Unverified Commit db326fd8 authored by Hong Minhee's avatar Hong Minhee
Browse files

Fix XHTML self-closing link tag parsing in ActivityPub Discovery 

The HTML/XHTML parser now correctly recognizes self-closing <link> tags
with whitespace before the closing slash (/>) in XHTML documents. This
improves compatibility with XHTML documents that follow proper self-closing
tag format for ActivityPub Discovery.
parent 70309729

CHANGES.md

+5 −0
Original line number Diff line number Diff line
@@ -8,6 +8,11 @@ Version 1.0.27 


To be released.



 -  Fixed a bug where ActivityPub Discovery failed to recognize XHTML

    self-closing `<link>` tags. The HTML/XHTML parser now correctly handles

    whitespace before the self-closing slash (`/>`), improving compatibility

    with XHTML documents that follow the self-closing tag format.





Version 1.0.26

--------------

src/runtime/docloader.test.ts

+30 −0
Original line number Diff line number Diff line
@@ -209,6 +209,36 @@ test("fetchDocumentLoader()", async (t) => { 
    });

  });



  mf.mock("GET@/xhtml-link", (_req) =>

    new Response(

      `<html>

          <head>

            <meta charset=utf-8>

            <link

              rel=alternate

              type="application/activity+json"

              href="https://example.com/object" />

          </head>

        </html>`,

      {

        status: 200,

        headers: { "Content-Type": "application/xhtml+xml; charset=utf-8" },

      },

    ));



  await t.step("XHTML <link>", async () => {

    assertEquals(await fetchDocumentLoader("https://example.com/xhtml-link"), {

      contextUrl: null,

      documentUrl: "https://example.com/object",

      document: {

        "@context": "https://www.w3.org/ns/activitystreams",

        id: "https://example.com/object",

        name: "Fetched object",

        type: "Object",

      },

    });

  });



  mf.mock("GET@/html-a", (_req) =>

    new Response(

      `<html>

src/runtime/docloader.ts

+2 −1
Original line number Diff line number Diff line
@@ -166,7 +166,8 @@ async function getRemoteDocument( 
      contentType === "application/xhtml+xml" ||

      contentType?.startsWith("application/xhtml+xml;"))

  ) {

    const p = /<(a|link)((\s+[a-z][a-z:_-]*=("[^"]*"|'[^']*'|[^\s>]+))+)\/?>/ig;

    const p =

      /<(a|link)((\s+[a-z][a-z:_-]*=("[^"]*"|'[^']*'|[^\s>]+))+)\s*\/?>/ig;

    const p2 = /\s+([a-z][a-z:_-]*)=("([^"]*)"|'([^']*)'|([^\s>]+))/ig;

    const html = await response.text();

    let m: RegExpExecArray | null;