18 Aug 10:08

8d35c3a

8.0.0

What's Changed

See the 8.0.0 changelog for a more comprehensive summary of changes. To see the full list of pull requests and issues, see the 8.0 milestone on GitHub, or the full list of changes since 7.x.

Users

Here are some highlights of user-visible changes in ipywidgets 8.0.

Date and time pickers

In addition to the existing DatePicker widget, we now have new DatetimePicker and TimePicker widgets. (#2715)

from ipywidgets import VBox, TimePicker, DatetimePicker
VBox([
  TimePicker(description='Time'),
  DatetimePicker(description='Date/Time')
])

Tags input widget

The new TagsInput widget provides an easy way to collect and manage tags in a widget. You can drag and drop tags to reorder them, limit them to a set of allowed values, or even prevent making duplicate tags. (#2591, #3272)

from ipywidgets import TagsInput
TagsInput(
    value=['pizza', 'fries'],
    allowed_tags=['pizza', 'fries', 'tomatoes', 'steak'],
    allow_duplicates=False
)

Similarly, the new ColorsInput widget provides a way to select colors as tags

from ipywidgets import ColorsInput
ColorsInput(
    value=['red', '#2f6d30'],
    # allowed_tags=['red', 'blue', 'green'],
    # allow_duplicates=False
)

Stack widget

The new Stack container widget shows only the selected child widget, while other child widgets remain hidden. This can be useful if you want to have a area that displays different widgets depending on some other interaction.

from ipywidgets import Stack, Button, IntSlider, Dropdown, VBox, link
s = Stack([Button(description='Click here'), IntSlider()], selected_index=0)
d = Dropdown(options=['button', 'slider'])
link((dropdown, 'index'), (stacked, 'selected_index'))
VBox([d, s])

File upload widget

The file upload widget has been overhauled to handle multiple files in a more useful format:

The .value attribute is now a list of dictionaries, rather than a dictionary mapping the uploaded name to the content.
The contents of each uploaded file is a memory view in the .content key, e.g., uploader.value[0].content.
The .data attribute has been removed.
The .metadata attribute has been removed.

See the user migration guide for details on how to migrate your code.

(#2767, #2724, #2666, #2480)

More styling options

Many style and layout options have been added to core widgets:

Tooltips are now supported for many core widgets, rather than just a few description tooltips (#2680)
```
from ipywidgets import Button
Button(description="Click me", tooltip='An action')
```
Borders can be styled independently with the layout's border_top, border_right, border_bottom, border_left attributes (#2757, #3269)
```
from ipywidgets import Button
Button(description="Click me", layout={'border_bottom': '3px solid blue'})
```
Descriptions are now plain text by default for security, but you can set them to allow HTML with the description_allow_html attribute (HTML content is still sanitized for security). (#2785)
```
from ipywidgets import Text
Text(description="<b>Name</b>", description_allow_html=True)
```
Many other styling attributes can be set on core widgets, such as font family, size, style, weight, text color, and text decoration. See the table in the documentation for a reference. (#2728)
The SelectionSlider now has a handle_color style attribute (#3142)
To control keyboard navigation, widgets can be set to be tabbable or not (i.e., that the tab key will traverse to the widget) (#2640)

Selection container titles

The Accordion, Tab, and Stack widgets now have a .titles attribute that you can use to get and set titles from the constructor or as an attribute. (#2746, #3296, #3477)

from ipywidgets import Tab, IntSlider, Text
Tab([IntSlider(), Text()], titles=('Slider', 'Text'))

Slider implementation

The slider implementation in the core widgets now uses nouislider. This enables us to fix long-standing bugs and introduce new features, like dragging the range in a RangeSlider. (#2712, #630, #3216, #2834)

Collaboration

By default, ipywidgets 8 enables a collaboration mode, where widget updates from one frontend are reflected back to other frontends, enabling a consistent state between multiple users and fixing synchronization between the kernel and frontend. You may want to disable these update echo messages if they are using too much bandwidth or causing slower interactivity. To disable echo update messages across ipywidgets, set the environment variable JUPYTER_WIDGETS_ECHO to 0. For widget authors, to opt a specific attribute of custom widget out of echo updates (for example, if the attribute contains a lot of data that does not need to be synchronized), tag the attribute with echo_update=False metadata (we do this in core for the FileUpload widget's data attribute). (#3195, #3343, #3394, #3407)

CDN for html manager

We have made it easier to load widgets from content delivery networks.

The default CDN is changed from unpkg to jsDelivr (#3121, #1627)
You can use the data-jupyter-widgets-cdn-only attribute to load modules only from CDN (#2792, #2786)
We have updated the webpack public path settings so the HTMLManager and the Jupyter Notebook extensions pull assets from wherever they are loaded, rather than only from CDN. If you author a custom widget, we highly encourage you to apply similar changes to your widget by adapting the changes at https://github.com/jupyter-widgets/widget-cookiecutter/pull/103/files. #3464, #3508

Other changes

Here is a short list of some of the other changes in ipywidgets 8.0.

Add a cookiecutter-based tutorial to build a custom widget (#2919)
Change media widgets to use memory views. (#2723)
Upgrade to FontAwesome 5 in html-manager (#2713)
Play widget now toggles between play and pause button as needed (#2703, #2671)
Drop support for mapping types as selection options (#2679, #1958)
Focus or blur a widget. (#2664, #2692, #2691, #2690)
Drop notebook dependency from widgetsnbextension (#2590)
Cast 'value' in range sliders to a tuple (#2441)
Play widget: expose playing and repeat (#2283, #1897)
Fix debouncing and throttling code (#3060)
Fix regression on spinning icons (#2685, #2477)
Fix selection container default index (#1823)
Remove deprecated overflow properties (#2688)
Select: Do not force a selec...

Contributors

agnat, qzchenwl, and 44 other contributors

Assets 2

27 Jul 10:28

jtpio

7.7.1

7ad9e62

7.7.1

To see the full list of pull requests, see the 7.7.1 milestone on GitHub.

Highlights include:

Fix broken link icon for FontAwesome 4 and 5 #3495
Fix message throttling bug #3494
Fix state message parsing to be more permissive #3486
Fix tests on Python 3.11 #3480
Add better front-page docs #3496

7.7

To see the full list of pull requests and issues, see the 7.7 milestone on GitHub.

Highlights include:

Fix installation on Python 3.10. #3368
Throw an error if we cannot render a widget, enabling the rendering system to fall back to rendering a different data type if available. #3290
Create a new widget control comm channel, enabling more efficient fetching of kernel widget state. #3201
Refactor logic for fetching kernel widget state to the manager base class. This logic first tries to use the new widget control comm channel, falling back to the existing method of requesting each widget's state individually. #3337
Enable HTMLManager output widgets to render state updates. #3372
Do not reset JupyterLab CSS variables if they are already defined. #3344
Fix variable inspector example. #3302
Introduce new widget manager has_model method for synchronously checking if a widget model is registered. #3377
Work around bug in Chrome rendering Combobox arrows. #3375
Optionally echo update messages from frontends to other frontends. This enables widget views in different frontends to maintain consistent state simultaneously, and also makes sure that simultaneous updates from the kernel and frontend resolve to a consistent state. This is off by default in ipywidgets 7.7, and it is anticipated this will be on by default in ipywidgets 8.0. To enable echo update messages across ipywidgets, set the environment variable JUPYTER_WIDGETS_ECHO to 1. To opt a specific attribute out of echo updates, tag the attribute with echo_update=False metadata (we do this in core for the FileUpload widget's data attribute). #3400, #3394

Assets 2

07 Feb 22:09

willingc

7.6.3

5532c91

7.6.3

A minor release in 7.6.x series.

From changelog.md:

7.6

To see the full list of pull requests and issues, see the 7.6.0 milestone on GitHub.

The main change in this release is that installing ipywidgets 7.6.0 will now automatically enable ipywidgets support in JupyterLab 3.0—a user has no extra JupyterLab installation step and no rebuild of JupyterLab, nor do they need Node.js installed. Simply install the python ipywidgets package with pip (pip install ipywidgets==7.6.0) or conda/mamba (conda install -c conda-forge ipywidgets=7.6.0) and ipywidgets will automatically work in classic Jupyter Notebook and in JupyterLab 3.0.

This is accomplished with the new python package jupyterlab_widgets version 1.0, on which ipywidgets 7.6.0 now depends (similar to how ipywidgets already depends on the widgetsnbextension package to configure ipywidgets for the classic Jupyter Notebook). The jupyterlab_widgets Python package is a JupyterLab 3.0 prebuilt extension, meaning that it can be installed into JupyterLab 3.0 without rebuilding JupyterLab and without needing Node.js installed.

Updates for Widget Maintainers

Custom widget maintainers will need to make two changes to update for JupyterLab 3:

Update the @jupyter-widgets/base dependency version to include ^4 to work in JupyterLab 3.0. For example, if you had a dependency on @jupyter-widgets/base version ^2 || ^3, update to ^2 || ^3 || ^4 for your widget to work in classic Jupyter Notebook, JupyterLab 1, JupyterLab 2, and JupyterLab 3. See #2472 for background.

In the package.json, add the following sharedPackages configuration inside the jupyterlab key. See the JupyterLab extension documentation for more information.

  "jupyterlab": {
    "sharedPackages": {
      "@jupyter-widgets/base": {
        "bundled": false,
        "singleton": true
      }
    }
  }

Separate from these two steps to update for JupyterLab 3, we also recommend that you make your widget's JupyterLab extension a prebuilt extension for JupyterLab 3.0. Users will be able to install your JupyterLab 3.0 prebuilt extension without rebuilding JupyterLab or needing Node.js. See the JupyterLab 3 extension developer documentation or the new widget extension cookiecutter for more details.

Assets 2