HADOOP-16455. ABFS: Implement FileSystem.access() method.

Contributed by Bilahari T H.

Author: bilaharith

hadoop-tools/hadoop-azure/src/main/java/org/apache/hadoop/fs/azurebfs/AbfsConfiguration.java

@@ -178,6 +178,10 @@ public class AbfsConfiguration{
       DefaultValue = DEFAULT_USE_UPN)
   private boolean useUpn;
 
+  @BooleanConfigurationValidatorAnnotation(ConfigurationKey =
+      FS_AZURE_ENABLE_CHECK_ACCESS, DefaultValue = DEFAULT_ENABLE_CHECK_ACCESS)
+  private boolean isCheckAccessEnabled;
+
   @BooleanConfigurationValidatorAnnotation(ConfigurationKey = FS_AZURE_ABFS_LATENCY_TRACK,
           DefaultValue = DEFAULT_ABFS_LATENCY_TRACK)
   private boolean trackLatency;
@@ -403,6 +407,10 @@ public long getAzureBlockSize() {
     return this.azureBlockSize;
   }
 
+  public boolean isCheckAccessEnabled() {
+    return this.isCheckAccessEnabled;
+  }
+
   public String getAzureBlockLocationHost() {
     return this.azureBlockLocationHost;
   }

hadoop-tools/hadoop-azure/src/main/java/org/apache/hadoop/fs/azurebfs/AzureBlobFileSystem.java

@@ -861,9 +861,14 @@ public AclStatus getAclStatus(final Path path) throws IOException {
    * @throws IOException                   see specific implementation
    */
   @Override
-  public void access(final Path path, FsAction mode) throws IOException {
-    // TODO: make it no-op to unblock hive permission issue for now.
-    // Will add a long term fix similar to the implementation in AdlFileSystem.
+  public void access(final Path path, final FsAction mode) throws IOException {
+    LOG.debug("AzureBlobFileSystem.access path : {}, mode : {}", path, mode);
+    Path qualifiedPath = makeQualified(path);
+    try {
+      this.abfsStore.access(qualifiedPath, mode);
+    } catch (AzureBlobFileSystemException ex) {
+      checkCheckAccessException(path, ex);
+    }
   }
 
   private FileStatus tryGetFileStatus(final Path f) {
@@ -976,6 +981,18 @@ <T> FileSystemOperation<T> execute(
     }
   }
 
+  private void checkCheckAccessException(final Path path,
+      final AzureBlobFileSystemException exception) throws IOException {
+    if (exception instanceof AbfsRestOperationException) {
+      AbfsRestOperationException ere = (AbfsRestOperationException) exception;
+      if (ere.getStatusCode() == HttpURLConnection.HTTP_FORBIDDEN) {
+        throw (IOException) new AccessControlException(ere.getMessage())
+            .initCause(exception);
+      }
+    }
+    checkException(path, exception);
+  }
+
   /**
    * Given a path and exception, choose which IOException subclass
    * to create.

hadoop-tools/hadoop-azure/src/main/java/org/apache/hadoop/fs/azurebfs/AzureBlobFileSystemStore.java

@@ -1072,6 +1072,25 @@ public AclStatus getAclStatus(final Path path) throws IOException {
     }
   }
 
+  public void access(final Path path, final FsAction mode)
+      throws AzureBlobFileSystemException {
+    LOG.debug("access for filesystem: {}, path: {}, mode: {}",
+        this.client.getFileSystem(), path, mode);
+    if (!this.abfsConfiguration.isCheckAccessEnabled()
+        || !getIsNamespaceEnabled()) {
+      LOG.debug("Returning; either check access is not enabled or the account"
+          + " used is not namespace enabled");
+      return;
+    }
+    try (AbfsPerfInfo perfInfo = startTracking("access", "checkAccess")) {
+      String relativePath =
+          AbfsHttpConstants.FORWARD_SLASH + getRelativePath(path, true);
+      final AbfsRestOperation op = this.client
+          .checkAccess(relativePath, mode.SYMBOL);
+      perfInfo.registerResult(op.getResult()).registerSuccess(true);
+    }
+  }
+
   public boolean isAtomicRenameKey(String key) {
     return isKeyForDirectorySet(key, azureAtomicRenameDirSet);
   }

hadoop-tools/hadoop-azure/src/main/java/org/apache/hadoop/fs/azurebfs/constants/AbfsHttpConstants.java

@@ -37,6 +37,7 @@ public final class AbfsHttpConstants {
   public static final String SET_PROPERTIES_ACTION = "setProperties";
   public static final String SET_ACCESS_CONTROL = "setAccessControl";
   public static final String GET_ACCESS_CONTROL = "getAccessControl";
+  public static final String CHECK_ACCESS = "checkAccess";
   public static final String GET_STATUS = "getStatus";
   public static final String DEFAULT_TIMEOUT = "90";
   public static final String TOKEN_VERSION = "2";

hadoop-tools/hadoop-azure/src/main/java/org/apache/hadoop/fs/azurebfs/constants/ConfigurationKeys.java

@@ -62,6 +62,10 @@ public final class ConfigurationKeys {
   public static final String FS_AZURE_DISABLE_OUTPUTSTREAM_FLUSH = "fs.azure.disable.outputstream.flush";
   public static final String FS_AZURE_USER_AGENT_PREFIX_KEY = "fs.azure.user.agent.prefix";
   public static final String FS_AZURE_SSL_CHANNEL_MODE_KEY = "fs.azure.ssl.channel.mode";
+  /** Provides a config to enable/disable the checkAccess API.
+   *  By default this will be
+   *  FileSystemConfigurations.DEFAULT_ENABLE_CHECK_ACCESS. **/
+  public static final String FS_AZURE_ENABLE_CHECK_ACCESS = "fs.azure.enable.check.access";
   public static final String FS_AZURE_USE_UPN = "fs.azure.use.upn";
   /** User principal names (UPNs) have the format “{alias}@{domain}”. If true,
    *  only {alias} is included when a UPN would otherwise appear in the output

hadoop-tools/hadoop-azure/src/main/java/org/apache/hadoop/fs/azurebfs/constants/FileSystemConfigurations.java

@@ -67,7 +67,8 @@ public final class FileSystemConfigurations {
   public static final boolean DEFAULT_ENABLE_HTTPS = true;
 
   public static final boolean DEFAULT_USE_UPN = false;
+  public static final boolean DEFAULT_ENABLE_CHECK_ACCESS = false;
   public static final boolean DEFAULT_ABFS_LATENCY_TRACK = false;
 
   private FileSystemConfigurations() {}
-}
+}

hadoop-tools/hadoop-azure/src/main/java/org/apache/hadoop/fs/azurebfs/constants/HttpQueryParams.java

@@ -32,6 +32,7 @@ public final class HttpQueryParams {
   public static final String QUERY_PARAM_RECURSIVE = "recursive";
   public static final String QUERY_PARAM_MAXRESULTS = "maxResults";
   public static final String QUERY_PARAM_ACTION = "action";
+  public static final String QUERY_FS_ACTION = "fsAction";
   public static final String QUERY_PARAM_POSITION = "position";
   public static final String QUERY_PARAM_TIMEOUT = "timeout";
   public static final String QUERY_PARAM_RETAIN_UNCOMMITTED_DATA = "retainUncommittedData";

hadoop-tools/hadoop-azure/src/main/java/org/apache/hadoop/fs/azurebfs/services/AbfsClient.java

@@ -524,6 +524,28 @@ public AbfsRestOperation getAclStatus(final String path, final boolean useUPN) t
     return op;
   }
 
+  /**
+   * Talks to the server to check whether the permission specified in
+   * the rwx parameter is present for the path specified in the path parameter.
+   *
+   * @param path  Path for which access check needs to be performed
+   * @param rwx   The permission to be checked on the path
+   * @return      The {@link AbfsRestOperation} object for the operation
+   * @throws AzureBlobFileSystemException in case of bad requests
+   */
+  public AbfsRestOperation checkAccess(String path, String rwx)
+      throws AzureBlobFileSystemException {
+    AbfsUriQueryBuilder abfsUriQueryBuilder = createDefaultUriQueryBuilder();
+    abfsUriQueryBuilder.addQuery(QUERY_PARAM_ACTION, CHECK_ACCESS);
+    abfsUriQueryBuilder.addQuery(QUERY_FS_ACTION, rwx);
+    URL url = createRequestUrl(path, abfsUriQueryBuilder.toString());
+    AbfsRestOperation op = new AbfsRestOperation(
+        AbfsRestOperationType.CheckAccess, this,
+        AbfsHttpConstants.HTTP_METHOD_HEAD, url, createDefaultHeaders());
+    op.execute();
+    return op;
+  }
+
   private URL createRequestUrl(final String query) throws AzureBlobFileSystemException {
     return createRequestUrl(EMPTY_STRING, query);
   }

hadoop-tools/hadoop-azure/src/main/java/org/apache/hadoop/fs/azurebfs/services/AbfsRestOperationType.java

@@ -39,5 +39,6 @@ public enum AbfsRestOperationType {
     Append,
     Flush,
     ReadFile,
-    DeletePath
+    DeletePath,
+    CheckAccess
 }

hadoop-tools/hadoop-azure/src/site/markdown/abfs.md

@@ -661,6 +661,10 @@ Hflush() being the only documented API that can provide persistent data
 transfer, Flush() also attempting to persist buffered data will lead to
 performance issues.
 
+### <a name="flushconfigoptions"></a> Access Options
+Config `fs.azure.enable.check.access` needs to be set true to enable
+ the AzureBlobFileSystem.access().
+
 ### <a name="perfoptions"></a> Perf Options
 
 #### <a name="abfstracklatencyoptions"></a> 1. HTTP Request Tracking Options