Eclipse XPath evaluation plugin

Add this URL to your Eclipse Installation to reach this solution's update site.

Oxygen (4.7)
Neon (4.6)
Mars (4.5)
Luna (4.4)
Kepler (4.3)
Juno (4.2, 3.8)
Previous to Juno (<=4.1)
Photon (4.8)
2018-09 (4.9)

Learn more...

Details Group Tabs

Monday, October 22, 2018 - 17:11
jkujawa's picture

I'm learning XPath and looked for an interpreter and this one gave me a first contact with it but apparently it has issues and interpreting XPath functions. Retyping the queries from the video I cannot get a result and in the video it appears well.

Now I've looked at the syntax of the concat function and I don't understand what's going on. Either it states that I can't use a single argument, Either it states that I am restricted on the first one, either it returns only the number of result but does not display the data.

So it seems it is buggy.
You can reproduce what I'm saying by downloading this XML file and trying that query for instance.
Look for Institute for Informatics Georg-August-Universität Göttingen mondial database. They provide an XML file

and the queries:

/mondial/country[population < 10000]/name/concat('Population: ', name)  <- returns no value
concat(/mondial/country[population < 10000]/name,' test ') <- Returns msg "A sequence of more than 1 item is not allowed as the first argument of concat()"
a working query from the video was

/mondial/country[population > 300000000]/name/concat(name,' -> ', population) <- Just returns the number of elements in the resultset but do not display the data.

muhlir's picture

Hi, thanks for reporting the issue. I've just found there is really a small bug regarding this. For now when evaluating the XPath query, please make sure the "Pretty Print" is turned off! Therefore uncheck the checkbox on the very right side next to the "Run Query" and "Load Namespaces" buttons. This is causing not showing the results, I will fix this soon.

Nevertheless, beside this small bug, there is another problem with your XPath query. I believe you were talking about this XML file

Lets have a look at your query: /mondial/country[population < 10000]/name/concat('Population: ', name)

1) Here you call the concat function at this level /mondial/country[population < 10000]/name and passing "name" as the second parameter of the concat function doesn't make sence at all. You may want to change the second parameter to "text()" and you'll get the country name instead (in fact that's the textual content of the "name" element.
2) I believe you want get the population of those selected countries. The problem in that XML file is that there are multiple population elements in each of the country elements. And you can only pass an atomic value to the concat function. Therefore you have to select only one "population" element (selecting it either by year attribute) or just the last one being listed in there.
You can try this query and start digging more from it. It gives you the country name and it's population (it selects only the last population element of the coutry element). /mondial/country[population > 300000000]/name/concat('Population: ', ../population[last()], ', Country name: ', text(), ' | ')

I hope, this helps.

muhlir's picture

This issue has been fixed in 1.4.5. Feel free to upgrade to the latest version.

fted's picture

using Eclipse Version: 4.4.2
(note: spaces in URL because URL was rejected as spam)
Bad HTTP Request: https : //

I checked this URL in a browser and it is in fact bad - returns 400 Bad Request

muhlir's picture

Hi, can you please go in your Eclipse to the Help->Install New Software and click on Available Software Sites and check which URL you have there for the installation of this plugin? In case you have more URLs for this plugin, please remove all of them. Then either install the plugin through the Eclipse marketplace again or add new site by yourself which will point to this URL
I don't know why the other URL (reported by you) doesn't work anymore, but I've just tried this new URL and I was able to install the plugin successfully. Let me know if you will be able to resolve the issue with this.

muhlir's picture

For those who have been using previous versions of this plugin, please uninstall it and install it again either through Eclipse Marketplace or use this new update site:

The reason for this is, that the plugin (and its update site) has been moved from Google to Github and it wouldn't get updated to this latest or future releases.

muhlir's picture

I forgot to mention that starting from this version the support for Java 5 HAS BEEN DROPPED! If you need to run the Eclipse with Java 5, please don't upgrade to this nor future versions, as it might cause problems in the functionality of the plugin.

hglathe's picture


I get always an exception when running the namespace loader.
Here are the details:
An internal error occurred during: "Namespace loader".
XPathFactory#newInstance() failed to create an XPathFactory for the default object model: hattp:// with the XPathFactoryConfigurationException: javax.xml.xpath.XPathFactoryConfigurationException: java.util.ServiceConfigurationError: javax.xml.xpath.XPathFactory: bundleresource://853.fwk1407675409/META-INF/services/javax.xml.xpath.XPathFactory:2: Illegal configuration-file syntax

Any hints for solving the problem?

Kind regards

muhlir's picture

Hi, I have already fixed the issue with the Namespace loader. Please update the plugin to the latest version 1.4.1. Please, see the release note above, as it's necessary to uninstall the plugin first and then install it newly again.

muhlir's picture

Hi, I believe you are using Java 8, right? It seems the issues is related to this bug in the Saxon library.

As a temporary workaround you can put following line into your eclipse.ini anywhere after the "-vmargs":

I've just tested that on my system and it works well with the modified eclipse.ini file. I will try to fix the issue soon though, so the workaround won't be needed. I hope, it helps!

ecornely's picture

When a file has a default namespace (xmlns="something") without prefix nothing is selectable.

Even when adding manualy the namespace value association with an empty string.

I have to define a prefix and use that prefix all arround...

muhlir's picture

Hi, thanks for reporting the issue. It has been fixed. Please, update to the latest version 1.4.1. Please, see the release note above to correctly update to the latest version.

szolotaryov's picture

- does not gather namespace info from the file itself, would be very easy to implement; it's quite tedious to fill it in by hand, esp. taking into account that this info is forgotten when you close the xml file
- would be nice to have options to ignore namespaces, etc.

muhlir's picture

Thanks for suggestion. I've just released new version 1.3.0 which contains a new button to load all the namespaces from the current file, so no need to enter those data manually anymore.

tbert21d's picture

A list of the returned XPath objects would in some cases be preferable, the raw xml is a lot of clutter if you are looking for all results of a complex query. A simple 'Number of results' in a Text widget would already be a big improvement.
In total: very nice plugin, also without the above request. Great that the namespaces work like they should (a feature that many online XPath evaluators miss or don't implement right).
Drag&Drop from the browser into Indigo also worked fine for me, chiq.

Kind regards, tw

muhlir's picture

Thanks for suggesting new enhancements. I've just implemented your request for showing "Number of results" of the executed query, so feel free to update to the newest version:-)
Can you, please, let me know more specifically what you mean by "list of returned XPath objects"?

wkeil's picture

Looks good, I also installed it via Drag&Drop from the browser successfully into Indigo. Why haven't you converted the Google Code Project into EclipseLabs btw., or started there in the first place?

muhlir's picture

Thanks. Yeah, I think I should have done that. Well, I'll check the possibilities of moving the project under EclipseLabs.