Indexing long text fields failing using arcpy AddIndex?

Question

Arcpy AddIndex_management won't index long text fields, is there a workaround?

I tried:

arcpy.AddIndex_management(r'C:\workspace\foo.dbf', "Text")

and got the error:

Runtime error Traceback (most recent call last): File "<string>", line 1, in <module> 
 File "c:\program files (x86)\arcgis\desktop10.3\arcpy\arcpy\management.py", line 5993, in AddIndex 
 raise e ExecuteError: 
 ERROR 999999: Error executing function. 
 Attribute indexes cannot be created for fields with a length that is greater than 80. 
Failed to execute (AddIndex).

I have also tried this with shapefiles as well as dbf's with the same error. I need to index a text field in order to perform a JoinField() more quickly with arcpy.

Is there a work around to do this that doesn't involve shortening text fields?

I haven't seen a reason not to index fields longer than 80 characters; especially since the text fields I'm indexing have been calculated and are therefore not very prone to typos. For example, if you had a series of field of species names and you wanted to join that to a table containing descriptions of those species, you would be joining on a potentially long text string.

• Midavalo: The problem here appears to be the length of the text field, rather than the fact that it's a text field. Have you tried it on fields shorter than 80 characters? Performing a join on a value longer than 80 characters seems unusual and risky (for typos) to me.
• Craig T: I have now and it does work. I haven't seen a reason not to index fields longer than 80 characters; especially since the text fields I'm indexing have been calculated and are therefore not very prone to typos. For example, if you had a series of field of species names and you wanted to join that to a table containing descriptions of those species, you would be joining on a potentially long text string.
• Midavalo: Could you create a new coded value field for those values in each table, and index and join on those?
• Craig T: I suppose I could. Ok, next question is if I am exporting an excel file with a text field to a dbf, how do I get the length of that field to not be the standard 256? Should I ask that as another question?
• Midavalo: Yes ask as a separate question

Answer 1

If you CAN edit the length of text fields the answer is as follows:

arcpy.AddField(r'C:\workspace\foo.dbf', "shortText", "#", "#", 75, "#", "#", "#", "#")
with arcpy.da.updateCursor(r'C:\workspace\foo.dbf', ["Text", "shortText"]) as cursor:
 for row in cursor:
 row[1] = row[0]
 cursor.updateCursor(row)
arcpy.AddIndex_management(r'C:\workspace\foo.dbf', "shortText")

and then perform your join on "shortText"