Compare commits

..

5 Commits

Author SHA1 Message Date
b8a920d0e7 Use upload-artifact@v3 (not v4)
All checks were successful
Test Workflow / Lint and test library (pull_request) Successful in 8m48s
Publish Workflow / Publish library (pull_request) Successful in 9m13s
2023-12-29 21:33:24 +06:00
fa49939a4d Improve comments in code 2023-12-29 21:33:05 +06:00
c802c76ace Test CI
Some checks failed
Test Workflow / Lint and test library (pull_request) Successful in 8m47s
Publish Workflow / Publish library (pull_request) Failing after 8m54s
2023-12-29 21:11:14 +06:00
9f6fad8466 Add usage in README
All checks were successful
Test Workflow / Lint and test library (pull_request) Successful in 8m11s
2023-12-29 21:10:47 +06:00
571e952897 Initial CI 2023-12-29 21:03:24 +06:00
20 changed files with 260 additions and 743 deletions

View File

@@ -1,19 +0,0 @@
root = true
[*]
charset = utf-8
end_of_line = lf
indent_size = 4
indent_style = space
insert_final_newline = false
max_line_length = 120
tab_width = 4
[{*.yaml,*.yml}]
indent_size = 2
[src/test/**/*]
ktlint_standard_no-wildcard-imports = disabled
[{*.kt,*.kts}]
ij_kotlin_packages_to_use_import_on_demand = org.junit.jupiter.api,aws.sdk.kotlin.services.s3,kotlinx.coroutines,java.io,ziputils

View File

@@ -1,8 +1,11 @@
name: Publish Workflow
on:
pull_request:
branches:
- main
push:
tags:
- v*
branches:
- main
jobs:
publish:
name: Publish library
@@ -22,12 +25,6 @@ jobs:
- name: Setup Gradle
uses: gradle/gradle-build-action@v2
- name: Setup AWS Credentials
run: |
mkdir ~/.aws
echo ${{ secrets.aws_config }} | base64 -d > ~/.aws/config
echo ${{ secrets.aws_credentials }} | base64 -d > ~/.aws/credentials
- name: Run checks
run: ./gradlew check

View File

@@ -25,11 +25,5 @@ jobs:
- name: Setup Gradle
uses: gradle/gradle-build-action@v2
- name: Setup AWS Credentials
run: |
mkdir ~/.aws
echo ${{ secrets.aws_config }} | base64 -d > ~/.aws/config
echo ${{ secrets.aws_credentials }} | base64 -d > ~/.aws/credentials
- name: Run checks
run: ./gradlew check

1
.idea/.name generated
View File

@@ -1 +0,0 @@
s3backup-tool

124
.idea/uiDesigner.xml generated
View File

@@ -1,124 +0,0 @@
<?xml version="1.0" encoding="UTF-8"?>
<project version="4">
<component name="Palette2">
<group name="Swing">
<item class="com.intellij.uiDesigner.HSpacer" tooltip-text="Horizontal Spacer" icon="/com/intellij/uiDesigner/icons/hspacer.svg" removable="false" auto-create-binding="false" can-attach-label="false">
<default-constraints vsize-policy="1" hsize-policy="6" anchor="0" fill="1" />
</item>
<item class="com.intellij.uiDesigner.VSpacer" tooltip-text="Vertical Spacer" icon="/com/intellij/uiDesigner/icons/vspacer.svg" removable="false" auto-create-binding="false" can-attach-label="false">
<default-constraints vsize-policy="6" hsize-policy="1" anchor="0" fill="2" />
</item>
<item class="javax.swing.JPanel" icon="/com/intellij/uiDesigner/icons/panel.svg" removable="false" auto-create-binding="false" can-attach-label="false">
<default-constraints vsize-policy="3" hsize-policy="3" anchor="0" fill="3" />
</item>
<item class="javax.swing.JScrollPane" icon="/com/intellij/uiDesigner/icons/scrollPane.svg" removable="false" auto-create-binding="false" can-attach-label="true">
<default-constraints vsize-policy="7" hsize-policy="7" anchor="0" fill="3" />
</item>
<item class="javax.swing.JButton" icon="/com/intellij/uiDesigner/icons/button.svg" removable="false" auto-create-binding="true" can-attach-label="false">
<default-constraints vsize-policy="0" hsize-policy="3" anchor="0" fill="1" />
<initial-values>
<property name="text" value="Button" />
</initial-values>
</item>
<item class="javax.swing.JRadioButton" icon="/com/intellij/uiDesigner/icons/radioButton.svg" removable="false" auto-create-binding="true" can-attach-label="false">
<default-constraints vsize-policy="0" hsize-policy="3" anchor="8" fill="0" />
<initial-values>
<property name="text" value="RadioButton" />
</initial-values>
</item>
<item class="javax.swing.JCheckBox" icon="/com/intellij/uiDesigner/icons/checkBox.svg" removable="false" auto-create-binding="true" can-attach-label="false">
<default-constraints vsize-policy="0" hsize-policy="3" anchor="8" fill="0" />
<initial-values>
<property name="text" value="CheckBox" />
</initial-values>
</item>
<item class="javax.swing.JLabel" icon="/com/intellij/uiDesigner/icons/label.svg" removable="false" auto-create-binding="false" can-attach-label="false">
<default-constraints vsize-policy="0" hsize-policy="0" anchor="8" fill="0" />
<initial-values>
<property name="text" value="Label" />
</initial-values>
</item>
<item class="javax.swing.JTextField" icon="/com/intellij/uiDesigner/icons/textField.svg" removable="false" auto-create-binding="true" can-attach-label="true">
<default-constraints vsize-policy="0" hsize-policy="6" anchor="8" fill="1">
<preferred-size width="150" height="-1" />
</default-constraints>
</item>
<item class="javax.swing.JPasswordField" icon="/com/intellij/uiDesigner/icons/passwordField.svg" removable="false" auto-create-binding="true" can-attach-label="true">
<default-constraints vsize-policy="0" hsize-policy="6" anchor="8" fill="1">
<preferred-size width="150" height="-1" />
</default-constraints>
</item>
<item class="javax.swing.JFormattedTextField" icon="/com/intellij/uiDesigner/icons/formattedTextField.svg" removable="false" auto-create-binding="true" can-attach-label="true">
<default-constraints vsize-policy="0" hsize-policy="6" anchor="8" fill="1">
<preferred-size width="150" height="-1" />
</default-constraints>
</item>
<item class="javax.swing.JTextArea" icon="/com/intellij/uiDesigner/icons/textArea.svg" removable="false" auto-create-binding="true" can-attach-label="true">
<default-constraints vsize-policy="6" hsize-policy="6" anchor="0" fill="3">
<preferred-size width="150" height="50" />
</default-constraints>
</item>
<item class="javax.swing.JTextPane" icon="/com/intellij/uiDesigner/icons/textPane.svg" removable="false" auto-create-binding="true" can-attach-label="true">
<default-constraints vsize-policy="6" hsize-policy="6" anchor="0" fill="3">
<preferred-size width="150" height="50" />
</default-constraints>
</item>
<item class="javax.swing.JEditorPane" icon="/com/intellij/uiDesigner/icons/editorPane.svg" removable="false" auto-create-binding="true" can-attach-label="true">
<default-constraints vsize-policy="6" hsize-policy="6" anchor="0" fill="3">
<preferred-size width="150" height="50" />
</default-constraints>
</item>
<item class="javax.swing.JComboBox" icon="/com/intellij/uiDesigner/icons/comboBox.svg" removable="false" auto-create-binding="true" can-attach-label="true">
<default-constraints vsize-policy="0" hsize-policy="2" anchor="8" fill="1" />
</item>
<item class="javax.swing.JTable" icon="/com/intellij/uiDesigner/icons/table.svg" removable="false" auto-create-binding="true" can-attach-label="false">
<default-constraints vsize-policy="6" hsize-policy="6" anchor="0" fill="3">
<preferred-size width="150" height="50" />
</default-constraints>
</item>
<item class="javax.swing.JList" icon="/com/intellij/uiDesigner/icons/list.svg" removable="false" auto-create-binding="true" can-attach-label="false">
<default-constraints vsize-policy="6" hsize-policy="2" anchor="0" fill="3">
<preferred-size width="150" height="50" />
</default-constraints>
</item>
<item class="javax.swing.JTree" icon="/com/intellij/uiDesigner/icons/tree.svg" removable="false" auto-create-binding="true" can-attach-label="false">
<default-constraints vsize-policy="6" hsize-policy="6" anchor="0" fill="3">
<preferred-size width="150" height="50" />
</default-constraints>
</item>
<item class="javax.swing.JTabbedPane" icon="/com/intellij/uiDesigner/icons/tabbedPane.svg" removable="false" auto-create-binding="true" can-attach-label="false">
<default-constraints vsize-policy="3" hsize-policy="3" anchor="0" fill="3">
<preferred-size width="200" height="200" />
</default-constraints>
</item>
<item class="javax.swing.JSplitPane" icon="/com/intellij/uiDesigner/icons/splitPane.svg" removable="false" auto-create-binding="false" can-attach-label="false">
<default-constraints vsize-policy="3" hsize-policy="3" anchor="0" fill="3">
<preferred-size width="200" height="200" />
</default-constraints>
</item>
<item class="javax.swing.JSpinner" icon="/com/intellij/uiDesigner/icons/spinner.svg" removable="false" auto-create-binding="true" can-attach-label="true">
<default-constraints vsize-policy="0" hsize-policy="6" anchor="8" fill="1" />
</item>
<item class="javax.swing.JSlider" icon="/com/intellij/uiDesigner/icons/slider.svg" removable="false" auto-create-binding="true" can-attach-label="false">
<default-constraints vsize-policy="0" hsize-policy="6" anchor="8" fill="1" />
</item>
<item class="javax.swing.JSeparator" icon="/com/intellij/uiDesigner/icons/separator.svg" removable="false" auto-create-binding="false" can-attach-label="false">
<default-constraints vsize-policy="6" hsize-policy="6" anchor="0" fill="3" />
</item>
<item class="javax.swing.JProgressBar" icon="/com/intellij/uiDesigner/icons/progressbar.svg" removable="false" auto-create-binding="true" can-attach-label="false">
<default-constraints vsize-policy="0" hsize-policy="6" anchor="0" fill="1" />
</item>
<item class="javax.swing.JToolBar" icon="/com/intellij/uiDesigner/icons/toolbar.svg" removable="false" auto-create-binding="false" can-attach-label="false">
<default-constraints vsize-policy="0" hsize-policy="6" anchor="0" fill="1">
<preferred-size width="-1" height="20" />
</default-constraints>
</item>
<item class="javax.swing.JToolBar$Separator" icon="/com/intellij/uiDesigner/icons/toolbarSeparator.svg" removable="false" auto-create-binding="false" can-attach-label="false">
<default-constraints vsize-policy="0" hsize-policy="0" anchor="0" fill="1" />
</item>
<item class="javax.swing.JScrollBar" icon="/com/intellij/uiDesigner/icons/scrollbar.svg" removable="false" auto-create-binding="true" can-attach-label="false">
<default-constraints vsize-policy="6" hsize-policy="0" anchor="0" fill="2" />
</item>
</group>
</component>
</project>

View File

@@ -3,90 +3,38 @@ This is a small backup utility for uploading/restoring a local directory to/from
an AWS S3 bucket.
## Usage
This tool is released as a JAR in the [releases page](https://git.koval.net/cyclane/teamcity-executors-test-task/releases).
Use `java -jar s3backup-tool-<version>.jar --help` for more detailed usage instructions.
### --help
```
Usage: s3backup-tool [<options>] <command> [<args>]...
A simple AWS S3 backup tool. This tool assumes credentials are properly configured using aws-cli.
Options:
-h, --help Show this message and exit
Commands:
create Create a backup of a file or directory.
restore Restore a backup from AWS S3.
restore-file Restore a single file from a backup from AWS S3.
```
#### Subcommands
```
Usage: s3backup-tool create [<options>] <source> <bucket>
Create a backup of a file or directory.
Options:
-h, --help Show this message and exit
Arguments:
<source> File or directory to backup
<bucket> Name of S3 bucket to backup to
```
```
Usage: s3backup-tool restore [<options>] <bucket> <backupkey> <destination>
Restore a backup from AWS S3.
Options:
-h, --help Show this message and exit
Arguments:
<bucket> Name of S3 bucket to restore the backup from
<backupkey> The S3 key of the backup to restore
<destination> Directory to restore to
```
```
Usage: s3backup-tool restore-file [<options>] <bucket> <backupkey> <filepath> <destination>
Restore a single file from a backup from AWS S3.
Options:
-h, --help Show this message and exit
Arguments:
<bucket> Name of S3 bucket to restore the backup from
<backupkey> The S3 key of the backup to restore
<filepath> File path within the backup
<destination> Directory to restore to
```
This tool is released as a JAR in the [release page](https://git.koval.net/cyclane/teamcity-executors-test-task/releases).
Use `java -jar <backup-jar-name>.jar --help` for more detailed usage instructions.
## Assumptions
1. The test task is not interested in re-implementations of common libraries (AWS SDK, Clikt, Gradle Shadow, ...)
2. The test task is more interested in Kotlin JVM (not Kotlin Native).
1. This test task is not interested in re-implementations of common libraries (AWS SDK, Clikt, Gradle Shadow, ...)
2. The last part (restoration of a single file) should be optimised so that only the part of the blob required for this
file is downloaded.
3. Only this tool is ever used to create backups, so S3 object keys are in the expected format, and ZIP files do not have
a comment in the *end of central directory* record (making it a predictable length of 22 bytes).
- EOCD64 should similarly not have a comment.
## Design decisions
- Backups may be large, so we want to use **multipart uploads** if possible (< 100mb is recommended).
- Backups may be large, so we want to use multipart uploads if possible (< 100mb is recommended).
https://docs.aws.amazon.com/AmazonS3/latest/userguide/mpuoverview.html
- The Java SDK has high-level support for this via [S3TransferManager](https://sdk.amazonaws.com/java/api/latest/software/amazon/awssdk/transfer/s3/S3TransferManager.html),
but unfortunately when the content is too small, the HTTP `Content-Length` is not automatically calculated resulting
in an error response from the API.
- I'm not sure whether this is intended behaviour or a bug, but decided to manually implement multipart uploads using
the Kotlin SDK instead anyway.
- **ZIP files** are used so that the backups can be stored in a very common format which also provides compression.
- Allows future expansion to allow for encryption as well.
- [Java ZIP specification](https://docs.oracle.com/en/java/javase/17/docs/api/java.base/java/util/zip/package-summary.html):
- **Note**: I could have just used a temporary file (with a known `Content-Length`), but I wanted to play around with
streams and kotlin concurrency a bit, which is why I went with the more scalable way using streams.
- Zip files are used so that the backups can be stored in a very common format which also provides compression.
- Java zip specification: https://docs.oracle.com/en/java/javase/17/docs/api/java.base/java/util/zip/package-summary.html
- ZIP64 implementation is optional, but possible, so we'll handle it.
- The End of Central Directory record is also useful for locating the exact positions of files in the blob, so that
single files can be downloaded using the HTTP `Range` header.
- End of Central Directory comment must be blank. Otherwise, the EOCD length is unpredictable, and so we
- End of Central Directory comment must be blank (assumption 3). Otherwise, the EOCD length is unpredictable and so we
cannot use just a single request the HTTP `Range` header to get the entire EOCD.
- *Alternative*: use S3 object tags to store the EOCD size, fallback to 22 bytes otherwise. This could be
interesting if we want the backup tool to be able to import existing ZIPs (which could potentially have a comment),
but that is beyond the scope of the instructions.
- Alternative: use S3 object tags to store the EOCD offset, but this way the blob itself would no longer contain all
the data required by this backup tool.
- Alternative: store the EOCD offset in the EOCD comment or the beginning of the file, but this makes a similar, but
more strict assumption anyway.
## Instructions
Create a backup utility that copies files to AWS S3. The utility should take a local directory with files and put it into AWS S3 in the form of one blob file. The reverse behavior should also be possible. We should be able to specify what backup we want to restore and where it should put the files on the local system. The utility should be able to restore one individual file from a backup.

View File

@@ -1,7 +1,7 @@
import com.github.jengelman.gradle.plugins.shadow.tasks.ShadowJar
plugins {
application
kotlin("jvm") version "1.9.21"
id("org.jlleitschuh.gradle.ktlint") version "12.0.3"
id("com.github.johnrengelman.shadow") version "8.1.1"
}
@@ -14,9 +14,8 @@ repositories {
dependencies {
implementation("aws.sdk.kotlin:s3:1.0.25")
implementation("org.slf4j:slf4j-simple:2.0.9") // AWS SDK wants a SLF4J provider.
implementation("org.slf4j:slf4j-simple:2.0.9")
implementation("org.jetbrains.kotlinx:kotlinx-coroutines-core:1.7.3")
implementation("com.github.ajalt.clikt:clikt:4.2.1")
testImplementation("org.jetbrains.kotlin:kotlin-test")
}
@@ -26,6 +25,8 @@ tasks.test {
kotlin {
jvmToolchain(17)
}
application {
mainClass.set("MainKt")
tasks.jar {
manifest {
attributes("Main-Class" to "backup.MainKt")
}
}

View File

@@ -1,4 +1,5 @@
plugins {
id("org.gradle.toolchains.foojay-resolver-convention") version "0.5.0"
}
rootProject.name = "s3backup-tool"
rootProject.name = "teamcity-executors-test-task"

View File

@@ -1,80 +0,0 @@
import aws.sdk.kotlin.services.s3.S3Client
import backup.BackupClient
import com.github.ajalt.clikt.core.CliktCommand
import com.github.ajalt.clikt.core.subcommands
import com.github.ajalt.clikt.parameters.arguments.argument
import com.github.ajalt.clikt.parameters.arguments.help
import com.github.ajalt.clikt.parameters.types.file
import kotlinx.coroutines.runBlocking
import kotlin.system.exitProcess
fun main(args: Array<String>) =
runBlocking {
S3Client.fromEnvironment().use { s3 ->
S3BackupTool()
.subcommands(
Create(s3),
Restore(s3),
RestoreFile(s3),
)
.main(args)
}
}
class S3BackupTool : CliktCommand(
help = "A simple AWS S3 backup tool. This tool assumes credentials are properly configured using aws-cli.",
) {
override fun run() {
shortHelp(currentContext)
}
}
class Create(val s3: S3Client) : CliktCommand(
help = "Create a backup of a file or directory.",
) {
val source by argument().file(mustExist = true).help("File or directory to backup")
val bucket by argument().help("Name of S3 bucket to backup to")
override fun run() =
runBlocking {
val backupKey = BackupClient(s3, bucket).upload(source)
echo("Successfully created backup with key '$backupKey'")
}
}
class Restore(val s3: S3Client) : CliktCommand(
help = "Restore a backup from AWS S3.",
) {
val bucket by argument().help("Name of S3 bucket to restore the backup from")
val backupKey by argument().help("The S3 key of the backup to restore")
val destination by argument().file(mustExist = true).help("Directory to restore to")
override fun run() =
runBlocking {
if (!destination.isDirectory) {
echo("Destination must be an existing directory", err = true)
exitProcess(1)
}
BackupClient(s3, bucket).restore(destination.toPath(), backupKey)
echo("Successfully restored backup '$backupKey' to '$destination'")
}
}
class RestoreFile(val s3: S3Client) : CliktCommand(
help = "Restore a single file from a backup from AWS S3.",
) {
val bucket by argument().help("Name of S3 bucket to restore the backup from")
val backupKey by argument().help("The S3 key of the backup to restore")
val filePath by argument().help("File path within the backup")
val destination by argument().file(mustExist = true).help("Directory to restore to")
override fun run() =
runBlocking {
if (!destination.isDirectory) {
echo("Destination must be an existing directory", err = true)
exitProcess(1)
}
BackupClient(s3, bucket).restoreFile(destination.toPath(), backupKey, filePath)
echo("Successfully restored '$filePath' from backup '$backupKey' to '$destination'")
}
}

View File

@@ -30,19 +30,17 @@ import kotlin.io.path.createDirectory
class BackupClient(
private val s3: S3Client,
private val bucketName: String,
private val bufSize: Int = 1024 * 1024 * 32,
private val bufSize: Int = 1024 * 1024 * 100
) {
/**
* Upload a file/directory backup to AWS S3.
* @param file The File object for the file or directory.
*/
suspend fun upload(file: File) =
coroutineScope {
val backupKey = "${file.canonicalFile.name}/${Instant.now()}.zip"
suspend fun upload(file: File) = coroutineScope {
val backupKey = "${file.name}/${Instant.now()}.zip"
PipedInputStream().use { inputStream ->
val outputStream = PipedOutputStream(inputStream)
val zipper =
launch(Dispatchers.IO) {
val zipper = launch(Dispatchers.IO) {
file.compressToZip(outputStream)
}
@@ -52,8 +50,7 @@ class BackupClient(
// Large upload, use multipart
// TODO: multipart uploads can be asynchronous, which would improve
// performance a little bit for big uploads.
val upload =
s3.createMultipartUpload {
val upload = s3.createMultipartUpload {
bucket = bucketName
key = backupKey
}
@@ -62,8 +59,7 @@ class BackupClient(
var number = 1
var bytesRead = initialRead
while (bytesRead > 0) {
val part =
s3.uploadPart {
val part = s3.uploadPart {
bucket = bucketName
key = backupKey
partNumber = number
@@ -78,8 +74,7 @@ class BackupClient(
bucket = bucketName
key = backupKey
uploadId = upload.uploadId
multipartUpload =
CompletedMultipartUpload {
multipartUpload = CompletedMultipartUpload {
parts = uploadParts
}
}
@@ -109,19 +104,15 @@ class BackupClient(
* @param destination The destination directory path for the backup contents.
* @param backupKey The S3 key of the backup.
*/
suspend fun restore(
destination: Path,
backupKey: String,
) = coroutineScope {
val req =
GetObjectRequest {
suspend fun restore(destination: Path, backupKey: String) = coroutineScope {
val req = GetObjectRequest {
bucket = bucketName
key = backupKey
}
s3.getObject(req) { resp ->
ZipInputStream(
resp.body?.toInputStream()
?: throw IOException("S3 response is missing body"),
?: throw IOException("S3 response is missing body")
).use { zipStream ->
zipStream.decompress { destination.resolve(it) }
}
@@ -134,14 +125,9 @@ class BackupClient(
* @param backupKey The S3 key of the backup.
* @param fileName The full name of the file to restore (including directories if it was under a subdirectory).
*/
suspend fun restoreFile(
destination: Path,
backupKey: String,
fileName: String,
) = coroutineScope {
suspend fun restoreFile(destination: Path, backupKey: String, fileName: String) = coroutineScope {
// For byte ranges refer to https://pkware.cachefly.net/webdocs/APPNOTE/APPNOTE-6.3.9.TXT
val eocdReq =
GetObjectRequest {
val eocdReq = GetObjectRequest {
bucket = bucketName
key = backupKey
// Assumption: EOCD has an empty comment
@@ -149,17 +135,14 @@ class BackupClient(
// in which case this function would error anyway, so it should be fine to have this edge-case.
range = "bytes=-${EndOfCentralDirectoryRecord.SIZE + EndOfCentralDirectoryLocator.SIZE}"
}
val eocdBytes =
s3.getObject(eocdReq) { resp ->
val eocdBytes = s3.getObject(eocdReq) { resp ->
val bytes = resp.body?.toByteArray() ?: throw IOException("S3 response is missing body")
bytes
}
val eocd = EndOfCentralDirectoryRecord.fromByteArray(eocdBytes, EndOfCentralDirectoryLocator.SIZE)
val eocd64 =
if (eocd.eocd64Required()) {
val eocd64 = if (eocd.eocd64Required()) {
val locator = EndOfCentralDirectoryLocator.fromByteArray(eocdBytes, 0)
val eocd64Req =
GetObjectRequest {
val eocd64Req = GetObjectRequest {
bucket = bucketName
key = backupKey
range = "bytes=${locator.endOfCentralDirectory64Offset}-"
@@ -168,25 +151,18 @@ class BackupClient(
val bytes = resp.body?.toByteArray() ?: throw IOException("S3 response is missing body")
EndOfCentralDirectoryRecord64.fromByteArray(bytes, 0)
}
} else {
null
}
val cenOffset =
if (eocd.centralDirectoryOffset == 0xffffffffU && eocd64 != null) {
} else null
val cenOffset = if (eocd.centralDirectoryOffset == 0xffffffffU && eocd64 != null) {
eocd64.centralDirectoryOffset
} else {
eocd.centralDirectoryOffset.toULong()
}
val censReq =
GetObjectRequest {
} else eocd.centralDirectoryOffset.toULong()
val censReq = GetObjectRequest {
bucket = bucketName
key = backupKey
// We only know where to fetch until if we've also fetched EOCD64 (which isn't always the case).
// So just over-fetch a little bit, these headers aren't that big anyway.
range = "bytes=$cenOffset-"
range = "bytes=${cenOffset}-"
}
val cen =
s3.getObject(censReq) { resp ->
val cen = s3.getObject(censReq) { resp ->
val bytes = resp.body?.toByteArray() ?: throw IOException("S3 response is missing body")
var p = 0
while (p < bytes.size) {
@@ -194,37 +170,36 @@ class BackupClient(
val cen = CentralDirectoryFileHeader.fromByteArray(bytes, p)
p += cen.size
if (cen.fileName == fileName) return@getObject cen
} catch (_: InvalidDataException) {
} catch (_: InvalidSignatureException) {
return@getObject null
}
}
null
} ?: throw FileNotFoundException("File '$fileName' not found in backup")
} ?: throw FileNotFoundException("File '${fileName}' not found in backup")
val localHeaderOffset =
cen.extraFieldRecords.firstNotNullOfOrNull {
val localHeaderOffset = cen.extraFieldRecords.firstNotNullOfOrNull {
if (it is Zip64ExtraFieldRecord && it.localHeaderOffset != null) it else null
}?.localHeaderOffset ?: cen.localHeaderOffset.toULong()
val compressedSize =
cen.extraFieldRecords.firstNotNullOfOrNull {
val compressedSize = cen.extraFieldRecords.firstNotNullOfOrNull {
if (it is Zip64ExtraFieldRecord && it.compressedSize != null) it else null
}?.compressedSize ?: cen.compressedSize.toULong()
val req =
GetObjectRequest {
val req = GetObjectRequest {
bucket = bucketName
key = backupKey
range = "bytes=$localHeaderOffset-${
// Over-fetch a bit so that ZipInputStream can see the next header (otherwise it EOFs, even though
// it knows exactly how much data is should read, so not sure why it reads ahead).
range = "bytes=${localHeaderOffset}-${
// Add CEN min size (46 bytes) so that the next CEN / LOC header is seen by the ZipInputStream
// and so it can see the current entry has stopped.
// Note: yes ZipInputStream should know the exact content length from the LOC, but it was still sending
// EOF errors. Perhaps due to fetching multiples of a power of two, or something else. But this helps.
localHeaderOffset + cen.size.toULong() + compressedSize + CentralDirectoryFileHeader.SIZE.toULong()
}"
}
s3.getObject(req) { resp ->
ZipInputStream(
resp.body?.toInputStream()
?: throw IOException("S3 response is missing body"),
?: throw IOException("S3 response is missing body")
).use { zipStream ->
zipStream.decompress(limit = 1) { name -> destination.resolve(name.takeLastWhile { it != '/' }) }
zipStream.decompress { name -> destination.resolve(name.takeLastWhile { it != '/' }) }
}
}
}
@@ -235,7 +210,7 @@ class BackupClient(
* @param number The part number that was used for this part upload.
* @return The CompletedPart object.
*/
internal fun UploadPartResponse.toCompletedPart(number: Int): CompletedPart {
private fun UploadPartResponse.toCompletedPart(number: Int): CompletedPart {
val part = this
return CompletedPart {
partNumber = number
@@ -252,24 +227,20 @@ internal fun UploadPartResponse.toCompletedPart(number: Int): CompletedPart {
* @param n The number of items to take.
* @return A ByteArray of the first `n` items.
*/
internal fun ByteArray.take(n: Int) =
if (n == size) {
this // No copy
} else {
asList().subList(0, n).toByteArray() // TODO: One copy (toByteArray()), not sure how to do 0 copies here
}
private fun ByteArray.take(n: Int) =
if (n == size) this // No copy
else asList().subList(0, n).toByteArray() // TODO: One copy (toByteArray()), not sure how to do 0 copies here
/**
* Compress a file or directory as a ZIP file to an `OutputStream`.
* @param outputStream The `OutputStream` to write the ZIP file contents to.
*/
internal fun File.compressToZip(outputStream: OutputStream) =
ZipOutputStream(outputStream).use { zipStream ->
val parentDir = this.canonicalFile.parent + "/"
private fun File.compressToZip(outputStream: OutputStream) = ZipOutputStream(outputStream).use { zipStream ->
val parentDir = this.absoluteFile.parent + "/"
val fileQueue = ArrayDeque<File>()
fileQueue.add(this)
fileQueue.forEach { subFile ->
val path = subFile.canonicalPath.removePrefix(parentDir)
val path = subFile.absolutePath.removePrefix(parentDir)
val subFiles = subFile.listFiles()
if (subFiles != null) { // Is a directory
val entry = ZipEntry("$path/")
@@ -292,13 +263,11 @@ internal fun File.compressToZip(outputStream: OutputStream) =
* @param bufSize The buffer size to use for writing the decompressed files.
* @param entryNameToPath A function to convert ZIP entry names to destination `Path`s.
*/
internal fun ZipInputStream.decompress(
private fun ZipInputStream.decompress(
bufSize: Int = 1024 * 1024,
limit: Int? = null,
entryNameToPath: (String) -> Path,
entryNameToPath: (String) -> Path
) {
var entry = this.nextEntry
var count = 1
while (entry != null) {
val path = entryNameToPath(entry.name)
if (entry.isDirectory) {
@@ -314,9 +283,6 @@ internal fun ZipInputStream.decompress(
}
}
applyZipAttributes(entry, path)
// This is here, not in while loop, since we do not want to read more from the input stream.
// But this.nextEntry will read from the input stream.
if (limit != null && count++ >= limit) return
entry = this.nextEntry
}
}
@@ -326,10 +292,7 @@ internal fun ZipInputStream.decompress(
* @param entry The `ZipEntry` to set attributes of.
* @param path The `Path` of the file to get the attributes from.
*/
internal fun setZipAttributes(
entry: ZipEntry,
path: Path,
) {
private fun setZipAttributes(entry: ZipEntry, path: Path) {
try {
val attrs = Files.getFileAttributeView(path, BasicFileAttributeView::class.java).readAttributes()
entry.setCreationTime(attrs.creationTime())
@@ -344,10 +307,7 @@ internal fun setZipAttributes(
* @param entry The `ZipEntry` to get the attributes from.
* @param path The `Path` of the file to set the attributes of.
*/
internal fun applyZipAttributes(
entry: ZipEntry,
path: Path,
) {
private fun applyZipAttributes(entry: ZipEntry, path: Path) {
try {
val attrs = Files.getFileAttributeView(path, BasicFileAttributeView::class.java)
attrs.setTimes(entry.lastModifiedTime, entry.lastAccessTime, entry.creationTime)

View File

@@ -0,0 +1,10 @@
package backup
import aws.sdk.kotlin.services.s3.S3Client
import kotlinx.coroutines.runBlocking
fun main() = runBlocking {
S3Client.fromEnvironment().use { s3 ->
val backupClient = BackupClient(s3, "teamcity-executors-test-task", 1024 * 1024 * 10)
}
}

View File

@@ -15,7 +15,7 @@ internal class CentralDirectoryFileHeader(
val disk: UShort,
val localHeaderOffset: UInt,
val fileName: String,
val extraFieldRecords: List<ExtraFieldRecord>,
val extraFieldRecords: List<ExtraFieldRecord>
) {
val size: Int
get() = SIZE + nameLength.toInt() + extraFieldLength.toInt() + commentLength.toInt()
@@ -32,85 +32,65 @@ internal class CentralDirectoryFileHeader(
* @return A `CentralDirectoryFileHeader`.
*/
@Throws(InvalidDataException::class)
fun fromByteArray(
data: ByteArray,
offset: Int,
): CentralDirectoryFileHeader {
fun fromByteArray(data: ByteArray, offset: Int): CentralDirectoryFileHeader {
if (data.size - offset < SIZE) {
throw InvalidDataException("CEN must be at least 46 bytes")
}
val buf = ByteBuffer.wrap(data, offset, 46).order(ByteOrder.LITTLE_ENDIAN)
if (buf.getInt().toUInt() != SIGNATURE) {
throw InvalidDataException("Invalid signature")
throw InvalidSignatureException("Invalid signature")
}
val extraFieldRecords = mutableListOf<ExtraFieldRecord>()
val nameLength = buf.getShort(offset + 28).toUShort()
buf.position(offset + 20)
val cen =
CentralDirectoryFileHeader(
val cen = CentralDirectoryFileHeader(
compressedSize = buf.getInt().toUInt(),
uncompressedSize = buf.getInt().toUInt(),
nameLength =
nameLength
nameLength = nameLength
.also { buf.position(offset + 30) },
extraFieldLength = buf.getShort().toUShort(),
commentLength = buf.getShort().toUShort(),
disk =
buf.getShort().toUShort()
disk = buf.getShort().toUShort()
.also { buf.position(offset + 42) },
localHeaderOffset = buf.getInt().toUInt(),
fileName = String(data.sliceArray(offset + SIZE..<offset + SIZE + nameLength.toInt())),
extraFieldRecords = extraFieldRecords,
extraFieldRecords = extraFieldRecords
)
if (data.size - offset < cen.size) {
throw InvalidDataException("CEN is too short")
}
// Parse extra field records
val extraFieldsBuf =
ByteBuffer.wrap(
data,
offset + SIZE + cen.nameLength.toInt(),
cen.extraFieldLength.toInt(),
val extraFieldsBuf = ByteBuffer.wrap(
data, offset + SIZE + cen.nameLength.toInt(), cen.extraFieldLength.toInt()
).order(ByteOrder.LITTLE_ENDIAN)
while (extraFieldsBuf.remaining() > 0) {
val id = extraFieldsBuf.getShort().toUShort()
val size = extraFieldsBuf.getShort().toUShort()
extraFieldRecords.add(
when (id) {
extraFieldRecords.add(when (id) {
Zip64ExtraFieldRecord.ID -> {
Zip64ExtraFieldRecord(
size,
if (cen.uncompressedSize == 0xffffffffU) {
extraFieldsBuf.getLong().toULong()
} else {
null
},
} else null,
if (cen.compressedSize == 0xffffffffU) {
extraFieldsBuf.getLong().toULong()
} else {
null
},
} else null,
if (cen.localHeaderOffset == 0xffffffffU) {
extraFieldsBuf.getLong().toULong()
} else {
null
},
} else null,
if (cen.disk == 0xffffU.toUShort()) {
extraFieldsBuf.getInt().toUInt()
} else {
null
},
} else null
)
}
else -> {
extraFieldsBuf.position(extraFieldsBuf.position() + size.toInt())
ExtraFieldRecord(id, size)
}
},
)
})
}
return cen

View File

@@ -7,12 +7,11 @@ import java.nio.ByteOrder
* Represents a partial ZIP64 end of central directory locator.
*/
internal class EndOfCentralDirectoryLocator(
val endOfCentralDirectory64Offset: ULong,
val endOfCentralDirectory64Offset: ULong
) {
companion object {
const val SIGNATURE = 0x07064b50U
const val SIZE = 20
/**
* Create `EndOfCentralDirectoryLocator` from raw byte data.
* @throws InvalidDataException Provided `ByteArray` is not a supported EOCD locator.
@@ -21,16 +20,13 @@ internal class EndOfCentralDirectoryLocator(
* @return A `EndOfCentralDirectoryLocator`.
*/
@Throws(InvalidDataException::class)
fun fromByteArray(
data: ByteArray,
offset: Int,
): EndOfCentralDirectoryLocator {
fun fromByteArray(data: ByteArray, offset: Int): EndOfCentralDirectoryLocator {
if (data.size - offset < SIZE) {
throw InvalidDataException("EOCD64 locator must be at least 20 bytes")
}
val buf = ByteBuffer.wrap(data, offset, SIZE).order(ByteOrder.LITTLE_ENDIAN)
if (buf.getInt().toUInt() != SIGNATURE) {
throw InvalidDataException("Invalid signature")
throw InvalidSignatureException("Invalid signature")
}
buf.position(offset + 8)
return EndOfCentralDirectoryLocator(buf.getLong().toULong())

View File

@@ -7,14 +7,14 @@ import java.nio.ByteOrder
* Represents a partial ZIP end of central directory record.
*/
internal class EndOfCentralDirectoryRecord(
val centralDirectoryOffset: UInt,
val centralDirectoryOffset: UInt
) {
fun eocd64Required(): Boolean = centralDirectoryOffset == 0xffffffffU
fun eocd64Required(): Boolean =
centralDirectoryOffset == 0xffffffffU
companion object {
const val SIGNATURE = 0x06054b50U
const val SIZE = 22
/**
* Create `EndOfCentralDirectoryRecord` from raw byte data.
* @throws InvalidDataException Provided `ByteArray` is not a supported EOCD64.
@@ -23,20 +23,17 @@ internal class EndOfCentralDirectoryRecord(
* @return A `EndOfCentralDirectoryRecord`.
*/
@Throws(InvalidDataException::class)
fun fromByteArray(
data: ByteArray,
offset: Int,
): EndOfCentralDirectoryRecord {
fun fromByteArray(data: ByteArray, offset: Int): EndOfCentralDirectoryRecord {
if (data.size - offset < SIZE) {
throw InvalidDataException("EOCD must be at least 22 bytes")
}
val buf = ByteBuffer.wrap(data, offset, SIZE).order(ByteOrder.LITTLE_ENDIAN)
if (buf.getInt().toUInt() != SIGNATURE) {
throw InvalidDataException("Invalid signature")
throw InvalidSignatureException("Invalid signature")
}
buf.position(offset + 16)
return EndOfCentralDirectoryRecord(
centralDirectoryOffset = buf.getInt().toUInt(),
centralDirectoryOffset = buf.getInt().toUInt()
)
}
}

View File

@@ -7,12 +7,11 @@ import java.nio.ByteOrder
* Represents a partial ZIP64 end of central directory record.
*/
internal class EndOfCentralDirectoryRecord64(
val centralDirectoryOffset: ULong,
val centralDirectoryOffset: ULong
) {
companion object {
const val SIGNATURE = 0x06064b50U
const val SIZE = 56
/**
* Create `EndOfCentralDirectoryRecord64` from raw byte data.
* @throws InvalidDataException Provided `ByteArray` is not a supported EOCD.
@@ -21,20 +20,17 @@ internal class EndOfCentralDirectoryRecord64(
* @return A `EndOfCentralDirectoryRecord64`.
*/
@Throws(InvalidDataException::class)
fun fromByteArray(
data: ByteArray,
offset: Int,
): EndOfCentralDirectoryRecord64 {
fun fromByteArray(data: ByteArray, offset: Int): EndOfCentralDirectoryRecord64 {
if (data.size - offset < SIZE) {
throw InvalidDataException("EOCD64 must be at least 56 bytes")
}
val buf = ByteBuffer.wrap(data, offset, SIZE).order(ByteOrder.LITTLE_ENDIAN)
if (buf.getInt().toUInt() != SIGNATURE) {
throw InvalidDataException("Invalid signature")
throw InvalidSignatureException("Invalid signature")
}
buf.position(offset + 48)
return EndOfCentralDirectoryRecord64(
centralDirectoryOffset = buf.getLong().toULong(),
centralDirectoryOffset = buf.getLong().toULong()
)
}
}

View File

@@ -0,0 +1,11 @@
package ziputils
/**
* Represents an invalid raw byte data exception.
*/
class InvalidDataException(message: String): Exception(message)
/**
* Represents an invalid raw byte signature exception.
*/
class InvalidSignatureException(message: String): Exception(message)

View File

@@ -5,5 +5,5 @@ package ziputils
*/
internal open class ExtraFieldRecord(
val id: UShort,
val size: UShort,
val size: UShort
)

View File

@@ -1,6 +0,0 @@
package ziputils
/**
* Represents an invalid raw byte data exception.
*/
class InvalidDataException(message: String) : Exception(message)

View File

@@ -8,7 +8,7 @@ internal class Zip64ExtraFieldRecord(
val uncompressedSize: ULong?,
val compressedSize: ULong?,
val localHeaderOffset: ULong?,
val disk: UInt?,
val disk: UInt?
): ExtraFieldRecord(ID, size) {
companion object {
const val ID: UShort = 0x0001U

View File

@@ -1,144 +0,0 @@
package backup
import aws.sdk.kotlin.services.s3.*
import kotlinx.coroutines.runBlocking
import org.junit.jupiter.api.*
import org.junit.jupiter.api.Assertions.*
import java.io.File
import kotlin.io.path.*
val bucketName = System.getenv("BACKUP_BUCKET") ?: "teamcity-executors-test-task"
class BackupClientTest {
lateinit var s3: S3Client
lateinit var backupClient: BackupClient
@BeforeEach
fun `before each`() =
runBlocking {
s3 = S3Client.fromEnvironment {}
backupClient = BackupClient(s3, bucketName)
}
@AfterEach
fun `after each`() {
s3.close()
}
@TestFactory
fun `round-trip tests`() =
listOf(
"empty directory" to {
listOf(
Path("_test").createDirectory(),
)
},
"single file" to {
listOf(
Path("_test.txt").apply { writeText("Hello World!") },
)
},
"directory structure" to {
listOf(
Path("_test").createDirectory(),
Path("_test/a.txt").apply { writeText("This is file A!\nAnother line here.") },
Path("_test/folder").createDirectory(),
Path("_test/another-folder").createDirectory(),
Path("_test/another-folder/b").apply { writeText("This is file B\n") },
Path("_test/another-folder/c.txt").createFile(),
Path("_test/README.md").apply { writeText("# This is a test directory structure.") },
)
},
).map { (name, pathsGen) ->
DynamicTest.dynamicTest(name) {
val paths = pathsGen()
val backupKey =
assertDoesNotThrow("should upload files") {
runBlocking {
backupClient.upload(paths.first().toFile())
}
}
val restoreDir = Path("_test_restore").createDirectory()
assertDoesNotThrow("should recover files") {
runBlocking {
backupClient.restore(restoreDir, backupKey)
}
}
assertEquals(
paths.size,
restoreDir.toFile().countEntries() - 1,
"number of files in backup restore should be equal to original",
)
val individualRestoreDir = Path("_test_restore_individual").createDirectory()
paths.forEach { path ->
if (path.isDirectory()) {
individualRestoreDir.resolve(path).createDirectory()
} else {
assertDoesNotThrow("should recover file '$path'") {
runBlocking {
backupClient.restoreFile(
individualRestoreDir.resolve(path).parent,
backupKey,
path.toString(),
)
}
}
}
}
assertEquals(
paths.size,
individualRestoreDir.toFile().countEntries() - 1,
"number of files in individual backup restore should be equal to original",
)
paths.asReversed().forEach { path ->
val restorePath = restoreDir.resolve(path)
val individualRestorePath = individualRestoreDir.resolve(path)
if (path.isDirectory()) {
assertTrue(restorePath.exists(), "'$path' should exist in backup")
assertTrue(restorePath.isDirectory(), "'$path' should be a directory in backup")
assertTrue(individualRestorePath.exists(), "'$path' should exist in backup (individual)")
assertTrue(
individualRestorePath.isDirectory(),
"'$path' should be a directory in backup (individual)",
)
} else {
val originalBytes = path.toFile().readBytes().asList()
assertEquals(
originalBytes,
restorePath.toFile().readBytes().asList(),
"File contents of '$path' should equal",
)
assertEquals(
originalBytes,
individualRestorePath.toFile().readBytes().asList(),
"File contents of '$path' should equal (individual)",
)
}
// cleanup
path.deleteExisting()
restorePath.deleteExisting()
individualRestorePath.deleteExisting()
}
// cleanup
restoreDir.deleteExisting()
individualRestoreDir.deleteExisting()
runBlocking {
s3.deleteObject {
bucket = bucketName
key = backupKey
}
}
}
}
}
internal fun File.countEntries(): Int {
val queue = ArrayDeque<File>()
queue.add(this)
return queue.count { file ->
if (file.isDirectory) {
queue.addAll(file.listFiles()!!) // shouldn't ever be null, since we know it's a directory
}
true
}
}