topic Re: AppendManagement failing because of field widths in Python Questions

AppendManagement failing because of field widths

Anonymous User — Wed, 23 Mar 2022 15:38:31 GMT

I have a script running that is updating an SDE feature class from a parcel fabric feature class..

The script is failing because two fields in the target database are not wide enough.

Field 1 - 50 in the source , 20 in the target

Field 2 - 2000 in the source 25 in the target

I've tested and verified that appendmanagement will run fine if I alter the field widths in the target database.

I'd rather not do that and was wondering if there is a way to have appendmanagment just truncate the values at 20 and 25 when it does the append.

Feedback is greatly appreciated!

Re: AppendManagement failing because of field widths

DonMorrison1 — Wed, 23 Mar 2022 16:54:37 GMT

Did you try schema_type="NO_TEST" and the arcpy.Append_management call? This allows appends when schemas don't match, however I suspect it will still fail if you have a value in one of those two fields that does not fit in the target column. But it might be worth a try.

Re: AppendManagement failing because of field widths

Anonymous User — Wed, 23 Mar 2022 20:31:52 GMT

Thanks for that suggestion. In a test database I altered the widths of the fields to match the source database. It's not ideal but i figured I could at least test the results.

Unfortunately the field calc part of the script is failing now.

The fields as exported from the parcel feature class ended up being called

SHAPE.STArea() and SHAPE.STLength()

I can't seem to get the correct syntax to add two new fields (AREA and PERMIETER) and calculate the new fields to these fields before deleting them.

When I call AddField it's naming the field Database_SCHEMA_featureclass.AREA instead of just AREA.

Here are the lines:
arcpy.AddField_management(feature_lyr_out , "AREA", "DOUBLE", "", "", "", "", "NULLABLE", "NON_REQUIRED", "")
arcpy.CalculateField_management(feature_lyr_out , "AREA", "[!SHAPE.STArea()!]", "PYTHON_9.3", "")
arcpy.DeleteField_management(feature_lyr_out , "![SHAPE.STArea()]!")

I'm wondering if the system is having trouble resolving the field names that include . and () characters.

Feedback is appreciated.

Re: AppendManagement failing because of field widths

Anonymous User — Wed, 23 Mar 2022 20:40:54 GMT

These are the fields:

Re: AppendManagement failing because of field widths

DonMorrison1 — Wed, 23 Mar 2022 21:19:44 GMT

Try using "!shape.area!" in your expression instead of "![SHAPE.STArea()]!".

I'm not sure why it is prepending the field with the schema name but perhaps "area" is a reserved name and you should try something else like "my_area".

Re: AppendManagement failing because of field widths

Anonymous User — Thu, 24 Mar 2022 21:52:27 GMT

I believe the NO_TEST will still fail at the DB level.

Re: AppendManagement failing because of field widths

Anonymous User — Thu, 24 Mar 2022 21:56:33 GMT

It's failing because "area" is a reserved keyword in an SDE table. You'll have to name it "my_area" (or something) to add that field without the inserted qualifications

Re: AppendManagement failing because of field widths

Anonymous User — Thu, 24 Mar 2022 22:22:49 GMT

To answer your original question - One way around this is something like this...

fld1 = "your field that's too long" fld2 = "another" desc = arcpy.Describe(table) for fld_obj in desc.fields() if fld.name == fld1: fld1_len = fld.length if fld.name == fld2: fld2_len = fld.length with arcpy.da.UpdateCursor("your table", ["fld1", "fld2"]) as uCur: for uRow in uCur: uRow[0] = value_too_long[:fld1_len] uRow[1] = value_too_long[:fld2_len] uCur.updateRow(uRow)

the value_too_long[:fld2_len] is a string slice that goes up to the length identified in the first part. (it may need a "fld2_len - 1" if it's still too long.

disclaimer: this is off the top of my head which is more 2.7 based. Also, getting the value_too_long value is hard for me to guess since I don't know where it's coming from. if it's another field inthe same table, you would add it to the list of fields in the UpdateCursor initialization

edit: since you already know their lengths, you could skip the first section and just put it in the value_too_long

for fld_obj in desc.fields()