how to replace a text on XML with lxml? - python-2.7

Question

I'm trying to troncate several element.text on a xml file. I succeed to get two list, the first one regroup the formers too long element.text as str (long_name) and the second regroup the same after a troncation (short_name).
Now i want to replace the element.text on my xml, i tried some script but i surrended to work with the function readlines(), i want to find a similar solution with lxml as this code :
txt = open('IF_Generic.arxml','r')
Lines = txt.readlines()
txt.close()
txt = open('IF_Genericnew.arxml','w')
for e in range(len(long_name)) :
for i in range(len(Lines)) :
if (long_name[e] in Lines[i]) == True :
Lines[i] = Lines[i].replace(long_name[e],short_name[e])
for i in Lines :
txt.write(i)
txt.close()
I tried this, but it doesn't work :
f = open('IF_Generic.arxml')
arxml = f.read()
f.close()
tree = etree.parse(StringIO(arxml))
for e,b in enumerate(long_name) :
context = etree.iterparse(StringIO(arxml))
for a,i in context:
if not i.text:
pass
else:
if (b in i.text) == True :
i.text = short_name[e]
obj_arxml = etree.tostring(tree,pretty_print=True)
f = open('IF_Genericnew.arxml','w')
f.write(obj_arxml)
f.close()
Let's say the first element of the list long_name is RoutineServices_EngMGslLim_NVMID03
<BALISE_A>
<BALISE_B>
<SHORT-NAME>RoutineServices_EngMGslLim_NVMID03</SHORT-NAME>
</BALISE_B>
</BALISE_A>
<BALISE_C>
<POSSIBLE-ERROR-REF DEST="APPLICATION-ERROR">/Interfaces/RoutineServices_EngMGslLim_NVMID03/E_NOT_OK</POSSIBLE-ERROR-REF>
<SHORT-NAME>Blah_Bleh_Bluh</SHORT-NAME>
</BALISE_C>
The first element of the list short_name is RoutineServices_EngMGslLim_NV
<BALISE_A>
<BALISE_B>
<SHORT-NAME>RoutineServices_EngMGslLim_NV</SHORT-NAME>
</BALISE_B>
</BALISE_A>
<BALISE_C>
<POSSIBLE-ERROR-REF DEST="APPLICATION-ERROR">/Interfaces/RoutineServices_EngMGslLim_NV/E_NOT_OK</POSSIBLE-ERROR-REF>
<SHORT-NAME>Blah_Bleh_Bluh</SHORT-NAME>
</BALISE_C>
I want this
P.S: I use python 2.7.9
Thanks in advance everyone !

Answer 1

Don't open XML files like text files. I have explained in this answer why this is a bad idea.
Simply let etree read and write the file. It's also less code to write.
from lxml import etree
# read the file and load it into a DOM tree
tree = etree.parse('IF_Generic.arxml')
for elem in tree.iterfind("//*"):
# find elements that contain only text
if len(elem) == 0 and elem.text and elem.text.strip() > '':
# do your replacements ...
elem.text = "new text"
# serialize the DOM tree and write it to file
tree.write('IF_Genericnew.arxml', pretty_print=True)
Instead of going over all elements, which is what "//*" does, you can use more specific XPath to narrow down the elements you want to work on.
For example, something like "//SHORT-NAME | //POSSIBLE-ERROR-REF" would help to reduce the overall work load.