Tools XML Inspector and XPath Generator/en

This tool presents an XML document's DOM tree elements. Its most useful function is the XPath input field and XPath generator.

Enter an XPath expression into the field to see which nodes are selected, or vice versa, select an element to get an XPath which finds that element.

In the following, a simple XML document is used as example input:

<DOC>
 <ITEM>
   <REVENUE_YEAR>2554-02</REVENUE_YEAR>
   <REGION>Central</REGION>
   <COST>100</COST>
 </ITEM>
 <ITEM>
   <REVENUE_YEAR>2552-02</REVENUE_YEAR>
   <REGION>Central</REGION>
 </ITEM>
 <ITEM>
   <REVENUE_YEAR>2552-03</REVENUE_YEAR>
   <REGION>Central</REGION>
   <COST>400</COST>
 </ITEM>
 <ITEM>
   <REVENUE_YEAR>2552-04</REVENUE_YEAR>
   <REGION>West</REGION>
   <COST>100</COST>
 </ITEM>
 <ITEM>
   <REVENUE_YEAR>2552-05</REVENUE_YEAR>
   <REGION>East</REGION>
   <COST>200</COST>
 </ITEM>
 <ITEM>
   <REVENUE_YEAR>2552-06</REVENUE_YEAR>
   <REGION>East</REGION>
   <COST>300</COST>
 </ITEM>
</DOC>

Opening the XML Inspector

As described elsewhere, the inspector can be opened via the pin-value popup menu, programmatically or via an action from the standard library. It is also embedded into a FileBrowser when an XML document file is double clicked or selected inside a zip archive there.

Finally, the regular inspector also contains a DOM tab if the inspected string represents an XML document.

Programmatically an inspector is opened by sending the "inspect" message to any object - in this case to a string (try it in a workspace or a code editor, by selecting the expression and executing it with doIt):

|text|
text := '<DOC> .... </DOC>'.
text inspect

or:

'<DOC>...</DOC>' inspect

or in JavaScript:

var text;
text = "<DOC> .... </DOC>".
text.inspect()

or:

"<DOC>...</DOC>".inspect()

Opening Inspector in Workspace

Or select the string only (incl. the quotes) and execute the "InspectIt" menu function,

Fehler beim Erstellen des Vorschaubildes: Die Miniaturansicht konnte nicht am vorgesehenen Ort gespeichert werden

XML Inspector showing the DOM

Trying XPath Expressions

Here are a few examples on how to use XPath on the following XML string:

<DOC>
 <ITEM>
   <REVENUE_YEAR>2554-02</REVENUE_YEAR>
   <REGION>Central</REGION>
   <COST>100</COST>
 </ITEM>
 <ITEM>
   <REVENUE_YEAR>2552-02</REVENUE_YEAR>
   <REGION>Central</REGION>
 </ITEM>
 <ITEM>
   <REVENUE_YEAR>2552-03</REVENUE_YEAR>
   <REGION>Central</REGION>
   <COST>400</COST>
 </ITEM>
 <ITEM>
   <REVENUE_YEAR>2552-04</REVENUE_YEAR>
   <REGION>West</REGION>
   <COST>100</COST>
 </ITEM>
 <ITEM>
   <REVENUE_YEAR>2552-05</REVENUE_YEAR>
   <REGION>East</REGION>
   <COST>200</COST>
 </ITEM>
 <ITEM>
   <REVENUE_YEAR>2552-06</REVENUE_YEAR>
   <REGION>East</REGION>
   <COST>300</COST>
 </ITEM>
</DOC>

Selecting and Processing Nodes

To get all COST elements (as a set of nodes), use:

//COST

to get their contents (as strings):

//COST[text()]

to get all Nodes for a particular REGION:

//REGION[text() = "East"]

to get the COST elements for a particular REGION

//REGION[text() = "East"]/following-sibling::COST

and those values:

//REGION[text() = "East"]/following-sibling::COST[text()]

the maximum of them:

max(//REGION[text() = "East"]/following-sibling::COST[text()])

the sum of them:

sum(//REGION[text()]/following-sibling::COST[text()])

Programmatically Operating on a NodeSet

In a real application, you would provide the XML document via an input pin; either as a string (and parsed as shown below), or as a filename, or as an already constructed DOM object. Notice, that the XML library already provides corresponding action blocks, so the following is presented to give you some background details (it may also be useful to process the nodeSet programmatically for performance reasons, or because a corresponding XPath function is not available or the XPath expression becomes too complex and unmaintanable).

Using an elementary Smalltalk action, the above operations can be implemented with:

|xml dom xpath nodes|

xml := '<DOC>...</DOC>'.
dom := XML::XMLParser parse:xml.
xpath := XML:XPathParser parse:'//REGION[text() = "East"]/following-sibling::COST[]'.
nodes := xpath nodesMatchingInDocument:dom.
nodes do:[:eachNode |
   "/ ... do something with eachNode ...
   "/ in this case, this will be each COST element;
   "/ extract its text, make it a number and show it
   Transcript showCR: (eachNode characterData asNumber).
   => prints 200 and 300 on the Transcript
].

or sum those values:

nodes sum:[:eachNode | eachNode characterData asNumber]
=> prints 500 on the Transcript

Programmatically Using XPath Functions

You can also use XPath functions which operate on a NodeSet. In this case, you will not get a NodeSet but a value such as a string or number from the XPath evaluation. For example:

...
xpath := XML:XPathParser parse:'max(//REGION[text() = "East"]/following-sibling::COST[])'.
maxValue := xpath xPathValueFor:dom.
=> maxValue will be 300

or:

...
xpath := XML:XPathParser parse:'sum(//REGION[text() = "East"]/following-sibling::COST[])'.
sum := xpath xPathValueFor:dom.
=> sum will be 500

Back to Online Documentation.
Back to Tools.

---

Copyright © 2014-2024 eXept Software AG