I have Python script to parse XML files into a more friendly format for another platform.
Every so often one of the data files contains no data - only the encoding info and no other tags, which is causing ElementTree to throw a ParseError when it finds them.
<?xml version="1.0" encoding="utf-8"?>
Is there a way of testing for the empty file before calling ElementTree?
Ta.
You should ask for forgiveness not permission here.
Handle the exception by wrapping the code in a try/except block.
import xml.etree.ElementTree as ET
...
try:
tree = ET.parse(fooxml)
except ET.ParseError:
# log error
pass
Of course have several ways, use:
try:
pass # delete this and add your parse code
except:
pass # write your exception when empty
or use if statement:
if (some code to evalue if xml is not empty):
# your code
elif (some code to check if .xml is empty):
# your code
let me know how it was!
Of course you could catch the exception that lxml throws. If you want to avoid parsing, you could check if the file contains only one < symbol:
with open("input.xml","rb") as f:
contents = f.read()
if contents.count(b"<")<=1:
# empty or only header: skip
pass
else:
x = etree.XML(contents)
of course this heuristic method doesn't protect from other parsing errors. So it's best to just protect the parsing by a try/except block.
But this method has the advantage of being extremely fast if you have lots of corrupt 1-line "header only" file.
<character in data? if you find just 1 then file is probably empty.endswith("?>")?