Language support #4
Conversation
| String url =href.substring("file:".length()); // some calculation from its parameters | ||
| InputStream is = this.getClass().getResourceAsStream(url); | ||
| return new StreamSource(is); | ||
| int index = href.lastIndexOf("/"); |
There was a problem hiding this comment.
Why remove the path? I guess it works now because the java package path and the resources path are the same?
There was a problem hiding this comment.
Yes, and also It did not work with the path. In any case "ddi-to-fo.xsl" also does not use path.
There was a problem hiding this comment.
Weird - I thought I had the i18n part (without the tika test) working at one point.
| <groupId>org.apache.tika</groupId> | ||
| <artifactId>tika-langdetect-optimaize</artifactId> | ||
| <version>2.9.2</version> | ||
| </dependency> |
There was a problem hiding this comment.
Have you looked at issue #3 - is fop still not included? Are the tika classes? (If so - then we can close that issue when this merges).
There was a problem hiding this comment.
fop is included. I did not have problems with it.
There was a problem hiding this comment.
Got it - guessing it's a build command problem - I made a note in the issue #3.
|
I changed version to 1.1.0 |
|
Can we publish it to Maven Central? And update the TBD at https://github.com/gdcc/dataverse-exporters ? |
|
I was able to test this in English and it worked fine, so I'll go ahead and merge. |
Language support was added. The code is checking xml:lang (Metadatalanguage) property in xml codebook and if it is present, it selects the language for pdf codebook based on it. If inc language file does not exists than it produces English pdf codebook.
If property xml:lang does not exist then code tries to guess the language of dataset based on Title or Description. It uses Apache Tika language detection. If it detects language with high probability then this language is used for generating pdf codebook, in all other cases English is used.