ScipyFlow-flowblocks

ScipyFlow-flowblocks is the home repo for flowblocks

How to use

Importing Libraries

In order to import libraries into ScipyFlow, the Flowblock URL in settings must be set to the raw github content url of the flowblock_source_files in the form of

https://raw.githubusercontent.com/{user}/{repo}/{branch}/{path_in_repo}/flowblock_source_files/

by default the url is set to https://raw.githubusercontent.com/asfadmin/ScipyFlow-flowblocks/main/flowblock_source_files/

Contributing

Contributions should be done through a pull request from a feature branch to the main branch. To complete the pull request, at least 1 approving review is required

Branching

For each new feature/minor change, a dedicated branch should be created under the format {identifier}/{branch-name} with the identifier being some set of characters to represent yourself such as your GitHub username. This is to help define ownership of features and branches.

Conventional Commit

All commits made should follow Conventional Commit. This allows for automatic changelog generation and acts as a standard for readability and commit size.

Adding Flowblocks

What is a flowblock?

Flowblocks are small snippets of python code that can be assembled into a larger workflow using ScipyFlow

Note: All FlowBlocks should be encoded in UTF-8

How to create a flowblock?

Flowblocks are just python files with a few required components

1. A yaml docstring at the top of the file
   • The docstring requirements and options are described in the Flowblock Specification section
2. main() function
   • Helper functions can be used, but only the main() function will be called by the program

Flowblock Specification

"""
name: "<< Enter Flowblock Name >>"
requirements:
    - add
    - requirements
    - here
inputs:
    name_of_input_1:
        type: << Any, Bool, Binary, Set, Sequence, Str, Map, None, Number, !CustomClass your.type.here >>
        default: << Value this variable should default to >>
        user_input: << None, Text, Dropdown, FileUpload >>
    name_of_input_N:
        type: << Any, Bool, Binary, Set, Sequence, Str, Map, None, Number, !CustomClass your.type.here >>
        default: << Value this variable should default to >>
        user_input: << None, Text, Dropdown, FileUpload >>
outputs:
    name_of_output:
        type: << Any, Bool, Binary, Set, Sequence, Str, Map, None, Number, !CustomClass your.type.here >>
description: "<< Describe what this flowblock does >>"
"""

def main(name_of_input_1, name_of_input_N):
    # python code 
    return <<your return>>

• name: The display name of your flowblock
• requirements: The packagages that need to be installed via micropip
• inputs: One or more variables that will be input into the flowblock
  • type: The type of the variable, options include
    • Any: Any of the following types
    • Bool: A Boolean value
    • Binary: Represents bytes, bytearray, and memoryview types
    • Set: Represents set and frozenset types
    • Sequence: Represents list, tuple, and range types
    • Str: Represents str type
    • Map: Represents dict type
    • None: Represents no type
    • Number: Represents int, float, or complex types
    • CustomClass: Allows for any arbitrary python class to be passed with the syntax !CustomClass your.custom.class
  • default: (OPTIONAL) The value that this input will default to
  • user_input: (OPTIONAL) Boolean value on if the user should be able to manually input a value, defaults to False
• outputs: One or more variables that will be returned from the flowblock
  • type: The type of the variable, options include
    • Any: Any of the following types
    • Bool: A Boolean value
    • Binary: Represents bytes, bytearray, and memoryview types
    • Set: Represents set and frozenset types
    • Sequence: Represents list, tuple, and range types
    • Str: Represents str type
    • Map: Represents dict type
    • None: Represents no type
    • Number: Represents int, float, or complex types
    • CustomClass: Allows for any arbitrary python class to be passed with the syntax !CustomClass your.custom.class
• description: A short text that describes what this flowblock does

Code Requirements

All python code must run in a WASM Emscripten environment. This means that all imported python packages must have a pure python wheel. All python code should also be able to run in a Desktop environment.
You can determine what system the code is currently running on using the platform library

import platform
system = platform.system()

if system == "Emscripten":
    # import web packages
    # Run web only code
else:
    # import desktop packages
    # Run desktop only code

# Run universal code

Location and Libraries

Flowblocks must be added to the /flowblock_source_files directory. Flowblocks should be in a "library", a library is a folder inside the flowblock_source_files directory so the path of a library is /flowblock_source_files/{library name}.

Libraries are meant to categorize the flowblocks. If there is an existing library relevant to your flowblock idea, put your new flowblock in there, otherwise create a new library with an appropriate name.