Status: Solved
Priority: Medium
Security: Public
Views: 229
Last Modified: 2016-08-17

MSSQL Parse XML Values Into SQL Fields

Hello...

We inherited a small project in MSSQL. The database has a field in an SQL table called "Address", which, literally, has the following sample values stored in it (6 rows follow as an example) :

Wilson Photography 333 54th Ave. North Nashville TN 37209
Wyatt Archaelogical Research 4313 Green Hills Rd Nashville TN 37174-5132
Ed Perdue 311 Hobbs Dr White House TN 37188
NULL
NULL
Maury Wahtera 407 Normandy Circle Nashville TN 37209

What we need...is to be able to have these values...parsed out into "proper" fields in a VIEW, as follows:
NAME, ADDRESS, CITY, STATE, ZIPCODE

...wherever the field value is NOT NULL (obviously :-)

We see that those values are separated by a " " string, but we don't know the syntax, to parse this (XML?) value out. Any help would be very much appreciated. Thank you!...Mark

datatechcorp

Asked: 2016-08-09

Advertise Here

1 Solution

LVL 41

SharathData EngineerCommented: 2016-08-09

What is the datatype of this Address field? XML or VARCHAR?

datatechcorpAuthor Commented: 2016-08-09

Hi Sharath...

It's an NVARCHAR(1000), and it allows NULLS. Thanks!...Mark

LVL 12

funwithdotnetCommented: 2016-08-09

Parsing strings in T-SQL is kind of a pita, unless you are using a version 13 (SQL Server 2016) database, which you could use just to do the parsing.

Using a v13 database, you can:

select * from STRING_SPLIT(replace('Wilson Photography<br>333 54th Ave. North<br>Nashville TN 37209', '<br>', '~') , '~' )

Open in new window

That will return three rows:
Wilson Photography
333 54th Ave. North
Nashville TN 37209

LVL 41

SharathData EngineerCommented: 2016-08-09

Can I assume that the max no.of columns expected are 5?

datatechcorpAuthor Commented: 2016-08-09

Yes, correct, 5-Columns max. Unfortunately, we are stuck with the MSSQL 2012 Iteration, and cannot go beyond that (sigh). Thoughts? Thanks!...Mark

LVL 41

SharathData EngineerCommented: 2016-08-10

Check this.

declare @table table (col varchar(100))
insert @table values 
('Wilson Photography<br>333 54th Ave. North<br>Nashville TN 37209'),
('Wyatt Archaelogical Research<br>4313 Green Hills Rd<br>Nashville TN 37174-5132'),
('Ed Perdue<br>311 Hobbs Dr<br>White House TN 37188'),
(NULL),
(NULL),
('Maury Wahtera<br>407 Normandy Circle<br>Nashville TN 37209')

;with cte1 as (select *,charindex('<br>',col) idx1, charindex('<br>', col, charindex('<br>',col)+1) idx2 from @table),
      cte2 as (select col,left(col,idx1-1) Name, substring(col,idx1+4,idx2-idx1-4) Address,substring(col,idx2+4,len(col)) str3 from cte1),
      cte3 as (select *,reverse(str3) rev_str3 from cte2),
	  cte4 as (select *,charindex(' ',rev_str3) idx1, charindex(' ',rev_str3,charindex(' ',rev_str3)+1) idx2 from cte3)
select Name, Address,
       reverse(rtrim(ltrim(substring(rev_str3,idx2,len(rev_str3))))) City,
	   reverse(rtrim(ltrim(substring(rev_str3,idx1,idx2-idx1)))) State,
	   reverse(left(rev_str3,idx1-1)) ZipCode
  from cte4
/*
Name	Address	City	State	ZipCode
Wilson Photography	333 54th Ave. North	Nashville	TN	37209
Wyatt Archaelogical Research	4313 Green Hills Rd	Nashville	TN	37174-5132
Ed Perdue	311 Hobbs Dr	White House	TN	37188
NULL	NULL	NULL	NULL	NULL
NULL	NULL	NULL	NULL	NULL
Maury Wahtera	407 Normandy Circle	Nashville	TN	37209
*/

Open in new window

datatechcorpAuthor Commented: 2016-08-10

Hi Sharath...

Thank you for responding. If my Existing TableName is "NSCustomer"...and that FieldName that we want to parse into our select statement is "NSCustomer.Address"...how should this query syntax look? Please help...and I apologize for our ignorance :-)

Thanks!...Mark

LVL 41

SharathData EngineerCommented: 2016-08-10

;with cte1 as (select *,charindex('<br>',Address) idx1, charindex('<br>', Address, charindex('<br>',Address)+1) idx2 from NSCustomer),
      cte2 as (select Address original_Address,left(Address,idx1-1) Name, substring(Address,idx1+4,idx2-idx1-4) Address,substring(Address,idx2+4,len(Address)) str3 from cte1),
      cte3 as (select *,reverse(str3) rev_str3 from cte2),
      cte4 as (select *,charindex(' ',rev_str3) idx1, charindex(' ',rev_str3,charindex(' ',rev_str3)+1) idx2 from cte3)
select Name, Address,
       reverse(rtrim(ltrim(substring(rev_str3,idx2,len(rev_str3))))) City,
       reverse(rtrim(ltrim(substring(rev_str3,idx1,idx2-idx1)))) State,
       reverse(left(rev_str3,idx1-1)) ZipCode
  from cte4

Open in new window

datatechcorpAuthor Commented: 2016-08-10

Hi Sharath...

That seemed to work beautifully...until it got to the "umpteenth" record in the NSCustomer table...and then we got the following error:

Invalid length parameter passed to the LEFT or SUBSTRING function.

How can we avoid this? Any thoughts here? The value of the FIELD is as follows, in the "next" record in the Table...and we didn't see anything "squirrely" with the data:

ArNells Art Photography 864 Joseph Ave. Nashville TN

LVL 41

SharathData EngineerCommented: 2016-08-10

You don't have ZipCode for this value. Do you have any other part (City/State) missing for any row?
Also, what is your SQL version?

datatechcorpAuthor Commented: 2016-08-10

Hi Sharath...

MSSQL 2012. And, yes, it's very conceivable...that ANY part of the parsed field values...may be "missing". It depends what the doggone Salespersons keyed in to the Customer's account record. There are over 45,000 records, so, yes, certainly, I'm sure "parts" may be missing. UUGH...I know,

Thoughts? And your assistance has been *very much* appreciated...Thanks!...Mark

LVL 41

SharathData EngineerCommented: 2016-08-11

Is it safe assumption that the STATE is always a 2-character string?
Are these assumptions fair?

If a 2-character string is missing, that means STATE part is missing.
If numerical string is missing, ZipCode is missing.

datatechcorpAuthor Commented: 2016-08-11

Hi Sharath...

Perhaps. Let me review the data...I should have an answer within the hour, OK? Thanks!...Mark

datatechcorpAuthor Commented: 2016-08-11

Hi Sharath...

OK...sooooo...the data is kind of...all over the place...where "CITY", "STATE", and/or "ZIPCODE" are concerned. That's because (a) in some cases the STATE was left out of the data entry, (b) in some cases there are International Addresses and the value of 'Canada' or 'Australia' exists, (c) in some cases there are NO ZipCodes at all...and so forth.

For the *most* part, they are all "in line" with the structure of the fields we require...but probably a good 15-20% or so of the data is 'whacky'.

I think as long as the data gets parsed, somehow, someway, we can do a "data scrubbing" exercise "after the fact". Seems to ALWAYS be the case, where "Addresses" are concerned, in a Customer Master table...make sense?

So, how should we proceed? Please let me know asap...and, again, I very much appreciate the assistance...Thanks!...Mark

LVL 41

SharathData EngineerCommented: 2016-08-13

Can you post some examples for international address and how do you want to parse?

datatechcorpAuthor Commented: 2016-08-17

Hi Sharath...

I've been busy doing some "data scrubbing"...and I believe we're all good for now. Thank you sooooo much for all your assistance. I'll let you know if I need anything further. I'm awarding you all the points.

Again, Thanks!...Mark

Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

MSSQL Parse XML Values Into SQL Fields

Join & Write a Comment Already a member? Login.

Featured Post