deploy content fails with non-ASCII file names on Windows

[fsnow]

With file "EntitéGénérique.xml" in my data directory, I get the following error. Paxton tested on Mac and did not get an error.

ERROR: 500 "Internal Server Error"
ERROR: <error:error xsi:schemaLocation="http://marklogic.com/xdmp/error error.xsd" xmlns:error="http
://marklogic.com/xdmp/error" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">
  <error:code>XDMP-URI</error:code>
  <error:name/>
  <error:xquery-version>1.0-ml</error:xquery-version>
  <error:message>Invalid URI format</error:message>
  <error:format-string>XDMP-URI: Invalid URI format: ""</error:format-string>
  <error:retryable>false</error:retryable>
  <error:expr/>
  <error:data>
    <error:datum>""</error:datum>
  </error:data>
  <error:stack>
    <error:frame>
      <error:xquery-version>1.0-ml</error:xquery-version>
    </error:frame>
  </error:stack>
</error:error>

[grtjn]

Interestingly, the error message is talking about an empty string. Wondering whether this is a client or a server-side issue. Client-side sounds more likely I guess..

[grtjn]

I get a different message, but not unlikely same cause. On mac filename is uploaded as:

%2FEntite%CC%81Ge%CC%81ne%CC%81rique.xml

On Windows though it is sent as:

%2FEntit%E9G%E9n%E9rique.xml

Sounds like an encoding issue. @paxtonhare do you know a trick to adjust encoding of a string to UTF-8 before applying uri encoding?

[grtjn]

path.encode("UTF-8") seems to do the trick..

[hunterwilliams]

I just came across the same problem with my US English Windows 10 when attempting to load files with Japanese Kanji in the filename.

Solution:

1. Type Region in the start menu. If that doesn't work navigate to Control Panel\All Control Panel Items\Language\Advanced settings and then press "Apply language settings for to the welcome screen ...."
2. Under Language for non-Unicode programs press Change system locale...
3. Select the language of the filenames causing the ingestion issue. Note this will cause require a restart and afterwards the command prompt may show weird characters instead of the standard \
4. Attempt to reload content. If you get a 400 error about a blank file then you may have picked the wrong language. If the load fails for another reason try starting command prompt by running cmd /u

[dmcassel]

@hunterwilliams, thanks for providing the info

[grtjn]

Hunter suggests it might be worthwhile to add these notes in the Wiki somewhere. Maybe create a page dedicated to Windows issues?

[grtjn]

Fixed in dev. Leaving open for the wiki edits..

[grtjn]

Wiki section added: https://github.com/marklogic/roxy/wiki/Troubleshooting#windows-issues

Closing issue