Rexml & nested loops

I'm a bit of a newbie to Ruby, and to xpath..... and hoping someone here can give me the one-liner version of why this is failing.

I have an XML structure like:

<foos>
  <foo>
  <fooID>Foo1</fooID>
  <subfoos>
    <subfoo>
      <subfooName>foobar1</subfooName>
    </subfoo>
    <subfoo>
      <subfooName>foobar2</subfooName>
    </subfoo>
  </subfoos>
  </foo>
  <foo>
  <fooID>Foo2</fooID>
  <subfoos>
    <subfoo>
      <subfooName>foobar3</subfooName>
    </subfoo>
    <subfoo>
      <subfooName>foobar4</subfooName>
    </subfoo>
  </subfoos>
  </foo>
</foos>

And the code

doc.each_element('//foo') { |foo|
  puts "*** Processing #{foo.elements['fooId'].text}"
  foo.each_element('//subFoo') { |subFoo|
    puts subFoo.elements["subfooName"].text
    }
  }

What I was expecting was a nested loop.. but what I get is:

doc.each_element('//foo') { |foo|

This iterator does what I expect.

  foo.each_element('//subFoo') { |subFoo|

This iterator seems to give me all instances of subFoo in doc, not just in the current instance of foo.

I'm sure this i just something about ruby that I don't understand... but can someone point me in the right direction?

Cheers

ash

No, it's something about XPath that you do not understand. :slight_smile: "//" means "from the root of the document", so you are iterating all "subfoo" elements each time. I can think of two remedies:

1. do a single loop with an XPath expression that selects only "subfoo" below "foo"'s.

2. keep your nested loop but change the XPath expression to not start at the root. (I believe it should be ".//subfoo".)

These are the two pages I usually consult when in doubt about XPath expressions:

http://www.w3schools.com/xpath/
http://www.zvon.org/xxl/XPathTutorial/General/examples.html

And then there's of course the standard:

Kind regards

  robert

···

On 08.06.2008 00:15, Paul Ash wrote:

I'm a bit of a newbie to Ruby, and to xpath..... and hoping someone here can give me the one-liner version of why this is failing.

I have an XML structure like:

<foos>
    <foo>
    <fooID>Foo1</fooID>
    <subfoos>
        <subfoo>
            <subfooName>foobar1</subfooName>
        </subfoo>
        <subfoo>
            <subfooName>foobar2</subfooName>
        </subfoo>
    </subfoos>
    </foo>
    <foo>
    <fooID>Foo2</fooID>
    <subfoos>
        <subfoo>
            <subfooName>foobar3</subfooName>
        </subfoo>
        <subfoo>
            <subfooName>foobar4</subfooName>
        </subfoo>
    </subfoos>
    </foo>
</foos>

And the code

doc.each_element('//foo') { |foo|
puts "*** Processing #{foo.elements['fooId'].text}"
foo.each_element('//subFoo') { |subFoo|
   puts subFoo.elements["subfooName"].text
   }
}

What I was expecting was a nested loop.. but what I get is:

doc.each_element('//foo') { |foo|

This iterator does what I expect.

foo.each_element('//subFoo') { |subFoo|

This iterator seems to give me all instances of subFoo in doc, not just in the current instance of foo.

I'm sure this i just something about ruby that I don't understand... but can someone point me in the right direction?

I get that part of the syntax - what I don't (or didn't) get is why
the two xpath operations I have are referring to the same document - I
assumed they would be scoped to only the data passed to the block.

···

On Jun 8, 2:35 am, Robert Klemme <shortcut...@googlemail.com> wrote:

On 08.06.2008 00:15, Paul Ash wrote:

> I'm a bit of a newbie to Ruby, and to xpath..... and hoping someone here
> can give me the one-liner version of why this is failing.

> I have an XML structure like:

> <foos>
> <foo>
> <fooID>Foo1</fooID>
> <subfoos>
> <subfoo>
> <subfooName>foobar1</subfooName>
> </subfoo>
> <subfoo>
> <subfooName>foobar2</subfooName>
> </subfoo>
> </subfoos>
> </foo>
> <foo>
> <fooID>Foo2</fooID>
> <subfoos>
> <subfoo>
> <subfooName>foobar3</subfooName>
> </subfoo>
> <subfoo>
> <subfooName>foobar4</subfooName>
> </subfoo>
> </subfoos>
> </foo>
> </foos>

> And the code

> doc.each_element('//foo') { |foo|
> puts "*** Processing #{foo.elements['fooId'].text}"
> foo.each_element('//subFoo') { |subFoo|
> puts subFoo.elements["subfooName"].text
> }
> }

> What I was expecting was a nested loop.. but what I get is:

> doc.each_element('//foo') { |foo|

> This iterator does what I expect.

> foo.each_element('//subFoo') { |subFoo|

> This iterator seems to give me all instances of subFoo in doc, not just
> in the current instance of foo.

> I'm sure this i just something about ruby that I don't understand... but
> can someone point me in the right direction?

No, it's something about XPath that you do not understand. :slight_smile: "//"
means "from the root of the document", so you are iterating all "subfoo"
elements each time. I can think of two remedies:

1. do a single loop with an XPath expression that selects only "subfoo"
below "foo"'s.

2. keep your nested loop but change the XPath expression to not start at
the root. (I believe it should be ".//subfoo".)

These are the two pages I usually consult when in doubt about XPath
expressions:

http://www.w3schools.com/xpath/http://www.zvon.org/xxl/XPathTutorial/General/examples.html

And then there's of course the standard:

Cover page | xpath | W3C standards and drafts | W3C

Kind regards

    robert

Well, they are because they use the node as basis. But if you give
the command "search everything from the root on" then no matter where
you start the root node of that document will determine the start
position. Hence you have "." to denote the current node.

Cheers

robert

···

2008/6/8 Dennis Edwards <dedward@gmail.com>:

I get that part of the syntax - what I don't (or didn't) get is why
the two xpath operations I have are referring to the same document - I
assumed they would be scoped to only the data passed to the block.

--
use.inject do |as, often| as.you_can - without end

Thanks a lot Robert.. makes perfect sense now.

Cheers.
Dennis

···

On Jun 9, 4:53 am, Robert Klemme <shortcut...@googlemail.com> wrote:

2008/6/8 Dennis Edwards <dedw...@gmail.com>:

> I get that part of the syntax - what I don't (or didn't) get is why
> the two xpath operations I have are referring to the same document - I
> assumed they would be scoped to only the data passed to the block.

Well, they are because they use the node as basis. But if you give
the command "search everything from the root on" then no matter where
you start the root node of that document will determine the start
position. Hence you have "." to denote the current node.

Cheers

robert

--
use.inject do |as, often| as.you_can - without end