This module provides a Fusepool P3 Transformer implementation for Apache Any23
The Any23 Transformer is an asynchronous transformer implementation.
The transformer can be started using the latest release that can be downloaded from the releases section. The executable jar file contains all the necessary dependencies. To start it, run the command
java -jar any23-transformer-any23-transformer-*.jar
The transformer will be started and listen on it's default port 8303.
As an example, a RDFa-document can be sent to the transformer (the document can be found here):
curl -v -X "POST" -H "Content-Type: text/html;charset=UTF-8" -T "rdfa11.html" \http://localhost:8303/
After submitting this request, the transformer will answer with an URL where the response can be fetched:
HTTP/1.1 202 Accepted
Date: Mon, 06 Jul 2015 11:07:07 GMT
Location: /job/a7e3d269-55c0-40d6-9513-58bb2a966bd5
Thus, by accessing http://localhost:8303/job/a7e3d269-55c0-40d6-9513-58bb2a966bd5
the resulting RDF data should be returned.
This module uses Maven as build system. After obtaining the sources, build using
mvn clean install
in the root directory of the sources, this will build a runnable jar. After the build succeeds you can find it in the ´/target´ folder.
Calling
java -Xmx1g -jar any23-transformer-{version}.jar
will run the Any23 Transformer on port 8303
with the default Any23 configuration
and a thread pool of min 3
and max 20
threads.
The command line tool provides the following configuration parameters:
java -Xmx{size} -jar {jar-name} [options]
Any23 Transformer:
-c,--config <arg> The Any23 configuration file. Will be applied on
top of the Any23 default configuration
-h,--help display this help and exit
-m,--mode <arg> The validation mode used by Any23 (options:
[None, Validate, ValidateAndFix], default:ValidateAndFix)
-p, -P,--port. --Port <arg> the port for the Any23 transformer (default: 8303)
-x,--core-pool <arg> The core pool size of the thread pool used to
transform parsed resources (default: 3)
-y,--max-pool <arg> The maximum pool size of the thread pool used to
transform parsed resources (default: 20)
-z,--keep-alive <arg> The maximum time that excess idle threads (default: 60)
As the Any23 transformer implements the [Fusepool Transfomer API] (https://github.com/fusepoolP3/overall-architecture/blob/master/transformer-api.md) communication is expected as specified by the Fusepool.
The capabilities of a transformer can be requested by a simple GET request at
the base URI. The following listing shows the response of the Any23 transformer
running at localhost at port 8303
curl http://localhost:8303/
<http://localhost:8303/>
<http://vocab.fusepool.info/transformer#supportedInputFormat>
"text/csv"^^xsd:string , "application/octet-stream"^^xsd:string ,
"text/turtle"^^xsd:string , "application/rdf+xml"^^xsd:string ,
"application/xhtml+xml"^^xsd:string , "text/html"^^xsd:string ,
"text/rdf+nt"^^xsd:string , "application/n-quads"^^xsd:string ,
"application/ld+json"^^xsd:string ;
<http://vocab.fusepool.info/transformer#supportedOutputFormat>
"text/turtle"^^xsd:string .
Based on that one now knows that HTML documents are supported. So next we want
to send an RDFa 1.1 example document and let the transformer extract the
contained knowledge. For that we need to POST the content with the correct
Content-Type
header - text/html;charset=UTF-8
in in this case the
RDFa is embedded in an HTML document (XHTML would be an other option).
curl -v -X "POST" -H "Content-Type: text/html;charset=UTF-8" \
-H "Content-Location: http://www.example.org/fusepool/rdfa-example.html"
-T "test/resources/rdfa11.html" \
http://localhost:8303/
NOTE that the above request does set the “Content-Location” header. The parsed value will be used as base URI for the extracted knowledge. If this parameter is not defined the generated job URI will get used as base URI
As Any23 is implemented as an asynchronous transformer what you will receive on
this request is a 202 Accept
with the location of the transformation job
< HTTP/1.1 202 Accepted
< Date: Fri, 07 Nov 2014 09:09:42 GMT
< Location: /job/1678699a-ed36-4282-aaf8-1823aea19970
< Transfer-Encoding: chunked
< Server: Jetty(9.2.z-SNAPSHOT)
Now one can send a GET request to the job to retrieve the current status or if finished the final result.
curl http://localhost:8303/job/1678699a-ed36-4282-aaf8-1823aea19970
In this case this will return the extracted Good Relation statements
serialized as text/turtle