Special join or other way

Question

Hi Forum,
I am having a problem with incosistent data of two columns with semicolon separated values. Column A and column B should have the exact same number of elements, which is not always the case.
In order to split the values into rows I use the approach from a previous thread
http://social.msdn.microsoft.com/Forums/en-US/06d2a674-c166-49fc-9a70-4fbc383190bb/split-multiple-columns-into-rows-using-xml?forum=transactsql
Joining by the row number I somehow need to normalize the tables that contain the split columns so that both columns are aligned before creating the row number.
If table A has one row and table B has 2, any kind of join would replicate the one row of table A. This gives me a wrong result. What I need is row 1 from table A match row 1 from table B and (non-existing) row 2 from table A with value null match row
2 from table B (vice versa). So:
Table A:
pk      manufacturer
1         man A
Table B:
pk      partnumber
1        part A
1        part B
should result in Table C:
pk      manufacturer   partnumber
1         man A             part A
1         null                 part B
how can I achieve this?
TIA
Alex

Answer 1

>> I am having a problem with inconsistent data of two columns with semicolon separated values. <<
This make no sense. A column has only scalar values by definition. What you have described is a mess that has nothing to do with RDBMS or correct SQL. 
>> In order to split the values into rows I use the approach from a previous thread
http://social.msdn.microsoft.com/Forums/en-US/06d2a674-c166-49fc-9a70-4fbc383190bb/split-multiple-columns-into-rows-using-xml?forum=transactsql <<
This should never be done in the database. XML is not SQL. 
>> Joining by the row number I somehow need to normalize the tables that contain the split columns so that both columns are aligned before creating the row number. <<
Row number? We have a ROW_NUMBER() function, but RDBMS is based on sets and sequential files, so there is no physical row number. 
Please post DDL, so that people do not have to guess what the keys, constraints, Declarative Referential Integrity, data types, etc. in your schema are. Learn how to follow ISO-11179 data element naming conventions and formatting rules. Temporal data should
use ISO-8601 formats. Code should be in Standard SQL as much as possible and not local dialect. 
This is minimal polite behavior on SQL forums. Without it, we cannot help you. What you want to do is non-relational garbage, not proper RDBMS. Your poorly named table “C” is not table! It is a report with display formatting. An SQL programmer would never destroy
information by turning a known value into a NULL. 
Why do you think that this is how SQL works? I have to teach people and I need to know what leads to such fundamental mistakes. It looks like you have no idea what a table is and have confused this concept with a deck of punch cards that can be shuffled together. 
Oh, there are horrible kludges, but why not learn to do it right instead? 
--CELKO-- Books in Celko Series for Morgan-Kaufmann Publishing: Analytics and OLAP in SQL / Data and Databases: Concepts in Practice Data / Measurements and Standards in SQL SQL for Smarties / SQL Programming Style / SQL Puzzles and Answers / Thinking
in Sets / Trees and Hierarchies in SQL