strange table behaviour

[more-urgent-jest]

Hi, I am trying to restrict the allowed tags by clearing the set and adding a restricted list like so:

        var sanitizer = new HtmlSanitizer();

        sanitizer.AllowedTags.Clear();
        sanitizer.AllowedTags.UnionWith(["ul", "li", "table", "tr", "td", "strong"]);
        sanitizer.AllowedAttributes.Clear();
        sanitizer.AllowedAttributes.Add("class");
        sanitizer.AllowedAtRules.Clear();
        sanitizer.AllowedClasses.Clear();
        sanitizer.AllowedClasses.Add("right");
        sanitizer.AllowedCssProperties.Clear();
        sanitizer.AllowedSchemes.Clear();

        services.AddSingleton<IHtmlSanitizer>(sanitizer);

but finding that it strips tr tags and their children and content, but just stripping td tags and leaving their content.

this:
"<ul><li>the item in this list containing this <tr> </tr> <td>should stay</td> <table class="right"> </table> as it is</li></ul>"
becomes:
"<ul><li>the item in this list containing this should stay <table class="right"> </table> as it is</li></ul>"

this:
"<ul><li>the item in this list containing this <tr> <td>should stay</td> </tr> <table class="right"> </table> as it is</li></ul>"
becomes:
"<ul><li>the item in this list containing this should stay <table class="right"> </table> as it is</li></ul>"

this:
"<ul><li>the item in this list containing this <table class="right"> <tr> <td>should stay</td> </tr> </table> as it is</li></ul>"
becomes:
"<ul><li>the item in this list containing this <table class="right"> </table> as it is</li></ul>"

this:
"<ul><li>the item in this list containing this <table><tr class="right"><td>should stay</td></tr></table> as it is</li></ul>"
becomes:
"<ul><li>the item in this list containing this <table></table> as it is</li></ul>"

I would expect all of these to not be changed.

am I not understanding how this works?

[more-urgent-jest]

seems to be happening, because the parser inserts <tbody>, but with it in there I am getting this even stranger behaviour:

this:
"<ul><li>the item in this list containing this <table><tbody><tr class="right"><td>should stay</td></tr><tbody></table> as it is</li></ul>"
becomes:
"<ul><li>the item in this list containing this <table><tbody><tr class="right"><td>should stay</td></tr></tbody><tbody></tbody></table> as it is</li></ul>"

which seems especially odd as tr is not required to be in a tbody, but can be contained directly within a table, as specified in the table content model:

In this order: optionally a caption element, followed by zero or more colgroup elements, followed optionally by a thead element, followed by either zero or more tbody elements or one or more tr elements, followed optionally by a tfoot element, optionally intermixed with one or more script-supporting elements.

the table element

and the tr contexts:

As a child of a thead element.
As a child of a tbody element.
As a child of a tfoot element.
As a child of a table element, after any caption, colgroup, and thead elements, but only if there are no tbody elements that are children of the table element.

the tr element

[more-urgent-jest]

actually, I had a second opening tbody instead of a closing one, so that behaviour isn't that strange after all, but it is odd that tbody is required when it shouldn't be.

I would have hoped that since AngleSharp, if indeed that is causing this, claims to be standards compliant that it wouldn't exhibit such eccentricities.

[tiesont]

I assume AngleSharp is following browser behavior, which is to insert a <tbody> (around all of the content) into the table if one is not present, which then gets stripped out if tbody isn't in the allowed tags list.

Beyond that, an MVCE would be helpful to verify what you're seeing.

Also, if you're trying to use code fences (```) in your code above, then the opening and closing fences need to be on their own lines (Github is currently parsing your text with inline code blocks, which are just a single backtick on either side of the code). Just FYI.