looking for.
Before i write down my solution, i must warn you: DON'T USE IT. Use a function, or do this in some other language. This code is probably buggy.
It doesn't handle stuff like escaped quotes etcetc.
The idea is to first remove the stuff inside brackets [] and quotes '' and then just do a "simple" charindex.
To remove the brackets, i'm using a recursive CTE that loops ever part of matching quotes and replaces their content with placeholder strings.
One important point is that quotes might be embedded in each other, so you have to try both variants and chose the one that is earliest.
WITH CTE AS (
SELECT *
FROM
(VALUES (N'''a, --''[, --], --[, --]', 14),
(N'[a, —-b]aaaaaaa_ , --'', --''', 18)) t(string, desired_pos)
)
, cte2 AS (
select x.start
, x.finish
, case when x.start > 0 THEN STUFF(string, x.start, x.finish - x.start + 1, REPLICATE('a', x.finish - x.start + 1)) ELSE string END AS newString
, 1 as level
, string as orig
, desired_pos
from cte
CROSS APPLY (
SELECT *
, ROW_NUMBER() OVER(ORDER BY case when start > 0 THEN 0 ELSE 1 END, start) AS sortorder
FROM (
SELECT charindex('[', string) AS start
, charindex(']', string) AS finish
UNION ALL
SELECT charindex('''', string) AS startQ
, charindex('''', string, charindex('''', string) + 1) AS finishQ
) x
) x
WHERE x.sortorder = 1
UNION ALL
select x.start
, x.finish
, STUFF(newString, x.start, x.finish - x.start + 1, REPLICATE('a', x.finish - x.start + 1))
, 1 as level
, orig
, desired_pos
from cte2
CROSS APPLY (
SELECT *
, ROW_NUMBER() OVER(ORDER BY case when start > 0 THEN 0 ELSE 1 END, start) AS sortorder
FROM (
SELECT charindex('[', newString) AS start
, charindex(']', newString) AS finish
UNION ALL
SELECT charindex('''', newString) AS startQ
, charindex('''', newString, charindex('''', newString) + 1) AS finishQ
) x
) x
WHERE x.sortorder = 1
AND x.start > 0
AND cte2.start > 0 -- Must have been a match
)
SELECT PATINDEX('%, --%', newString), *
from (
select *, row_number() over(partition by orig order by level desc) AS sort
from cte2
) x
where x.sort = 1
Try this approach. I'm replacing the strings you don't need for another string of the same length. Then look for the position of the interested string.
SELECT string, desired_pos,
CHARINDEX(', --', REPLACE(REPLACE(string, ''', --''', '******'), '[, --]', '******')
) start_index
FROM (VALUES (N''', --''[, --], --[, --]', 13),
(N'[, --]aaaaaaa_ , --'', --''', 16)) t(string, desired_pos)
I don't know if it makes sense with a C# solution, but this class for CVS is a nice little parcer: TextFieldParser
Then you just define Delimeters etc. and assuming the input is escaped consistently then all is good.
Im late the game here but This kind of thing is simple in SQL Server when leveraging NGrams8k. Not only do you not need REGEX, a CLR, C# required. Furthermore, NGrams8k will be the fastest by far. In 8 years nobody has produced anything remotely as fast. Furthermore, this code will be faster and far less complex than a recursive CTE solution (which are almost always slow in SQL Server)
;--==== Sample Data
DECLARE #T Table (String VARCHAR(100))
INSERT #T
VALUES (N'''a, --''[, --], --[, --]'),
(N'[a, —-b]aaaaaaa_ , --'', --''');
;--==== Solution
SELECT
t.String, ng.Position
FROM #t AS t
CROSS APPLY (VALUES(REPLACE(t.String,'[',CHAR(1)))) AS f(S)
CROSS APPLY samd.NGrams8k(f.S,4) AS ng
CROSS APPLY (VALUES(SUBSTRING(f.S,ng.Position-2,7))) AS g(String)
WHERE ng.Token = ', --'
AND g.String NOT LIKE '%''%''%'
AND g.String NOT LIKE '%'+CHAR(1)+'%]%';
Results:
String Position
----------------------------- --------------------
'a, --'[, --], --[, --] 14
[a, —-b]aaaaaaa_ , --', --' 18
I am trying to extract the data between two underscore characters. In some situations, the 2nd underscore may not exist.
MyFld
P_36840
U_216137
C_203134_H
C_203134_W
I tried this:
substring(i.[MyFld],
CHARINDEX ('_',i.[MyFld])+1,len(i.[MyFld])
-CHARINDEX ('_',i.[MyFld])
) [DerivedPrimaryKey]
And I get this:
DerivedPrimaryKey
36840
216137
203134_H
203134_W
https://dbfiddle.uk/uPKC6oX4
I want to remove the second underscore and data that follows it. I'm trying to combine it with a trim right, but I'm unsure where to start.
How can I do this?
We can start by simplifying what you have so far. I will also add enough to make this a complete query, so we can see it in context for later steps:
SELECT
right(i.MyFld, len(i.MyFld) - charindex('_', i.MyFld)) [DerivedPrimaryKey]
FROM I
With this much done, we can now use it as the source for removing the trailing portion of the field:
SELECT
reverse(substring(reverse(step1)
, charindex('_', reverse(step1))+1
, len(step1)
)) [DerivedPrimaryKey]
FROM (
SELECT right(i.MyFld, len(i.MyFld) - charindex('_', i.MyFld)) [step1]
FROM I
) T
Notice the layer of nesting. You can, of course, remove the nesting, but it means replicating the entire inner expression every time you see step1 (good thing I took the time to simplify it):
SELECT
reverse(substring(reverse(right(i.MyFld, len(i.MyFld) - charindex('_', i.MyFld)))
, charindex('_', reverse(right(i.MyFld, len(i.MyFld) - charindex('_', i.MyFld))))+1
, len(right(i.MyFld, len(i.MyFld) - charindex('_', i.MyFld)))
))
FROM I
And now back to just the expression:
reverse(substring(reverse(right(i.MyFld, len(i.MyFld) - charindex('_', i.MyFld)))
, charindex('_', reverse(right(i.MyFld, len(i.MyFld) - charindex('_', i.MyFld))))+1
, len(right(i.MyFld, len(i.MyFld) - charindex('_', i.MyFld)))
))
See it work here:
https://dbfiddle.uk/nFO4Vwhm
There is also this alternate expression that saves one function call:
left( right(i.MyFld,len(i.MyFld)-charindex('_',i.MyFld)),
coalesce(
nullif(
charindex('_',
right(i.MyFld,len(i.MyFld)-charindex('_',i.MyFld))
) -1, -1,
),
len( right(i.MyFld,len(i.MyFld)-charindex('_',i.MyFld)) )
)
)
Just a two more options. One using parsename() provided your data does not have more than 4 segments. The second using a JSON array
Example
Declare #YourTable Table ([MyFld] varchar(50)) Insert Into #YourTable Values
('P_36840')
,('U_216137')
,('C_203134_H')
,('C_203134_W')
Select *
,UsingParseName = reverse(parsename(reverse(replace(MyFld,'_','.')),2))
,UsingJSONValue = json_value('["'+replace(MyFld,'_','","')+'"]','$[1]')
From #You
Results
MyFld UsingParseName UsingJSONValue
P_36840 36840 36840
U_216137 216137 216137
C_203134_H 203134 203134
C_203134_W 203134 203134
We can do this:
Declare #testData Table ([MyFld] varchar(50));
Insert Into #testData (MyFld)
Values ('P_36840')
, ('U_216137')
, ('C_203134_H')
, ('C_203134_W');
Select *
, second_element = substring(v.MyFld, p1.pos, p2.pos - p1.pos - 1)
From #testData As td
Cross Apply (Values (concat(td.MyFld, '__'))) As v(MyFld) -- Make sure we have at least 2 delimiters
Cross Apply (Values (charindex('_', v.MyFld, 1) + 1)) As p1(pos) -- First Position
Cross Apply (Values (charindex('_', v.MyFld, p1.pos) + 1)) As p2(pos) -- Second Position
If you actually have a fixed number of characters in the first element, then it could be simplified to:
Select *
, second_element = substring(v.MyFld, 3, charindex('_', v.MyFld, 4) - 3)
From #testData td
Cross Apply (Values (concat(td.MyFld, '_'))) As v(MyFld)
Often I try to fake out SQL if an expected character isn't always present and I don't need the resulting value:
SELECT SUBSTRING(field_Calculated, 1, CHARINDEX('_', field_Calculated) - 1)
FROM (SELECT SUBSTRING(MyFld, CHARINDEX('_', MyFld) + 1, LEN(MyFld)) + '_' As field_Calculated
FROM MyTable) T
I think this is clear, but I really like the ParseName solution #JohnCappalletti suggests.
If it's only ever one numeric value you can use string_split:
SELECT * FROM MyTable
CROSS APPLY string_split(MyFld, '_')
WHERE ISNUMERIC(value) = 1
Either way you have to be careful of the data before deciding the best approach.
your data
Declare #Table Table ([MyFld] varchar(100))
Insert Into #Table
([MyFld] ) Values
('P_36840')
,('U_216137')
,('C_203134_H')
,('C_203134_W')
use SubString,Left and PatIndex
select
Left(
SubString(
[MyFld],
PatIndex('%[0-9.-]%', [MyFld]),
8000
),
PatIndex(
'%[^0-9.-]%',
SubString(
[MyFld],
PatIndex('%[0-9.-]%', [MyFld]),
8000
) + 'X'
)-1
) as DerivedPrimaryKey
from
#Table
I am trying to replace SQL results with all the middle values with asterix, *. All results are words. I am using SSMS.
The words that are 4-5 letters, it should only show 1 letter in the beginning, one to the end.
6 letters and more, it it should only show 2 letter in the beginning, 2 letters in the end.
1-3 letters, no replacement.
For example:
(I am now using - instead of * so it does not make the text bold).
"Banana" 6 letters should become ba--na
"False" 5 letters should become F---e
"a" stays the same
"Selin is a vegetable and banana is a fruit" becomes "S---n is a ve-----le and ba--na is a f---t."
What I have done so far, is to make this for emails, after the #. But now I want it to happen with every word of the result.
What I've done:
DECLARE #String VARCHAR(100) = 'sample#gmail.com'
SELECT STUFF(STUFF(#STring,
CHARINDEX('#',#String)+2,
(CHARINDEX('.',#String, CHARINDEX('#',#String))-CHARINDEX('#',#String)-3),
REPLICATE('*',CHARINDEX('.',#String, CHARINDEX('#',#String))-CHARINDEX('#',#String)))
,2
,CHARINDEX('#',#String)-3
,REPLICATE('*',CHARINDEX('#',#String)-3))```
With result s----e#g------l.com
instead of -
And I tried the mask method
Select
--select first character from Email and use replicate
SUBSTRING(Sxolia,1,1) + REPLICATE('*',5)+
--function to put asterisks
SUBSTRING(Sxolia,CHARINDEX('#',Sxolia),len(Sxolia)-CHARINDEX('#',Sxolia)+1)
--at this statement i select this part #gmail,com and to first part to become like this A*****#gmail.com
as Emailmask
From [mytable]
With result
B***** Bana is a fruit
And
declare #str nvarchar(max)
select #str = '123456'
select '****' + substring(#str, 5, len(#str) - 3)
Result: ****56
Not what I am looking for.
How should I look into this?
If I had to deal with this in SQL Server I'd operate on each word as a row, however using string_split is not (currently) an option since it does not guarantee ordering.
The following uses json to split the string as an array and provides a key value for ordering, which allows the words to be aggregated in the correct order:
select t.Sentence,
String_Agg( masked, ' ') within group(order by seq) Masked
from t
cross apply (
select seq, [value] word,
case
when l<=3 then [value]
when l<=5 then Stuff([value],2,l-2,Replicate('*',l-2))
else
Stuff([value],3,l-4,Replicate('*',l-4))
end Masked
from (
select j.[value], 1 + Convert(tinyint,j.[key]) Seq
from OpenJson(Concat('["',replace(t.Sentence,' ', '","'),'"]')) j
)w
cross apply (values(Len([value])))x(l)
)w
group by t.Sentence;
See working demo
Result:
I'm not sure how e-mail fits into all this because you're asking for word masks, so I'm going to assume you actually want this. Use divide and conquer to implement this, so first implement an expression that would do this for simplest cases (e.g. single words). Then if you need it for e-mails, just split the e-mails however you see fit and then apply the same expression.
The expression itself is rather simple:
SELECT *
FROM (VALUES
('banana'),
('selin'),
('vegetable')
) words(word)
CROSS
APPLY (SELECT CASE
WHEN ln BETWEEN 4 AND 5
THEN LEFT(word, 1) + REPLICATE('*', ln-2) + RIGHT(word, 1)
WHEN ln >= 6
THEN LEFT(word, 2) + REPLICATE('*', ln-4) + RIGHT(word, 2)
ELSE word
END as result
FROM (VALUES (LEN(words.word))) x(ln)
) calc
This already provides the expected result. You could define a function out of this, if you have the permissions, and use it like so:
SELECT *
FROM (VALUES
('banana'),
('selin'),
('vegetable')
) words(word)
CROSS
APPLY fnMaskWord(word)
Here's a working demo on dbfiddle, it includes the statement to create the function.
Expanding on a few answers:
select case when len(#String) <= 3 then #String
when len(#String) > 3 AND len(#String) <= 5 then
substring(#String, 1, 2) +
REPLICATE('*', Len(#String) - 2) +
substring(#String, Len(#String) - 1, 2)
when len(#String) >= 6 then
substring(#String, 1, 2) +
REPLICATE('*', Len(#String) - 2) +
substring(#String, Len(#String) - 1, 2)
else 'unrecognized length!'
If the length of the string is less than or equal to 3, return the string.
If the length of the string is more than 3 and less than or equal to 5 then create a substring starting at position 1, then replicate * by the length of the string -2 and finally add another substring -1 from the end of the string.
Similar for if the result is over 6 characters.
Else unrecognized length!
Hope this helps understand what's going on!
Maybe this can help
declare #t table (word varchar(50))
insert into #t values ('banana'), ('selin'), ('vegetable')
select case when len(t.word) < 3 then t.word
else left(t.word, 1) + -- take first char from left
replicate('*', Len(t.word) - 2) + -- fill middle with *
right(t.word, 1) -- take last char from right
end
from #t t
this returns
COLUMN1
b****a
s***l
v*******e
If you want to keep 2 chars left and right when the len > 5 then maybe this
select case when len(t.word) < 3 then t.word
when len(t.word) < 6 then
left(t.word, 1) +
replicate('*', len(t.word) - 2) +
right(t.word, 1)
else left(t.word, 2) +
replicate('*', len(t.word) - 4) +
right(t.word, 2)
end
from #t t
The result
COLUMN1
ba**na
s***l
ve*****le
EDIT: What if there is a whole sentence ?
Well then we first split the sentence in words,
and then concat the individual words back together while putting the ** in them
declare #t table (word varchar(50))
insert into #t values ('banana'), ('selin'), ('vegetable'), ('Banana is a fruit')
select t.word,
-- put the words back togheter into the sentence, and ** them while we are at it
( select string_agg(case when len(value) < 3 then value
when len(value) < 6 then
left(value, 1) +
replicate('*', len(value) - 2) +
right(value, 1)
else left(value, 2) +
replicate('*', len(value) - 4) +
right(value, 2)
end,
' ')
)
from #t t
cross apply string_split(t.word, ' ') s -- split the sentence into words
group by t.word
the result is
word COLUMN1
---- -------
banana ba**na
Banana is a fruit Ba**na is a f***t
selin s***n
vegetable ve*****le
In a certain SQL table I'm working on, there's a column that contains data formatted like:
"year-text~year-text~year-text~year-text~year-text~year-text~year-text~" and so on and so forth.
(year is in 'yyyy' format)
(for example):
"2012-10000~2013-5000~2014-500~2015-50000~2016-100~"
How, using SQL might I extract, say, the value "50000" based on having the year, "2015"
Things to note/clarify:
The "-" and "~" characters can be trusted as delimiters. That is, they do not exist within any of the values or, of course, the years.
No year exists without a value. In other words, if the value becomes blank, the year is stripped out, as well (In other words, the stored string will never have an "-" and a "~" right next to each other, such as 2016 in the string "2015-200~2016-~2014-1000", for example).
The years in the string may not be in chronological order from left to right.
There could be virtually any number of years (each with a value) in the string or, indeed, none, at all. If no year/value pair exists for the column, the value becomes NULL
Please note that after each value for each year the character "~" is applied even if it is the last year/value pair. Any string value that is not NULL will therefore always end with a "~".
Perhaps this can help
With the aid of a parser and cross apply
Declare #String varchar(max) = '012-10000~2013-5000~2014-500~2015-50000~2016-100~'
Select A.*
,B.*
From [dbo].[udf-Str-Parse](#String,'~') A
Cross Apply (Select Val1=max(IIF(Key_PS=1,Key_Value,NULL))
,Val2=max(IIF(Key_PS=2,Key_Value,NULL))
From [dbo].[udf-Str-Parse](A.Key_Value,'-')) B
Where A.Key_Value<>''
Returns
Key_PS Key_Value Val1 Val2
1 012-10000 012 10000
2 2013-5000 2013 5000
3 2014-500 2014 500
4 2015-50000 2015 50000
5 2016-100 2016 100
My Parser if needed
CREATE FUNCTION [dbo].[udf-Str-Parse] (#String varchar(max),#Delimeter varchar(10))
--Usage: Select * from [dbo].[udf-Str-Parse]('Dog,Cat,House,Car',',')
-- Select * from [dbo].[udf-Str-Parse]('John Cappelletti was here',' ')
Returns #ReturnTable Table (Key_PS int IDENTITY(1,1), Key_Value varchar(max))
As
Begin
Declare #XML xml;Set #XML = Cast('<x>' + Replace(#String,#Delimeter,'</x><x>')+'</x>' as XML)
Insert Into #ReturnTable Select ltrim(rtrim(String.value('.', 'varchar(max)'))) FROM #XML.nodes('x') as T(String)
Return
End
Here is an option which uses SUBSTRING and CHARINDEX to get the job done:
SUBSTRING(col,
CHARINDEX('2015', col) + 5,
CHARINDEX('~', col, CHARINDEX('2015', col) + 5) - (CHARINDEX('2015', col) + 5))
In the sample input
2012-10000~2013-5000~2014-500~2015-50000~2016-100~
CHARINDEX('2015', col) + 5 would start at the 5 in the number 50000 after the occurrence of 2015.
The term
CHARINDEX('~', col, CHARINDEX('2015', col) + 5) - (CHARINDEX('2015', col) + 5)
yields the length of the number 50000, which in this case would be 5.
you can use substr and charindex
select substr( charindex( 'your_val', your_column) + length('your__val') +1, 4);
in your case
select substr( charindex( '2016', your_column) + length('2016') +1, 4);
I have a field that holds an account code. I've managed to extract the first 2 parts OK but I'm struggling with the last 2.
The field data is as follows:
812330/50110/0-0
812330/50110/BDG001-0
812330/50110/0-X001
I need to get the string between the second "/" and the "-" and after the "-" .Both fields have variable lengths, so I would be looking to output 0 and 0 on the first record, BDG001 and 0 on the second record and 0 and X001 on the third record.
Any help much appreciated, thanks.
You can use CHARINDEX and LEFT/RIGHT:
CREATE TABLE #tab(col VARCHAR(1000));
INSERT INTO #tab VALUES ('812330/50110/0-0'),('812330/50110/BDG001-0'),
('812330/50110/0-X001');
WITH cte AS
(
SELECT
col,
r = RIGHT(col, CHARINDEX('/', REVERSE(col))-1)
FROM #tab
)
SELECT col,
r,
sub1 = LEFT(r, CHARINDEX('-', r)-1),
sub2 = RIGHT(r, LEN(r) - CHARINDEX('-', r))
FROM cte;
LiveDemo
EDIT:
or even simpler:
SELECT
col
,sub1 = SUBSTRING(col,
LEN(col) - CHARINDEX('/', REVERSE(col)) + 2,
CHARINDEX('/', REVERSE(col)) -CHARINDEX('-', REVERSE(col))-1)
,sub2 = RIGHT(col, CHARINDEX('-', REVERSE(col))-1)
FROM #tab;
LiveDemo2
EDIT 2:
Using PARSENAME SQL SERVER 2012+ (if your data does not contain .):
SELECT
col,
sub1 = PARSENAME(REPLACE(REPLACE(col, '/', '.'), '-', '.'), 2),
sub2 = PARSENAME(REPLACE(REPLACE(col, '/', '.'), '-', '.'), 1)
FROM #tab;
LiveDemo3
...Or you can do this, so you only go from left side to right, so you don't need to count from the end in case you have more '/' or '-' signs:
SELECT
SUBSTRING(columnName, CHARINDEX('/' , columnName, CHARINDEX('/' , columnName) + 1) + 1,
CHARINDEX('-', columnName) - CHARINDEX('/' , columnName, CHARINDEX('/' , columnName) + 1) - 1) AS FirstPart,
SUBSTRING(columnName, CHARINDEX('-' , columnName) + 1, LEN(columnName)) AS LastPart
FROM table_name
One method way is to download a split() function off the web and use it. However, the values end up in separate rows, not separate columns. An alternative is a series of nested subqueries, CTEs, or outer applies:
select t.*, p1.part1, p12.part2, p12.part3
from table t outer apply
(select t.*,
left(t.field, charindex('/', t.field)) as part1,
substring(t.field, charindex('/', t.field) + 1) as rest1
) p1 outer apply
(select left(p1.rest1, charindex('/', p1.rest1) as part2,
substring(p1.rest1, charindex('/', p1.rest1) + 1, len(p1.rest1)) as part3
) p12
where t.field like '%/%/%';
The where clause guarantees that the field value is in the right format. Otherwise, you need to start sprinkling the code with case statements to handle misformated data.