Python XML parsing, extraction and renaming files

Question

I've written some code to:

1. Parse an XML file.
2. Extract some strings.
3. Make copies of the original XML whose names are suffixed with the extracted strings.

Please review this.

from lxml import etree as ET
import re

def create_paramsuffix_copies(xmlfile):

    NSMAP = {"c": "http://www.copasi.org/static/schema"}

    parsed = ET.parse(xmlfile)

    list_of_params = []

    for a in parsed.xpath("//c:Reaction", namespaces=NSMAP):
        for b in a.xpath(".//c:Constant", namespaces=NSMAP):
            list_of_params.append((a.attrib['name'])+'_'+(b.attrib['name']))

    return list_of_params


def add_suffix(xmlfile, name_suffix):

    parsed = ET.parse(xmlfile) #parse xml

    xmlfile = re.sub('.cps', '', xmlfile) #remove suffix

    parsed.write(xmlfile+'_'+name_suffix+'.cps', 
                 xml_declaration = True,
                 encoding = 'utf-8',
                 method = 'xml')


modelfile = 'ct.cps'

for a in create_paramsuffix_copies(modelfile):
    add_suffix(modelfile,a)

Answer 1

Some style changes can happen, most are described here.

• Using single letter variable names can be hard to understand, and don't explain what the data is well. It is good not to use them.
• str.format is the best way to put variables into strings. E.G. '{}_{}.cps'.format(xmlfile, name_suffix)
• CONSTANTS should be all caps. MODEL_FILE = 'ct.cps'
• When passing 'kwargs' don't put spaces around the = operator. E.G. encoding='utf-8'
• Use new-lines sparingly.
• In-line comments are harder to read then comments before the code.
• Comments should have a space between the '#' and the comment.

---

It may be good to change create_paramsuffix_copies to a generator, in this case it should be faster than making a list.

def create_paramsuffix_copies(xmlfile):
    NSMAP = {"c": "http://www.copasi.org/static/schema"}
    parsed = ET.parse(xmlfile)
    for a in parsed.xpath("//c:Reaction", namespaces=NSMAP):
        for b in a.xpath(".//c:Constant", namespaces=NSMAP):
            yield '{}_{}'.format(a.attrib['name'], b.attrib['name'])

---

You could add docstrings to your program so people know what the functions do at a later date.

But it looks really good!