feat(experimental): authorship

demo/demo.js

@@ -34,9 +34,10 @@ window.parseHtml = () => {
   const lang = document.getElementById("lang").checked;
   const textContent = document.getElementById("textContent").checked;
   const metaformats = document.getElementById("metaformats").checked;
+  const authorship = document.getElementById("authorship").checked;
 
   return parse(html, {
     baseUrl,
-    experimental: { lang, textContent, metaformats },
+    experimental: { lang, textContent, metaformats, authorship },
   });
 };

demo/index.tpl.html

@@ -83,6 +83,16 @@ <h3>Experimental options</h3>
             />
             <span>Metaformats parsing</span>
           </label>
+          <label>
+            <input
+              type="checkbox"
+              name="authorship"
+              id="authorship"
+              value="true"
+              checked
+            />
+            <span>Authorship discovery</span>
+          </label>
         </p>
 
         <div class="submit">

src/helpers/authorship.ts

@@ -0,0 +1,108 @@
+import {
+  Author,
+  Image,
+  MicroformatProperty,
+  MicroformatRoot,
+  Rels,
+} from "../types";
+
+function getPlainText(values: MicroformatProperty[]): string | null {
+  if (values.length === 0) {
+    return null;
+  }
+
+  const value = values[0] as Author;
+  let plainText: string | null;
+  if (value.value !== undefined && typeof value.value === "string") {
+    plainText = value.value;
+  } else if (typeof value === "string") {
+    plainText = value;
+  } else {
+    plainText = null;
+  }
+
+  return plainText && plainText.trim();
+}
+
+const parseAuthor = (hCard: MicroformatRoot) => {
+  // TODO: Figure out how to stop TypeScript complaining about missing `value`
+  const result: Author = {};
+
+  if (hCard.properties !== undefined) {
+    // Use first (or only) name
+    const names = hCard.properties.name as string[];
+    if (names?.length > 0) {
+      result.name = names[0];
+    }
+
+    // Use first (or only) photo
+    const photos = hCard.properties.photo as Image[];
+    if (photos?.length > 0) {
+      const photo = getPlainText(photos);
+      if (photo) {
+        result.photo = photo;
+      }
+    }
+
+    // Use first (or only) URL
+    const urls = hCard.properties.url as string[];
+    if (urls?.length > 0) {
+      result.url = urls[0];
+    }
+  } else if (hCard) {
+    if (URL.canParse(String(hCard))) {
+      result.url = String(hCard);
+    } else {
+      result.name = String(hCard);
+    }
+  }
+
+  return result as Author;
+};
+
+const findEntryAuthor = (hEntry: MicroformatRoot) => {
+  const values = hEntry.properties.author || [];
+
+  if (Object.keys(values).length === 0) {
+    return;
+  }
+
+  return parseAuthor(values[0] as MicroformatRoot);
+};
+
+const findFeedAuthor = () => false;
+
+export const findAuthor = async (item: MicroformatRoot, rels: Rels) => {
+  // 1. If no `h-entry` then there’s no post to find authorship for.
+  const itemIsEntry = item.type && item.type[0] === "h-entry";
+  if (!itemIsEntry) {
+    return false;
+  }
+
+  // 2. Parse the `h-entry`
+  const entryAuthor = findEntryAuthor(item);
+  const feedAuthor = findFeedAuthor(); // TODO
+
+  // 3 & 4. Return author in `h-entry`, else find author in parent `h-feed`
+  const author = entryAuthor ? entryAuthor : feedAuthor;
+
+  // 5. Return `author` if `h-card`
+  const authorIsCard = author && author.type[0] === "h-card";
+  if (authorIsCard) {
+    return author;
+  }
+
+  // 6. Use `h-card` fetched from rel=author
+  const authorPage = author.properties?.url || rels.author;
+  if (authorPage) {
+    // Fetch `authorPage` and parse result using `parseMicroformat`
+    // This is an async function, which would bubble up to the parent function
+  }
+
+  // 7. From the parsed `authorPage`, return the first `h-card` that either:
+  //    * Has a value for `u-url` (or `u-uid`) that matches the `authorPage` URL
+  //    * Has a value for `u-url` that matches a `rel=me` on the `authorPage`
+
+  // 8. Else, no deterministic author can be found
+  return author;
+};

src/index.ts

@@ -8,6 +8,7 @@ export interface Options {
     lang?: boolean;
     textContent?: boolean;
     metaformats?: boolean;
+    authorship?: boolean;
   };
 }
 

src/microformats/parse.ts

@@ -5,6 +5,7 @@ import {
   Element,
 } from "../types";
 import { microformatProperties } from "./properties";
+import { findAuthor } from "../helpers/authorship";
 import { textContent } from "../helpers/textContent";
 import { getAttributeValue, getClassNames } from "../helpers/attributes";
 import { findChildren } from "../helpers/findChildren";
@@ -66,6 +67,15 @@ export const parseMicroformat = (
     item.lang = lang;
   }
 
+  if (isEnabled(options, "authorship")) {
+    const author = findAuthor(item, options.rels);
+
+    if (author) {
+      console.log("author", author);
+      // item.properties.author = author;
+    }
+  }
+
   if (children.length) {
     item.children = children.map((child) =>
       parseMicroformat(child, { ...options, inherited }),
@@ -86,7 +96,7 @@ export const parseMicroformat = (
   }
 
   /**
-   * There is some ambigutity on how this should be handled.
+   * There is some ambiguity on how this should be handled.
    * At the moment, we're following other parsers and keeping `value` a string
    * and adding `html` as an undocumented property.
    */

src/parser.ts

@@ -23,6 +23,7 @@ export const parser = (
     baseUrl,
     idRefs,
     inherited: { roots: [], lang },
+    rels,
   };
   let items = findChildren(doc, isMicroformatRoot).map((mf) =>
     parseMicroformat(mf, parsingOptions),

src/types.ts

@@ -14,6 +14,7 @@ export interface ParserOptions {
     lang?: boolean;
     textContent?: boolean;
     metaformats?: boolean;
+    authorship?: boolean;
   };
 }
 
@@ -26,6 +27,7 @@ export interface ParsingOptions extends ParserOptions {
     roots: BackcompatRoot[];
     lang?: string;
   };
+  rels: Rels;
 }
 
 export interface ParsedDocument {
@@ -45,6 +47,13 @@ export interface MicroformatRoot {
   value?: MicroformatProperty;
 }
 
+export interface Author {
+  name?: string;
+  value: string;
+  photo?: string;
+  url?: string;
+}
+
 export interface Image {
   alt: string;
   value?: string;
@@ -56,7 +65,12 @@ export interface Html {
   lang?: string;
 }
 
-export type MicroformatProperty = MicroformatRoot | Image | Html | string;
+export type MicroformatProperty =
+  | MicroformatRoot
+  | Author
+  | Image
+  | Html
+  | string;
 
 export type Rels = Record<string, string[]>;
 

src/validator.ts

@@ -81,7 +81,7 @@ export const validator = (
   if ("experimental" in options) {
     const experimental = assertIsObject(
       options.experimental,
-      ["lang", "textContent", "metaformats"],
+      ["lang", "textContent", "metaformats", "authorship"],
       "experimental",
     );
 
@@ -96,6 +96,10 @@ export const validator = (
     if ("metaformats" in experimental) {
       assertIsBoolean(experimental.metaformats, "experimental.metaformats");
     }
+
+    if ("authorship" in experimental) {
+      assertIsBoolean(experimental.authorship, "experimental.authorship");
+    }
   }
 };
 

test/scenarios.spec.ts

@@ -67,7 +67,12 @@ describe("mf2() // experimental scenarios", () => {
     it(`should correctly parse ${name}`, () => {
       const result = mf2(input, {
         ...options,
-        experimental: { lang: true, textContent: true, metaformats: true },
+        experimental: {
+          lang: true,
+          textContent: true,
+          metaformats: true,
+          authorship: true,
+        },
       });
       expect(result).to.deep.equal(expected);
     });

test/suites/experimental/authorship-h-card-with-rel-author-to-rel-me.html

@@ -0,0 +1,16 @@
+<!-- This next entry is by Virginia Woolf. Her URL is https://virginia.example, and her photo is at https://virginia.example/photo.jpg. -->
+
+<html lang="en">
+  <head>
+    <meta charset="utf-8" />
+    <title>Entry with rel=author (links to h-card with rel=me)</title>
+  </head>
+  <body class="h-entry">
+    <div class="e-content">
+      <p>A woman must have money and a room of her own if she is to write fiction.</p>
+    </div>
+    <footer>
+      <a href="https://virginia.woolf" rel="author">About Virginia Woolf</a>
+    </footer>
+  </body>
+</html>

test/suites/experimental/authorship-h-card-with-rel-author-to-rel-me.json

@@ -0,0 +1,38 @@
+{
+  "items": [
+    {
+      "type": ["h-entry"],
+      "lang": "en",
+      "properties": {
+        "author": [
+          {
+            "type": ["h-card"],
+            "lang": "en",
+            "properties": {
+              "name": ["Virginia Woolf"],
+              "photo": ["https://virginia.example/photo.jpg"],
+              "url": ["https://virginia.example"]
+            },
+            "value": "Virginia Woolf"
+          }
+        ],
+        "content": [
+          {
+            "value": "A woman must have money and a room of her own if she is to write fiction.",
+            "lang": "en",
+            "html": "<p>A woman must have money and a room of her own if she is to write fiction.</p>"
+          }
+        ]
+      }
+    }
+  ],
+  "rel-urls": {
+    "https://virginia.woolf": {
+      "rels": ["author"],
+      "text": "About Virginia Woolf"
+    }
+  },
+  "rels": {
+    "author": ["https://virginia.woolf"]
+  }
+}

test/suites/experimental/authorship-h-card-with-rel-author-to-u-url.html

@@ -0,0 +1,16 @@
+<!-- This is a haiku by Basho. His URL is https://basho.example, and his photo is at https://basho.example/photo.jpg. -->
+
+<html lang="en">
+  <head>
+    <meta charset="utf-8" />
+    <title>Entry with rel=author (links to h-card with u-url and u-uid)</title>
+  </head>
+  <body class="h-entry">
+    <div class="e-content" lang="jp">
+      <p>古池や<br />蛙飛び込む<br />水の音</p>
+    </div>
+    <footer>
+      <a href="https://basho.example" rel="author">About Basho</a>
+    </footer>
+  </body>
+</html>

test/suites/experimental/authorship-h-card-with-rel-author-to-u-url.json

@@ -0,0 +1,38 @@
+{
+  "items": [
+    {
+      "type": ["h-entry"],
+      "lang": "en",
+      "properties": {
+        "author": [
+          {
+            "type": ["h-card"],
+            "lang": "en",
+            "properties": {
+              "name": ["Basho"],
+              "photo": ["https://basho.example/photo.jpg"],
+              "url": ["https://basho.example"]
+            },
+            "value": "Basho"
+          }
+        ],
+        "content": [
+          {
+            "value": "古池や\n蛙飛び込む\n水の音",
+            "lang": "jp",
+            "html": "<p>古池や<br>蛙飛び込む<br>水の音</p>"
+          }
+        ]
+      }
+    }
+  ],
+  "rel-urls": {
+    "https://basho.example": {
+      "rels": ["author"],
+      "text": "About Basho"
+    }
+  },
+  "rels": {
+    "author": ["https://basho.example"]
+  }
+}

test/suites/experimental/authorship-h-card-with-rel-author.html

@@ -0,0 +1,21 @@
+<!-- This is an entry by Patañjali. His URL is https://patanjali.example, and he uses no photo. -->
+
+<html lang="en">
+  <head>
+    <meta charset="utf-8" />
+    <title>Entry with separate h-card and rel=author</title>
+  </head>
+  <body class="h-entry">
+    <div class="e-content">
+      <p>For one who sees the distinction, there is no further confusing of the mind with the self.</p>
+    </div>
+    <p class="p-author h-card">
+      <a href="https://patanjali.example" class="u-url">
+        <span class="p-name">Patañjali</span>
+      </a>
+    </p>
+    <footer>
+      <a href="https://patanjali.example" rel="author">About Patañjali</a>
+    </footer>
+  </body>
+</html>