Error 999999 on FeatureClass To FeatureClass conversion in ArcPy?

Question

I am currently working on a python script where I am trying to copy selected features into a feature dataset.

Basically I have an mxd containing some layers, and for each layer I select features by their attributes, this query is based on the featuredataset name. The script I have written get a list of the feature datasets of my geodatabase. With the name of the featuredataset, I can build a query filter (arcpy.SelectLayerByAttribute_management) for each layers of the mxd. If my selection return something, I export the selected feature in the proper dataset (arcpy.FeatureClassToFeatureClass_conversion). The script is processing the first dataset properly, but once it jumps to the next one it fails... I cannot figuring out why I am receiving the following message error(arcgisscripting.ExecuteError: ERROR 999999: Error executing function. Failed to execute (FeatureClassToFeatureClass).)

#-----Work-on-the-Geodatabase-----
##path of the geodatabase 
gdb_hull="D:\\As_Con_Plans\\As_Con_Hull_part1.gdb"
##set the workspace
arcpy.env.workspace=gdb_hull
##get the list of the featuredataset
fdlist =arcpy.ListDatasets()
#-----Work-on-the-mxd-----
##get the mxd path and initialise it
path_mxd="D:\\As_Con_Plans\\As_constructed.mxd"
doc_mxd = arcpy.mapping.MapDocument(path_mxd)
##list of layers
fclist=arcpy.mapping.ListLayers(doc_mxd)
#-----Loop-on-each-featuredataset----
for fds in fdlist:
 print "________","\n ", fds
 ##work on the featuredataset name, for suiting the query
 get_fds= fds.replace("As_con_","")
 fds_name=get_fds.replace("_","-")
 ##path of the featuredataset
 fds_path=gdb_hull+"\\"+fds
 ##prepare the query
 SQLexpression= "\"As_Constructed_Plan\"='"+fds_name+"'"
 ##path where we will store the selection
 fds_path_test=str(fds_path)
 #-----Loop-on-each-layers-of-the-mxd-----
 for fc in fclist:
 fc_to_copy=fc.dataSource
 #fc_to_copy2=arcpy.mapping.Layer(fc_to_copy)
 ##Select the featureclasses
 arcpy.SelectLayerByAttribute_management (fc,"NEW_SELECTION",SQLexpression)
 ##Count the selected features
 count = int(arcpy.GetCount_management(fc).getOutput(0))
 print fc,count
 ##if the query succeed, we will copy the selection into the proper featuredataset
 if count>0:
 ##print some info
 fc_path2=fds_path+"\\"+str(fc)
 fc_path=str(fc_path2)
 fc_name=str(fc)
 fc_copy_from=str(fc_to_copy)
 ##check if the featureclass does not exist
 if not arcpy.Exists(fc_path):
 print "\n copying...",count," entities , on ",fc_path
 print (fc_copy_from,fds_path_test,fc_name,SQLexpression)
 arcpy.FeatureClassToFeatureClass_conversion(fc_to_copy,fds_path_test,fc_name,SQLexpression)
 print "export ok"

• Michael Stimson: Are you familiar with try..except blocks? Is it possible that the next feature class has the same name as the first? You can test with arcpy.exists(fds_path_test+"\\"+fc_name) see resources.arcgis.com/en/help/main/10.2/index.html#//…. A copy of the outputs may help (turn on quick edit in the command window, select the area and right click then paste into notepad) or screen grab.
• Michael Stimson: It looks like you are looping through the output feature datasets and storing the feature class more than once (once in each feature dataset) hence the error - names must be unique in the database, not just the dataset. If there is only one feature dataset in the output database gdb_hull = "D:\\As_Con_Plans\\As_Con_Hull_part1.gdb" then I'll need to look harder. Check also for spaces and non-alphanumeric characters in the layer names, that will also cause that error.

Answer 1

Finally, I figured out that my problem was linked to the name of the feature class' output. It is the fc_name of the following line:

arcpy.FeatureClassToFeatureClass_conversion(fc_to_copy,fds_path_test,fc_name,SQLexpression).

I have run some test and discovered that I can't add 2 similar output name... even if it is processing a different dataset... So to fix the problem, I have created an incremented index that I linked to the feature class name (new_fc_name=fc_name+str(i) with I incremented after a loop). It may not be the best way to process, but at least, it is working...

Answer 2

You can have 2 check:

1. Check if name is valid

   vname = arcpy.ValidateTableName(fc_name,wksp)

2. Check if name exist

   def validUniqueName(name, wk):
    """ Create a unique name not exist in GDB
    Note: you can fix limit to number of pas and add test if name is too long"""
    check = True
    while check:
    check = arcpy.Exists(os.path.join(wk, name))
    if check:
    name = "{}_1".format(name)
    return name
   

Add 2 lines in function:

vname = arcpy.ValidateTableName(fc_name,wksp)
v_unique_name = validUniqueName(vname,wksp)

Test

>>> name = u"Voirie etroite"
>>> name
u'Voirie etroite'
>>> vname = arcpy.ValidateTableName(name,wk)
>>> vname
u'Voirie_etroite'
>>> vuname = validUniqueName(vname,wk)
>>> vuname
u'Voirie_etroite_1'