Fix autolinking with highlighted search results

The current logic assumes that all spans in the text layer contain
only one text node, and thus that the position information
returned by `highlighter._convertMatches` can be directly used
on the element's only child.

This is not true in case of highlighted search results: they will be
injected in the DOM as `<span>` elements, causing the `<span>`s
in the text layer to have more than one child.

This patch fixes the problem by properly converting the (span, offset)
pair in a (textNode, offset) pair that points to the right text node.

test/integration/autolinker_spec.mjs

@@ -13,10 +13,46 @@
  * limitations under the License.
  */
 
-import { closePages, createPromise, loadAndWait } from "./test_utils.mjs";
+import {
+  awaitPromise,
+  closePages,
+  createPromise,
+  loadAndWait,
+} from "./test_utils.mjs";
 
-function waitForLinkAnnotations(page) {
+function waitForLinkAnnotations(page, pageNumber) {
+  return page.evaluateHandle(
+    number => [
+      new Promise(resolve => {
+        const { eventBus } = window.PDFViewerApplication;
+        eventBus.on("linkannotationsadded", function listener(e) {
+          if (number === undefined || e.pageNumber === number) {
+            resolve();
+            eventBus.off("linkannotationsadded", listener);
+          }
+        });
+      }),
+    ],
+    pageNumber
+  );
+}
+
+function recordInitialLinkAnnotationsEvent(eventBus) {
+  globalThis.initialLinkAnnotationsEventFired = false;
+  eventBus.on(
+    "linkannotationsadded",
+    () => {
+      globalThis.initialLinkAnnotationsEventFired = true;
+    },
+    { once: true }
+  );
+}
+function waitForInitialLinkAnnotations(page) {
   return createPromise(page, resolve => {
+    if (globalThis.initialLinkAnnotationsEventFired) {
+      resolve();
+      return;
+    }
     window.PDFViewerApplication.eventBus.on("linkannotationsadded", resolve, {
       once: true,
     });
@@ -178,4 +214,49 @@ describe("autolinker", function () {
       );
     });
   });
+
+  describe("when highlighting search results", function () {
+    let pages;
+
+    beforeAll(async () => {
+      pages = await loadAndWait(
+        "issue3115r.pdf",
+        ".annotationLayer",
+        null,
+        { eventBusSetup: recordInitialLinkAnnotationsEvent },
+        { enableAutoLinking: true }
+      );
+    });
+
+    afterAll(async () => {
+      await closePages(pages);
+    });
+
+    it("must find links that overlap with search results", async () => {
+      await Promise.all(
+        pages.map(async ([browserName, page]) => {
+          await awaitPromise(await waitForInitialLinkAnnotations(page));
+
+          const linkAnnotationsPromise = await waitForLinkAnnotations(page, 36);
+
+          // Search for "rich.edu"
+          await page.click("#viewFindButton");
+          await page.waitForSelector("#viewFindButton", { hidden: false });
+          await page.type("#findInput", "rich.edu");
+          await page.waitForSelector(".textLayer .highlight");
+
+          await awaitPromise(linkAnnotationsPromise);
+
+          const urls = await page.$$eval(
+            ".page[data-page-number='36'] > .annotationLayer > .linkAnnotation > a",
+            annotations => annotations.map(a => a.href)
+          );
+
+          expect(urls)
+            .withContext(`In ${browserName}`)
+            .toContain(jasmine.stringContaining("rich.edu"));
+        })
+      );
+    });
+  });
 });

web/autolinker.js

@@ -69,13 +69,55 @@ function calculateLinkPosition(range, pdfPageView) {
   return { quadPoints, rect };
 }
 
+/**
+ * Given a DOM node `container` and an index into its text contents `offset`,
+ * returns a pair consisting of text node that the `offset` actually points
+ * to, together with the offset relative to that text node.
+ * When the offset points at the boundary between two node, the result will
+ * point to the first text node in depth-first traversal order.
+ *
+ * For example, given this DOM:
+ * <p>abc<span>def</span>ghi</p>
+ *
+ * textPosition(p, 0) -> [#text "abc", 0] (before `a`)
+ * textPosition(p, 2) -> [#text "abc", 2] (between `b` and `c`)
+ * textPosition(p, 3) -> [#text "abc", 3] (after `c`)
+ * textPosition(p, 5) -> [#text "def", 2] (between `e` and `f`)
+ * textPosition(p, 6) -> [#text "def", 3] (after `f`)
+ */
+function textPosition(container, offset) {
+  let currentContainer = container;
+  do {
+    if (currentContainer.nodeType === Node.TEXT_NODE) {
+      const currentLength = currentContainer.textContent.length;
+      if (offset <= currentLength) {
+        return [currentContainer, offset];
+      }
+      offset -= currentLength;
+    } else if (currentContainer.firstChild) {
+      currentContainer = currentContainer.firstChild;
+      continue;
+    }
+
+    while (!currentContainer.nextSibling && currentContainer !== container) {
+      currentContainer = currentContainer.parentNode;
+    }
+    if (currentContainer !== container) {
+      currentContainer = currentContainer.nextSibling;
+    }
+  } while (currentContainer !== container);
+  throw new Error("Offset is bigger than container's contents length.");
+}
+
 function createLinkAnnotation({ url, index, length }, pdfPageView, id) {
   const highlighter = pdfPageView._textHighlighter;
   const [{ begin, end }] = highlighter._convertMatches([index], [length]);
 
   const range = new Range();
-  range.setStart(highlighter.textDivs[begin.divIdx].firstChild, begin.offset);
-  range.setEnd(highlighter.textDivs[end.divIdx].firstChild, end.offset);
+  range.setStart(
+    ...textPosition(highlighter.textDivs[begin.divIdx], begin.offset)
+  );
+  range.setEnd(...textPosition(highlighter.textDivs[end.divIdx], end.offset));
 
   return {
     id: `inferred_link_${id}`,