Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
144 changes: 144 additions & 0 deletions Rules/AvoidUsingArrayList.cs
Original file line number Diff line number Diff line change
@@ -0,0 +1,144 @@
// Copyright (c) Microsoft Corporation. All rights reserved.
// Licensed under the MIT License.

using Microsoft.Windows.PowerShell.ScriptAnalyzer.Generic;
using System;
using System.Collections.Generic;
using System.Globalization;
using System.Management.Automation.Language;
using System.Text.RegularExpressions;
using System.ComponentModel;
Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

System.ComponentModel is unused and can be removed



#if !CORECLR
using System.ComponentModel.Composition;
#endif

namespace Microsoft.Windows.PowerShell.ScriptAnalyzer.BuiltinRules
{
#if !CORECLR
[Export(typeof(IScriptRule))]
#endif

/// <summary>
/// Rule that warns when the ArrayList class is used in a PowerShell script.
/// </summary>
public class AvoidUsingArrayListAsFunctionNames : IScriptRule
Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The class name has a little copy-pasta 😊

Also, I think this should be a configurable rule, disabled by default.

Instead of directly implementing IScriptRule, implement ConfigurableRule. See AvoidExclaimOperator as an example.

{

/// <summary>
/// Analyzes the PowerShell AST for uses of the ArrayList class.
/// </summary>
/// <param name="ast">The PowerShell Abstract Syntax Tree to analyze.</param>
/// <param name="fileName">The name of the file being analyzed (for diagnostic reporting).</param>
/// <returns>A collection of diagnostic records for each violation.</returns>

public IEnumerable<DiagnosticRecord> AnalyzeScript(Ast ast, string fileName)
{
if (ast == null) { throw new ArgumentNullException(Strings.NullAstErrorMessage); }

// If there is an using statement for the Collections namespace, check for the full typename.
// Otherwise also check for the bare ArrayList name.
Regex ArrayListName = null;
Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

C# convention is camelCase for local variables (arrayListName).

var sbAst = ast as ScriptBlockAst;
foreach (UsingStatementAst usingAst in sbAst.UsingStatements)
{
if (
usingAst.UsingStatementKind == UsingStatementKind.Namespace &&
(
usingAst.Name.Value.Equals("Collections", StringComparison.OrdinalIgnoreCase) ||
usingAst.Name.Value.Equals("System.Collections", StringComparison.OrdinalIgnoreCase)
)
)
{
ArrayListName = new Regex(@"^((System\.)?Collections\.)?ArrayList$", RegexOptions.IgnoreCase);
break;
}
}
if (ArrayListName == null) { ArrayListName = new Regex(@"^(System\.)?Collections\.ArrayList", RegexOptions.IgnoreCase); }
Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think this is missing the closing anchor?

Suggested change
if (ArrayListName == null) { ArrayListName = new Regex(@"^(System\.)?Collections\.ArrayList", RegexOptions.IgnoreCase); }
if (ArrayListName == null) { ArrayListName = new Regex(@"^(System\.)?Collections\.ArrayList$", RegexOptions.IgnoreCase); }


// Find all type initializers that create a new instance of the ArrayList class.
IEnumerable<Ast> typeAsts = ast.FindAll(testAst =>
(
testAst is ConvertExpressionAst convertAst &&
convertAst.StaticType != null &&
convertAst.StaticType.FullName == "System.Collections.ArrayList"
) ||
(
testAst is TypeExpressionAst typeAst &&
typeAst.TypeName != null &&
ArrayListName.IsMatch(typeAst.TypeName.Name) &&
typeAst.Parent is InvokeMemberExpressionAst parentAst &&
parentAst.Member != null &&
parentAst.Member is StringConstantExpressionAst memberAst &&
memberAst.Value.Equals("new", StringComparison.OrdinalIgnoreCase)
),
true
);

foreach (Ast typeAst in typeAsts)
{
yield return new DiagnosticRecord(
string.Format(
CultureInfo.CurrentCulture,
Strings.AvoidUsingArrayListError,
typeAst.Parent.Extent.Text),
typeAst.Extent,
GetName(),
DiagnosticSeverity.Warning,
fileName
);
}

// Find all New-Object cmdlets that create a new instance of the ArrayList class.
var newObjectCommands = ast.FindAll(testAst =>
testAst is CommandAst cmdAst &&
cmdAst.GetCommandName() != null &&
cmdAst.GetCommandName().Equals("New-Object", StringComparison.OrdinalIgnoreCase),
true);

foreach (CommandAst cmd in newObjectCommands)
{
// Use StaticParameterBinder to reliably get parameter values
var bindingResult = StaticParameterBinder.BindCommand(cmd, true);

// Check for -TypeName parameter
if (
bindingResult.BoundParameters.ContainsKey("TypeName") &&
ArrayListName.IsMatch(bindingResult.BoundParameters["TypeName"].ConstantValue as string)
Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

ConstantValue could be null here - i.e. if it's a variable expression and not a compile-time constant.

$type = "System.Collections.ArrayList"
New-Object -TypeName $type

If it were, IsMatch() would not be a happy camper.

Take a look at AvoidGlobalAliases and UseConstrainedLanguageMode for how they handle it.

)
{
yield return new DiagnosticRecord(
string.Format(
CultureInfo.CurrentCulture,
Strings.AvoidUsingArrayListError,
cmd.Extent.Text),
bindingResult.BoundParameters["TypeName"].Value.Extent,
GetName(),
DiagnosticSeverity.Warning,
fileName
);
}

}


}

public string GetCommonName() => Strings.AvoidUsingArrayListCommonName;

public string GetDescription() => Strings.AvoidUsingArrayListDescription;

public string GetName() => string.Format(
CultureInfo.CurrentCulture,
Strings.NameSpaceFormat,
GetSourceName(),
Strings.AvoidUsingArrayListName);

public RuleSeverity GetSeverity() => RuleSeverity.Warning;

public string GetSourceName() => Strings.SourceName;

public SourceType GetSourceType() => SourceType.Builtin;
}
}
12 changes: 12 additions & 0 deletions Rules/Strings.resx
Original file line number Diff line number Diff line change
Expand Up @@ -932,6 +932,18 @@
</data>
<data name="AvoidSemicolonsAsLineTerminatorsError" xml:space="preserve">
<value>Line ends with a semicolon</value>
</data>
<data name="AvoidUsingArrayListCommonName" xml:space="preserve">
<value>Avoid using the ArrayList class</value>
</data>
<data name="AvoidUsingArrayListDescription" xml:space="preserve">
<value>Avoid using the ArrayList class in PowerShell scripts. Consider using generic collections or fixed arrays instead.</value>
</data>
<data name="AvoidUsingArrayListName" xml:space="preserve">
<value>AvoidUsingArrayList</value>
</data>
<data name="AvoidUsingArrayListError" xml:space="preserve">
<value>The ArrayList class is used in '{0}'. Consider using a generic collection or a fixed array instead.</value>
</data>
<data name="PlaceOpenBraceName" xml:space="preserve">
<value>PlaceOpenBrace</value>
Expand Down
18 changes: 18 additions & 0 deletions Tests/Rules/AvoidUsingArrayList.ps1
Original file line number Diff line number Diff line change
@@ -0,0 +1,18 @@
using namespace system.collections

# Using New-Object
$List = New-Object ArrayList
$List = New-Object 'ArrayList'
$List = New-Object "ArrayList"
$List = New-Object -Type ArrayList
$List = New-Object -TypeName ArrayLIST
$List = New-Object Collections.ArrayList
$List = New-Object System.Collections.ArrayList

# Using type initializer
$List = [ArrayList](1,2,3)
$List = [ArrayLIST]@(1,2,3)
$List = [ArrayList]::new()
$List = [Collections.ArrayList]::New()
$List = [System.Collections.ArrayList]::new()
1..3 | ForEach-Object { $null = $List.Add($_) }
26 changes: 26 additions & 0 deletions Tests/Rules/AvoidUsingArrayList.tests.ps1
Original file line number Diff line number Diff line change
@@ -0,0 +1,26 @@
# Copyright (c) Microsoft Corporation. All rights reserved.
# Licensed under the MIT License.

BeforeAll {
$ruleName = "PSAvoidArrayList"
$ruleMessage = "The ArrayList class is used in '*'. Consider using a generic collection or a fixed array instead."
}

Describe "AvoidUsingWriteHost" {
Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Suggested change
Describe "AvoidUsingWriteHost" {
Describe "AvoidArrayList" {

Context "When there are violations" {
$violations = Invoke-ScriptAnalyzer -ScriptDefinition $scriptDefinition -IncludeRule @($ruleName)
Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

You've clearly put a lot of thought into the various edge cases, but there's some issues with how the tests are set up - they don't currently run.

Take a look at something like AvoidExclaimOperator.tests.ps1 as an example of how to structure the tests.

With the current approach, a file of lots of violations and another with none, you're really only writing 2 tests.

If something changes in the future that breaks your rule, CI will just tell you that one (or both) of those tests no longer passes. e.g. That you got 11 violations instead of 12. Some troubleshooting would then be needed to work out what case no longer works.

I'd really recommend writing more scoped tests.

Describe "AvoidArrayList" {
    Context "When using New-Object with ArrayList passed to TypeName" {
        It "Should find a violation" {
            $def = '$List = New-Object -TypeName ArrayLIST'
            $violations = Invoke-ScriptAnalyzer -ScriptDefinition $def
            $violations.Count | Should -Be 1
        }
    }
}

LLMs are fairly good at writing them - they just need the right input and examples.

It "has ArrayList violations" {
$violations.Count | Should -Be 12
}

It "has the correct description message" {
$violations[0].Message | Should -Like $ruleMessage
}
}

Context "When there are no violations" {
It "returns no violations" {
$noViolations.Count | Should -Be 0
}
}
}
19 changes: 19 additions & 0 deletions Tests/Rules/AvoidUsingArrayListNoViolations.ps1
Original file line number Diff line number Diff line change
@@ -0,0 +1,19 @@
using namespace System.Collections.Generic

# Using a generic List
$List = New-Object List[Object]
1..3 | ForEach-Object { $List.Add($_) } # This will not return anything

$List = [List[Object]]::new()
1..3 | ForEach-Object { $List.Add($_) } # This will not return anything

# Creating a fixed array by using the PowerShell pipeline
$List = 1..3 | ForEach-Object { $_ }

# This should not violate because there isn't a
# `using namespace System.Collections` directive
# and ArrayList could belong to another namespace
$List = New-Object ArrayList
$List = [ArrayList](1,2,3)
$List = [ArrayList]@(1,2,3)
$List = [ArrayList]::new()
52 changes: 52 additions & 0 deletions docs/Rules/AvoidUsingArrayList.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,52 @@
---
description: Avoid using ArrayList
ms.date: 04/16/2025
ms.topic: reference
title: AvoidUsingArrayList
---
# AvoidUsingArrayList

**Severity Level: Warning**

## Description

Per dotnet best practices, the
[`ArrayList` class](https://learn.microsoft.com/dotnet/api/system.collections.arraylist)
is not recommended for new development, the same recommendation applies to PowerShell:

Avoid the ArrayList class for new development.
The `ArrayList` class is a non-generic collection that can hold objects of any type. This is inline with the fact
that PowerShell is a weakly typed language. However, the `ArrayList` class does not provide any explicit type
safety and performance benefits of generic collections. Instead of using an `ArrayList`, consider using either a
[`System.Collections.Generic.List[Object]`](https://learn.microsoft.com/dotnet/api/system.collections.generic.list-1)
class or a fixed PowerShell array.
Besides, the `ArrayList.Add` method returns the index of the added element which often unintendedly pollutes the
PowerShell pipeline and therefore might cause unexpected issues.

## How to Fix

In cases where only the `Add` method is used, you might just replace the `ArrayList` class with a generic
`List[Object]` class but you could also consider using the idiomatic PowerShell pipeline syntax instead.

## Example

### Wrong

```powershell
# Using an ArrayList
$List = [System.Collections.ArrayList]::new()
1..3 | ForEach-Object { $List.Add($_) } # Note that this will return the index of the added element
```

### Correct

```powershell
# Using a generic List
$List = [System.Collections.Generic.List[Object]]::new()
1..3 | ForEach-Object { $List.Add($_) } # This will not return anything
```

```PowerShell
# Creating a fixed array by using the PowerShell pipeline
$List = 1..3 | ForEach-Object { $_ }
```
1 change: 1 addition & 0 deletions docs/Rules/README.md
Original file line number Diff line number Diff line change
Expand Up @@ -28,6 +28,7 @@ The PSScriptAnalyzer contains the following rule definitions.
| [AvoidShouldContinueWithoutForce](./AvoidShouldContinueWithoutForce.md) | Warning | Yes | |
| [AvoidTrailingWhitespace](./AvoidTrailingWhitespace.md) | Warning | Yes | |
| [AvoidUsingAllowUnencryptedAuthentication](./AvoidUsingAllowUnencryptedAuthentication.md) | Warning | Yes | |
| [AvoidUsingArrayList](./AvoidUsingArrayList.md) | Warning | Yes | |
Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I don't think that this rule should be enabled by default

| [AvoidUsingBrokenHashAlgorithms](./AvoidUsingBrokenHashAlgorithms.md) | Warning | Yes | |
| [AvoidUsingCmdletAliases](./AvoidUsingCmdletAliases.md) | Warning | Yes | Yes<sup>2</sup> |
| [AvoidUsingComputerNameHardcoded](./AvoidUsingComputerNameHardcoded.md) | Error | Yes | |
Expand Down