Take the 2-minute tour ×
Geographic Information Systems Stack Exchange is a question and answer site for cartographers, geographers and GIS professionals. It's 100% free, no registration required.

I want to select all of the features by attribute where the attribue in myattrib is like...

Selection "criteria H". That is not a problem.

However then I need to replace (myattrib," H", "") except I only want to replace the H at the end of the attribute, not within, like "toast is Hot H".

In this record I want the value to end up as "toast is HotH"

There are 15k records but only 4800 have the value I am replacing. And it is always the same value.

edit response:

No. All values I want to fix end in H with a space in front of it. I am depending on search replace but not sure how to leave the space in front of Hot.

In my example I am able to select with the "% H" which specifies there is nothing after the H so the select won't get Hot even though there is a space before it. However when it comes to replace it does a literal replace on " H" and doesn't care about the ot.

I think if I knew some regex and "IF" arcmap supports that, It might do the trick (from what I have seen on the FMEtalk forum). I'm sure someone can talk me through this.

share|improve this question
what is your data store? shapefile, sde (sql server, oracle, postgres), fgdb? I think you might be looking at some regexp work. –  Jay Cummins Dec 8 '10 at 23:08
Have you tried using the field calculator in 10.0, with a python pre-logic code block that leverages the EndsWith function? –  Kirk Kuykendall Dec 8 '10 at 23:54
1. shapefile, I wish I knew how regex works. –  Brad Nesom Dec 9 '10 at 2:50
2. I am using the calculator in 10. but the pyth pre-logic I haven't looked for prior. I guess I will dig and see if I can figure anything out. BTW I am not a real programmer (pay no attention to that man behind the curtain!) –  Brad Nesom Dec 9 '10 at 2:52
add comment

2 Answers

up vote 10 down vote accepted

Yes Regular expressions are probably the best way to do this, and yes you can use regular expressions in the Calculate Field tool

First the regular expression

>>> import re
>>> val = 'National and Provincial capital'
>>> print re.sub('al$', 'X', val)
National and Provincial capitX

I'm subsituting "a" followed by "l" followed by "$" (which means the end of the sting) with the string 'X'. So 'capital' becomes 'capiX', while 'National' and 'Provincial' are unchanged.

Now use this same logic inside the code block for CalculateField. Set the calculateField parameters as follows -'Field Name" param to the output field name (i recommend calc'ing into a new field at least until you're sure the output is correct).

-'Expression' param to this

  • 'Expressioin Type' param to PYTHON

  • 'Code block' param to this


import re
def myroutine(val):
    return re.sub('al$', 'X', val)

Some docs if you want to read up.

share|improve this answer
The more generic form of this following the Brad's example would be re.sub(r'\s(H)$', \1, val). Which isn't really necessary as he states that value is always the same but I thought I'd throw that out there. –  Sean Dec 9 '10 at 15:44
Thanks gotchula for making it so easy for me. I did try it and found a bug in arcmap 10 (don't have a nim3 yet), I will work on getting whatever the problem is unkinked. I wish I had more time (I would spend it on regex, python, and gdal). –  Brad Nesom Dec 9 '10 at 16:44
I found that if you use the expression param in the code block and the code block in the expression area you get an error. then the resulting message continues popping up not sure how many times. When I did it right it worked the CHARM. Thanks gotchula –  Brad Nesom Dec 9 '10 at 20:24
add comment

So we 100% sure good old search & replace won't cut the mustard? Is the main issue, you also have values like:

'toast is Hot H and burnt'

  • And these kinds of attributes you want to ignore, but the search and replace can't be tweaked to exclude these attributes?

alt text

share|improve this answer
add comment

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.