alfresco-transform-core/alfresco-transformer-base
Kristian Dimitrov a1b6283a4c
ATS-669: Parameterize T-Engines transformer execution locations (#203)
* ATS-669: Implement cmd line arguments for ImageMagick, PdfRenderer and LibreOffice

* ATS-669: Remove unnecessary test ImageMagick line

* ATS-669: Implement Spring boot properties via application.yaml

* ATS-669: Implement Spring config binds and utilize new functionality in pdfRender

* ATS-669: Wire externalProps for ImageMagick

* ATS-669: Wire externalProps for LibreOffice

* ATS-669: Fix failing tests

* ATS-669: Implement parameterized execution for All-In-One transform module

* ATS-669: Use string values instead of GlobalProperties class

* ATS-669: Change pdfrenderer property format

* ATS-669: Add validation to executor constructors

* ATS-669: Fix failing LibreOffice tests

* ATS-669: Add missing license

* ATS-669: Update LibreOffice version

* ATS-669: Remove unnecessary annotation

* ATS-669: Standardise properties

* ATS-669: Change field variable names

* ATS-669: Change field variable values

* ATS-669: Add unit tests for passing system properties

* ATS-669: Standardise yaml properties

* ATS-669: Remove unnecessary super() calls

* ATS-669: Change CRLF to LF

* ATS-669: Change LF to CRLF

* ATS-669: Fix yaml indentation

* ATS-669: Update tika and misc yaml file with new sub-property

* ATS-669: Remove unused import

* ATS-669: Update TransformRegistryImpl property location
2020-04-16 16:32:01 +01:00
..
2019-05-22 11:49:35 +03:00

Common code for Docker based ACS transformers

This project contains code that is common between all the ACS transformers that run within their own Docker containers. It performs common actions such as logging, throttling requests and handling the streaming of content to and from the container. It also provides structure and hook points to allow specific transformers to simply check request parameter and perform the transformation using either files or a pair of InputStream and OutputStream.

A transformer project is expected to provide the following files:

src/main/resources/templates/transformForm.html
src/main/java/org/alfresco/transformer/<XXX>Controller.java
src/main/java/org/alfresco/transformer/Application.java
  • transformerForm.html - A simple test page using thymeleaf that gathers request parameters so they may be used to test the transformer.
<html xmlns:th="http://www.thymeleaf.org">
<body>
  <div>
    <h2>Test Transformation</h2>
    <form method="POST" enctype="multipart/form-data" action="/transform">
      <table>
        <tr><td><div style="text-align:right">file *</div></td><td><input type="file" name="file" /></td></tr>
        <tr><td><div style="text-align:right">targetFilename *</div></td><td><input type="text" name="targetFilename" value="" /></td></tr>
        <tr><td><div style="text-align:right">width</div></td><td><input type="text" name="width" value="" /></td></tr>
        <tr><td><div style="text-align:right">height</div></td><td><input type="text" name="height" value="" /></td></tr>
        <tr><td><div style="text-align:right">allowPdfEnlargement</div></td><td><input type="checkbox" name="allowPdfEnlargement" value="true" /></td></tr>
        <tr><td><div style="text-align:right">maintainPdfAspectRatio</div></td><td><input type="checkbox" name="maintainPdfAspectRatio" value="true" /></td></tr>
        <tr><td><div style="text-align:right">page</div></td><td><input type="text" name="page" value="" /></td></tr>
        <tr><td><div style="text-align:right">timeout</div></td><td><input type="text" name="timeout" value="" /></td></tr>
        <tr><td></td><td><input type="submit" value="Transform" /></td></tr>
	  </table>
	</form>
  </div>
  <div>
    <a href="/log">Log entries</a>
  </div>
</body>
</html>
  • TransformerNameController.java - A Spring Boot Controller that extends AbstractTransformerController to handel a POST request to "/transform".
...
@Controller
public class AlfrescoPdfRendererController extends AbstractTransformerController
{
    ...

    @PostMapping("/transform")
    public ResponseEntity<Resource> transform(HttpServletRequest request,
                                              @RequestParam("file") MultipartFile sourceMultipartFile,
                                              @RequestParam("targetFilename") String targetFilename,
                                              @RequestParam(value = "width", required = false) Integer width,
                                              @RequestParam(value = "height", required = false) Integer height,
                                              @RequestParam(value = "allowPdfEnlargement", required = false) Boolean allowPdfEnlargement,
                                              @RequestParam(value = "maintainPdfAspectRatio", required = false) Boolean maintainPdfAspectRatio,
                                              @RequestParam(value = "page", required = false) Integer page,
                                              @RequestParam(value = "timeout", required = false) Long timeout)
    {
        try
        {
            File sourceFile = createSourceFile(request, sourceMultipartFile);
            File targetFile = createTargetFile(request, targetFilename);
            // Both files are deleted by TransformInterceptor.afterCompletion

            StringJoiner args = new StringJoiner(" ");
            if (width != null)
            {
                args.add("--width=" + width);
            }
            if (height != null)
            {
                args.add("--height=" + height);
            }
            if (allowPdfEnlargement != null && allowPdfEnlargement)
            {
                args.add("--allow-enlargement");
            }
            if (maintainPdfAspectRatio != null && maintainPdfAspectRatio)
            {
                args.add("--maintain-aspect-ratio");
            }
            if (page != null)
            {
                args.add("--page=" + page);
            }
            String options = args.toString();
            LogEntry.setOptions(options);

            Map<String, String> properties = new HashMap<>();
            properties.put("options", options);
            properties.put("source", sourceFile.getAbsolutePath());
            properties.put("target", targetFile.getAbsolutePath());

            executeTransformCommand(properties, targetFile, timeout);

            return createAttachment(targetFilename, targetFile);
        }
        catch (UnsupportedEncodingException e)
        {
            throw new TransformException(500, "Filename encoding error", e);
        }
    }
}
  • TransformerNameController#processTransform(File sourceFile, File targetFile, Map<String, String> transformOptions, Long timeout)

/transform (Consumes: application/json, Produces: application/json)

The new consumes and produces arguments have been specified in order to differentiate this endpoint from the previous one (which consumes multipart/form-data)

The endpoint should always receive a TransformationRequest and should always respond with a TransformationReply.

As specific transformers require specific arguments (e.g. transform for the Tika transformer) the request body should include this in the transformRequestOptions via the Map<String,String> transformRequestOptions.

Example request body

var transformRequest = {
	"requestId": "1",
	"sourceReference": "2f9ed237-c734-4366-8c8b-6001819169a4",
	"sourceMediaType": "pdf",
	"sourceSize": 123456,
	"sourceExtension": "pdf",
	"targetMediaType": "txt",
	"targetExtension": "txt",
	"clientType": "ACS",
	"clientData": "Yo No Soy Marinero, Soy Capitan, Soy Capitan!",
	"schema": 1,
	"transformRequestOptions": {
		"targetMimetype": "text/plain",
		"targetEncoding": "UTF-8",
		"transform": "PdfBox"
	}
}

Example response body

var transformReply = {
    "requestId": "1",
    "status": 201,
    "errorDetails": null,
    "sourceReference": "2f9ed237-c734-4366-8c8b-6001819169a4",
    "targetReference": "34d69ff0-7eaa-4741-8a9f-e1915e6995bf",
    "clientType": "ACS",
    "clientData": "Yo No Soy Marinero, Soy Capitan, Soy Capitan!",
    "schema": 1
}

processTransform method

public abstract class AbstractTransformerController 
{
    void processTransform(File sourceFile, File targetFile, Map<String, String> transformOptions, Long timeout) { /* Perform the transformation*/ }
}

The abstract method is declared in the AbstractTransformerController and must be implemented by the specific controllers.

This method is called by the AbstractTransformerController directly in the new /transform endpoint which consumes application/json and produces application/json.

The method is responsible for performing the transformation. Upon a successful transformation it updates the targetFile parameter.

  • Application.java - Spring Boot expects to find an Application in a project's source files. The following may be used:
package org.alfresco.transformer;

import org.springframework.boot.SpringApplication;
import org.springframework.boot.autoconfigure.SpringBootApplication;

@SpringBootApplication
public class Application
{
    public static void main(String[] args)
    {
        SpringApplication.run(Application.class, args);
    }
}

Building and testing

The project can be built by running the Maven command:

mvn clean install

Artifacts

The artifacts can be obtained by:

<dependency>
  <groupId>org.alfresco</groupId>
  <artifactId>alfresco-transformer-base</artifactId>
  <version>1.0</version>
</dependency>

and the Alfresco Maven repository:

<repository>
  <id>alfresco-maven-repo</id>
  <url>https://artifacts.alfresco.com/nexus/content/groups/public</url>
</repository>

The build plan is available in TravisCI.

Contributing guide

Please use this guide to make a contribution to the project.