Strip Font Tags from XML with nokogiri

I'm trying to clean up some xml by removing Font Tags. Here's what I start with:

<?xml version="1.0"?>
 <Worksheet ss:Name="Subtitles">
  <Table ss:ExpandedColumnCount="3" ss:ExpandedRowCount="53" x:FullColumns="1"
   x:FullRows="1" ss:DefaultRowHeight="13.5">
   <Column ss:StyleID="s62" ss:Width="80.25" ss:Span="1"/>
   <Column ss:Index="3" ss:StyleID="s63" ss:Width="249.75"/>
   <Row ss:AutoFitHeight="0">
    <Cell><Data ss:Type="String">00:00:11:09</Data></Cell>
    <Cell><Data ss:Type="String">00:00:13:06</Data></Cell>
    <Cell><ss:Data ss:Type="String" xmlns="http://www.w3.org/TR/REC-html40">안녕하세요<Font
       html:Face="Arial" html:Color="#000000">, </Font><Font html:Face="돋움"
       html:Color="#000000">저는</Font><Font html:Face="Arial" html:Color="#000000"> </Font><Font
       html:Face="돋움" html:Color="#000000">잭</Font><Font html:Face="Arial"
       html:Color="#000000">, 9</Font><Font html:Face="돋움" html:Color="#000000">살</Font><Font
       html:Face="Arial" html:Color="#000000"> </Font><Font htm开发者_JAVA百科l:Face="돋움"
       html:Color="#000000">입니다</Font><Font html:Face="Arial" html:Color="#000000">. </Font></ss:Data></Cell>
   </Row>
  </Table>
 </Worksheet>
</Workbook>

And here's what I want:

<?xml version="1.0"?>
 <Worksheet ss:Name="Subtitles">
  <Table ss:ExpandedColumnCount="3" ss:ExpandedRowCount="53" x:FullColumns="1"
   x:FullRows="1" ss:DefaultRowHeight="13.5">
   <Column ss:StyleID="s62" ss:Width="80.25" ss:Span="1"/>
   <Column ss:Index="3" ss:StyleID="s63" ss:Width="249.75"/>
   <Row ss:AutoFitHeight="0">
    <Cell><Data ss:Type="String">00:00:11:09</Data></Cell>
    <Cell><Data ss:Type="String">00:00:13:06</Data></Cell>
    <Cell><Data ss:Type="String">안녕하세요, 저는잭, 9살 입니다.</Data></Cell>
   </Row>
  </Table>
 </Worksheet>
</Workbook>

I'm trying to accomplish this with nokogiri. Here's what I've got so far:

require 'nokogiri'
f = File.open("junk_from_excel.xml")
doc = Nokogiri::XML(f)

# <Font html:Face="돋움" html:Color="#000000">, </Font>
doc.xpath('//@Face').each(&:remove)
# becomes: <Font html:Color="#000000">, </Font>
doc.xpath('//@Color').each(&:remove)
# becomes: <Font>, </Font>

puts doc

I'm not sure that helps at all. I still need to turn:

<Font>, </Font>

into:

, 

Any help is much appreciated!

Thank you all for your help. I got it working with this hackish gsub stuff at the end:

require 'nokogiri'

f = File.open("full_file_from_excel.xml")
doc = Nokogiri::XML(f)

def font_killer(children)
    children.each do |c|
        if(c.name == 'Font')
            c.replace(font_killer(c.children))
        else
            font_killer(c.children)
        end
    end
    children
end
doc = Nokogiri::XML(open('full_file_from_excel.xml').read)
doc.encoding = 'utf-8'
doc.xpath('//ss:Data[@ss:Type="String"]|//Data[@ss:Type="String"]').each { |n| font_killer(n.children) }

#save to file:
open('full_file_from_excel_fixed.xml', 'w') { |f|
  f.puts doc
}

# few more file cleanup find and replace
text = File.read("full_file_from_excel_fixed.xml")
replace = text.gsub("  </ss:Data>","</Data>")
replace = replace.gsub(" </ss:Data>","</Data>")
replace = replace.gsub("</ss:Data>","</Data>")
replace = replace.gsub("<ss:Data xmlns=\"http://www.w3.org/TR/REC-html40\" ss:Type=\"String\">", "<Data ss:Type=\"String\">")
File.open("full_file_from_excel_fixed.xml", "w") {|file| file.puts replace}

If the "junk from excel" is that simple and the ss:Type="String" elements only contain text or the <Font> nonsense then:

doc.encoding = 'utf-8'
doc.xpath('//Data[@Type="String"]').each { |n| n.content = n.text }

should remove the ugliness. That gives me this output:

<?xml version="1.0" encoding="utf-8"?>
<Worksheet Name="Subtitles">
  <Table ExpandedColumnCount="3" ExpandedRowCount="53" FullColumns="1" FullRows="1" DefaultRowHeight="13.5">
   <Column StyleID="s62" Width="80.25" Span="1"/>
   <Column Index="3" StyleID="s63" Width="249.75"/>
   <Row AutoFitHeight="0">
    <Cell><Data Type="String">00:00:11:09</Data></Cell>
    <Cell><Data Type="String">00:00:13:06</Data></Cell>
    <Cell><Data xmlns="http://www.w3.org/TR/REC-html40" Type="String">안녕하세요, 저는 잭, 9살 입니다. </Data></Cell>
   </Row>
  </Table>
 </Worksheet>

I don't know how to convince Nokogiri to remove that stray xmlns though.

If you're not sure about what's inside the <Data>, then you can remove only the <Font> stuff with something like this:

def font_killer(children)
    children.each do |c|
        if(c.name == 'Font')
            c.replace(font_killer(c.children))
        else
            font_killer(c.children)
        end
    end
    children
end
doc = Nokogiri::XML(open('junk_from_excel.xml').read)
doc.encoding = 'utf-8'
doc.xpath('//Data[@Type="String"]').each { |n| font_killer(n.children) }

I tested this a bit but I'd recommend that you test it a bit more on your real data.

In either case, you lose the ss and x namespace prefixes; this is fair since the XML doesn't doesn't properly declare namespaces so Nokogiri pretends that they don't exist. If you want the namespaces, then add these attributes

xmlns:x="urn:schemas-microsoft-com:office:excel"  
xmlns:ss="urn:schemas-microsoft-com:office:spreadsheet"

to the <Worksheet> element before the ss:Name attribute and then update the XPath expressions to include the namespace:

doc.xpath('//ss:Data[@ss:Type="String"]|//Data[@ss:Type="String"]')

There's probably a better way of expressing that but my XPath-Fu isn't that strong; that should get the job done though.

You could extract the contents before removing them like this:

doc.xpath('//*[@Face]').each do |node|
  node.children.each do |child|
    node.parent << child
  end
  node.remove
end

source: http://rubyforge.org/pipermail/nokogiri-talk/2009-June/000333.html