reffy 16.1.0 → 16.1.2

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "reffy",
-  "version": "16.1.0",
+  "version": "16.1.2",
   "description": "W3C/WHATWG spec dependencies exploration companion. Features a short set of tools to study spec references as well as WebIDL term definitions and references found in W3C specifications.",
   "repository": {
     "type": "git",
@@ -32,20 +32,20 @@
   "main": "index.js",
   "bin": "./reffy.js",
   "dependencies": {
-    "ajv": "8.16.0",
+    "ajv": "8.17.1",
     "ajv-formats": "3.0.1",
     "commander": "12.1.0",
     "fetch-filecache-for-crawling": "5.1.1",
-    "puppeteer": "22.13.0",
+    "puppeteer": "22.13.1",
     "semver": "^7.3.5",
     "web-specs": "3.13.1",
     "webidl2": "24.4.1"
   },
   "devDependencies": {
-    "mocha": "10.6.0",
+    "mocha": "10.7.0",
     "respec": "35.1.1",
     "respec-hljs": "2.1.1",
-    "rollup": "4.18.1",
+    "rollup": "4.19.0",
     "undici": "^6.1.0"
   },
   "overrides": {

package/schemas/files/extracts/algorithms.json

@@ -4,9 +4,9 @@
 
   "type": "object",
   "additionalProperties": false,
-  "required": ["spec", "dfns"],
+  "required": ["spec", "algorithms"],
   "properties": {
     "spec": { "$ref": "../../common.json#/$defs/specInExtract" },
-    "dfns": { "$ref": "../../browserlib/extract-algorithms.json" }
+    "algorithms": { "$ref": "../../browserlib/extract-algorithms.json" }
   }
 }

package/src/browserlib/extract-algorithms.mjs

@@ -371,8 +371,22 @@ function getDefinedNameIn(el) {
     else {
       name += getTextContent(dfn);
     }
-    const href = dfn.id ? getAbsoluteUrl(dfn) : null;
-    return { name, href };
+    if (dfn.id) {
+      return { name, href: getAbsoluteUrl(dfn) };
+    }
+    else {
+      // Two known exceptions to the rule:
+      // - one due to CSS 2.1 not following the definitions data model:
+      // https://www.w3.org/TR/CSS21/visudet.html#containing-block-details
+      // - the other due to HTML still containing dfns without IDs as well,
+      // including one for an algorithm:
+      // https://html.spec.whatwg.org/multipage/server-sent-events.html#processField
+      // It's possible to find an ID in both cases. But it's not clear that
+      // CSS 2.1 algorithms are real algorithms; and it seems doable to fix the
+      // HTML spec. Let's just return the name without href, not to end up
+      // with a null `href` that the JSON schema forbids.
+      return { name };
+    }
   }
   else {
     const heading = el.querySelector('h2[id],h3[id],h4[id],h5[id],h6[id]');
@@ -390,14 +404,17 @@ function getDefinedNameIn(el) {
  */
 function findIntroParagraph(algo) {
   let paragraph;
-  let container = algo.root.closest('.algorithm');
+  let container = algo.root.closest('li,.algorithm');
   while (container) {
     const dfn = container.querySelector('dfn');
-    if (dfn) {
-      paragraph = dfn.closest('p,div');
+    if (dfn && !algo.root.contains(dfn)) {
+      paragraph = dfn.closest('p,div,li');
       break;
     }
-    container = container.parentElement.closest('.algorithm');
+    if (container.nodeName === 'LI') {
+      break;
+    }
+    container = container.parentElement.closest('li,.algorithm');
   }
 
   if (!paragraph) {
@@ -427,9 +444,10 @@ function getAlgorithmInfo(algo, context) {
   // Note some specs add the "algorithm" class to the `<ol>` and to the
   // wrapping container, and define the name in the wrapping container.
   let info = {};
+
   let container = algo.root.closest('.algorithm');
-  while (container) {
-    if (container && !context?.nested) {
+  if (!context?.nested) {
+    while (container) {
       if (container.getAttribute('data-algorithm')) {
         info.name = normalize(container.getAttribute('data-algorithm'));
         if (container.getAttribute('data-algorithm-for')) {
@@ -448,13 +466,15 @@ function getAlgorithmInfo(algo, context) {
             info.href = dfn.href;
           }
         }
-        break;
       }
       else {
         info = getDefinedNameIn(container);
+        if (info.name || info.href) {
+          break;
+        }
       }
+      container = container.parentElement.closest('.algorithm');
     }
-    container = container.parentElement.closest('.algorithm');
   }
 
   // Get the introductory prose from the previous paragraph