Skip to content

Latest commit

 

History

History
82 lines (60 loc) · 4.08 KB

architecture.md

File metadata and controls

82 lines (60 loc) · 4.08 KB
Raw

Assembly Line Weaver Architecture

The AssemblyLine Weaver is a collection of Python modules and Docassemble YAML interviews that provides a step-by-step graphical process for building draft Docassemble interviews. It is optimized for quickly building a skeleton project. The assumption is that the most common use cases will be for building court forms or court-form like automations.

For a high-level overview of the components of the AssemblyLine project, see: https://suffolklitlab.org/docassemble-AssemblyLine-documentation/docs/al_project_architecture

Overview

YAML files contain interactive questions and answers. One Python module contains a list of strings that are handled specially when recognizing fields in PDF and Word documents. The remaining modules are used as the engine of the draft interview automation process. This package also contains tests and templates that are used to standardize the automation process.

Key files

YAML files (in docassemble/ALWeaver/data/questions)

File you should run in the browser:

  • assembly_line.yml can be run directly on a Docassemble server and references the other YAML files in the same directory

Other files:

  • config.yml: frontend to help developers continue building the configuration system (not for end users)
  • docx_field_tester.yml: logic for DOCX template validation
  • feedback.yml: Feedback form that lets users report issues on GitHub in-context
  • pdf_field_test.yml: logic for PDF template validation
  • visual.yml: controls visual elements in the Weaver - essentially a theme

Source files in docassemble/ALWeaver/data/sources

  • configuration_capabilities.yml: currently describes the list of optional packages someone can install in a generated package with the Weaver--envisioned to allow more flexible configuration in the future
  • output_patterns.yml: this file contains small Mako templates that are used to build the YAML file that the Weaver produces. It is set up as a series of small templates right now rather a single large template.

This directory also contains test files for unit testing with ALKiln (see below for more information.)

Static files in docassemble/ALWeaver/data/static

These files are primarily the front-end interface files, including images and CSS.

The next_steps DOCX files are templates for "next steps" documents that a user can print and read after using an interview. They are associated with different kinds of interviews that the Weaver can produce.

Python modules

  • interview_generator.py is the primary module containing most of the Python code used by the Weaver
  • advertise_capabilities.py is part of the plugin-able configuration system - it tells the server what optional dependencies this Weaver can add to a generated interview file
  • custom_values.py is used to scan the server for packages that contain custom configuration settings for the Weaver
  • draggable_table.py is used by the Weaver frontend to allow rearranging long lists of fields
  • field_grouping.py is a copy of some features from FormyFyxer that power the "I'm feeling lucky" button (should be deprecated)
  • generator_constants.py contains several lists of rules for how to transform PDF field names like users_name_full into Docassemble objects like users[0].name, as well as indicating reserved DOCX variable names that are handled by questions in the AssemblyLine's question library

Testing

The Weaver has two kinds of test that are currently configured to run on push to GitHub:

  1. Standard unit tests of pure-Python modules
  2. Integration tests using the ALKiln testing framework

Unit tests can be found in docassemble/ALWeaver/. Filenames begin with test_.

The integration tests are located in docassemble/ALWeaver/data/sources/ and filenames ending with .feature will be run as ALKiln tests.

In addition, generator_test.yml is an interactive Docassemble interview that will test the map_raw_to_final_display() function from interview_generator.py. This is designed for quick in-browser testing.