I am trying to write any error messages that arise in my script and record it in a text file. When I check my errorLog, the error messages are not recorded. Below is my code: Note the parameter 'filePath' is the file path to the file that produced the error.

    addError = arcpy.GetMessages(2)
    with open(errorLog, 'a') as errorMsg:
        errorMsg.write('%s, %s\n' % (filePath,addError))

Can anyone offer insight?

  • so errorLog is a file path to the log file and filePath is a string that contains some useful information... seems right to me, the using statement ensures the file is open and closed so you shouldn't need a flush()... the only thing I can think of is perhaps there's an override for open that is unintentional (like GDAL.open or os.open), what imports do you have? Could the file be locked? – Michael Stimson Feb 19 '15 at 0:55
  • Hi, the imports are arcpy, os, time and datetime. mmm now that you have mentioned imports I think I am missing one - poss. tracebak. – TsvGis Feb 19 '15 at 2:20

I've written a python log handler that I'm more than happy to share. The idea being that you can use the standard Python logging framework and also have the messages reflected back to ArcGIS through the arcpy messages.

import logging
import logging.handlers

import arcpy

class ArcPyLogHandler(logging.handlers.RotatingFileHandler):
    Custom logging class that bounces messages to the arcpy tool window as well
    as reflecting back to the file.

    def emit(self, record):
        Write the log message
            msg = record.msg.format(record.args)
            msg = record.msg

        if record.levelno >= logging.ERROR:
        elif record.levelno >= logging.WARNING:
        elif record.levelno >= logging.INFO:

        super(ArcPyLogHandler, self).emit(record)

You can then use the handler with a standard python logging object:

logger = logging.getLogger("LoggerName")
handler = ArcPyLogHandler(
    maxBytes=1024 * 1024 * 2, #2MB log files
formatter = logging.Formatter("%(levelname)-8s %(message)s")

And then you can write log messages like so:

logger.debug("A debug message")
logger.info("An info message")
logger.warning("A warning message")
logger.error("An error message")
logger.critical("A critical error message")

Which will write the messages of levels info and above through arcpy to the processing window, and all messages from debug and above to the log file specified with the format specified above (see the documentation for other formatting options).

Lastly you can use logger.exception in an except block which will give you a logging message at ERROR level, and also dump the stacktrace to the log.

    #Something that might fail
except Exception as e:
  • 1
    Your codes worked the first time but failed subsequently in ArcMap. It threw out "NameError: global name 'logging' is not defined" on line "if record.levelno >= logging.ERROR:". I think the log file was being locked by ArcMap after the first run. – alextc Jan 21 '16 at 22:48

A simple test script:

errorLog = r'd:\log.txt'
filePath = errorLog
    with open(errorLog,'a') as errorMsg:
        errorMsg.write("%s,%s\n" % (errorLog,filePath))
except RuntimeError:
    print("Unable to log")
    print RuntimeError.message

Note: no imports, it's all internal..

This works from a pythonic sense which makes me wonder what's up with yours not working.

A more complicated script using arcpy:

import arcpy

errorLog = r'd:\log.txt'
filePath = errorLog
    arcpy.Delete_management("Data doesn't exist, lets' just force the exception")
    arcpy.AddMessage(arcpy.GetMessages(2)) # see if there is some error there
        with open(errorLog,'a') as errorMsg:
            errorMsg.write("%s,%s\n" % (errorLog,arcpy.GetMessages(2)))
    except RuntimeError:
        arcpy.AddMessage("Unable to log")

puts this in the text file:

d:\log.txt,Failed to execute. Parameters are not valid. ERROR 000732: Input Data Element: Dataset Data doesn't exist, lets' just force the exception does not exist or is not supported Failed to execute (Delete).

So, there's nothing wrong with your code, you will need to look elsewhere. Note the 2nd try in the except, it may not log to a file but at least you should get an error message on screen; from here it's likely to be 'file not found' or 'unable to access' that's causing the logging to fail.


thanks for the responses guys. I am still not 100% why the original code didn't work but I ended up going with a modified version of the trace back code shown here. In the end my coded ended up looking like this (see below), which produced the results that I was after:

import sys
errorLog = r'd:\log.txt'
filePath = errorLog
'''Code not included in example'''
    with open(LOG_FILE, 'a') as errorMsg:
        errorMsg.write('%s, %s\n' % (filePath, str(sys.exc_info()[1])))

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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