Feat/add grobid

backend/app.js

@@ -8,6 +8,7 @@ const indexRouter = require('./routes/index');
 const usersRouter = require('./routes/users');
 const testAPIRouter = require("./routes/testAPI");
 const databaseRouter = require("./routes/database");
+const extractRouter = require("./routes/extract");
 const bodyParser = require('body-parser');
 const config = require("./config.json");
 
@@ -40,6 +41,8 @@ app.use('/', indexRouter);
 app.use('/users', usersRouter);
 app.use("/testAPI", testAPIRouter);
 app.use("/database", databaseRouter);
+app.use("/extract", extractRouter);
+
 
 // catch 404 and forward to error handler
 app.use(function (req, res, next) {

backend/package.json

@@ -8,18 +8,21 @@
     "dev-docker": "nodemon --legacy-watch ./bin/www"
   },
   "dependencies": {
+    "axios": "^1.6.8",
     "body-parser": "^1.20.2",
     "cookie-parser": "~1.4.4",
     "cors": "^2.8.5",
     "debug": "~2.6.9",
     "express": "~4.18.2",
+    "form-data": "^4.0.0",
     "http-errors": "~1.6.3",
     "jade": "^1.9.2",
     "morgan": "~1.10.0",
     "multer": "^1.4.5-lts.1",
     "pg": "^8.8.0",
     "wink-eng-lite-web-model": "^1.5.2",
-    "wink-nlp": "^1.14.3"
+    "wink-nlp": "^1.14.3",
+    "xml2js": "^0.6.2"
   },
   "devDependencies": {
     "nodemon": "^3.0.2"

backend/routes/extract.js

@@ -0,0 +1,47 @@
+const express = require('express');
+const axios = require('axios');
+const FormData = require('form-data');
+const multer = require('multer'); // Import multer
+const xml2js = require('xml2js');
+const router = express.Router();
+
+const upload = multer({ storage: multer.memoryStorage() });
+
+router.post("/", upload.single('file'), async (req, res) => {
+    // Ensure a file is actually provided
+    if (!req.file) {
+        return res.status(400).send('No file uploaded.');
+    }
+
+    const formData = new FormData();
+    formData.append('input', req.file.buffer, {
+        filename: req.file.originalname,
+        contentType: req.file.mimetype,
+        knownLength: req.file.size,
+    });
+
+    try {
+        const grobidResponse = await axios.post('http://grobid:8070/api/processFulltextDocument', formData, {
+            headers: {
+                ...formData.getHeaders(),
+            },
+            responseType: 'text', // Changed to 'text' to handle XML response correctly
+        });
+
+        // Convert XML response to JSON
+        xml2js.parseString(grobidResponse.data, (err, result) => {
+            if (err) {
+                console.error('Error parsing XML: ', err);
+                return res.status(500).send('Error parsing XML response');
+            }
+
+            res.json(result);
+        });
+
+    } catch (error) {
+        console.error('Error when calling GROBID: ', error);
+        res.status(500).send('Error when processing the PDF file');
+    }
+});
+
+module.exports = router;

docker-compose.dev.yml

@@ -42,7 +42,7 @@ services:
       - backend
 
   grobid:
-    image: lfoppiano/grobid:0.7.0
+    image: lfoppiano/grobid:0.8.0
     ports:
       - "8070:8070"
       - "8071:8071"

docker-compose.yml

@@ -36,10 +36,7 @@ services:
       - backend
 
   grobid:
-    image: lfoppiano/grobid:0.7.0
-    ports:
-      - "8070:8070"
-      - "8071:8071"
+    image: lfoppiano/grobid:0.8.0
 
 volumes:
   postgres_data: