[Artifacts] Save md5 hash for each artifact upload (#1494)

* Hash artifact upload using md5

* Add imports

* Small tweaks

* PR feedback

* PR Feedback

packages/artifact/src/internal/upload/blob-upload.ts

@@ -3,6 +3,8 @@ import {TransferProgressEvent} from '@azure/core-http'
 import {ZipUploadStream} from './zip'
 import {getUploadChunkSize} from '../shared/config'
 import * as core from '@actions/core'
+import * as crypto from 'crypto'
+import * as stream from 'stream'
 
 export interface BlobUploadResponse {
   /**
@@ -14,6 +16,11 @@ export interface BlobUploadResponse {
    * The total reported upload size in bytes. Empty if the upload failed
    */
   uploadSize?: number
+
+  /**
+   * The MD5 hash of the uploaded file. Empty if the upload failed
+   */
+  md5Hash?: string
 }
 
 export async function uploadZipToBlobStorage(
@@ -41,15 +48,31 @@ export async function uploadZipToBlobStorage(
     onProgress: uploadCallback
   }
 
+  let md5Hash: string | undefined = undefined
+  const uploadStream = new stream.PassThrough()
+  const hashStream = crypto.createHash('md5')
+
+  zipUploadStream.pipe(uploadStream) // This stream is used for the upload
+  zipUploadStream.pipe(hashStream).setEncoding('hex') // This stream is used to compute a hash of the zip content that gets used. Integrity check
+
   try {
+    core.info('Beginning upload of artifact content to blob storage')
+
     await blockBlobClient.uploadStream(
-      zipUploadStream,
+      uploadStream,
       bufferSize,
       maxBuffers,
       options
     )
+
+    core.info('Finished uploading artifact content to blob storage!')
+
+    hashStream.end()
+    md5Hash = hashStream.read() as string
+    core.info(`MD5 hash of uploaded artifact zip is ${md5Hash}`)
+
   } catch (error) {
-    core.info(`Failed to upload artifact zip to blob storage, error: ${error}`)
+    core.warning(`Failed to upload artifact zip to blob storage, error: ${error}`)
     return {
       isSuccess: false
     }
@@ -62,12 +85,9 @@ export async function uploadZipToBlobStorage(
     }
   }
 
-  core.info(
-    `Successfully uploaded all artifact file content. Total reported size: ${uploadByteCount}`
-  )
-
   return {
     isSuccess: true,
-    uploadSize: uploadByteCount
+    uploadSize: uploadByteCount,
+    md5Hash: md5Hash
   }
 }

packages/artifact/src/internal/upload/upload-artifact.ts

@@ -10,9 +10,13 @@ import {
   validateRootDirectory
 } from './upload-zip-specification'
 import {getBackendIdsFromToken} from '../shared/util'
-import {CreateArtifactRequest} from 'src/generated'
 import {uploadZipToBlobStorage} from './blob-upload'
 import {createZipUploadStream} from './zip'
+import {
+  CreateArtifactRequest,
+  FinalizeArtifactRequest,
+  StringValue
+} from '../../generated'
 
 export async function uploadArtifact(
   name: string,
@@ -39,7 +43,9 @@ export async function uploadArtifact(
   // get the IDs needed for the artifact creation
   const backendIds = getBackendIdsFromToken()
   if (!backendIds.workflowRunBackendId || !backendIds.workflowJobRunBackendId) {
-    core.warning(`Failed to get backend ids`)
+    core.warning(
+      `Failed to get the necessary backend ids which are required to create the artifact`
+    )
     return {
       success: false
     }
@@ -77,7 +83,10 @@ export async function uploadArtifact(
   }
 
   // Upload zip to blob storage
-  const uploadResult = await uploadZipToBlobStorage(createArtifactResp.signedUploadUrl, zipUploadStream)
+  const uploadResult = await uploadZipToBlobStorage(
+    createArtifactResp.signedUploadUrl,
+    zipUploadStream
+  )
   if (uploadResult.isSuccess === false) {
     return {
       success: false
@@ -85,12 +94,24 @@ export async function uploadArtifact(
   }
 
   // finalize the artifact
-  const finalizeArtifactResp = await artifactClient.FinalizeArtifact({
+  const finalizeArtifactReq: FinalizeArtifactRequest = {
     workflowRunBackendId: backendIds.workflowRunBackendId,
     workflowJobRunBackendId: backendIds.workflowJobRunBackendId,
     name: name,
     size: uploadResult.uploadSize!.toString()
-  })
+  }
+
+  if (uploadResult.md5Hash) {
+    finalizeArtifactReq.hash = StringValue.create({
+      value: `md5:${uploadResult.md5Hash!}`
+    })
+  }
+
+  core.info(`Finalizing artifact upload`)
+
+  const finalizeArtifactResp = await artifactClient.FinalizeArtifact(
+    finalizeArtifactReq
+  )
   if (!finalizeArtifactResp.ok) {
     core.warning(`Failed to finalize artifact`)
     return {
@@ -98,9 +119,14 @@ export async function uploadArtifact(
     }
   }
 
+  const artifactId = BigInt(finalizeArtifactResp.artifactId)
+  core.info(
+    `Artifact ${name}.zip successfully finalized. Artifact ID ${artifactId}`
+  )
+
   return {
     success: true,
     size: uploadResult.uploadSize,
-    id: parseInt(finalizeArtifactResp.artifactId) // TODO - will this be a problem due to the id being a bigint?
+    id: Number(artifactId)
   }
 }

packages/artifact/src/internal/upload/zip.ts

@@ -76,7 +76,9 @@ const zipErrorCallback = (error: any): void => {
 // eslint-disable-next-line @typescript-eslint/no-explicit-any
 const zipWarningCallback = (error: any): void => {
   if (error.code === 'ENOENT') {
-    core.warning('ENOENT warning during artifact zip creation. No such file or directory')
+    core.warning(
+      'ENOENT warning during artifact zip creation. No such file or directory'
+    )
     core.info(error)
   } else {
     core.warning(