Nokogiri: How to get node name with namespace prefix - ruby

Question

i trying (for testing purpose) to parse Google merchant XML feed, defined as:
<?xml version="1.0" encoding="UTF-8"?>
<feed xml:lang="cs" xmlns="http://www.w3.org/2005/Atom" xmlns:g="http://base.google.com/ns/1.0">
<link rel="alternate" type="text/html" href="http://www.example.com"/>
<link rel="self" type="application/atom+xml" href="http://www.example.com/cs/feed/google.xml"/>
<title>EasyOptic</title>
<updated>2014-08-01T16:31:11Z</updated>
<entry>
<title>Sluneční Brýle Producer 1 133a code_color_1 Color 1 133a RayBan</title>
<link href="http://www.example.com/cs/katalog/price-category-1-style-1-optical-glasses-producer-1-rayban-133a-code_color_1-color-1"/>
<summary>Moc krásný a velmi levný produkt</summary>
<updated>2014-08-01T16:31:11Z</updated>
<g:id>EO111</g:id>
<g:condition>new</g:condition>
<g:price>100 Kč</g:price>
<g:availability>in stock</g:availability>
<g:image_link>http://www.example.com/images/fallback/default.png</g:image_link>
<g:additional_image_link>http://www.example.com/images/fallback/default.png</g:additional_image_link>
<g:brand>Producer 1</g:brand>
<g:mpn>EO111</g:mpn>
<g:gender>female</g:gender>
<g:google_product_category>Apparel & Accessories > Clothing Accessories > Sunglasses</g:google_product_category>
<g:product_type>Sluneční Brýle </g:product_type>
</entry>
<entry>
<title>Sluneční Brýle Producer 1 133a code_color_1 Color 1 133a RayBan</title>
<link href="http://www.example.com/cs/katalog/price-category-1-style-1-optical-glasses-producer-1-rayban-133a-code_color_1-color-1"/>
<summary>Moc krásný a velmi levný produkt</summary>
<updated>2014-08-01T16:31:10Z</updated>
<g:id>EO111</g:id>
<g:condition>new</g:condition>
<g:price>100 Kč</g:price>
<g:availability>in stock</g:availability>
<g:image_link>http://www.example.com/images/fallback/default.png</g:image_link>
<g:additional_image_link>http://www.example.com/images/fallback/default.png</g:additional_image_link>
<g:brand>Producer 1</g:brand>
<g:mpn>EO111</g:mpn>
<g:gender>female</g:gender>
<g:google_product_category>Apparel & Accessories > Clothing Accessories > Sunglasses</g:google_product_category>
<g:product_type>Sluneční Brýle </g:product_type>
</entry>
</feed>
with this ruby script:
require 'nokogiri'
def have_node_with_children(body, path_type, path, children_names)
doc = Nokogiri::XML(body)
case path_type
when :xpath
nodes = doc.xpath(path)
when :css
nodes = doc.css(path)
else
nodes = doc.xpath(path)
end
nodes.each do |node|
nchildren_names=[]
for child in node.children
nchildren_names << child.name unless child.to_s.strip =="" #nokogiri takes formating spaces as blank node with name "text"
end
puts("demanded_nodes: #{children_names.sort.join(", ")} , nodes found: #{nchildren_names.sort.join(", ")} ")
missing = children_names - nchildren_names
over = nchildren_names - children_names
puts("Missing: #{missing.sort.join(", ")} , Over: #{over.sort.join(", ")} ")
end
end
EXPECTED_ENTRY_NODES=[
'title',
'link',
'summary',
'updated',
'g:id',
'g:condition',
'g:price',
'g:availability',
'g:image_link',
'g:additional_image_link',
'g:brand',
'g:mpn',
'g:gender',
'g:google_product_category',
'g:product_type'
]
file=File.open('google.xml')
have_node_with_children(file.read,:xpath,'//xmlns:entry',EXPECTED_ENTRY_NODES)
It find node 'entry' (thanks for this tip ).
But when collecting it's children method child.name returns name without namespace prefix (e.g.: <'g:brand'>.name => 'brand'.
So comparsion with demanded fields fail.
Do anybody have tip hot to get node name with/and it's namespace prefix?
If I delete namespace definitions all work fine, but I cannot change the original XML.
I use this test in rspec request test, so another namespaces with maybe indentical base node names can appear.

Answer 1

xml_doc = Nokogiri::XML(xml)
xml_doc.xpath("//xmlns:entry").each do |entry|
entry.xpath("./*").each do |element| #Step through all Element nodes that are direct children of <entry>
prefix = element.namespace.prefix
puts prefix ? "#{element.namespace.prefix}:#{element.name}"
: element.name
end
break #only show output for the first <entry>
end
--output:--
title
link
summary
updated
g:id
g:condition
g:price
g:availability
g:image_link
g:additional_image_link
g:brand
g:mpn
g:gender
g:google_product_category
g:product_type
Now about this:
for child in node.children
A well grounded rubyist does not ever use a for-loop...because a for_loop just calls each(), so rubyists call each() directly:
node.children.each do |child|