Splitting string ID code into various parts

Question

I have a series of identification codes that I need to split out. The format of these codes is [region(letter)][district(number)] - [place(number)][subdistrict(letter)].

An example of some codes includes S22-201, TT100-12, and V6-1B. Often there is no subdistrict, and all points fall within the same larger district (so no As or Cs or whatever at the end of the string.

I can do parts of it, like splitting at the hyphen.

!Original_ID!.split('-')[0]

and then extracting the district

!Split_ID![1:3]

But it seems like two steps for this are unnecessary, and only works when I know the specific number of characters in the string, which isn't realistic for a large data set.

I'd like to be able to grab each piece at once:

• letters on the left of the hyphen
• numbers on the left of the hyphen
• numbers on the right of the hyphen
• letters (if any) on the right of the hyphen.

I'd need the numeric fields to be integers (or I guess possibly floats in some rare cases maybe).

---

I am still not doing something correctly. I may need to start smaller and brush up on my Python before I do this, I just assumed this would be a good place to start learning. Here's where I am at, in the Python window in ArcMap.

with arcpy.da.UpdateCursor("Wet_Sub",['Flag_ID','District','Split_ID']) as uCur:
for sRow in uCur:
    OrigID  = sRow[0].split('-')[0] # first element in the Original_ID
    charRng = range(len(OrigID)) # a range to iterate over
    Chars   = ''
    Numbers = ''
    for Idx in charRng:
        if OrigID[Idx].isnumeric():
            Numbers += OrigID[Idx]
        else:
            chars += OrigID[Idx]
    sRow[1] = float(Numbers)
    sRow[2] = Chars
    uCur.updateRow(sRow)

"Wet_Sub" and 'Flag_ID' are the names of the feature class and actual original field. I also tried to follow along with user2856's suggestion. It looks like I may need to be using both of those code blocks, one pasted into another, but I wasn't sure how to fit them together and what parts to change/remove (e.g. "etc... from code block above").

Answer 1

You're not going to be able to calculate two fields in one go.. though you can split it up into two calcs. I would do this with an update cursor:

with arcpy.da.UpdateCursor(YourFeatureClass,['Original_ID','District','Split_ID']) as uCur:
    for sRow in uCur:
        OrigID  = sRow[0].split('-')[0] # first element in the Original_ID
        charRng = range(len(OrigID)) # a range to iterate over
        Chars   = ''
        Numbers = ''
        for Idx in charRng:
            if OrigID[Idx].isnumeric():
                Numbers += OrigID[Idx]
            else:
                chars += OrigID[Idx]
        sRow[1] = float(Numbers)
        sRow[2] = Chars
        uCur.updateRow(sRow)

This shows how to break up a string into numbers and not numbers and put the values into a row, it should give you some ideas where to start from.

Answer 2

Assuming you have four fields, region, district, place and subdistrict already added and you want to use the field calculator to populate them. You would have to run the calculator four times using an expression like:

Code Block

import re
def parse(s):
    """The format of these codes is [region(letter)][district(number)] - [place(number)][subdistrict(letter)].
    An example of a some codes include S22-201, TT100-12, and V6-1B.
    Often there is no subdistrict, and all points fall within the same larger district
    (so no As or Cs or whatever at the end of the string)."""

    letters = re.findall(r'[a-z A-Z]+', s)
    numbers = re.findall(r'[0-9]+', s)

    region = letters[0]
    district, place = [int(n) for n in numbers]
    try:
        subdistrict = letters[1]
    except IndexError:
        subdistrict = None

    return region, district, place, subdistrict

Then for the region field, use:

parse(!Original_ID!)[0]

For district:

parse(!Original_ID!)[1]

For place:

parse(!Original_ID!)[2]

For subdistrict:

parse(!Original_ID!)[3]

However, I would use the update cursor approach suggested by Michael Stimson so you could update all four fields in one hit. Use the following in the python window of ArcMap/ArcGIS Pro:

import re 
def parse(s): 
    etc... from code block above

with arcpy.da.UpdateCursor(YourFeatureClass, ['Original_ID','Region', 'District', 'Place', 'Subdistrict']) as rows:
    for row in rows:
        region, district, place, subdistrict = parse(row[0])
        row = [row[0], region, district, place, subdistrict]
        rows.updateRow(row)