Merge pull request #143 from hoangr1010/Autofill-profile

pdf parsing file from s3
hoangr1010 · May 28, 2024 · ada4f01 · ada4f01
2 parents 98fc7ee + 8e85ba7
commit ada4f01
Show file tree

Hide file tree

Showing 3 changed files with 48 additions and 0 deletions.
diff --git a/server/package-lock.json b/server/package-lock.json
diff --git a/server/package.json b/server/package.json
@@ -29,6 +29,7 @@
     "mongoose": "^8.2.3",
     "multer": "^1.4.5-lts.1",
     "nodemon": "^3.1.0",
+    "pdf-parse": "^1.1.1",
     "serverless": "^3.38.0",
     "serverless-http": "^3.2.0",
     "supertest": "^6.3.4",

diff --git a/server/src/utils/extractPDF.js b/server/src/utils/extractPDF.js
@@ -0,0 +1,21 @@
+import pdf from "pdf-parse";
+import axios from "axios";
+export const getResumeFromS3 = async (resume_url) => {
+  const pdfUrl = resume_url;
+
+  const response = await axios.get(pdfUrl, { responseType: "arraybuffer" });
+  return response.data;
+};
+
+const parsePdf = async (filePath) => {
+  try {
+    const dataBuffer = await getResumeFromS3(filePath);
+    const data = await pdf(dataBuffer);
+
+    return data.text;
+  } catch (error) {
+    console.error("Error parsing PDF:", error);
+  }
+};
+
+export default parsePdf;