Muokkaa

Jaa

about_Trap

  • Artikkeli

Short description

Describes a keyword that handles a terminating error.

Long description

A terminating error stops a statement from running. If PowerShell doesn't handle a terminating error in some way, PowerShell also stops running the function or script in the current pipeline. In other languages, such as C#, terminating errors are known as exceptions.

The trap keyword specifies a list of statements to run when a terminating error occurs. trap statements can handle the terminating errors in the following ways:

  • Display the error after processing the trap statement block and continuing execution of the script or function containing the trap. This behavior is the default.

    Note

    When the terminating error occurs in a subordinate script block, such as an if statement or foreach loop, the statements in the trap block are run and execution continues at the next statement outside the subordinate script block.

  • Display the error and abort execution of the script or function containing the trap using break in the trap statement.

  • Silence the error, but continue execution of the script or function containing the trap by using continue in the trap statement.

The statement list of the trap can include multiple conditions or function calls. A trap can write logs, test conditions, or even run another program.

Syntax

The trap statement has the following syntax:

trap [[<error type>]] {<statement list>}

The trap statement includes a list of statements to run when a terminating error occurs. A trap statement consists of the trap keyword, optionally followed by a type expression, and the statement block containing the list of statements to run when an error is trapped. The type expression refines the types of errors the trap catches.

A script or command can have multiple trap statements. trap statements can appear anywhere in the script or command.

Trapping all terminating errors

When a terminating error occurs that isn't handled in another way in a script or command, PowerShell checks for a trap statement that handles the error. If a trap statement is present, PowerShell continues running the script or command in the trap statement.

The following example is a minimal trap statement:

trap { 'Error found.' }

This trap statement traps any terminating error.

In the following example, the function includes a nonsense string that causes a runtime error.

function TrapTest {
    trap { 'Error found.' }
    nonsenseString
}

TrapTest

Running this function returns the following output:

Error found.
nonsenseString : The term 'nonsenseString' is not recognized as the name
of a cmdlet, function, script file, or operable program. Check the
spelling of the name, or if a path was included, verify that the path is
correct and try again.
At line:3 char:5
+     nonsenseString
+     ~~~~~~~~~~~~~~
    + CategoryInfo          : ObjectNotFound: (nonsenseString:String) []
   , CommandNotFoundException
    + FullyQualifiedErrorId : CommandNotFoundException

The following example includes a trap statement that displays the error by using the $_ automatic variable:

function TrapTest {
    trap { "Error found: $_" }
    nonsenseString
}

TrapTest

Running this version of the function returns the following output:

Error found: The term 'nonsenseString' is not recognized as the name of a
cmdlet, function, script file, or operable program. Check the spelling of
the name, or if a path was included, verify that the path is correct and
try again.
nonsenseString : The term 'nonsenseString' is not recognized as the name
of a cmdlet, function, script file, or operable program. Check the
spelling of the name, or if a path was included, verify that the path is
correct and try again.
At line:3 char:5
+     nonsenseString
+     ~~~~~~~~~~~~~~
    + CategoryInfo          : ObjectNotFound: (nonsenseString:String) []
   , CommandNotFoundException
    + FullyQualifiedErrorId : CommandNotFoundException

Important

trap statements may be defined anywhere within a given script block, but always apply to all statements in that script block. At runtime, trap statements in a block are defined before any other statements are executed. In JavaScript, this is known as hoisting. This means that trap statements apply to all statements in that block even if execution hasn't advanced past the point at which they're defined. For example, defining a trap at the end of a script and throwing an error in the first statement still triggers that trap.

Trapping specific errors

A script or command can have multiple trap statements. A trap can be defined to handle specific errors.

The following example is a trap statement that traps the specific error CommandNotFoundException:

trap [System.Management.Automation.CommandNotFoundException] {
    'Command error trapped'
}

When a function or script encounters a string that doesn't match a known command, this trap statement displays the Command error trapped string. After running the trap statement list, PowerShell writes the error object to the error stream and then continues the script.

PowerShell uses .NET exception types. The following example specifies the System.Exception error type:

trap [System.Exception] { 'An error trapped' }

The CommandNotFoundException error type inherits from the System.Exception type. This statement traps any errors raised by unknown commands. It also traps other error types.

You can find the exception type for an error by inspecting the error object. The following example shows how to get the full name of the exception for the last error in a session:

nonsenseString
$Error[0].Exception.GetType().FullName

nonsenseString : The term 'nonsenseString' is not recognized as the name
of a cmdlet, function, script file, or operable program. Check the
spelling of the name, or if a path was included, verify that the path is
correct and try again.
At line:1 char:1
+ nonsenseString
+ ~~~~~~~~~~~~~~
    + CategoryInfo          : ObjectNotFound: (nonsenseString:String) []
   , CommandNotFoundException
    + FullyQualifiedErrorId : CommandNotFoundException

System.Management.Automation.CommandNotFoundException

You can have more than one trap statement in a script. Only one trap statement can trap each error type. When a terminating error occurs, PowerShell searches for the trap with the most specific match, starting in the current script block of execution.

The following script example contains an error. The script includes a general trap statement that traps any terminating error and a specific trap statement that specifies the CommandNotFoundException type.

trap { 'Other terminating error trapped' }
trap [System.Management.Automation.CommandNotFoundException] {
  'Command error trapped'
}
nonsenseString

Running this script produces the following result:

Command error trapped
nonsenseString : The term 'nonsenseString' is not recognized as the name
of a cmdlet, function, script file, or operable program. Check the
spelling of the name, or if a path was included, verify that the path is
correct and try again.
At line:5 char:1
+ nonsenseString}
+ ~~~~~~~~~~~~~~
    + CategoryInfo          : ObjectNotFound: (nonsenseString:String) []
   , CommandNotFoundException
    + FullyQualifiedErrorId : CommandNotFoundException

Because PowerShell doesn't recognize "nonsenseString" as a cmdlet or other item, it returns a CommandNotFoundException error. The specific trap statement traps this terminating error.

The following script example contains the same trap statements with a different error:

trap { 'Other terminating error trapped' }
trap [System.Management.Automation.CommandNotFoundException] {
    'Command error trapped'
}
1/$null

Running this script produces the following result:

Other terminating error trapped
Attempted to divide by zero.
At line:5 char:1
+ 1/$null}
+ ~~~~~~~
    + CategoryInfo          : NotSpecified: (:) [], RuntimeException
    + FullyQualifiedErrorId : RuntimeException

The attempt to divide by zero doesn't create a CommandNotFoundException error. The other trap statement, which traps any terminating error, traps the divide by zero error.

Trapping errors in a script block

By default, when a terminating error is thrown, execution transfers to the trap statement. After the trap block is run, control returns to the next statement block after the location of the error.

For example, when a terminating error occurs in an foreach statement, the trap statement is run and execution continues at the next statement after the foreach block, not within the foreach block.

trap { 'An error occurred!'}
foreach ($x in 3..-1) {
       "1/$x = "
       "`t$(1/$x)"
}
'after loop'

1/3 =
        0.333333333333333
1/2 =
        0.5
1/1 =
        1
1/0 =
An error occurred!
Attempted to divide by zero.
At line:3 char:4
+    1/$x
+    ~~~~
    + CategoryInfo          : NotSpecified: (:) [], RuntimeException
    + FullyQualifiedErrorId : RuntimeException

after loop

In the output, you can see the loops continue until the last iteration. When the script tries to divide 1 by 0, PowerShell throws a terminating error. The script skips the rest of the foreach script block, runs the try statement, and continues after the foreach script block.

Trapping errors and scope

If a terminating error occurs in the same script block as the trap statement, PowerShell runs the list of statements defined by the trap. Execution continues at the statement after the error. If the trap statement is in a different script block from the error, execution continues at the next statement that's in the same script block as the trap statement.

For example, if an error occurs in a function, and the trap statement is in the function, the script continues at the next statement. The following script contains an error and a trap statement:

function function1 {
    trap { 'An error: ' }
    NonsenseString
    'function1 was completed'
}

function1

Running this script produces the following result:

An error:
NonsenseString : The term 'NonsenseString' is not recognized as the name
of a cmdlet, function, script file, or operable program. Check the
spelling of the name, or if a path was included, verify that the path is
correct and try again.
At line:3 char:5
+     NonsenseString
+     ~~~~~~~~~~~~~~
    + CategoryInfo          : ObjectNotFound: (NonsenseString:String) []
   , CommandNotFoundException
    + FullyQualifiedErrorId : CommandNotFoundException

function1 was completed

The trap statement in the function traps the error. After displaying the message, PowerShell resumes running the function. Notice that Function1 completed after the trap statement.

Compare this behavior with the following example, which has the same error and trap statement. In this example, the trap statement occurs outside the function:

function function2 {
    NonsenseString
    'function2 was completed'
}

trap { 'An error:' }

function2

Running the Function2 function produces the following result:

An error:
NonsenseString : The term 'NonsenseString' is not recognized as the name
of a cmdlet, function, script file, or operable program. Check the
spelling of the name, or if a path was included, verify that the path is
correct and try again.
At line:2 char:5
+     NonsenseString
+     ~~~~~~~~~~~~~~
    + CategoryInfo          : ObjectNotFound: (NonsenseString:String) []
   , CommandNotFoundException
    + FullyQualifiedErrorId : CommandNotFoundException

In this example, the function2 was completed command wasn't run. In both examples, the terminating error occurs within the function. In this example, however, the trap statement is outside the function. PowerShell doesn't go back into the function after the trap statement runs.

Caution

When multiple traps are defined for the same error condition, the first trap defined lexically (highest in the script block) is used.

In the following example, only the trap with whoops 1 runs.

Remove-Item -ErrorAction Stop ThisFileDoesNotExist
trap { 'whoops 1'; continue }
trap { 'whoops 2'; continue }

Important

A trap statement is scoped to where it compiles. If you have a trap statement inside a function or dot sourced script, when the function or dot sourced script exits, all trap statements inside are removed.

Using the break and continue keywords

You can use the break and continue keywords in a trap statement to determine whether a script or command continues to run after a terminating error.

If you include a break statement in a trap statement list, PowerShell stops the function or script. The following sample function uses the break keyword in a trap statement:

function break_example {
    trap {
        'Error trapped'
        break
    }
    1/$null
    'Function completed.'
}

break_example

Error trapped
Attempted to divide by zero.
At line:6 char:5
+     1/$null
+     ~~~~~~~
    + CategoryInfo          : NotSpecified: (:) [], ParentContainsErrorR
   ecordException
    + FullyQualifiedErrorId : RuntimeException

Because the trap statement included the break keyword, the function doesn't continue to run, and the Function completed line isn't run.

If you include a continue keyword in a trap statement, PowerShell resumes after the statement that caused the error, just as it would without break or continue. With the continue keyword, however, PowerShell doesn't write an error to the error stream.

The following sample function uses the continue keyword in a trap statement:

function ContinueExample {
    trap {
        'Error trapped'
        continue
    }
    foreach ($x in 3..-1) {
       "1/$x = "
       "`t$(1/$x)"
    }
    'End of function'
}

ContinueExample

1/3 =
        0.333333333333333
1/2 =
        0.5
1/1 =
        1
1/0 =
Error trapped
End of function

The function resumes after the error is trapped, and the End of function statement runs. No error is written to the error stream.

Notes

trap statements provide a way to ensure all terminating errors within a script block are handled. For more finer-grained error handling, use try/catch blocks where traps are defined using catch statements. The catch statements only apply to the code inside the associated try statement. For more information, see about_Try_Catch_Finally.

See also