I would do that with regexprep (REGular EXPression REPlace), which is not the easiest thing to pick up. Definitely look at the documentation for that function if you're not familiar.
My command for a single entry would be the following:
>> regexprep('Production.K07L890E4Y00','(Production.)(\w{4})(\w*)','$2')
ans = K07L
What this is saying is I'm partitioning my search text into three expressions:
- (Production.) -- literally the text "Production."
- (\w{4}) -- four alphanumeric/underscore characters, i.e., the ones I want to keep
- (\w*) -- the remaining alphanumeric/underscore characters, i.e., the ones I can throw away
By choosing $2 as the third argument, I'm saying replace all the strings that match my above criteria with the second expression, which is those four alphanumeric characters after the period (\w{4}).
Doing this on a table variable of character arrays would then be:
>> regexprep(myTable.VarName1,'(Production.)(\w{4})(\w*)','$2')
ans =
'K02L'
'K07L'
'K09L'
'Heating.Cr_Cleaning'
'Heating.Cr_67kev_Cleaning'
'Top_OFF.IT4_K57L'
'K09L'
'Top_OFF.IT4_K57L'
'DummyTest.G078K342E5T6'
Now the problem with this logic is that Top_OFF.IT4_Production.K57LUY01_Try4 became Top_OFF.IT4_K57L and didn't retain the _Try4 at the end.
You can get around this with more complex logic that stops the replacement at any underscores which may be there. I will defer to the documentation to explain how this one works :)
>> regexprep(myTable.VarName1,'(Production.)(\w{4})([a-zA-Z0-9]*)(\_?)','$2$4')
ans =
'K02L'
'K07L'
'K09L'
'Heating.Cr_Cleaning'
'Heating.Cr_67kev_Cleaning'
'Top_OFF.IT4_K57L'
'K09L'
'Top_OFF.IT4_K57L_Try4'
'DummyTest.G078K342E5T6'
- Sebastian
