π Parliament Watch
Citizens are watching
Thank you for every contibutors who participated in the HACKTOBERFEST 2023 π
Table of Content
β Goal
We want to record and visualise the Thai parliament information including politicians, assemblies, bills, voting processes, and promises.
This project can be seen as a renovated combination of They Work for Us, Law Watch, and Promise Tracker which aim to support several election era.
π Environments
π± Tech Stack
Front-end
Local development
- Yarn v1 as a package manager
- Husky and lint-staged will
- Lint (ESLint) and format (Prettier) code before commiting
- Validate that commit message is align with conventional commit using commitlint
- Run svelte-check before pushing
- For VSCode user, format on save is enabled and prettier-vscode extension will be recommended when open the project.
- Hygen for a code generation
Deployment pipeline
- Staging: Each push will trigger the Github Actions Workflow to build the site, upload the build artifact, and deploy on Cloudflare Pages. Can also be triggered manually.
- Production: The Github Actions Workflow can only be manually triggered to download the latest build artifact and upload to our server through SSH.
πͺ Useful Commands
Start SvelteKit
Start the project in development mode
yarn dev
Start Histoire
to see/develop custom components from Histoire's stories
yarn story:dev
Generate a new component
For a shared component
yarn gen:component
src/components/ComponentName/ directory will be created with the following files:
- ComponentName.svelte for the component source code.
- ComponentName.story.svelte for the Histoire's story file. Follow a guide on writing stories.
Log
Server-side logging for data warning and SvelteKit error can be enabled via environment variable process.env.LOG_TARGET
by setting it to stdout
or file
. More details in logger.ts.
ποΈ Directory Structure
- /_templates Hygen's code generation templates
- /.husky Husky's git hooks
- /src main source codes
- /components Svelte's components
- /mocks Mock data, while we still don't have backend
- /models Main data structure defined with TypeScript interface
- /routes Sveltekit's routes
- /styles Stylesheets, including custom Carbon Design System, tailwind and fonts
- /static static assets such as logos
π Design System
The project design system is based on Carbon Design System v10 with some modification. Custom theme is defined with scss in src/styles/carbon/. To reduce overhead on development, we compile Carbon related stylesheet into src/styles/carbon/precompiled.css with yarn sass:build
command.
Typography
Colors
- tailwind.config.js define utility classes based on color function name according to the Carbon's theme (see Figma file)
- scss variable (need to be imported where you want to use)
Components
- Use Carbon Components Svelte
- We have custom shared component available in src/components/.
- To see shared components' story, open Histoire in local with
yarn story:dev
or browse it in the staging/stories
- See all planned component in Figma file.
- If the component is not yet developed:
- If the component is used by only a specific route, create it in src/components/route-name-and-sub-route-if-exist/
- If the component is shared, run
yarn gen:component
to generate a new component. Don't forget to update a story file for the component documentation.
Icons
πΎ Data Pipeline
Data is pre-process server-side during the build step as following
flowchart TD
A[Google Sheets] -->|d3 fetch: fetch and parse| B(JS objects)
B --> |Zod: validate and transform| C(Type-safe objects)
C --> |used in| D(Svelte's routes)
D --> |Svelte SSR/SSG| E(dev/prod website)
C --> |used in| G(Scheduled GitHub Action)
F(External data source) --> |fetched by| G
G --> |build| H(JSON on GitHub Page)
H --> |fetched by| E(GitHub Page)
- Original data is avaiable at our public Google Sheets
- lib/datasheets provides fetch functions for each tables wrapping d3-fetch and Zod's validation.
- Zod's schema for each data table are defined in src/models which also contains ER Diagram and other TypeScript's interfaces.
- Some data, such as politician ranking from external source, will be updated periodically through scheduled Github Action to reduce unnessesary build-time. The output JSON data is served through Github Pages and can be fetch from the client-side.
π€ Contributing Guideline
Please read CONTRIBUTING.md
π License and Terms of Use
Read more at WeVis's Terms of Use (in Thai)