feat: rts file validation

src/fileValidation/ai.ts

@@ -0,0 +1,142 @@
+import OpenAI from "openai";
+import fsp from "fs/promises";
+import { PdfReader } from "pdfreader";
+
+const openaiClient = new OpenAI({
+  apiKey: process.env.OPENAI_API_KEY,
+});
+
+async function queryChatGPT(prompt: string) {
+  let res = await openaiClient.chat.completions.create({
+    model: "gpt-4o",
+    messages: [
+      {
+        role: "system",
+        content: prompt,
+      },
+    ],
+    response_format: { type: "json_object" },
+  });
+
+  return JSON.parse(res.choices[0].message.content);
+}
+
+async function parseBufferAsync(buffer: Buffer): Promise<string> {
+  let parsedData = "";
+
+  return new Promise((resolve, reject) => {
+    new PdfReader().parseBuffer(buffer, (error, item) => {
+      if (error) return reject(error);
+      if (!item) return resolve(parsedData);
+      if (item.text) parsedData += item.text;
+    });
+  });
+}
+
+export default async function generateNote(
+  folderPath: string
+): Promise<string> {
+  let energyData = {};
+  let checklistData = {};
+  let energyDataFlag = false;
+  let checklistDataFlag = false;
+  let energyFilePath = "";
+  let checklistFilePath = "";
+
+  const files = await fsp.readdir(folderPath);
+
+  for (const file of files) {
+    const input = await fsp.readFile(folderPath + `/${file}`);
+    const pdfData = await parseBufferAsync(input);
+
+    if (
+      Object.keys(checklistData).length == 0 &&
+      pdfData.toLowerCase().includes("general contractors")
+    ) {
+      const prompt = `
+        Extract the following fields and contractor info from the given data and return the response in JSON.
+        Fields: ["Site Address", "Lot Number", "Estimated Construction Cost", "Total Square Footage", "A/C Square Footage", "Number of Bedrooms", "Number of Bathrooms", "Construction Type", "Number of Stories", "Height", "General Contractors License", "Mechanical Contractors License", "Electrical Contractors License", "Plumbing Contractors License", "Roofing Contractors License", "Concrete/Mason Contractors License", "Confirm the following"]
+
+        Data: ${pdfData}
+      `;
+
+      checklistDataFlag = true;
+      checklistFilePath = file;
+      checklistData = await queryChatGPT(prompt);
+    }
+
+    if (
+      Object.keys(energyData).length == 0 &&
+      pdfData.toLowerCase().includes("energy efficiency")
+    ) {
+      const prompt = `
+        Extract the following fields form the given data and return the response in JSON.
+        Fields: [Project Name, address, county]
+
+        Data: ${pdfData}
+      `;
+
+      energyDataFlag = true;
+      energyFilePath = file;
+      energyData = await queryChatGPT(prompt);
+    }
+  }
+
+  const addressCheck = {};
+
+  if (checklistDataFlag && Object.keys(checklistData).length > 0) {
+    for (const file of files) {
+      if (file == checklistFilePath) continue;
+
+      const input = await fsp.readFile(folderPath + `${file}`);
+      const pdfData = await parseBufferAsync(input);
+
+      const prompt = `
+        Check if the address in the data given below matches this ${checklistData["Site Address"]}. Return the response in JSON with schema {match: Boolean}.
+
+        Data: ${pdfData}
+      `;
+
+      const { match } = await queryChatGPT(prompt);
+      addressCheck[file] = match;
+    }
+  }
+
+  let note = "";
+
+  if (!energyDataFlag) {
+    note += "Energy Efficiency document not found.\n";
+  }
+
+  if (!checklistDataFlag) {
+    note += `Checklist document not found.\n\n`;
+  }
+
+  let filesList = "";
+  for (const file in addressCheck) {
+    if (!addressCheck[file]) {
+      filesList += file.split("/").pop() + "\n";
+    }
+  }
+
+  if (filesList != "") {
+    note +=
+      "Below files don't have address or the address doesn't match with the address in checklist\n\n";
+    note += filesList;
+    note += "\n";
+  }
+
+  if (energyDataFlag || checklistDataFlag) {
+    note += "\nExtracted Data:\n";
+  }
+
+  if (Object.keys(energyData).length > 0) {
+    note += `\n${JSON.stringify(energyData, null, 2)}\n`;
+  }
+
+  if (Object.keys(checklistData).length > 0) {
+    note += `\n${JSON.stringify(checklistData, null, 2)}\n`;
+  }
+
+  return note;
+}