Many resources are needed to download a project. Please understand that we have to compensate our server costs. Thank you in advance.
Project price only 1 $
You can buy this project and download/modify it how often you want.
io.cdap.plugin.gcp.gcs.StorageClient Maven / Gradle / Ivy
/*
* Copyright © 2019-2020 Cask Data, Inc.
*
* Licensed under the Apache License, Version 2.0 (the "License"); you may not
* use this file except in compliance with the License. You may obtain a copy of
* the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
* License for the specific language governing permissions and limitations under
* the License.
*/
package io.cdap.plugin.gcp.gcs;
import com.google.api.gax.paging.Page;
import com.google.cloud.http.HttpTransportOptions;
import com.google.cloud.kms.v1.CryptoKeyName;
import com.google.cloud.storage.Blob;
import com.google.cloud.storage.BlobId;
import com.google.cloud.storage.BlobInfo;
import com.google.cloud.storage.Bucket;
import com.google.cloud.storage.CopyWriter;
import com.google.cloud.storage.Storage;
import com.google.cloud.storage.StorageException;
import com.google.cloud.storage.StorageOptions;
import com.google.common.annotations.VisibleForTesting;
import io.cdap.plugin.gcp.common.GCPConnectorConfig;
import io.cdap.plugin.gcp.common.GCPUtils;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import java.io.IOException;
import java.nio.file.FileSystems;
import java.nio.file.PathMatcher;
import java.nio.file.Paths;
import java.util.ArrayList;
import java.util.HashSet;
import java.util.Iterator;
import java.util.List;
import java.util.Map;
import java.util.Set;
import java.util.function.Consumer;
import java.util.regex.Pattern;
import javax.annotation.Nullable;
/**
* A wrapper around the GCS storage client that has extended logic around recursively copying a "directory" on GCS.
*/
public class StorageClient {
private static final Logger LOG = LoggerFactory.getLogger(StorageClient.class);
private final Storage storage;
public StorageClient(Storage storage) {
this.storage = storage;
}
/**
* Picks one blob that has the path prefix and is not ending with '/'
* @param path
* @return
*/
public Blob pickABlob(String path) {
if (path == null || path.isEmpty()) {
return null;
}
GCSPath gcsPath = GCSPath.from(path);
Page blobPage = storage.list(gcsPath.getBucket(), Storage.BlobListOption.prefix(gcsPath.getName()));
Iterator iterator = blobPage.getValues().iterator();
while (iterator.hasNext()) {
Blob blob = iterator.next();
if (blob.getName().endsWith("/")) {
continue;
}
return blob;
}
return null;
}
/**
* Updates the metadata for the blob
* @param blob
* @param metaData
*/
public void setMetaData(Blob blob, Map metaData) {
if (blob == null || metaData == null || metaData.isEmpty()) {
return;
}
storage.update(BlobInfo.newBuilder(blob.getBlobId()).setMetadata(metaData).build());
}
/**
* Applies the given function with metadata of each blobs in the path
* @param path
* @param function
*/
public void mapMetaDataForAllBlobs(String path, Consumer> function) {
if (path == null || path.isEmpty() || function == null) {
return;
}
GCSPath gcsPath = GCSPath.from(path);
Page blobPage = storage.list(gcsPath.getBucket(), Storage.BlobListOption.prefix(gcsPath.getName()));
Iterator blobIterator = blobPage.iterateAll().iterator();
while (blobIterator.hasNext()) {
Blob blob = blobIterator.next();
Map metadata = blob.getMetadata();
if (metadata == null) {
continue;
}
function.accept(metadata);
}
}
/**
* Creates the given bucket if it does not exists.
*
* @param path the path of the bucket
* @param location the location of bucket
* @param cmekKeyName the name of the cmek key
*/
public void createBucketIfNotExists(GCSPath path, @Nullable String location, @Nullable CryptoKeyName cmekKeyName) {
try {
GCPUtils.createBucket(storage, path.getBucket(), location, cmekKeyName);
LOG.info("Bucket {} has been created successfully", path.getBucket());
} catch (StorageException e) {
// Don't throw error if bucket already exists
// https://cloud.google.com/storage/docs/json_api/v1/status-codes#409_Conflict
if (e.getCode() == 409) {
LOG.warn("Getting 409 Conflict: {} Bucket at destination path {} may already exist.",
e.getMessage(), path.getUri());
} else {
throw new RuntimeException(
String.format("Unable to create bucket %s. Ensure you entered the correct bucket path and " +
"have permissions for it.", path.getBucket()), e);
}
}
}
/**
* Copy objects from the source path to the destination path. If the source path is a single object, that object
* will be copied to the destination. If the source path represents a directory, objects within the directory
* will be copied to the destination directory.
*
* @param sourcePath the path to copy objects from
* @param destPath the path to copy objects to
* @param recursive whether to copy objects in all subdirectories
* @param overwrite whether to overwrite existing objects
* @throws IllegalArgumentException if overwrite is false and copying would overwrite an existing object
*/
public void copy(GCSPath sourcePath, GCSPath destPath, boolean recursive, boolean overwrite) {
pairTraverse(sourcePath, destPath, recursive, overwrite, BlobPair::copy);
}
/**
* Move objects from the source path to the destination path. If the source path is a single object, that object
* will be moved to the destination. If the source path represents a directory, objects within the directory
* will be moved to the destination directory.
*
* @param sourcePath the path to move objects from
* @param destPath the path to move objects to
* @param recursive whether to move objects in all subdirectories
* @param overwrite whether to overwrite existing objects
* @throws IllegalArgumentException if overwrite is false and moving would overwrite an existing object
*/
public void move(GCSPath sourcePath, GCSPath destPath, boolean recursive, boolean overwrite) {
pairTraverse(sourcePath, destPath, recursive, overwrite, BlobPair::move);
}
/**
* Get all the matching wildcard paths given the regex input.
*/
public List getMatchedPaths(GCSPath sourcePath, boolean recursive, Pattern wildcardRegex) {
Page blobPage = storage.list(sourcePath.getBucket(), Storage.BlobListOption.prefix(
getWildcardPathPrefix(sourcePath, wildcardRegex)
));
List blobPageNames = new ArrayList<>();
blobPage.getValues().forEach(blob -> blobPageNames.add(blob.getName()));
return getFilterMatchedPaths(sourcePath, blobPageNames, recursive);
}
static String getWildcardPathPrefix(GCSPath sourcePath, Pattern wildcardRegex) {
String pattern = sourcePath.getName();
String[] patternSplits = pattern.split(wildcardRegex.pattern());
// prefix may be empty
return patternSplits.length >= 1 ? patternSplits[0] : "";
}
static List getFilterMatchedPaths(GCSPath sourcePath, List blobPageNames, boolean recursive) {
Set matchedPaths = new HashSet<>();
String globPattern = "glob:" + sourcePath.getName();
PathMatcher matcher = FileSystems.getDefault().getPathMatcher(globPattern);
for (String blobName : blobPageNames) {
if (matcher.matches(Paths.get(blobName))) {
LOG.debug("Blob name {} matches the glob pattern {}", blobName, globPattern);
String gcsPath = String.format("gs://%s/%s", sourcePath.getBucket(), blobName);
matchedPaths.add(GCSPath.from(gcsPath));
}
}
if (!recursive) {
matchedPaths.removeIf(path -> path.getName().endsWith("/"));
}
return new ArrayList<>(matchedPaths);
}
/**
* Gets source and destination pairs by traversing the source path. Consumes each pair after the directory structure
* is completely traversed.
*/
private void pairTraverse(GCSPath sourcePath, GCSPath destPath, boolean recursive, boolean overwrite,
Consumer consumer) {
Bucket sourceBucket = null;
try {
sourceBucket = storage.get(sourcePath.getBucket());
} catch (StorageException e) {
// Add more descriptive error message
throw new RuntimeException(
String.format("Unable to access source bucket %s. ", sourcePath.getBucket())
+ "Ensure you entered the correct bucket path.", e);
}
if (sourceBucket == null) {
throw new IllegalArgumentException(
String.format("Source bucket '%s' does not exist.", sourcePath.getBucket()));
}
Bucket destBucket = null;
try {
destBucket = storage.get(destPath.getBucket());
} catch (StorageException e) {
// Add more descriptive error message
throw new RuntimeException(
String.format("Unable to access destination bucket %s. ", destPath.getBucket())
+ "Ensure you entered the correct bucket path.", e);
}
if (destBucket == null) {
throw new IllegalArgumentException(
String.format("Destination bucket '%s' does not exist. Please create it first.", destPath.getBucket()));
}
boolean destinationBaseExists;
String baseDestName = destPath.getName();
if (destPath.isBucket() || storage.get(BlobId.of(destPath.getBucket(), baseDestName)) != null) {
destinationBaseExists = true;
} else {
// if gs://bucket2/subdir doesn't exist, check if gs://bucket2/subdir/ exists
// similarly, if gs://bucket2/subdir/ doesn't exist, check if gs://bucket2/subdir exists
// this is because "cp dir0 subdir" and "cp dir0 subdir/" are equivalent if the 'subdir' directory exists
String modifiedName = baseDestName.endsWith("/") ?
baseDestName.substring(0, baseDestName.length() - 1) : baseDestName + "/";
destinationBaseExists = storage.get(BlobId.of(destPath.getBucket(), modifiedName)) != null;
}
List copyList = new ArrayList<>();
traverse(BlobId.of(sourcePath.getBucket(), sourcePath.getName()), recursive, sourceBlob -> {
BlobId destBlobID = resolve(sourcePath.getName(), sourceBlob.getBlobId().getName(),
destPath, destinationBaseExists);
if (!overwrite) {
Blob destBlob = storage.get(destBlobID);
// we can't just use Blob's isDirectory() because the cloud console will create a 'directory' by creating
// a 0 size placeholder blob that ends with '/'. This placeholder blob's isDirectory() method returns false,
// but we don't want the overwrite check to fail on it. So we explicitly ignore the check for these 0 size
// placeholder blobs.
if (destBlob != null && !destBlob.getName().endsWith("/") && destBlob.getSize() != 0) {
throw new IllegalArgumentException(String.format("%s already exists.", toPath(destBlobID)));
}
}
copyList.add(new BlobPair(sourceBlob, destBlobID));
});
LOG.debug("Found {} objects.", copyList.size());
for (BlobPair blobPair : copyList) {
consumer.accept(blobPair);
}
}
/**
* Resolves what the destination blob id should be when copying/moving the source blob.
*
* Suppose gs://bucket0/dir1/dir2 is being recursively copied to gs://bucket1/subdir and the following object exists:
*
* gs://bucket0/dir1/dir2/a/b/c
*
* In this example, baseName = dir1/dir2, sourceName = dir1/dir2/a/b/c, and dest = gs://bucket1/subdir.
*
* If gs://bucket1/subdir already exists, 'dir2' should be copied into the 'subdir' directory,
* resolving to gs://bucket1/subdir/dir2/a/b/c.
* If gs://bucket1/subdir does not already exist, 'dir2' should become the 'subdir' directory,
* resolving to gs://bucket1/subdir/a/b/c.
*
* @param baseName the base object that is being copied or moved
* @param sourceName the actual object that is being copied or moved
* @param dest the object destination
* @param destExists whether the destination exists
* @return the full destination
*/
@VisibleForTesting
static BlobId resolve(String baseName, String sourceName, GCSPath dest, boolean destExists) {
// the relative part is the part of the sourceName that comes after the baseName.
// if baseName = dir1/dir2/ and sourceName = dir1/dir2/a/b/c, the relative part is /a/b/c
String relativePart = sourceName.substring(baseName.length());
if (dest.isBucket()) {
// if the destination is a bucket, just use the source name with that bucket
return BlobId.of(dest.getBucket(), sourceName);
}
// if the destination exists or ends in '/', take the last part of the baseName and append that to the destination,
// ex: subdir -> subdir/dir2
// after that, append the relative part
// ex: subdir/dir2 -> subdir/dir2/a/b/c
// also do this if the destination ends with '/'.
if (destExists || dest.getName().endsWith("/")) {
int lastDirIndex = baseName.lastIndexOf("/");
String lastPart = lastDirIndex > 0 ? baseName.substring(lastDirIndex) : baseName;
return BlobId.of(dest.getBucket(), append(append(dest.getName(), lastPart), relativePart));
}
// if the destination doesn't exist and doesn't end in '/', append the relative part to the destination
return BlobId.of(dest.getBucket(), append(dest.getName(), relativePart));
}
// appends a part to a base, making sure there is one '/' separating them, assuming the base does not end with more
// than one '/' and part does not start with more than one '/'.
@VisibleForTesting
static String append(String base, String part) {
boolean baseEndsWithDivider = base.endsWith("/");
boolean partStartWithDivider = part.startsWith("/");
if (baseEndsWithDivider && partStartWithDivider) {
return base.substring(0, base.length() - 1) + part;
} else if (!baseEndsWithDivider && !base.isEmpty() && !partStartWithDivider && !part.isEmpty()) {
return base + "/" + part;
} else {
return base + part;
}
}
/**
* Add all objects (non-directory blobs) that exist for the given blob id. If the id is an object itself, that blob
* is added to the collection. If it represents a directory, all objects within that directory are added.
* If recursive is true, all subdirectories will also be searched.
* If the blob does not exist and does not represent a directory, nothing happens.
*
* @param blobId the blob id to traverse
* @param recursive whether to recursively traverse subdirectories
* @param consumer the blob consumer
*/
private void traverse(BlobId blobId, boolean recursive, Consumer consumer) {
Page blobList = storage.list(blobId.getBucket(), Storage.BlobListOption.currentDirectory(),
Storage.BlobListOption.prefix(blobId.getName()));
for (Blob blob : blobList.iterateAll()) {
if (!blob.isDirectory()) {
consumer.accept(blob);
} else if (recursive) {
traverse(blob.getBlobId(), true, consumer);
}
}
}
private static String toPath(BlobId blobId) {
return String.format("gs://%s/%s", blobId.getBucket(), blobId.getName());
}
public static StorageClient create(String project, @Nullable String serviceAccount,
Boolean isServiceAccountFilePath, @Nullable Integer readTimeout)
throws IOException {
StorageOptions.Builder builder = StorageOptions.newBuilder().setProjectId(project);
if (serviceAccount != null) {
builder.setCredentials(GCPUtils.loadServiceAccountCredentials(serviceAccount, isServiceAccountFilePath));
}
if (readTimeout != null) {
builder.setTransportOptions(HttpTransportOptions.newBuilder().setReadTimeout(readTimeout * 1000).build());
}
Storage storage = builder.build().getService();
return new StorageClient(storage);
}
public static StorageClient create(GCPConnectorConfig config) throws IOException {
return create(config.getProject(), config.getServiceAccount(), config.isServiceAccountFilePath(), null);
}
/**
* Represents a blob to be copied or moved.
*/
private static class BlobPair {
private final Blob sourceBlob;
private final BlobId destination;
private BlobPair(Blob sourceBlob, BlobId destination) {
this.sourceBlob = sourceBlob;
this.destination = destination;
}
private Blob copy() {
LOG.debug("Copying {} to {}.", toPath(sourceBlob.getBlobId()), toPath(destination));
CopyWriter copyWriter = sourceBlob.copyTo(destination);
Blob copied = copyWriter.getResult();
LOG.debug("Successfully copied {} to {}.", toPath(sourceBlob.getBlobId()), toPath(destination));
return copied;
}
private Blob move() {
Blob moved = copy();
LOG.debug("Deleting {}.", toPath(sourceBlob.getBlobId()));
sourceBlob.delete();
LOG.debug("Successfully deleted {}.", toPath(sourceBlob.getBlobId()));
return moved;
}
}
}
