" Format Error: Num Dot Decimal,000,000" raised intermittently within Viewer

[gjk-corefiling]

Dear all,

I wanted to report a bug with Edgar Renderer.
Transformation error is being raised incorrectly and inconsistently, for example, I open the fact list on the right-hand side, I see the wrong errors, I go to a different page, come back to the original page and the errors have disappeared.

Further, in preparation software, I can delete several tags, and tags which previously have caused issues, would not have the error anymore. No modifications would happen to the tags that were kept.

I am attaching an example iXBRL-this issue would happen with real filings as well.

Video: https://github.com/user-attachments/assets/b91f95db-6954-439e-bfc7-4986c9a6ccef
iXBRL tested: Microsoft-2023-12-31.zip

Reproduction steps:
*upload an attached document
*go to the fact list
*see errors
*go to a different page

• go back to the original page
• see no errors occur

We have tested this with other software and could not reproduce the issues as well as inspecting iXBRL. But I might have missed some expectations of your viewer.

Kind regards,
Gabi

[austinmatherne-wk]

Hi @gjk-corefiling,

As far as I'm aware, the SEC team that builds the EDGAR plugins doesn't monitor issues posted here. They do have a support email address that you can contact though: StructuredData@sec.gov

[gjk-corefiling]

Hi @austinmatherne-wk thank you!

[gjk-corefiling]

I have emailed the below information to the mentioned email. For the sake of completion, I am pasting it here as well.

Turns out I was testing the issue with sometimes too small of documents, where we would not apply the issue causing CSS. Apologies if I have misdirected communication.

1. We produce iXBRL with “content-visibility: auto” It improves performance, especially notable in Annual Reports, and this is being recommended now.[1]
2. EDGAR viewer is using “innerText”[2]
3. Inner text and content visibility set to auto combination means the initial return is empty and “Format Error: Num Dot Decimal” is shown as an error. On the second loading, the inner text is not empty anymore and you get the right value returned.
4. This would not happen if “textContent” would be used instead of “innerText”

The spec does refer to the value of ix:nonFraction being the text content [3]

Some info about the differences between the two can be found here: Node: textContent property - Web APIs | MDN

Please let us know if you would like any assistance :)

[1] https://www.xbrl.org/WGN/ixbrl-rendering-performance-wgn/WGN-2024-12-17/ixbrl-rendering-performance-wgn-2024-12-17.html#sec-content-visibility
[2] https://github.com/Arelle/EDGAR/blob/master/view/ixviewer/js/app/filters/number.js#L156
[3]Inline XBRL Part 1: Specification 1.1