[dyad] Added FAQ structured data extraction - wrote 4 file(s)
This commit is contained in:
@@ -2,6 +2,11 @@
|
||||
|
||||
import * as cheerio from "cheerio";
|
||||
|
||||
interface FaqItem {
|
||||
question: string;
|
||||
answer: string;
|
||||
}
|
||||
|
||||
export async function extractMetaData(url: string) {
|
||||
if (!url) {
|
||||
return { error: "URL is required." };
|
||||
@@ -37,7 +42,45 @@ export async function extractMetaData(url: string) {
|
||||
"No description found";
|
||||
const image = $('meta[property="og:image"]').attr("content") || null;
|
||||
|
||||
return { data: { title, description, image } };
|
||||
const faqData: FaqItem[] = [];
|
||||
$('script[type="application/ld+json"]').each((i, el) => {
|
||||
const jsonContent = $(el).html();
|
||||
if (!jsonContent) return;
|
||||
|
||||
try {
|
||||
const data = JSON.parse(jsonContent);
|
||||
const graph = data["@graph"] || [data];
|
||||
|
||||
for (const item of graph) {
|
||||
if (item["@type"] === "FAQPage" && Array.isArray(item.mainEntity)) {
|
||||
item.mainEntity.forEach((qa: any) => {
|
||||
if (
|
||||
qa["@type"] === "Question" &&
|
||||
qa.name &&
|
||||
qa.acceptedAnswer &&
|
||||
qa.acceptedAnswer.text
|
||||
) {
|
||||
faqData.push({
|
||||
question: qa.name,
|
||||
answer: qa.acceptedAnswer.text,
|
||||
});
|
||||
}
|
||||
});
|
||||
}
|
||||
}
|
||||
} catch (e) {
|
||||
// Ignore parsing errors
|
||||
}
|
||||
});
|
||||
|
||||
return {
|
||||
data: {
|
||||
title,
|
||||
description,
|
||||
image,
|
||||
faq: faqData.length > 0 ? faqData : null,
|
||||
},
|
||||
};
|
||||
} catch (error) {
|
||||
console.error(error);
|
||||
if (error instanceof Error && error.message.includes("Invalid URL")) {
|
||||
|
||||
Reference in New Issue
Block a user