stdlib

stdlib_system

The stdlib_system module provides interface for interacting with external processes, enabling the execution and monitoring of system commands or applications directly from Fortran.

:local:
:depth: 2

run - Execute an external process synchronously

Status

Experimental

Description

The run interface allows execution of external processes using a single command string or a list of arguments.
Processes run synchronously, meaning execution is blocked until the process finishes.
Optional arguments enable the collection of standard output and error streams, as well as sending input via standard input. Additionally, a callback function can be specified to execute upon process completion, optionally receiving a user-defined payload.

Syntax

process = run (args [, stdin] [, want_stdout] [, want_stderr] [, callback] [, payload])

Arguments

args: Shall be a character(*) string (for command-line execution) or a character(*), dimension(:) array (for argument-based execution). It specifies the command and arguments to execute. This is an intent(in) argument.

stdin (optional): Shall be a character(*) value containing input to send to the process via standard input (pipe). This is an intent(in) argument.

want_stdout (optional): Shall be a logical flag. If .true., the standard output of the process will be captured; if .false. (default), it will be lost. This is an intent(in) argument.

want_stderr (optional): Shall be a logical flag. If .true., the standard error output of the process will be captured. If .false. (default), it will be lost. This is an intent(in) argument.

callback (optional): Shall be a procedure conforming to the process_callback interface. If present, this function will be called upon process completion with the process ID, exit state, and optionally collected standard input, output, and error streams. This is an intent(in) argument.

payload (optional): Shall be a generic (class(*)) scalar that will be passed to the callback function upon process completion. It allows users to associate custom data with the process execution. This is an intent(inout), target argument.

Return Value

Returns an object of type process_type that contains information about the state of the created process.

Example

! Example usage with command line or list of arguments
type(process_type) :: p

! Run a simple command line synchronously
p = run("echo 'Hello, world!'", want_stdout=.true.)

runasync - Execute an external process asynchronously

Status

Experimental

Description

The runasync interface allows execution of external processes using a single command string or a list of arguments.
Processes are run asynchronously (non-blocking), meaning execution does not wait for the process to finish.
Optional arguments enable the collection of standard output and error streams, as well as sending input via standard input. Additionally, a callback function can be specified to execute upon process completion, optionally receiving a user-defined payload.

Syntax

process = runasync (args [, stdin] [, want_stdout] [, want_stderr] [, callback] [, payload])

Arguments

args: Shall be a character(*) string (for command-line execution) or a character(*), dimension(:) array (for argument-based execution). It specifies the command and arguments to execute. This is an intent(in) argument.

stdin (optional): Shall be a character(*) value containing input to send to the process via standard input (pipe). This is an intent(in) argument.

want_stdout (optional): Shall be a logical flag. If .true., the standard output of the process will be captured; if .false. (default), it will be lost. This is an intent(in) argument.

want_stderr (optional): Shall be a logical flag. If .true., the standard error output of the process will be captured. Default: .false.. This is an intent(in) argument.

callback (optional): Shall be a procedure conforming to the process_callback interface. If present, this function will be called upon process completion with the process ID, exit state, and optionally collected standard input, output, and error streams. This is an intent(in) argument.

payload (optional): Shall be a generic (class(*)) scalar that will be passed to the callback function upon process completion. It allows users to associate custom data with the process execution. This is an intent(inout), target argument.

Return Value

Returns an object of type process_type that contains information about the state of the created process.

Example

```{literalinclude} ../../example/system/example_process_1.f90 :language: fortran


## `is_running` - Check if a process is still running

### Status

Experimental

### Description

The `is_running` interface provides a method to check if an external process is still running.  
This is useful for monitoring the status of asynchronous processes created with the `run` interface.

### Syntax

`status = ` `is_running` `(process)`

### Arguments

`process`:  Shall be a `type(process_type)` object representing the external process to check. This is an `intent(inout)` argument.


### Return Value

Returns a `logical` value: `.true.` if the process is still running, or `.false.` if the process has terminated.
After a call to `is_running`, the `type(process_type)` structure is also updated to the latest process state.

### Example

```{literalinclude} ../../example/system/example_process_2.f90
:language: fortran

is_completed - Check if a process has completed execution

Status

Experimental

Description

The is_completed interface provides a method to check if an external process has finished execution.
This is useful for determining whether asynchronous processes created with the run interface have terminated.

Syntax

status = is_completed (process)

Arguments

process: Shall be a type(process_type) object representing the external process to check. This is an intent(inout) argument.

Return Value

Returns a logical value:

After a call to is_completed, the type(process_type) structure is updated to reflect the latest process state.

Example

```{literalinclude} ../../example/system/example_process_1.f90 :language: fortran


## `elapsed` - Return process lifetime in seconds

### Status

Experimental

### Description

The `elapsed` interface provides a method to calculate the total time that has elapsed since a process was started.  
This is useful for tracking the duration of an external process or for performance monitoring purposes.  

The result is a real value representing the elapsed time in seconds, measured from the time the process was created.

### Syntax

`delta_t = ` `elapsed` `(process)`

### Arguments

`process`: Shall be a `type(process_type)` object representing the external process. It is an `intent(in)` argument.

### Return Value

Returns a `real(real64)` value that represents the elapsed time (in seconds) since the process was started.  
If the process is still running, the value returned is the time elapsed until the call to this function. 
Otherwise, the total process duration from creation until completion is returned.

### Example

```{literalinclude} ../../example/system/example_process_3.f90
:language: fortran

wait - Wait until a running process is completed

Status

Experimental

Description

The wait interface provides a method to block the calling program until the specified process completes.
If the process is running asynchronously, this subroutine will pause the workflow until the given process finishes.
Additionally, an optional maximum wait time can be provided. If the process does not finish within the specified time, the subroutine will return without waiting further.

On return from this routine, the process state is accordingly updated. This is useful when you want to wait for a background task to complete, but want to avoid indefinite blocking in case of process hang or delay.

Syntax

call {f:subr}wait (process [, max_wait_time])

Arguments

process: Shall be a type(process_type) object representing the external process to monitor.
This is an intent(inout) argument, and its state is updated upon completion.

max_wait_time (optional): Shall be a real value specifying the maximum wait time in seconds.
If not provided, the subroutine will wait indefinitely until the process completes.

Example

```{literalinclude} ../../example/system/example_process_2.f90 :language: fortran


## `update` - Update the internal state of a process

### Status

Experimental

### Description

The `update` interface allows the internal state of a process object to be updated by querying the system.  
After the process completes, the standard output and standard error are retrieved, if they were requested, and loaded into the `process%stdout` and `process%stderr` string variables, respectively.

This is especially useful for monitoring asynchronous processes and retrieving their output after they have finished.

### Syntax

`call ` {f:subr}`update` `(process)`

### Arguments

`process`: Shall be a `type(process_type)` object representing the external process whose state needs to be updated.  
This is an `intent(inout)` argument, and its internal state is updated on completion.

### Example

```{literalinclude} ../../example/system/example_process_5.f90
:language: fortran

kill - Terminate a running process

Status

Experimental

Description

The kill interface is used to terminate a running external process. It attempts to stop the process and returns a boolean flag indicating whether the operation was successful. This interface is useful when a process needs to be forcefully stopped, for example, if it becomes unresponsive or if its execution is no longer required.

Syntax

call {f:subr}kill (process, success)

Arguments

process: Shall be a type(process_type) object representing the external process to be terminated.
This is an intent(inout) argument, and on return is updated with the terminated process state.

success: Shall be a logical variable. It is set to .true. if the process was successfully killed, or .false. otherwise.

Example

```{literalinclude} ../../example/system/example_process_4.f90 :language: fortran


## `sleep` - Pause execution for a specified time in milliseconds

### Status

Experimental

### Description

The `sleep` interface pauses the execution of a program for a specified duration, given in milliseconds. 
This routine acts as a cross-platform wrapper, abstracting the underlying platform-specific sleep implementations. 
It ensures that the requested sleep duration is honored on both Windows and Unix-like systems.

### Syntax

`call ` {f:subr}`sleep` `(millisec)`

### Arguments

`millisec`: Shall be an `integer` representing the number of milliseconds to sleep. This is an `intent(in)` argument.

### Example

```{literalinclude} ../../example/system/example_sleep.f90
:language: fortran

is_windows - Check if the system is running on Windows

Status

Experimental

Description

The is_windows interface provides a quick, compile-time check to determine if the current system is Windows. It leverages a C function that checks for the presence of the _WIN32 macro, which is defined in C compilers when targeting Windows. This function is highly efficient and works during the compilation phase, avoiding the need for runtime checks.

Syntax

result = is_windows ()

Return Value

Returns a logical flag: .true. if the system is Windows, or .false. otherwise.

Example

```{literalinclude} ../../example/system/example_process_1.f90 :language: fortran


## `get_runtime_os` - Determine the OS type at runtime

### Status

Experimental

### Description

`get_runtime_os` inspects the runtime environment to identify the current OS type. It evaluates environment variables (`OSTYPE`, `OS`) and checks for specific files associated with known operating systems.
The supported OS types are `integer, parameter` variables stored in the `stdlib_system` module:

- **Linux** (`OS_LINUX`)
- **macOS** (`OS_MACOS`)
- **Windows** (`OS_WINDOWS`)
- **Cygwin** (`OS_CYGWIN`)
- **Solaris** (`OS_SOLARIS`)
- **FreeBSD** (`OS_FREEBSD`)
- **OpenBSD** (`OS_OPENBSD`)

If the OS cannot be identified, the function returns `OS_UNKNOWN`.

### Syntax

`os = ` `get_runtime_os` `()`

### Class

Function

### Arguments

None.

### Return Value

Returns one of the `integer` `OS_*` parameters representing the OS type, from the `stdlib_system` module, or `OS_UNKNOWN` if undetermined.

### Example

```{literalinclude} ../../example/system/example_get_runtime_os.f90
:language: fortran

OS_TYPE - Cached OS type retrieval

Status

Experimental

Description

OS_TYPE provides a cached result of the get_runtime_os function. The OS type is determined during the first invocation and stored in a static variable. Subsequent calls reuse the cached value, making this function highly efficient.

This caching mechanism ensures negligible overhead for repeated calls, unlike get_runtime_os, which performs a full runtime inspection.

Syntax

os = OS_TYPE()

Class

Function

Arguments

None.

Return Value

Returns one of the integer OS_* parameters representing the OS type, from the stdlib_system module, or OS_UNKNOWN if undetermined.

Example

```{literalinclude} ../../example/system/example_os_type.f90 :language: fortran


---

## `FS_ERROR` - Helper function for error handling

### Status

Experimental

### Description

A helper function for returning the `type(state_type)` with the flag `STDLIB_FS_ERROR` set.

### Syntax

`err = FS_ERROR([a1,a2,a3,a4...... a20])`

### Class
Pure Function

### Arguments

`a1,a2,a3.....a20`(optional): They are of type `class(*), dimension(..), optional, intent(in)`. 
An arbitrary list of `integer`, `real`, `complex`, `character` or `string_type` variables. Numeric variables may be provided as either scalars or rank-1 (array) inputs.

### Behavior

Formats all the arguments into a nice error message, utilizing the constructor of `state_type`

### Return values

`type(state_type)`

### Example

```{literalinclude} ../../example/system/example_fs_error.f90
:language: fortran

FS_ERROR_CODE - Helper function for error handling (with error code)

Status

Experimental

Description

A helper function for returning the type(state_type) with the flag STDLIB_FS_ERROR set. It also formats and prefixes the code passed to it as the first argument.

Syntax

err = FS_ERROR_CODE(code [, a1,a2,a3,a4...... a19])

Class

Pure Function

Arguments

code: An integer code.

a1,a2,a3.....a19(optional): They are of type class(*), dimension(..), optional, intent(in). An arbitrary list of integer, real, complex, character or string_type variables. Numeric variables may be provided as either scalars or rank-1 (array) inputs.

Behavior

Formats all the arguments into a nice error message, utilizing the constructor of state_type

Return values

type(state_type)

Example

```{literalinclude} ../../example/system/example_fs_error.f90 :language: fortran


---

## `is_file` - Test if a path is a regular file

### Status

Experimental

### Description

This function checks if a specified file system path is a regular file. 
It follows symbolic links and returns the status of the `target`.
It is designed to work across multiple platforms. On Windows, paths with both forward `/` and backward `\` slashes are accepted.

### Syntax

`result = ` `is_file``(path)`

### Class

Function

### Arguments

`path`: Shall be a character string containing the file system path to evaluate. It is an `intent(in)` argument.

### Return values

The function returns a `logical` value:

- `.true.` if the path matches an existing regular file.
- `.false.` otherwise, or if path does not exist.

### Example

```{literalinclude} ../../example/system/example_is_file.f90
:language: fortran

is_directory - Test if a path is a directory

Status

Experimental

Description

This function checks if a specified file system path is a directory. It follows symbolic links and returns the status of the target. It is designed to work across multiple platforms. On Windows, paths with both forward / and backward \ slashes are accepted.

Syntax

result = is_directory(path)

Class

Function

Arguments

path: Shall be a character string containing the file system path to evaluate. It is an intent(in) argument.

Return values

The function returns a logical value:

Example

```{literalinclude} ../../example/system/example_is_directory.f90 :language: fortran


---

## `is_symlink` - Test if a path is a symbolic link.

### Status

Experimental

### Description

This function checks if a specified file system path is a symbolic link to either a file or a directory.
Use `is_file` and `is_directory` functions
to check further if the link is to a file or a directory respectively.
It is designed to work across multiple platforms. On Windows, paths with both forward `/` and backward `\` slashes are accepted.

### Syntax

`result = ` `is_symlink``(path)`

### Class

Function

### Arguments

`path`: Shall be a character string containing the file system path to evaluate. It is an `intent(in)` argument.

### Return values

The function returns a `logical` value:

- `.true.` if the path matches an existing regular file.
- `.false.` otherwise, or if the path does not exist.

### Example

```{literalinclude} ../../example/system/example_is_symlink.f90
:language: fortran

make_directory - Creates an empty directory

Status

Experimental

Description

It creates an empty directory with default permissions. It is designed to work across multiple platforms. On Windows, paths with both forward / and backward \ slashes are accepted.

Syntax

call {f:subr}make_directory (path [,err])`

Class

Subroutine

Arguments

path: Shall be a character string containing the path of the directory to create. It is an intent(in) argument.

err(optional): Shall be of type state_type, and is used for error handling. It is an optional, intent(out) argument.

Return values

err is an optional state return flag. If not requested and an error occurs, an FS_ERROR will trigger an error stop.

Example

```{literalinclude} ../../example/system/example_make_directory.f90 :language: fortran


---

## `make_directory_all` - Creates an empty directory with all its parent directories

### Status

Experimental

### Description

It creates an empty directory with default permissions.
It also creates all the necessary parent directories in the path if they do not exist already.

### Syntax

`call ` {f:subr}`make_directory_all` (path [,err])`

### Class

Subroutine

### Arguments

`path`: Shall be a character string containing the path of the directory to create. It is an `intent(in)` argument.

`err`(optional): Shall be of type `state_type`, and is used for error handling. It is an `optional, intent(out)` argument.

### Return values

`err` is an optional state return flag. If not requested and an error occurs, an `FS_ERROR` will trigger an error stop.

### Example

```{literalinclude} ../../example/system/example_make_directory.f90
:language: fortran

remove_directory - Removes an empty directory

Status

Experimental

Description

It deletes an empty directory. It is designed to work across multiple platforms. On Windows, paths with both forward / and backward \ slashes are accepted.

Syntax

call {f:subr}remove_directory (path, err)`

Class

Subroutine

Arguments

path: Shall be a character string containing the path of the directory to create. It is an intent(in) argument.

err(optional): Shall be of type state_type, and is used for error handling. It is an optional, intent(out) argument.

Return values

err is an optional state return flag. On error if not requested, an FS_ERROR will trigger an error stop.

Example

```{literalinclude} ../../example/system/example_remove_directory.f90 :language: fortran


---

## `get_cwd` - Gets the current working directory

### Status

Experimental

### Description

This subroutine retrieves the current working directory the running process is executing from.
It is designed to work across multiple platforms. On Windows, paths with both forward `/` and backward `\` slashes are accepted.

### Syntax

`call ` {f:subr}`get_cwd` (cwd [, err])`

### Class

Subroutine

### Arguments

`cwd`: Shall be a character string for receiving the path of the current working directory (cwd). It is an `intent(out)` argument.

`err`(optional): Shall be of type `state_type`, and is used for error handling. It is an `intent(out)` argument.

### Return values

`err` is an optional state return flag. On error if not requested, an `FS_ERROR` will trigger an error stop.

### Example

```{literalinclude} ../../example/system/example_cwd.f90
:language: fortran

set_cwd - Sets the current working directory

Status

Experimental

Description

This subrotine sets the current working directory the process is executing from. It is designed to work across multiple platforms. On Windows, paths with both forward / and backward \ slashes are accepted.

Syntax

call {f:subr}set_cwd (path [, err])`

Class

Subroutine

Arguments

path: Shall be a character string containing the path of the directory. It is an intent(in) argument.

err(optional): Shall be of type state_type, and is used for error handling. It is an intent(out) argument.

Return values

err is an optional state return flag. On error if not requested, an FS_ERROR will trigger an error stop.

Example

```{literalinclude} ../../example/system/example_cwd.f90 :language: fortran


---

## `set_environment_variable` - Sets an environment variable

### Status

Experimental

### Description

This subroutine creates an environment variable for the calling process, or changes the value of one that already exists.

The environment variable is visible to this process and to any process it starts afterwards. It is not exported back to the parent shell; that is a property of the operating system rather than of this implementation.

`overwrite` selects whether an existing variable is replaced, and defaults to `.true.`. Windows has no equivalent of the POSIX flag -- `_putenv_s` always replaces -- so the argument has no effect under the Windows environment.

### Syntax

`call ` {f:subr}`set_environment_variable` `(name, value [, overwrite] [, err])`

### Class

Subroutine

### Arguments

`name`: Shall be a character string holding the name of the variable. It must not be empty and must not contain `=`; both are rejected before the call reaches the operating system, which reports them only as `EINVAL`. It is an `intent(in)` argument.

`value`: Shall be a character string holding the value of `name`. It is an `intent(in)` argument.

`overwrite`(optional): Shall be a `logical` selecting whether an existing variable is replaced. Defaults to `.true.`. If the variable exists and `overwrite` is `.false.`, its value is left unchanged and no error is returned. Has no effect on Windows, where the value is always replaced. It is an `intent(in)` argument.

`err`(optional): Shall be of type `state_type`, and is used for error handling. It is an `intent(out)` argument.

### Return values

`err` is an optional state return flag. On error if not requested, an `FS_ERROR` will trigger an error stop.

### Example

```{literalinclude} ../../example/system/example_environment_variable.f90
:language: fortran

delete_environment_variable - Removes an environment variable

Status

Experimental

Description

This subroutine removes an environment variable from the calling process.

Deleting a variable that is not set is not an error, which matches the behaviour of unsetenv.

As with set_environment_variable, the name must not be empty and must not contain =.

Syntax

call {f:subr}delete_environment_variable (name [, err])

Class

Subroutine

Arguments

name: Shall be a character string holding the name of the variable. It is an intent(in) argument.

err(optional): Shall be of type state_type, and is used for error handling. It is an intent(out) argument.

Return values

err is an optional state return flag. On error if not requested, an FS_ERROR will trigger an error stop.

Example

```{literalinclude} ../../example/system/example_environment_variable.f90 :language: fortran


---

## `exists` - Checks if a path exists in the filesystem

### Status

Experimental

### Description

This function makes a system call (syscall) to retrieve metadata for the specified path and determines its type.
It can distinguish between the following path types:

- Regular File
- Directory
- Symbolic Link

It returns a constant representing the detected path type, or `type_unknown` if the type cannot be determined. 
Any encountered errors are handled using `state_type`.

### Syntax

`fs_type = `exists` (path [, err])`

### Class

Function

### Arguments

`path`: Shall be a character string containing the path. It is an `intent(in)` argument.

`err`(optional): Shall be of type `state_type`, and is used for error handling. It is an `optional, intent(out)` argument.

### Return values

`fs_type`: An `intent(out), integer` parameter indicating the type. The possible values are:
- `fs_type_unknown`: 0      => an unknown type
- `fs_type_regular_file`: 1 => a regular file
- `fs_type_directory`: 2    => a directory 
- `fs_type_symlink`: 3      => a symbolic link

`err`(optional): It is an optional state return flag. If not requested and an error occurs, an `FS_ERROR` will trigger an error stop.

```{literalinclude} ../../example/system/example_exists.f90
:language: fortran

null_device - Return the null device file path

Status

Experimental

Description

This function returns the file path of the null device, which is a special file used to discard any data written to it. It reads as an empty file. The null device’s path varies by operating system:

Syntax

path = null_device()

Class

Function

Arguments

None.

Return Value

Example

```{literalinclude} ../../example/system/example_null_device.f90 :language: fortran


---

## `delete_file` - Delete a file

### Status

Experimental

### Description

This subroutine deletes a specified file from the filesystem. It ensures that the file exists and is not a directory before attempting deletion.
If the file cannot be deleted due to permissions, being a directory, or other issues, an error is raised. 
The function provides an optional error-handling mechanism via the `state_type` class. If the `err` argument is not provided, exceptions will trigger an `error stop`.

### Syntax

`call ` {f:subr}`delete_file` (path [, err])`

### Class
Subroutine

### Arguments

`path`: Shall be a character string containing the path to the file to be deleted. It is an `intent(in)` argument.

`err` (optional): Shall be a `type(state_type)` variable for error handling. If provided, errors are returned as a state object. If not provided, the program stops execution on error.

### Behavior

- Checks if the file exists. If not, an error is raised.
- Ensures the path is not a directory before deletion.
- Attempts to delete the file, raising an error if unsuccessful.

### Return values

The file is removed from the filesystem if the operation is successful. If the operation fails, an error is raised.

### Example

```{literalinclude} ../../example/system/example_delete_file.f90
:language: fortran

join_path - Joins the provided paths according to the OS

Status

Experimental

Description

This interface joins the paths provided to it according to the platform specific path-separator. i.e \ for windows and / for others

Syntax

res = {f:func}join_path ` (p1, p2)`

res = {f:func}join_path ` (p)`

Class

Pure function

Arguments

p1, p2: Shall be a character string or type(string_type). It is an intent(in) argument. or p: Shall be a list of character strings or list of type(string_type). It is an intent(in) argument.

Return values

The resultant path, either a character string or type(string_type).

operator(/)

Alternative syntax tojoin_path using an overloaded operator. Join two paths according to the platform specific path-separator.

Status

Experimental

Syntax

p = lval / rval

Class

Pure function.

Arguments

lval: A character string or type(string_type). It is an intent(in) argument.

rval: A character string or type(string_type). It is an intent(in) argument.

Result value

The result is an allocatable character string or type(string_type)

Example

```{literalinclude} ../../example/system/example_path_join.f90 :language: fortran


---

## `split_path` - splits a path immediately following the last separator

### Status

Experimental

### Description

This subroutine splits a path immediately following the last separator after removing the trailing separators
splitting it into most of the times a directory and a file name.

### Syntax

`call ` {f:subr}`split_path` `(p, head, tail)`

### Class
Subroutine

### Arguments

`p`: A character string or `type(string_type)` containing the path to be split. It is an `intent(in)` argument.
`head`: The first part of the path. Either a character string or `type(string_type)`. It is an `intent(out)` argument.
`tail`: The rest part of the path. Either a character string or `type(string_type)`. It is an `intent(out)` argument.

### Behavior

- If `p` is empty, `head` is set to `.` and `tail` is left empty.
- If `p` consists entirely of path-separators, `head` is set to the path-separator and `tail` is left empty.
- `head` ends with a path-separator if and only if `p` appears to be a root directory or child of one.

### Return values

The splitted path. `head` and `tail`.

### Example

```{literalinclude} ../../example/system/example_path_split_path.f90
:language: fortran

base_name - The last part of a path

Status

Experimental

Description

This function returns the last part of a path after removing trailing path separators.

Syntax

res = base_name(p)

Class

Function

Arguments

p: the path, a character string or type(string_type). It is an intent(in) argument.

Behavior

Return values

A character string or type(string_type).

Example

```{literalinclude} ../../example/system/example_path_base_name.f90 :language: fortran


---

## `dir_name` - Everything except the last part of the path

### Status

Experimental

### Description

This function returns everything except the last part of a path.

### Syntax

`res = ` `dir_name``(p)`

### Class
Function

### Arguments

`p`: the path, a character string or `type(string_type)`. It is an `intent(in)` argument.

### Behavior

- The `head` of ``split_path`` is exactly what is returned. Same Behavior.

### Return values

A character string or `type(string_type)`.

### Example

```{literalinclude} ../../example/system/example_path_dir_name.f90
:language: fortran
This site is open source. Improve this page.