Skip to content

Latest commit

 

History

History
233 lines (165 loc) · 8.32 KB

README.rst

File metadata and controls

233 lines (165 loc) · 8.32 KB

open-humans-api

This package aims to provide some tools to facilitate working with the Open Humans APIs.

In particular, this package provides some command line tools for data file downloads and uploads. These tools are listed below.

Installation

This package is distributed via PyPI. We recommend you install it using pip, e.g. pip install open-humans-api.

Command line tools

Command line tools aim to facilitate one-off operations by users (for example, one-off data upload by a project).

These tools might also be helpful for programmers seeking to use the API in non-Python programmatic contexts.

ohpub-download

Usage: ohpub-download [OPTIONS]

  Download public data from Open Humans.

Options:
  -s, --source TEXT     the source to download files from
  -u, --username TEXT   the user to download files from
  -d, --directory TEXT  the directory for downloaded files
  -m, --max-size TEXT   the maximum file size to download
  --help                Show this message and exit.

Examples

# download all 23andMe files to 23andme/
$ mkdir 23andme
$ ohpub-download --source twenty_three_and_me --directory 23andme
# download all of beau's files to the current directory
$ ohpub-download --username beau

ohproj-download

Usage: ohproj-download [OPTIONS]

  Download data from project members to the target directory.

  Unless this is a member-specific download, directories will be created for
  each project member ID. Also, unless a source is specified, all shared
  sources are downloaded and data is sorted into subdirectories according to
  source.

  Projects can optionally return data to Open Humans member accounts. If
  project_data is True (or the "--project-data" flag is used), this data
  (the project's own data files, instead of data from other sources) will be
  downloaded for each member.

Options:
  -d, --directory TEXT     Target directory for downloaded files.  [required]
  -T, --master-token TEXT  Project master access token.
  -m, --member TEXT        Project member ID.
  -t, --access-token TEXT  OAuth2 user access token.
  -s, --source TEXT        Only download files from this source.
  --project-data TEXT      Download this project's own data.
  --max-size TEXT          Maximum file size to download.  [default: 128m]
  -v, --verbose            Report INFO level logging to stdout
  --debug                  Report DEBUG level logging to stdout.
  --memberlist TEXT        Text file with whitelist IDs to retrieve
  --excludelist TEXT       Text file with blacklist IDs to avoid
  --help                   Show this message and exit.

ohproj-download-metadata

Usage: ohproj-download-metadata [OPTIONS]

  Draft or review metadata files for uploading files to Open Humans.

  The target directory should either represent files for a single member (no
  subdirectories), or contain a subdirectory for each project member ID.

  -T, --master-token TEXT  Project master access token.  [required]
  --output-csv TEXT        Output project metedata CSV  [required]
  --max-size TEXT       Maximum file size to consider.  [default: 128m]
  -v, --verbose            Show INFO level logging
  --debug                  Show DEBUG level logging.
  --help                Show this message and exit.

ohproj-upload-metadata

Usage: ohproj-upload-metadata [OPTIONS]

  Draft or review metadata files for uploading files to Open Humans.

  The target directory should either represent files for a single member (no
  subdirectories), or contain a subdirectory for each project member ID.

Options:
  -d, --directory TEXT  Target directory  [required]
  --create-csv TEXT     Create draft CSV metadata  [required]
  --max-size TEXT       Maximum file size to consider.  [default: 128m]
  -v, --verbose         Show INFO level logging
  --debug               Show DEBUG level logging.
  --help                Show this message and exit.

Example usage: creating metadata for data upload

Create directory containing data for project members. For example it might look like the following example (two project members with IDs '01234567' and '12345678').

  • member_data/
  • 01234567/
    • testdata.json
    • testdata.txt
  • 12345678/
    • testdata.json
    • testdata.txt

Draft metadata file:

$ ohproj-metadata -d member_data --create-csv member_data_metadata.csv

Initially it looks like this:

project_member_id,filename,tags,description,md5,creation_date
01234567,testdata.txt,,,fa61a92e21a2597900cbde09d8ddbc1a,2016-08-23T15:23:22.277060+00:00
01234567,testdata.json,json,,577da9879649acaf17226a6461bd19c8,2016-08-23T16:06:16.415039+00:00
12345678,testdata.txt,,,fa61a92e21a2597900cbde09d8ddbc1a,2016-09-20T10:10:59.863201+00:00
12345678,testdata.json,json,,577da9879649acaf17226a6461bd19c8,2016-09-20T10:10:59.859201+00:00

You can use a spreadsheet editor to edit it. Make sure to save the result as CSV! For example, it might look like this if you add descriptions and more tags:

1234567,testdata.txt,"txt, verbose-data",Complete test data in text format.,fa61a92e21a2597900cbde09d8ddbc1a,2016-08-23T15:23:22.277060+00:00
1234567,testdata.json,"json, metadata",Summary metadata in JSON format.,577da9879649acaf17226a6461bd19c8,2016-08-23T16:06:16.415039+00:00
12345678,testdata.txt,"txt, verbose-data",Complete test data in text format.,fa61a92e21a2597900cbde09d8ddbc1a,2016-09-20T10:10:59.863201+00:00
12345678,testdata.json,"json, metadata",Summary test data JSON.,577da9879649acaf17226a6461bd19c8,2016-09-20T10:10:59.859201+00:00

ohproj-upload

Usage: ohproj-upload [OPTIONS]

  Upload files for the project to Open Humans member accounts.

  If using a master access token and not specifying member ID:

  (1) Files should be organized in subdirectories according to project
  member ID, e.g.:

      main_directory/01234567/data.json
      main_directory/12345678/data.json
      main_directory/23456789/data.json

  (2) The metadata CSV should have the following format:

      1st column: Project member ID
      2nd column: filenames
      3rd & additional columns: Metadata fields (see below)

  If uploading for a specific member:
      (1) The local directory should not contain subdirectories.
      (2) The metadata CSV should have the following format:
          1st column: filenames
          2nd & additional columns: Metadata fields (see below)

  The default behavior is to overwrite files with matching filenames on Open
  Humans, but not otherwise delete files. (Use --safe or --sync to change
  this behavior.)

  If included, the following metadata columns should be correctly formatted:
      'tags': should be comma-separated strings
      'md5': should match the file's md5 hexdigest
      'creation_date', 'start_date', 'end_date': ISO 8601 dates or datetimes

  Other metedata fields (e.g. 'description') can be arbitrary strings.

Options:
  -d, --directory TEXT     Target directory for downloaded files.  [required]
  --metadata-csv TEXT      CSV file containing file metadata.  [required]
  -T, --master-token TEXT  Project master access token.
  -m, --member TEXT        Project member ID.
  -t, --access-token TEXT  OAuth2 user access token.
  --safe                   Do not overwrite files in Open Humans.
  --sync                   Delete files not present in local directories.
  --max-size TEXT          Maximum file size to download.  [default: 128m]
  -v, --verbose            Report INFO level logging to stdout
  --debug                  Report DEBUG level logging to stdout.
  --help                   Show this message and exit.

Example usage: uploading data

For organizing the data files and creating a metadata file, see the example usage for the ohproj-metadata command line tool.

Uploading that data with a master access token:

$ ohproj-upload -T MASTER_ACCESS_TOKEN --metadata-csv member_data_metadata.csv -d member_data