Cloud Object Storage file upload and download utility

Table of content:

• Getting Started
• Uploading files to a Cloud Object Storage bucket
• Downloading files from a Cloud Object Storage bucket

---

Getting started

The utility requires Python 3.6 or above.

Installation

You can install the utility from PyPI or from the source.

Install from pypi.org

$ pip install cos-utils --upgrade

Install from source code

$ git clone https://github.com/CODAIT/cos-utils.git
$ cd cos-utils
$ pip install .

Configuration

Set the AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEYenvironment variables based on your Cloud Object Storage HMAC credentials.

$ export AWS_ACCESS_KEY_ID=...
$ export AWS_SECRET_ACCESS_KEY=...

Uploading files to a Cloud Object Storage bucket

You can run the upload utility in a terminal window using the generated console script

$ upload_files --help

or explicitly

$ python -m cos_utils.upload_files --help

The help lists required and optional parameters. The examples listed below explain them in detail.

usage: upload_files [-h] [-p PREFIX] [-r] [-s] [-w] bucket source

Upload files to a Cloud Object Storage bucket.

positional arguments:
  bucket                Bucket name
  source                File or directory spec (supported wildcards: * and ?)

optional arguments:
  -h, --help            show this help message and exit
  -p PREFIX, --prefix PREFIX
                        Key name prefix
  -r, --recursive       Include files in subdirectories
  -s, --squash          Exclude subdirectory name from key name
  -w, --wipe            Clear bucket prior to upload

Environment variables AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY must be
defined to run the utility.

Example scenario

The </path/to/local/directory> contains the following directories and files:

file1.png
file2.png
file3.jpg
file4.txt
dir1/file5.gif
dir1/file6.png
dir1/dir2/file7.png
dir1/dir3/file8.jpg
dir1/dir3/file1.png

In the examples given below <bucket-name> refers to an existing bucket in Cloud Object Storage.

Upload directories

You can upload the content of any directory.

Upload the content of </path/to/local/directory> to bucket <bucket-name>

$ upload_files <bucket-name> </path/to/local/directory>

Bucket <bucket-name> contains the following objects:

file1.png
file2.png
file3.jpg
file4.txt

Same as before but clear the bucket first before uploading

Specify the optional --wipe parameter to clear the bucket before upload.

$ upload_files <bucket-name> </path/to/local/directory> --wipe

Bucket <bucket-name> contains the following objects:

file1.png
file2.png
file3.jpg
file4.txt

Same as before but include subdirectories

Specify the optional --recursive parameter include files in subdirectories.

$ upload_files <bucket-name> </path/to/local/directory> --wipe --recursive

Bucket <bucket-name> contains the following objects:

file1.png
file2.png
file3.jpg
file4.txt
dir1/file5.gif
dir1/file6.png
dir1/dir2/file7.png
dir1/dir3/file8.jpg
dir1/dir3/file1.png

Same as before but don't use subdirectory names during object key generation

Specify the optional --squash parameter to ignore subdirectory names during object key generation.

$ upload_files <bucket-name> </path/to/local/directory> --wipe --recursive --squash

Bucket <bucket-name> contains the following objects. Note that </path/to/local/directory> contains two files named file1.png. First file1.png is uploaded and later overwritten with the content of dir1/dir3/file1.png.

file2.png
file3.jpg
file4.txt
file5.gif
file6.png
file7.png
file8.jpg
file1.png

Same as before but include a static key name prefix

Specify the optional --prefix <prefix> parameter to add <prefix> to the object key for every file.

$ upload_files <bucket-name> </path/to/local/directory> --wipe --recursive --squash --prefix data

Bucket <bucket-name> contains the following objects:

data/file2.png
data/file3.jpg
data/file4.txt
data/file5.gif
data/file6.png
data/file7.png
data/file8.jpg
data/file1.png

Upload files

You can upload a single file by specifying </path/to/local/directory/filename>.

$ upload_files <bucket-name> /path/to/local/directory/file1.png --wipe 

Bucket <bucket-name> contains the following object:

file1.png

You can upload multiple files by specifying a pattern </path/to/local/directory/filename-pattern>

$ upload_files <bucket-name> /path/to/local/directory/*.png --wipe 

On Linux, Unix and MacOS wildcards need to be escaped to prevent shell expansion: /path/to/local/directory/\*.png.

Bucket <bucket-name> contains the following objects:

file1.png
file2.png

Use the --recursive parameter to extend the search to subdirectories of /path/to/local/directory/.

$ upload_files <bucket-name> /path/to/local/directory/*.png --wipe --recursive

file1.png
file2.png
dir1/file6.png
dir1/dir2/file7.png
dir1/dir3/file1.png

Downloading files from a Cloud Object Storage bucket

You can run the download utility in a terminal window using the generated console script

$ download_files --help

or explicitly

$ python -m cos_utils.dowload_files --help

The help lists required and optional parameters. The examples listed below explain them in detail.

usage: download_files [-h] [-d TARGET_DIR] bucket source

Download objects from a Cloud Object Storage bucket.

positional arguments:
  bucket                Bucket name
  source                Object key spec (supported wildcards: * and ?)

optional arguments:
  -h, --help            show this help message and exit
  -d TARGET_DIR, --target_dir TARGET_DIR
                        Local target directory. Defaults to the current
                        directory.

Environment variables AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY must be
defined to run the utility.

Download complete bucket content

You can download the complete content of a bucket to the current directory:

$ download_files <bucket-name> *

On Linux, Unix and MacOS wildcards need to be escaped to prevent shell expansion: download_files <bucket-name> \*.

Same as before but specify a target directory

Use the --target_dir </path/to/local/dir> parameter to specify an existing directory where the downloaded files will be stored:

$ download_files <bucket-name> * --target_dir /tmp/downloads

Use wildcards to selectively download files

Use the * (any character) and ? (one character) wildcards to define a filter condition.

Download only png files

$ download_files <bucket-name> *.png

Download files that contain a certain string in their name

$ download_files <bucket-name> *fil*

License

Apache-2.0