Build your personal knowledge base with Trilium Notes
TypeScript electronelectron-appknowledge-baseknowledge-graphknowledge-managementknowledge-management-graphlocal-firstnote-managmentnote-takernote-takingnotebooknotesnotes-apppersonal-knowledge-basepersonal-wikiscriptableself-hostedself-hostingwiki
fix(clipper): save a selection's lazy images from their real address
Elian Doranc86a3eb028a0
diff --git a/.claude/skills/developing-web-clipper/SKILL.md b/.claude/skills/developing-web-clipper/SKILL.mdindex 668aa172b9..66fb341010 100644--- a/.claude/skills/developing-web-clipper/SKILL.md+++ b/.claude/skills/developing-web-clipper/SKILL.md@@ -93,6 +93,10 @@ against older versions of the other.** Prefer changes that need nothing new from20-character URLs on purpose, so **an image left with its placeholder is a broken image**.`postProcessImages()` therefore restores the original URL (escaped with `escapeHtml()`) for everyimage it cannot download and reports the count in the toast.+- A selection copies the page's DOM as it is, so an image not yet scrolled into view still has a+ placeholder (often a `data:image/svg+xml` URL) in `src`. `getImages()` lists it under the web+ address in `data-src`, `data-original` or `data-lazy-src` (the attributes Defuddle reads), and+ ignores one that is not `http(s)` or does not parse.- **A `replaceAll()`/`replace()` replacement built from page data is a callback**(`replaceAll(id, () => escapeHtml(src))`). A string replacement reads `$&`, `$$`, `` $` `` and`$'` as patterns, so an image URL containing `$&` came back with the placeholder inside it.diff --git a/apps/web-clipper/entrypoints/content/index.spec.ts b/apps/web-clipper/entrypoints/content/index.spec.tsindex 2c686ae5b2..c005006d5d 100644--- a/apps/web-clipper/entrypoints/content/index.spec.ts+++ b/apps/web-clipper/entrypoints/content/index.spec.ts@@ -67,6 +67,36 @@ describe("content script", () => {]);});+ it("takes the address a lazy image has not loaded yet from where the page keeps it", async () => {+ const placeholder = "data:image/svg+xml,%3Csvg xmlns='http://www.w3.org/2000/svg'%3E%3C/svg%3E";+ document.body.innerHTML = `+ <p id="start">Photos</p>+ <img src="${placeholder}" data-src="/photos/one.jpg">+ <img data-original="https://cdn.example.com/two.jpg">+ <img src="${placeholder}" data-lazy-src="three.jpg">+ <img src="/photos/one.jpg" data-src="/photos/one.jpg">+ <img src="/loaded.png" data-src="javascript:alert(1)">+ <img src="/kept.png" data-src="">+ <img src="/malformed.png" data-src="http://[">+ <p id="end">End</p>`;+ selectRange("start", "end");++ const response = await send<Clipping>({ name: "trilium-save-selection" });++ expect(response.images.map((image) => image.src)).toEqual([+ "https://example.com/photos/one.jpg",+ "https://cdn.example.com/two.jpg",+ "https://example.com/blog/three.jpg",+ "https://example.com/loaded.png",+ "https://example.com/kept.png",+ "https://example.com/malformed.png"+ ]);+ const container = document.createElement("div");+ container.innerHTML = response.content;+ expect(attributes(container, "img", "src")).toEqual(+ [ 0, 1, 2, 0, 3, 4, 5 ].map((index) => response.images[index]?.imageId));+ });+it("keeps links to schemes other than the web as they are", async () => {const links = ["mailto:[email protected]",diff --git a/apps/web-clipper/utils.ts b/apps/web-clipper/utils.tsindex 3d60d00510..7ffcc8daa6 100644--- a/apps/web-clipper/utils.ts+++ b/apps/web-clipper/utils.ts@@ -86,16 +86,20 @@ export function makeLinksAbsolute(container: HTMLElement) {}}-/** Replaces the `src` of each image in `container` with an `imageId`, the same one for the same image. */+/**+ * Replaces the `src` of each image in `container` with an `imageId`, the same one for the same image.+ * An image waiting to be lazy-loaded is listed under the address the page keeps for later.+ */export function getImages(container: HTMLElement) {const images: ClippedImage[] = [];for (const img of container.getElementsByTagName('img')) {- if (!img.src) {+ const src = lazyImageSource(img) ?? img.src;+ if (!src) {continue;}- const existingImage = images.find(image => image.src === img.src);+ const existingImage = images.find(image => image.src === src);if (existingImage) {img.src = existingImage.imageId;@@ -105,7 +109,7 @@ export function getImages(container: HTMLElement) {images.push({imageId,- src: img.src+ src});img.src = imageId;@@ -114,3 +118,25 @@ export function getImages(container: HTMLElement) {return images;}++/** The attributes lazy-loading scripts keep an image's address in, the same ones Defuddle reads. */+const LAZY_SOURCE_ATTRIBUTES = [ "data-src", "data-original", "data-lazy-src" ];++/** The web address in the first lazy-loading attribute of `img` that holds one. */+function lazyImageSource(img: HTMLImageElement) {+ for (const attribute of LAZY_SOURCE_ATTRIBUTES) {+ const value = img.getAttribute(attribute)?.trim();+ if (!value) continue;++ let url: URL;+ try {+ url = new URL(value, img.baseURI);+ } catch {+ continue;+ }+ if (url.protocol === "http:" || url.protocol === "https:") {+ return url.href;+ }+ }+ return undefined;+}