CODEWITHSUNDEEP
xpath
jhon.smith
jhon.smith

Reputation: 2043

path and operator and wild card searches

Have many xmlfiles like this whose structure is not same and may vary.

<ratings>
  <rating>
    <agency>SP</agency>
    <provider>SP</provider>
    <type>LONG TERM</type>
    <currencyType>LOCAL</currencyType>
    <description>SP Standard LT LC rating</description>
    <crating by = "Moodys">
        <code>BBB+</code>
    </crating>
    <crating by = "S&P">
        <code>AAA-</code>
    </crating>
    <date>2011-09-07</date>
  </rating>
</ratings>

I would like to find all xml documents that have a text of BBB+ and AAA- Since the xml structures vary i have to resort to wild cards and cannot rely on element names any ideas how would i do this please.

My try is but it doesnt work

//*[. = "BBB+" and . = "AAA-"]

If i don't use the "and" it works but then it wouldn't meet my criteria.

Upvotes: 1

Views: 65

Answers (2)

Robin
Robin

Reputation: 9644

EDIT: This actually answers the case where you want only one criteria met, ie there's at least one node with the text 'AAA-' or 'BBB+'

You can use

//*[normalize-space(text()) = 'AAA-' or normalize-space(text()) = 'BBB+']

which will return the node where the text is found. You can directly put a condition on the text() attribute, and the normalize-space function trims the text from trailing whitespaces.

This will obviously be quite slow as it requires to check all nodes. If you can change the * into a code, this would greatly increase performance.

Upvotes: 1

har07
har07

Reputation: 89285

You can try this way :

/*[//*[normalize-space(text()) = 'AAA-'] and //*[normalize-space(text()) = 'BBB+']]

That will return element if there are "text of BBB+ and AAA-", and return nothing/null otherwise. In addition to this, check answer from @Robin, it contains explanation and good advice to consider.

Upvotes: 1

Related Questions