How to get the index of 'th' element using Nokogiri - ruby

Question

I have the following HTML code and need to determine the index of "Number of Strings" using the <span> id. I'm using Nokogiri to parse the HTML and get the row.
doc = Nokogiri::parse(myfile.html)
table = doc.xpath("//span[#id='NumStrs']/../../..")
row = table.xpath["tr[1]"]
Here is the HTML:
<tr>
<th id ="langframe">
<span id="cabinet">
Cabinet</span>
</th>
<th id ="langbb1">
<span id="bb1">
BB1</span>
</th>
<th id ="langbb2">
<span id="bb2">
BB2</span>
</th>
<th id ="langtemp">
<span id="Temp">
Temperature</span>
</th>
<th id="langstrs">
<span id="StringsPresent">
Strings Present</span>
</th>
<th id="langmstrQty">
<span id="NumStrs">
Number of Strings</span>
</th>
</tr>

Answer 1

I'd do it using Ruby's with_index combined with a select:
require 'nokogiri' # => true
doc = Nokogiri::HTML(<<EOT)
<tr>
<th id ="langframe">
<span id="cabinet">
Cabinet</span>
</th>
<th id ="langbb1">
<span id="bb1">
BB1</span>
</th>
<th id ="langbb2">
<span id="bb2">
BB2</span>
</th>
<th id ="langtemp">
<span id="Temp">
Temperature</span>
</th>
<th id="langstrs">
<span id="StringsPresent">
Strings Present</span>
</th>
<th id="langmstrQty">
<span id="NumStrs">
Number of Strings</span>
</th>
</tr>
EOT
th_idx = doc.search('th').to_enum.with_index.select { |th, idx| th.text['Number of Strings'] }.first
That returns:
th_idx
# => [#(Element:0x3fe72d83cd3c {
# name = "th",
# attributes = [
# #(Attr:0x3fe72d4440f4 { name = "id", value = "langmstrQty" })],
# children = [
# #(Text "\n"),
# #(Element:0x3fe72d43c3e0 {
# name = "span",
# attributes = [
# #(Attr:0x3fe72d439b04 { name = "id", value = "NumStrs" })],
# children = [ #(Text "\nNumber of Strings")]
# }),
# #(Text "\n")]
# }),
# 5]
The index is:
th_idx.last # => 5
Once you have th_idx, you can easily access parent or child nodes to find out about its surroundings:
th_node = th_idx.first
th_node['id'] # => "langmstrQty"
th_node.at('span')
# => #(Element:0x3fd5110286d8 {
# name = "span",
# attributes = [
# #(Attr:0x3fd511021b6c { name = "id", value = "NumStrs" })],
# children = [ #(Text "\nNumber of Strings")]
# })
th_node.at('span')['id'] # => "NumStrs"
with_index adds a 0-based index to each element passed to it. to_enum is required because search returns a NodeSet, which isn't an enumerator so to_enum returns that.
If you want a 1-based index use with_index(1).

Answer 2

Got it working, not sure if this is the efficient way to do it.. but it works
header = table.xpath("tr[1]")
value = header.xpath("//span[#id='#{id}']").text
index = header.search('th//text()').collect {|text| text.to_s.strip}.reject(&:empty?).index(value)+1