/smugcli

Command line tool for SmugMug

Primary LanguagePythonMIT LicenseMIT

smugcli

Command line tool for SmugMug, useful for automatically synchronizing a local folder hierarchy with a SmugMug account.

Implemented using the Smugmug V2 API.

Tested with Python 3.5, 3.6, 3.7, 3.8 and 3.9.

Installation

pip install smugcli

Sample usage

To use this command line tool, you will need to request your own API key by visiting https://api.smugmug.com/api/developer/apply. Using your key and secret, loging to smugcli using the following command:

$ ./smugcli.py login --key=<key> --secret=<secret>

This is a one time operation. From this point on, smugcli will be able to access your SmugMug account. To logout, run the command: $ ./smugcli.py logout

You can list the content of your SmugMug account by doing:

$ ./smugcli.py ls
 Photography
 Portfolio
 Other

$ ./smugcli.py ls Photography
 2014
 2015
 2016

$ ./smugcli.py ls Photography/2015
 Photoshoot with Dave

Note that smugcli can also be used to access other public SmugMug account by using the --user argument:

$ ./smugcli.py ls -u <username>

Folders can be created by using the mkdir command:

$ ./smugcli.py mkdir Photography/2017

Similarily, albums can be created by doing:

$ ./smugcli.py mkalbum 'Photography/2017/My new album'

To upload photos to an album, run:

$ ./smugcli.py upload local/folder/*.jpg 'Photography/2017/My new album'

Finally, the nicest feature of all, you can synchronize a whole local folder hierarchy to your SmugMug account using the sync command:

$ ./smugcli.py sync local/folder remote/folder
Creating Folder "remote/folder/2015".
Creating Album "remote/folder/2015/2015-08-03, Mary's Wedding".
Uploading "local/folder/2015/2015-08-03, Mary's Wedding/DSC_0001.JPG"
Uploading "local/folder/2015/2015-08-03, Mary's Wedding/DSC_0002.JPG"
Uploading "local/folder/2015/2015-08-03, Mary's Wedding/DSC_0003.JPG"
...
Creating Album "remote/folder/2015/2015-09-10, Andy's Photoshoot"
Uploading "local/folder/2015/2015-09-10, Andy's Photoshoot/DSC_0043.JPG"
Uploading "local/folder/2015/2015-09-10, Andy's Photoshoot/DSC_0052.JPG"
...

The sync command can be re-executed to update the remote Albums in the event that the local files might have been updated. Only the files that changed will be re-uploaded.

Multiple sources can be synced in the same operation, the last argument being the destination folder and the others being the sources:

$ ./smugcli.py sync 2016 2017 2018 remote/folder
Syncing:
  2016
  2017
  2018
to SmugMug folder "remote/folder"',
...

$ ./smugcli.py sync 201* remote/folder
...

Source files are synced to the destination SmugMug album and source folders are recursively synced to the destination SmugMug folder. For source folders with a trailing path delimitor ('/' or '\' depending on OS), only the content of the folder is synced, skipping the folder itself (equivalent of doing folder/*) . This means that smugcli.py sync src/album dst is equivalent to smugcli.py sync src/album/ dst/album. If no sources or destinations are specified, smugcli defaults to syncing the current folder to the SmugMug user's root.

The sync command uses multiple threads to speed-up the file scanning and upload. You may want to tune the number of threads used by SmugCLI depending on your machine's performance. For instance:

$ ./smugcli.py sync local/folder remote/folder --folder_threads=4 --file_threads=8 --upload_threads=2

folder_threads control the number of album folders that are processed in parallel. file_threads specifies the number fo files that are read from disk and compared with the server side version in parallel. upload_threads controls the number of parallel upload operations allowed when sending content to SmugMug. Keep in mind that too many or too few threads can be harmful to performance. Also keep in mind that increasing file_threads or upload_threads means that more files will be loaded in memory at the same time. If you have many large video files, loading too many in parallel could hog your system's resources.

When you are happy with the performance using certain thread counts, you may save these preferences so that they'd be used as defaults next time:

$ ./smugcli.py sync --set_defaults --folder_threads=4 --file_threads=8 --upload_threads=2

To exclude paths from the sync operation, run the command:

$ ./smugcli.py ignore local/folder/export-tmp

To undo this operation, you can run:

$ ./smugcli.py include local/folder/export-tmp

Running the tests

PLEASE READ, RUN UNIT TESTS AT YOUR OWN RISKS: smugcli's unit-tests use the logged-in user account to do run actual commands on SmugMug. All operations should be performed in a __smugcli_unit_tests__ subfolder in the SmugMug account's root. This folder should be deleted automatically when/if the test completes. If in doubt, do smugcli.py logout && smugcli.py login and use a test account.

SmugCLI uses tox to run tests using all supported Python interpreter versions. Run all tests with all Python versions by running:

$ tox

To run with only one specific Python version, for instance Python 3.7, do:

$ tox -e py39

Individual tests can be ran by doing:

$ tox -- tests/module[.class_name[.test_name]]

for instance:

$ tox -e py39 -- tests/end_to_end_test.py  # Runs all tests in tests/end_to_end_test.py.
$ tox -e py39 -- tests/end_to_end_test.py::EndToEndTest  # Runs all tests in class EndToEndTest.
$ tox -e py39 -- tests/end_to_end_test.py::EndToEndTest::test_sync  # Runs a single test.

Since the unit tests do actual operations on SmugMug, they are fairly slow. To speed things up during development, an HTTP request cache can be enabled so that responses from the previous run are replayed instead of re-doing the actual HTTP requests to SmugMug. To enable this cache, set the REUSE_RESPONSES environment variable to True:

$ REUSE_RESPONSES=True tox -e py39

Windows users can do the equivalent by doing:

C:\smugcli> cmd /C "set REUSE_RESPONSES=True && tox -e py39"

Note that if you change the code such that different HTTP requests are done, you will have to set REUSE_RESPONSES to False on the next run to update the cache.