joshuaberetta / kobomedia

Setup

git clone https://github.com/joshuaberetta/kobomedia
cd kobomedia

# optional
chmod +x kobomedia.py
sudo ln -s $(pwd)/kobomedia.py /usr/local/bin/kobomedia

Create kobo.json config file with the following settings:

{
    "token": "",
    "kf_url": "",
    "kc_url": ""
}

Usage

# download media
python3 kobomedia.py --asset-uid agBMEh8GWxTrCSWQuWyE5d

# or
./kobomedia.py --asset-uid agBMEh8GWxTrCSWQuWyE5d

# or
kobomedia --asset-uid agBMEh8GWxTrCSWQuWyE5d

Optional fields

• limit: Limit number of submissions per query, paginate until complete
• query: Set a custom query in the Mongo query syntax
• question-names: Specify question names to download media for, comma separated
• chunk-size: Set chunk size for saving data to files
• throttle: Control time between each download to reduce server strain
• verbosity: Control verbosity of stdout

./kobomedia.py --asset-uid agBMEh8GWxTrCSWQuWyE5d \
  --limit 10 \
  --query '{"_submission_time": {"$gt": "2021-08-04"}}' \
  --question-names group1/q1,group2/q2 \
  --chunk-size 2048 \
  --throttle 2 \
  --verbosity 2 \

Output

Media downloads will be in the following directory structure:

{asset_uid}
├── {submission_uid}
│   ├── {filename}
│   └── {filename}
├── {submission_uid}
│   └── {filename}
├── {submission_uid}
│   └── {filename}
├── {submission_uid}
│   └── {filename}
└── {submission_uid}
    ├── {filename}
    └── {filename}

If a file has previously been downloaded, it will be skipped (not downloaded again) on subsequent runs. Therefore the script can be run periodically to keep a local sync of submitted media on the server.