3

I am trying to merge files that are nested in different folders by their file names. For example folder 1 contains up to 100 subfolders that each contain a set of shapefiles: subfolder1_A.shp, subfolder1_B.shp, subfolder1_C.shp, etc.

I want to merge shapefiles from each subfolder with the same file name ending (e.g. merge subfolder1_A, subfolder2_A, subfolder3_A, ..., subfolder100_A), so the final output is a series of merged shapefiles with the same file name ending (A, B, C, etc.). So for example, A_merged.shp, B_merged.shp, C_merged.shp, etc.

I am not familiar with python, and would prefer to do this using model builder. I have tried using iterate feature class and merge tools, but not sure how to make this work like I want. Possibly using inline substitution or the iterator wildcard somehow?

Any idea what I should do?

2
  • I'd try an os crawl with a list feature class and a name test. Are you wanting to only work with shapefiles? or do you have fGDB with featuredatasets?
    – GISI
    May 25, 2015 at 8:26
  • only shapefiles.
    – macdonaw
    May 25, 2015 at 13:06

1 Answer 1

2

Do an OS walk to iterate all folders and sub folders. For each folder list all the feature class. Split the name of the feature class at your underscore and compose a list, then process that list so the names are unique. Use those unique names as a wild card while relisting all the feature class in the same folder. Lastly run it through the merge tool. I wired the below py into a toolbox script tool and it works OK for a single like geometry.

import arcpy
import os
import math
import sys

FolderOrWorkspace = arcpy.GetParameterAsText(0)
arcpy.AddMessage("======= START")

for dirname, dirnames, filenames in os.walk(FolderOrWorkspace): #Iterate through folder
    for subdirname in dirnames:
        arcpy.env.workspace = os.path.join(dirname, subdirname) # Workspace set to folder
        UList =[]
        arcpy.AddMessage(" ")
        arcpy.AddMessage("      - Folder: "+ subdirname)
        FeatureClassList = arcpy.ListFeatureClasses()
        arcpy.AddMessage("      - FEATURE CLASS LIST: ")
        arcpy.AddMessage(" ")
        arcpy.AddMessage(FeatureClassList)
        arcpy.AddMessage(" ")
        for ShortName in FeatureClassList: # Make a list of all the feature names after the the first "_"
            LastName = ShortName.split("_")[1]
            UList.append(LastName)
        arcpy.AddMessage("      - Short Name List: "+ str(UList))
        UniqueShortName = set(UList) # Process the list to be unique short names only
        arcpy.AddMessage("      - Unique Short Name List: "+ str(UniqueShortName))

        for mergetest in UList: # create a list of feature class ending in the short name
            NameWildCard = ("*_"+mergetest)
            MergeFeatureClassList = arcpy.ListFeatureClasses(NameWildCard,'',FeatureClassList)
            MFCLCount = len(MergeFeatureClassList)
            arcpy.AddMessage("      - Unique Short Name Count: "+ str(MFCLCount))

            if MFCLCount <= 1:
                arcpy.AddMessage("      - Only 1 Feature Class: Not Merged")
            else:
                MergeName = ("Merged"+mergetest) # the name of the merge shapefile

                try:
                    arcpy.AddMessage("      - Merging: " + str(MergeFeatureClassList))
                    arcpy.AddMessage(" ")
                    arcpy.Merge_management (MergeFeatureClassList, MergeName,"")    # Saves it in the same folder                    # NB Merge is an Advanced tool
                except Exception as e:
                    arcpy.AddError("       "+e.message)

arcpy.AddMessage("======= END")

You might want to add a geom type check..

Screen Grab: Shapefiles

enter image description here

Script Tools: Adding Script Tools

2
  • Thanks for this response. Exactly what I need except is there any way to do this without using python? i.e. Model Builder?
    – macdonaw
    May 27, 2015 at 15:49
  • I don't think so. The key problem is the iteration and processing of the names. But I could be wrong.
    – GISI
    May 27, 2015 at 21:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

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