I'm working on a Python script that involves selecting parcels from a feature class and I've run into some odd behavior in my sanity checks. In the code below I loop through the list of specified tax ID numbers and use the MakeTableView trick to get a count of elements matching a specific "where" clause. If I only give it one tax ID, the script finishes in a couple of seconds. If I give it two, it gets stuck on MakeTableView for about 4 minutes, and this seems to increase by about 4 minutes for each additional tax ID. I've tried explicitly deleting the resulting table view as soon as I'm finished with it, but that made no difference.
Ultimately, I changed the script to use a search cursor and the "where" clause to give me the same functionality, but this is just bouncing around the back of my head—why would the subsequent MakeTableView calls perform so much slower than the first?
Edit for clarity: TID is a list of tax ID's obtained by splitting a multi-value GetParameterAsText().
for tid in TID:
if tid and tid != "
if not re.match(pattern, tid):
raise GPException("Input Parcel IDs must be in the format " +
"YY-YYY-YYYY, where Y is a single digit number." +
" For example, 06-019-0009.")
where = TID_field + " = '" + tid + "'"
arcpy.MakeTableView_management(parcels, "in_memory/temp", where)
if arcpy.GetCount_management("in_memory/temp") < 1:
raise GPException("Cannot find parcel ID " + tid + " in parcel " +
"list.")
arcpy.Delete_management("in_memory/temp")
arcpy.SelectLayerByAttribute_management(parcels, "ADD_TO_SELECTION", where)
