I'm using the PythonCaller to write lines into the log message.
I use logMessage String and with data that has mutated vowels I get following error message:
Python Exception <UnicodeEncodeError>: 'ascii' codec can't encode character u'\\xfc' in position 67: ordinal not in range(128)
What can I do that I can write also data with ä,ü,ö to the log message?
Thanks for your help!
Isabell
Hi Isabell,
I also often encounter such a situation. My environment is Japanese locale.
My workaround is to test whether the character string is a unicode instance or not, and to encode the string by the default encoding if it's unicode.
# Example
import fmeobjects
import locale
class FeatureProcessor(object):
def __init__(self):
loc = locale.getdefaultlocale()
self.enc = loco1] # save the defualt encoding
self.logger = fmeobjects.FMELogFile()
def input(self, feature):
s = feature.getAttribute('attr')
if isinstance(s, unicode):
self.logger.logMessageString(s.encode(self.enc))
else:
self.logger.logMessageString(str(s))
def close(self):
passI'm not sure whether this workaround works fine in your locale.
Hope more suitable solution will be provided.
Reference link: Unicode HOWTO http://docs.python.org/2/howto/unicode.html
Takashi
it's a bit tricky, because text attributes in FME are usual encoded as Unicode, where as the logMessageString() function seems to expect Latin1 encoding, for some unknown reason.
But try this:
log_text = feature.getAttribute("my_text").encode("latin1")
FMELogFile().logMessageString(log_text, fmeobjects.FME_WARN)
The important part is in bold. Please let us know if this works for you.
David
I like your more general approach. Thanks for sharing!
David
I tried Takashi's suggestion and it works fine!!!
Thank you both very much for your help!
Regards
Isabell
Start your python script with the coding and for each variable you can add (@var).encode('utf-8')
# -*- coding: utf-8 -*-Hi there, @isabell, @david_r, @julien,
I found this description in the FME 2016.1.1.0 "What's New" list.
-----------------------------
FME 2016.1.1.0 16593 20160620
-----------------------------
Python: Modified logMessage() and getMessage() to allow mixed unicode and str values in the message parameters. Fixed encoding of source code passed from PythonFactory to Python interpreter. (PR#70683)
import fmeobjects
class FeatureProcessor(object):
def __init__(self):
self.logger = fmeobjects.FMELogFile()
def input(self, feature):
msg = feature.getAttribute("attr")
if isinstance(msg, unicode): # check if the value is a unicode instance
self.logger.logMessageString("the value is a unicode instance")
self.logger.logMessageString(msg)
Hi there, @isabell, @david_r, @julien,
I found this description in the FME 2016.1.1.0 "What's New" list.
-----------------------------
FME 2016.1.1.0 16593 20160620
-----------------------------
Python: Modified logMessage() and getMessage() to allow mixed unicode and str values in the message parameters. Fixed encoding of source code passed from PythonFactory to Python interpreter. (PR#70683)
import fmeobjects
class FeatureProcessor(object):
def __init__(self):
self.logger = fmeobjects.FMELogFile()
def input(self, feature):
msg = feature.getAttribute("attr")
if isinstance(msg, unicode): # check if the value is a unicode instance
self.logger.logMessageString("the value is a unicode instance")
self.logger.logMessageString(msg)