Awesome
dash-resumable-upload
Dash Resumable Upload component for large files.
This is an alternate (to dcc.Upload) dash upload component which uses resumable.js rather than just the HTML5 File API.
To use it, you must wrap your app.server
with the dash_resumable_upload(app.server, 'path/where/data/goes')
method. This creates a flask endpoint that the uploader sends data to, and puts all uploaded files into this folder.
usage.py in action
Dash
Go to this link to learn about Dash.
Getting started
# Install dependencies
$ npm install
# Watch source for changes and build to `lib/`
$ npm start
Development
Demo server
You can start up a demo development server to see a demo of the rendered components:
$ builder run demo
$ open http://localhost:9000
You have to maintain the list of components in demo/Demo.react.js
.
Code quality and tests
To run lint and unit tests:
$ npm test
To run unit tests and watch for changes:
$ npm run test-watch
To debug unit tests in a browser (Chrome):
$ npm run test-debug
- Wait until Chrome launches.
- Click the "DEBUG" button in the top right corner.
- Open up Chrome Devtools (
Cmd+opt+i
). - Click the "Sources" tab.
- Find source files
- Navigate to
webpack:// -> . -> spec/components
to find your test source files. - Navigate to
webpack:// -> [your/repo/path]] -> dash-resumable-upload -> src
to find your component source files.
- Now you can set breakpoints and reload the page to hit them.
- The test output is available in the "Console" tab, or in any tab by pressing "Esc".
To run a specific test
In your test, append .only
to a describe
or it
statement:
describe.only('Foo component', () => {
// ...
})l
Testing your components in Dash
-
Build development bundle to
lib/
and watch for changes# Once this is started, you can just leave it running. $ npm start
-
Install module locally (after every change)
# Generate metadata, and build the JavaScript bundle $ npm run install-local # Now you're done. For subsequent changes, if you've got `npm start` # running in a separate process, it's enough to just do: $ python setup.py install
-
Run the dash layout you want to test
# Import dash-resumable-upload to your layout, then run it: $ python my_dash_layout.py
TODO: There is a workflow that links your module into site-packages
which would
make it unnecessary to re-run 2.
on every change: python setup.py develop
.
Unfortunately, this doesn't seem to work with resources defined in
package_data
.
See https://github.com/plotly/dash-components-archetype/issues/20
Installing python package locally
Before publishing to PyPi, you can test installing the module locally:
# Install in `site-packages` on your machine
$ npm run install-local
Uninstalling python package locally
$ npm run uninstall-local
Publishing
For now, multiple steps are necessary for publishing to NPM and PyPi, respectively. TODO: #5 will roll up publishing steps into one workflow.
Ask @chriddyp to get NPM / PyPi package publishing accesss.
-
Preparing to publish to NPM
# Bump the package version $ npm version major|minor|patch # Push branch and tags to repo $ git push --follow-tags
-
Preparing to publish to PyPi
# Bump the PyPi package to the same version $ vi setup.py # Commit to github $ git add setup.py $ git commit -m "Bump pypi package version to vx.x.x"
-
Publish to npm and PyPi
$ npm run publish-all
Builder / Archetype
We use Builder to centrally manage build configuration, dependencies, and scripts.
To see all builder
scripts available:
$ builder help
See the dash-components-archetype repo for more information.