Python XML DOM Minidom

Usage

Parsing a file

from xml.dom import minidom
document = minidom.parse(filename)

If the XML file uses namespaces, it can be easier to disable that feature in the parser.

from xml.dom import minidom, expatbuilder
document = expatbuilder.parse(filename, False)

Traverse all nodes

def recurse_print(node):
    if node.nodeType == minidom.Node.TEXT_NODE:
        print(node.data)
    else:
        for child in node.childNodes:
            recurse_print(child)

recurse_print(document)

Scrape HTML tables

def recurse_text(node):
    buffer = ""
    for child in node.childNodes:
        if child.nodeType == minidom.Node.TEXT_NODE:
            buffer += child.data
        else:
            buffer += recurse_text(child)
    return buffer

for table in document.getElementsByTagName("table"):
    for row in table.getElementsByTagName("tr"):
        data.append([])
        for header_cell in row.getElementsByTagName("th"):
            data[0].append(recurse_text(header_cell))
        for cell in row.getElementsByTagName("td"):
            data[-1].append(recurse_text(cell))

Scrubbing the DOM

It can be useful to scrub the DOM of unhelpful or useless components.

To remove attributes, try:

if node.hasAttribute("hidden"):
    node.removeAttribute("hidden")

To remove nodes, try:

for child in node.childNodes:
    if child.hasAttribute("hidden"):
        node.removeChild(child)
        child.unlink()

To replace nodes, as with comments, try:

replacement = document.createComment("scrubbed useless node")
# alternatively, createTextNode or createElement
for child in node.childNodes:
    if child.hasAttribute("hidden"):
        node.replaceChild(child, replacement)

CategoryRicottone