Skip to main content

AIO

Summary

Creates an ArcIO (AIO) object to handle local and cloud files, and data stores in a unified way that can be used in Python.

Using this object, you can perform tasks such as copying files from one location to another and renaming or deleting a folder, as well as filing related operations including read and write. The AIO object can be used for cloud stores by specifying the path to an .acs file as the input. For a local data store, create an AIO object by specifying a local folder path, which will then be used as the current working directory, or by not specifying any parameter values.

Discussion

The AIO object supports file operations agnostic of the file system. The object currently supports Windows, Linux and major cloud storage implementations.

Apart from a local file system and UNC paths, the AIO object supports the following virtual cloud URIs for class initialization:

  • ACS Connection files (.acs)—Stores connection information in binary persisted encrypted format. See the Create Cloud Storage Connection File tool for more information.

  • /cloudStores json—Stores connection information in JSON format for ArcGIS Enterprise.

  • /vsi paths—GDAL is the preferred URI for file systems in agnostic file operations. See GDAL Virtual File Systems for more information.

  • http[s] paths—The cloud provider URL for a resource, which may contain connection information such as Microsoft Azure SAS.

  • A cloud provider-specific URI schema prefix such as s3:// or abfs://.

Note:

Connection information can also be used through IAM/RBAC, GDAL environment options, and provider-specific CLI configuration files in the .aws, .azure, or bin/gdalrc folders.

All cloud-specific methods can be accessed using the cloud property.

When working with cloud operations, class methods support relative paths to the current working directory (aio.getcwd). If absolute paths are used, only vsi paths are supported. These types of operations return vsi paths by default. The type of output path can be overridden by providing the CloudPathType enum.

These methods follow the same conventions and behavior as Python native file system methods.

Syntax

AIO({path})

Name Explanation Data type

path

(Optional)

To construct an AIO object to handle a cloud store, use the .acs file path.

To construct an AIO object to handle a local file store, this parameter can be a local folder path or it can be left empty. If a local folder path is provided, it will be used as the current working directory.

String

Properties

Name Explanation Data type

cloud

(Read only)

An instance of a class that has cloud-specific methods. It is used to retrieve details of the cloud store that was used when the AIO object was created.

Returns a CloudOp object.

cloud_io = AIO(r"C:\data\datacloud.acs")
cloud_io.cloud.getprovidername()

Object

Methods

chdir(path)

Changes the working directory relative to the current working directory.

Name Explanation Data type

path

A relative path from the current working directory or an absolute path (absolute vsi path for cloud).

String

cloud_io = AIO(r"C:\data\datacloud.acs")
cloud_io.chdir(r"aio")  # Relative path
cloud_io = AIO(r"C:\data\datacloud.acs")
cloud_io.chdir(r"/vsis3/data/aio")  # Absolute VSI path
local_io = AIO(r"C:\data")
local_io.chdir(r"aio")

close(handle)

Closes the open file handle.

Name Explanation Data type

handle

An open file handle.

The data type is AIOFile.

Note:

An instance of this class is returned by the AIO object's open method.

Object

cloud_io = AIO(r"C:\data\datacloud.acs")
rcsfile = cloud_io.open(r'testfile.txt', 'r')
rcsfile.read(size=4)
cloud_io.close(rcsfile)

copy(src, dst, {options})

Copies a file from the source to the destination.

For a cloud-based AIO object:

  • If the src or dst value is a local path, use an absolute local path.

  • If the src or dst value is a cloud path, use a relative path from the current working directory or an absolute vsi cloud path.

  • If the src and dst values are cloud paths, use absolute vsi cloud paths.

For a local AIO object:

  • The src and dst values can be a relative path from the current working directory or an absolute path.
Name Explanation Data type

src

The source path of the file.

String

dst

The target path of the file.

String

options

(Optional)

Specifies the options that will be used.

  • RECURSIVE—Subdirectories and their contents will be recursively copied. This option is only applicable for folder operations.

  • SYNC_STRATEGY—The criteria that will be used for overwriting if the target file exists.

  • NUM_THREADS—The number of threads that will be used for parallel file copying.

  • CHUNK_SIZE—The maximum chunk size (in bytes) that will be used to split large objects when uploading or downloading.

Syntax—RECURSIVE=[YES|NO], SYNC_STRATEGY=[TIMESTAMP|ETAG|OVERWRITE], NUM_THREADS=(integer) CHUNK_SIZE=(integer) [x-amz-*|x-goog-*|x-ms-*=(value)]

Example—{'RECURSIVE':'YES', 'SYNC_STRATEGY':'OVERWRITE', 'NUM_THREADS':'10', 'CHUNK_SIZE':'8'}

The default value is None.

Dictionary

Return value

Data type Explanation

String

The destination path.

cloud_io = AIO(r"C:\data\datacloud.acs")
cloud_io.copy(r'/vsis3/data/list/utf8_test.json',
              r'/vsis3/data/list/new_dir_new/utf8_test.json')
local_io = AIO(r"C:\data")
local_io.copy(r"C:\data\datatest.json", r"C:\data_temp\datatest.json")

copytree(src, dst, {options})

Copies the contents from one directory to another.

For a cloud-based AIO object:

  • If the src or dst value is a local path, use an absolute local path.

  • If the src or dst value is a cloud path, use a relative path from the current working directory or an absolute vsi cloud path.

  • If the src and dst values are cloud paths, use absolute vsi cloud paths.

For a local AIO object:

  • The src and dst values can be a relative path from the current working directory or an absolute path.
Name Explanation Data type

src

The source path of the file.

String

dst

The target path of the file.

String

options

(Optional)

Specifies the options that will be used.

  • RECURSIVE—Subdirectories and their contents will be recursively copied. This option is only applicable for folder operations.

  • SYNC_STRATEGY—The criteria that will be used for overwriting if the target file exists.

  • NUM_THREADS—The number of threads that will be used for parallel file copying.

  • CHUNK_SIZE—The maximum chunk size (in bytes) that will be used to split large objects when uploading or downloading.

Syntax—RECURSIVE=[YES|NO], SYNC_STRATEGY=[TIMESTAMP|ETAG|OVERWRITE], NUM_THREADS=(integer) CHUNK_SIZE=(integer) [x-amz-*|x-goog-*|x-ms-*=(value)]

Example—{'RECURSIVE':'YES', 'SYNC_STRATEGY':'OVERWRITE', 'NUM_THREADS':'10', 'CHUNK_SIZE':'8'}

The default value is None.

Dictionary

Return value

Data type Explanation

String

The destination path.

cloud_io = AIO(r"C:\data\datacloud.acs")
cloud_io.copytree(r'/vsis3/data/list', r'/vsis3/data/data_new_dir')
local_io = AIO(r"C:\data")
local_io.copytree(r"C:\data\list", r"C:\data\data_new_dir")

exists(path)

Verifies whether a path exists.

Use the CloudFileOp object's clearcache method before calling the exists method so that it generates correct results even if an external process updates the cloud storage.

Name Explanation Data type

path

A relative path from the current working directory or an absolute path (absolute vsi path for cloud).

String

Return value

Data type Explanation

Boolean

True if the directory or file exists; otherwise, False.

cloud_io = AIO(r"C:\data\datacloud.acs")
cloud_io.exists(r"C:\data\datacloud.acs\data")

getatime(path)

Gets the file access time for a file.

This method is not available for cloud storage.

Name Explanation Data type

path

The path for which the access time will be obtained. The value can be a relative path from the current working directory or an absolute path.

String

Return value

Data type Explanation

Double

The access time, in number of seconds since epoch.

local_io = AIO(r"C:\data")
local_io.getatime(r"data.json")

getctime(path)

Gets the file creation time for a file.

This method is not available for cloud storage.

Name Explanation Data type

path

The path for which the creation time will be obtained. The value can be a relative path from the current working directory or an absolute path.

String

Return value

Data type Explanation

Double

The creation time, in number of seconds since epoch.

local_io = AIO(r"C:\data")
local_io.getctime(r"data.json")

getcwd({type})

Gets the current working directory.

Name Explanation Data type

type

(Optional)

Specifies the path style for the returned URI (for cloud only). The data type is CloudPathType.

Available options are CloudPathType.VSI, CloudPathType.ACS, CloudPathType.HTTP, and CloudPathType.CLOUDSTORES.

The default value is CloudPathType.VSI.

Object

Return value

Data type Explanation

String

The current working directory.

local_io = AIO()
local_io.getcwd()
cloud_io = AIO(r"C:\data\datacloud.acs")
cloud_io.getcwd()

getmtime(path)

Gets the file modification time for a file.

Name Explanation Data type

path

The relative path from the current working directory or an absolute path (absolute vsi path for cloud).

String

Return value

Data type Explanation

Double

The file modification time in number of seconds since epoch.

local_io = AIO(r"C:\data")
local_io.getmtime(r"data.json")
cloud_io = AIO(r"C:\data\datacloud.acs")
cloud_io.getmtime(r"data.json")

getpath(path, {type})

Constructs a URI path.

Name Explanation Data type

path

The relative path from the current working directory or an absolute path (absolute vsi path for cloud).

String

type

(Optional)

Specifies the path style for the returned URI (for cloud only). The data type is CloudPathType.

Available options are CloudPathType.VSI, CloudPathType.ACS, CloudPathType.HTTP, and CloudPathType.CLOUDSTORES.

The default value is CloudPathType.VSI.

Object

Return value

Data type Explanation

String

The constructed URI path.

from arcpy import CloudPathType
cloud_io = AIO(r"C:\data\datacloud.acs")
cloud_io.getpath('cog.tif', CloudPathType.HTTP)

getsize(path)

Gets the size of a file in bytes.

Name Explanation Data type

path

The relative path from the current working directory or an absolute path (absolute vsi path for cloud).

String

Return value

Data type Explanation

Integer

The size of the file in bytes.

cloud_io = AIO(r"C:\data\datacloud.acs")
cloud_io.getsize(r'cog.tif')

isdir(path)

Verifies whether a path is a directory.

Name Explanation Data type

path

A relative path from the current working directory or an absolute path (absolute vsi path for cloud).

String

Return value

Data type Explanation

Boolean

True if the directory exists; otherwise, False.

cloud_io = AIO(r"C:\data\datacloud.acs")
cloud_io.isdir(r'cog')

isfile(path)

Verifies whether a path is a file.

Name Explanation Data type

path

A relative path from the current working directory or an absolute path (absolute vsi path for cloud).

String

Return value

Data type Explanation

Boolean

True if the file exists; otherwise, False.

cloud_io = AIO(r"C:\data\datacloud.acs")
cloud_io.isfile(r'cog')

listdir({path}, {type})

Lists the contents of a directory nonrecursively.

Name Explanation Data type

path

(Optional)

The relative path from the current working directory or an absolute path (absolute vsi path for cloud).

The default value is None.

String

type

(Optional)

Specifies the path style for the returned URI (for cloud only). The data type is CloudPathType.

Available options are CloudPathType.VSI, CloudPathType.ACS, CloudPathType.HTTP, and CloudPathType.CLOUDSTORES.

The default value is CloudPathType.VSI.

Object

Return value

Data type Explanation

String

A list of file names in the folder.

cloud_io = AIO(r"C:\data\datacloud.acs")
cloud_io.listdir('data_folder')
from arcpy import CloudPathType
cloud_io = AIO(r"C:\data\datacloud.acs")
cloud_io.listdir('data_folder', CloudPathType.ACS)
local_io = AIO(r"C:\data")
local_io.listdir()

makedirs(path)

Creates a directory recursively.

Name Explanation Data type

path

A relative path from the current working directory or an absolute path (absolute vsi path for cloud).

String

cloud_io = AIO(r"C:\data\datacloud.acs")
cloud_io.makedirs(r"datasensor\datatest\data")

mkdir(path)

Creates a directory.

Name Explanation Data type

path

A relative path from the current working directory or an absolute path (absolute vsi path for cloud).

String

cloud_io = AIO(r"C:\data\datacloud.acs")
cloud_io.mkdir(r'cog')

open(path, {mode}, {encoding}, {mime})

Opens a file handle.

The handle must be explicitly closed.

Name Explanation Data type

path

A relative path from the current working directory or an absolute path (absolute vsi path for cloud).

String

mode

(Optional)

Specifies the mode that will be used for opening a file.

  • r—The file will be opened in read mode.

  • w—The file will be opened in write mode.

  • rb—The file will be opened in binary format for reading.

  • wb—The file will be opened in binary format for writing.

Note:

Only read mode honors the encoding parameter.

The default value is None.

String

encoding

(Optional)

The type of encoding that will be used to open a file.

  • utf-8—UTF-8 encoding will be used. This is the default.

  • utf-16—UTF-16 encoding will be used.

  • ascii—ASCII encoding will be used.

The default value is utf-8.

String

mime

(Optional)

The Multipurpose Internet Mail Extensions (MIME) headers. It is recommended that you set MIME headers when creating the file for faster creation and setting of metadata.

The default value is None.

Dictionary

Return value

Data type Explanation

File

Returns the AIOFile object.

from arcpy import AIO
local_io = AIO(r"C:\data")
rcsfile = local_io.open(r'testfile.txt', 'r')
rcsfile.write("This is a test file.")
rcsfile.close()
from arcpy import AIO
cloud_io = AIO(r"C:\data\datacloud.acs")
rcsfile = cloud_io.open(r"C:\data\info\datafile.txt", 'w', mime={'Content-Type': 'text/plain'})
rcsfile.write("This is a test file.")
rcsfile.close()
# Using with statement (context manager to close the opened file)
cloud_io = AIO(r"C:\data\datacloud.acs")
with cloud_io.open(r"C:\data\info\datafile.txt", 'w', mime={'Content-Type': 'text/plain'}) as rcsfile:
    rcsfile.write("This is a test file.")

remove(path)

Deletes a file.

Name Explanation Data type

path

A relative path from the current working directory or an absolute path (absolute vsi path for cloud).

String

cloud_io = AIO(r"C:\data\datacloud.acs")
cloud_io.remove(r'rcsfile.txt')
local_io = AIO(r"c:\data")
local_io.remove(r'rcsfile.txt')

removefiles(path[path,...])

Delete multiple files.

Name Explanation Data type

path[path,...]

The relative path from the current working directory or an absolute path (absolute vsi path for cloud).

String

Return value

Data type Explanation

List

Returns the list of paths that were not removed in case of failure.

cloud_io = AIO(r"C:\data\datacloud.acs")
cloud_io.removefiles(['/vsis3/data/list/to_be_deleted/datatest.json','/vsis3/u-agu/list/to_be_deleted/data_new.json'])
local_io = AIO(r"c:\data")
local_io.removefiles([r'C:\data\datatest.json',r'C:\data\data_new.json'])

rename(src, dst)

Renames or moves a file or a directory. Renaming cloud files is computationally intensive, as copy and delete are performed on the server side.

For a cloud based AIO object:

  • If the src or dst value is a local path, use an absolute local path.

  • If the src or dst value is a cloud path, use a relative path from the current working directory or an absolute vsi cloud path.

  • If the src and dst values are cloud paths, use absolute vsi cloud paths.

For a local AIO object:

  • The src and dst values can be a relative path from the current working directory or an absolute path.
Name Explanation Data type

src

The source path of the file.

String

dst

The target path of the file.

String

cloud_io = AIO(r"C:\data\datacloud.acs")
cloud_io.rename(r'to_be_renamed/cog.tif',r'to_be_renamed/cognew.tif')
local_io = AIO(r"C:\data")
local_io.rename(r'test.json',r'new_test.json')

rmtree(path)

Removes a directory and its contents recursively.

Name Explanation Data type

path

The relative path from the current working directory or an absolute path (absolute vsi path for cloud).

String

cloud_io = AIO(r"C:\data\datacloud.acs")
cloud_io.rmtree(r"aio")  # Relative path
cloud_io = AIO(r"C:\data\datacloud.acs")
cloud_io.rmtree(r"/vsis3/data/arcio")  # Absolute VSI path
local_io = AIO(r"C:\data")
local_io.rmtree(r"aio")

scandir(path, {depth}, {type}, {prefix})

Returns an iterable of entries in a directory.

Name Explanation Data type

path

The relative path from the current working directory or an absolute path (absolute vsi path for cloud).

String

depth

(Optional)

The recursion depth.

  • A value of -1 is for recursive scanning.

  • A value of 0 is for scanning the current directory.

  • A value greater than 0 (>0) is for an arbitrary level.

The default value is 0.

Integer

type

(Optional)

Specifies the path style for the returned URI (for cloud only). The data type is CloudPathType.

Available options are CloudPathType.VSI, CloudPathType.ACS, CloudPathType.HTTP, and CloudPathType.CLOUDSTORES.

The default value is CloudPathType.VSI.

Object

prefix

(Optional)

A filter to select filenames that start with the specified prefix.

This is optimized for the cloud, as the filtering is handled by the BLOB storage API.

The default value is None.

String

Return value

Data type Explanation

Iterable

An iterable of the AIODirEntry object.

cloud_io = AIO(r"C:\data\datacloud.acs")
# Depth = 0, performs cur dir scanning only
from arcpy import CloudPathType
for item in cloud_io.scandir(r'list', depth=0, type=CloudPathType.HTTP):
    print(item.path)
    print(item.is_dir())
    print(item.is_file())
    ob = item.stat()
    print(ob.st_mode)
    # Cloud specific operations through cloud property
    print(item.cloud.getvsipath())
    print(item.cloud.getpath(CloudPathType.ACS))
local_io = AIO(r"c:\data")
# Depth = -1, performs recursive scanning
for item in local_io.scandir(r'aio', -1):
    print(item.name)
    print(item.path)
    print(item.is_dir())
    print(item.is_file())
    ob = item.stat()
    print(ob.st_mode)

Code sample

# To create an AIO object based on cloud store
cloud_io = AIO(r"C:\data\datacloud.acs")
cloud_io.getcwd()
cloud_io.listdir(r'aio')
# aio is a virtual folder in the cloud store referred in datacloud.acs
# To create an AIO object based on local file system
local_io = AIO()
local_io.getcwd()  # Returns the current working directory
# To create an AIO object based on local file system with path parameter
local_io = AIO(r"C:\data")  # Set the current working directory as C:\data
local_io.getcwd()  # Return the current working directory that is C:\data