Inverted font icons in release 4.5.0

[imalcolm1]

monday.com sync: #12611592408

Check existing issues

• I have checked for existing issues to avoid duplicates

Which Component

N/A

Actual Behavior

We are seeing the same behavior as in this issue for the publish, review map, and show oriented imagery font icons in release 4.5.0. The icons have the same style="fill:none" rather than fill="none" tag that seemed to be the cause before.

Expected Behavior

The font icons should render the same as the source svgs.

Reproduction Sample

https://github.com/Esri/calcite-design-system/releases/tag/%40esri%2Fcalcite-ui-icons%404.5.0

Reproduction Steps

1. Download the 4.5.0 font icons
2. Unzip the fonts from the TGZ file located in \package\fonts
3. Open up https://fontdrop.info/ and drag one of the fonts to the drop-area
4. Notice that the three mentioned fonts render in inverted colors

Reproduction Version

calcite-ui-icons 4.5.0

Relevant Info

No response

Regression?

No since these are new icons.

Priority impact

impact - p2 - want for an upcoming milestone

Impact

Unable to use the affected items from the font.

Calcite package

• @esri/calcite-components
• @esri/calcite-components-react
• @esri/calcite-design-tokens
• @esri/calcite-ui-icons
• @esri/eslint-plugin-calcite-components

Esri team

ArcGIS Maps SDK for JavaScript

[stephenquan]

@sfrizzell @nicklu10 I sent you an email regarding this issue on 25th May titled "Issues with new "buttons" and "entity types" ui-icons" in which I wrote that some of the SVG icons have switched to using CSS style for the backing rectangle which is being misinterpreted by Fantasticon causing Fantasicon to invert the icon unintentionally. This is a critical blocker preventing us from using the https://github.com/Esri/calcite-dotnet-toolkit. We had to make manual edits to copy and make the many edits swapping out the icon as follows:

 -    <path d="M0 0h16v16H0z" style="fill:none"/>
 +    <path d="M0 0h16v16H0z" fill="none"/>

This was also the same conclusion that @dotMorten came to in this comment: #13116 (comment)