♻️ rework URL inputs (#318)

Author: ianardee

src/main/java/com/mindee/input/InputSourceUtils.java

@@ -3,7 +3,6 @@
 import com.mindee.MindeeException;
 import java.io.ByteArrayInputStream;
 import java.io.IOException;
-import java.net.URL;
 import org.apache.pdfbox.Loader;
 import org.apache.pdfbox.io.RandomAccessReadBuffer;
 import org.apache.pdfbox.pdmodel.PDDocument;
@@ -79,15 +78,6 @@ public static boolean isPdf(byte[] fileBytes) {
     return true;
   }
 
-  /**
-   * Ensures the URL can be sent to the Mindee server.
-   */
-  public static void validateUrl(URL inputUrl) {
-    if (!"https".equalsIgnoreCase(inputUrl.getProtocol())) {
-      throw new MindeeException("Only HTTPS source URLs are allowed");
-    }
-  }
-
   /**
    * Returns true if the source PDF has source text inside. Returns false for images.
    *

src/main/java/com/mindee/input/URLInputSource.java

@@ -1,11 +1,13 @@
 package com.mindee.input;
 
+import com.mindee.MindeeException;
 import java.io.File;
 import java.io.FileOutputStream;
 import java.io.IOException;
 import java.io.InputStream;
 import java.io.OutputStream;
 import java.net.HttpURLConnection;
+import java.net.MalformedURLException;
 import java.net.URL;
 import java.nio.file.Files;
 import java.nio.file.Path;
@@ -19,7 +21,7 @@
  */
 public class URLInputSource {
   @Getter
-  private final String url;
+  private final URL url;
   private final String username;
   private final String password;
   @Getter
@@ -43,20 +45,21 @@ public class URLInputSource {
    * @param url URL to fetch the file from.
    * @return An instance of {@link URLInputSource}.
    */
-  public static Builder builder(String url) {
-    return new Builder(url);
+  public static Builder builder(String url) throws MalformedURLException {
+    return new Builder(new URL(url));
   }
 
-  private HttpURLConnection prepareConnection() throws IOException {
-    HttpURLConnection connection = createConnection(url);
-    connection = handleRedirects(connection);
+  public static Builder builder(URL url) {
+    return new Builder(url);
+  }
 
-    int responseCode = connection.getResponseCode();
-    if (responseCode != HttpURLConnection.HTTP_OK) {
-      throw new IOException("Failed to fetch file: " + responseCode);
+  /**
+   * Ensures the URL can be sent to the Mindee server.
+   */
+  public void validateSecure() {
+    if (!"https".equalsIgnoreCase(this.url.getProtocol())) {
+      throw new MindeeException("Only HTTPS source URLs are allowed");
     }
-
-    return connection;
   }
 
   /**
@@ -72,8 +75,20 @@ public void fetchFile() throws IOException {
     }
   }
 
-  protected HttpURLConnection createConnection(String urlString) throws IOException {
-    HttpURLConnection connection = (HttpURLConnection) new URL(urlString).openConnection();
+  private HttpURLConnection prepareConnection() throws IOException {
+    HttpURLConnection connection = createConnection(url);
+    connection = handleRedirects(connection);
+
+    int responseCode = connection.getResponseCode();
+    if (responseCode != HttpURLConnection.HTTP_OK) {
+      throw new IOException("Failed to fetch file: " + responseCode);
+    }
+
+    return connection;
+  }
+
+  protected HttpURLConnection createConnection(URL url) throws IOException {
+    HttpURLConnection connection = (HttpURLConnection) url.openConnection();
     connection.setInstanceFollowRedirects(true);
 
     if (username != null && password != null) {
@@ -101,7 +116,7 @@ private HttpURLConnection handleRedirects(HttpURLConnection connection) throws I
       String newUrl = connection.getHeaderField("Location");
       connection.disconnect();
 
-      HttpURLConnection newConnection = createConnection(newUrl);
+      HttpURLConnection newConnection = createConnection(new URL(newUrl));
       return handleRedirects(newConnection); // Recursive call to handle multiple redirects
     }
     return connection;
@@ -189,18 +204,27 @@ public void cleanup() {
    * Builder class for an URLInputSource.
    */
   public static class Builder {
-    private final String url;
+    private final URL url;
     private String username;
     private String password;
     private String localFilename;
     private String token;
 
     /**
-     * Default constructor.
+     * String constructor.
+     *
+     * @param url Remote URL resource.
+     */
+    public Builder(String url) throws MalformedURLException {
+      this.url = new URL(url);
+    }
+
+    /**
+     * URL constructor.
      *
      * @param url Remote URL resource.
      */
-    public Builder(String url) {
+    public Builder(URL url) {
       this.url = url;
     }
 

src/main/java/com/mindee/v1/MindeeClient.java

@@ -1,9 +1,9 @@
 package com.mindee.v1;
 
 import com.mindee.MindeeException;
-import com.mindee.input.InputSourceUtils;
 import com.mindee.input.LocalInputSource;
 import com.mindee.input.PageOptions;
+import com.mindee.input.URLInputSource;
 import com.mindee.pdf.PDFBoxApi;
 import com.mindee.pdf.PDFOperation;
 import com.mindee.v1.clientOptions.PollingOptions;
@@ -203,7 +203,6 @@ public <T extends Inference> AsyncPredictResponse<T> enqueue(
       Class<T> type,
       URL sourceUrl
   ) throws IOException {
-    InputSourceUtils.validateUrl(sourceUrl);
     return this.enqueue(type, new Endpoint(type), null, null, null, sourceUrl);
   }
 
@@ -222,7 +221,6 @@ public <T extends Inference> AsyncPredictResponse<T> enqueue(
       URL sourceUrl,
       PredictOptions predictOptions
   ) throws IOException {
-    InputSourceUtils.validateUrl(sourceUrl);
     return this.enqueue(type, new Endpoint(type), null, null, predictOptions, sourceUrl);
   }
 
@@ -232,8 +230,12 @@ private <T extends Inference> AsyncPredictResponse<T> enqueue(
       byte[] file,
       String filename,
       PredictOptions predictOptions,
-      URL urlInputSource
+      URL url
   ) throws IOException {
+    URLInputSource urlInputSource = null;
+    if (url != null) {
+      urlInputSource = new URLInputSource.Builder(url).build();
+    }
     RequestParameters params = RequestParameters
       .builder()
       .file(file)
@@ -402,7 +404,6 @@ public <T extends Inference> AsyncPredictResponse<T> enqueueAndParse(
       Class<T> type,
       URL sourceUrl
   ) throws IOException, InterruptedException {
-    InputSourceUtils.validateUrl(sourceUrl);
     return this.enqueueAndParse(type, new Endpoint(type), null, null, null, null, sourceUrl);
   }
 
@@ -441,11 +442,12 @@ private <T extends Inference> AsyncPredictResponse<T> enqueueAndParse(
       byte[] file,
       String filename,
       PredictOptions predictOptions,
-      URL urlInputSource
+      URL url
   ) throws IOException, InterruptedException {
     if (pollingOptions == null) {
       pollingOptions = PollingOptions.builder().build();
     }
+
     this.validateAsyncParams(pollingOptions);
     final int initialDelaySec = (int) (pollingOptions.getInitialDelaySec() * 1000);
     final int intervalSec = (int) (pollingOptions.getIntervalSec() * 1000);
@@ -456,7 +458,7 @@ private <T extends Inference> AsyncPredictResponse<T> enqueueAndParse(
       file,
       filename,
       predictOptions,
-      urlInputSource
+      url
     );
 
     String jobId = enqueueResponse.getJob().getId();
@@ -648,7 +650,6 @@ public <T extends Inference> PredictResponse<T> parse(
       Class<T> type,
       URL urlInputSource
   ) throws IOException {
-    InputSourceUtils.validateUrl(urlInputSource);
     return this.parse(type, new Endpoint(type), null, null, null, urlInputSource);
   }
 
@@ -667,7 +668,6 @@ public <T extends Inference> PredictResponse<T> parse(
       URL urlInputSource,
       PredictOptions predictOptions
   ) throws IOException {
-    InputSourceUtils.validateUrl(urlInputSource);
     return this.parse(type, new Endpoint(type), null, null, predictOptions, urlInputSource);
   }
 
@@ -677,8 +677,12 @@ private <T extends Inference> PredictResponse<T> parse(
       byte[] file,
       String filename,
       PredictOptions predictOptions,
-      URL urlInputSource
+      URL url
   ) throws IOException {
+    URLInputSource urlInputSource = null;
+    if (url != null) {
+      urlInputSource = new URLInputSource.Builder(url).build();
+    }
     RequestParameters params = RequestParameters
       .builder()
       .file(file)
@@ -760,7 +764,6 @@ public <T extends GeneratedV1> AsyncPredictResponse<T> enqueue(
       Endpoint endpoint,
       URL sourceUrl
   ) throws IOException {
-    InputSourceUtils.validateUrl(sourceUrl);
     return this.enqueue(type, endpoint, null, null, null, sourceUrl);
   }
 
@@ -781,7 +784,6 @@ public <T extends GeneratedV1> AsyncPredictResponse<T> enqueue(
       URL sourceUrl,
       PredictOptions predictOptions
   ) throws IOException {
-    InputSourceUtils.validateUrl(sourceUrl);
     return this.enqueue(type, endpoint, null, null, predictOptions, sourceUrl);
   }
 
@@ -893,7 +895,6 @@ public <T extends GeneratedV1> AsyncPredictResponse<T> enqueueAndParse(
       Endpoint endpoint,
       URL sourceUrl
   ) throws IOException, InterruptedException {
-    InputSourceUtils.validateUrl(sourceUrl);
     return this.enqueueAndParse(type, endpoint, null, null, null, null, sourceUrl);
   }
 
@@ -1024,7 +1025,6 @@ public <T extends GeneratedV1> PredictResponse<T> parse(
       Endpoint endpoint,
       URL documentUrl
   ) throws IOException {
-    InputSourceUtils.validateUrl(documentUrl);
     return this.parse(type, endpoint, null, null, null, documentUrl);
   }
 
@@ -1045,7 +1045,6 @@ public <T extends GeneratedV1> PredictResponse<T> parse(
       URL documentUrl,
       PredictOptions predictOptions
   ) throws IOException {
-    InputSourceUtils.validateUrl(documentUrl);
     return this.parse(type, endpoint, null, null, predictOptions, documentUrl);
   }
 

src/main/java/com/mindee/v1/http/RequestParameters.java

@@ -1,5 +1,6 @@
 package com.mindee.v1.http;
 
+import com.mindee.input.URLInputSource;
 import com.mindee.v1.clientOptions.PredictOptions;
 import com.mindee.v1.clientOptions.WorkflowOptions;
 import java.net.URL;
@@ -20,7 +21,7 @@ public class RequestParameters {
 
   @Builder
   private RequestParameters(
-      URL urlInputSource,
+      URLInputSource urlInputSource,
       byte[] file,
       PredictOptions predictOptions,
       WorkflowOptions workflowOptions,
@@ -39,7 +40,12 @@ private RequestParameters(
     } else {
       this.workflowOptions = workflowOptions;
     }
-    this.fileUrl = urlInputSource;
+    if (urlInputSource != null) {
+      urlInputSource.validateSecure();
+      this.fileUrl = urlInputSource.getUrl();
+    } else {
+      this.fileUrl = null;
+    }
     this.file = file;
     this.fileName = fileName;
   }

src/main/java/com/mindee/v2/MindeeClient.java

@@ -54,6 +54,7 @@ public JobResponse enqueue(
    * @param params The parameters to send along with the file.
    */
   public JobResponse enqueue(URLInputSource inputSource, BaseParameters params) throws IOException {
+    inputSource.validateSecure();
     return mindeeApi.reqPostEnqueue(inputSource, params);
   }
 
@@ -141,8 +142,12 @@ public <TResponse extends CommonResponse> TResponse enqueueAndGetResult(
       URLInputSource inputSource,
       BaseParameters params
   ) throws IOException, InterruptedException {
-    JobResponse job = enqueue(inputSource, params);
-    return pollAndFetch(responseClass, job, PollingOptions.builder().build());
+    return enqueueAndGetResult(
+      responseClass,
+      inputSource,
+      params,
+      PollingOptions.builder().build()
+    );
   }
 
   /**
@@ -162,6 +167,7 @@ public <TResponse extends CommonResponse> TResponse enqueueAndGetResult(
       BaseParameters params,
       PollingOptions pollingOptions
   ) throws IOException, InterruptedException {
+    inputSource.validateSecure();
     JobResponse job = enqueue(inputSource, params);
     return pollAndFetch(responseClass, job, pollingOptions);
   }

src/main/java/com/mindee/v2/http/MindeeHttpApiV2.java

@@ -101,7 +101,7 @@ public JobResponse reqPostEnqueue(URLInputSource inputSource, BaseParameters opt
 
     var builder = MultipartEntityBuilder.create();
     builder.setMode(HttpMultipartMode.EXTENDED);
-    builder.addTextBody("url", inputSource.getUrl());
+    builder.addTextBody("url", inputSource.getUrl().toString());
     post.setEntity(options.buildHttpBody(builder).build());
     return executeEnqueue(post);
   }