Add OpenAI API integration and document processing features

- Added OpenAI API key configuration to .env.example.
- Integrated OpenAI for document processing, including markdown conversion and data extraction.
- Implemented new API routes for fetching document processing status and handling various processing tasks.
- Enhanced the App component to manage document status and processing states with user feedback via Snackbar.
- Updated CSVImportPanel and TableManagement components to support navigation to specific tabs based on processing results.
- Introduced transaction handling in the database configuration for improved error management during document processing.

.cursor/rules/devdatabase.mdc

@@ -1,6 +1,4 @@
 ---
 alwaysApply: true
 ---
-sqlcmd -C -S tcp:192.168.56.1,1497 -U app -P 'readonly' -d eazybusiness -W
-
 sqlcmd -C -S tcp:192.168.56.1,1497 -U sa -P 'sa_tekno23' -d eazybusiness -W

.cursor/rules/devserver.mdc

@@ -0,0 +1,7 @@
+---
+alwaysApply: true
+---
+pm2 restart 10 -> restart backend (configured as "npm run dev:backend")
+pm2 restart 11 -> restart backend (configured as "npm run dev:frontend")
+
+(both should rarely neer restart because in dev mode HMR for frontend, and nodemon for backend should already do that)

.env.example

@@ -8,6 +8,9 @@ REACT_APP_GOOGLE_CLIENT_ID=your_google_client_id_here
 # JWT Secret
 JWT_SECRET=your_jwt_secret_here
 
+# OpenAI API Configuration
+OPENAI_API_KEY=your_openai_api_key_here
+
 # Authorized Email Addresses (comma-separated)
 AUTHORIZED_EMAILS=admin@example.com,user1@example.com,user2@example.com
 

client/src/App.js

@@ -1,11 +1,16 @@
 import React, { Component } from 'react';
 import { ThemeProvider, createTheme } from '@mui/material/styles';
 import CssBaseline from '@mui/material/CssBaseline';
-import { Container, AppBar, Toolbar, Typography, Button, Box, Tabs, Tab } from '@mui/material';
+import { Container, AppBar, Toolbar, Typography, Button, Box, Tabs, Tab, Badge, Chip, Divider, Snackbar, Alert } from '@mui/material';
 import LoginIcon from '@mui/icons-material/Login';
 import DashboardIcon from '@mui/icons-material/Dashboard';
 import DownloadIcon from '@mui/icons-material/Download';
 import TableChart from '@mui/icons-material/TableChart';
+import PlayArrowIcon from '@mui/icons-material/PlayArrow';
+import DocumentScannerIcon from '@mui/icons-material/DocumentScanner';
+import ExtractIcon from '@mui/icons-material/TextSnippet';
+import EmailIcon from '@mui/icons-material/Email';
+import UploadIcon from '@mui/icons-material/Upload';
 import AuthService from './services/AuthService';
 import DataViewer from './components/DataViewer';
 import Login from './components/Login';
@@ -31,6 +36,18 @@ class App extends Component {
       loading: true,
       exportData: null, // { selectedMonth, canExport, onExport }
       currentView: 'dashboard', // 'dashboard' or 'tables'
+      documentStatus: null,
+      processingStatus: {
+        markdown: false,
+        extraction: false,
+        datevSync: false,
+        datevUpload: false
+      },
+      snackbar: {
+        open: false,
+        message: '',
+        severity: 'info' // 'success', 'error', 'warning', 'info'
+      }
     };
     this.authService = new AuthService();
   }
@@ -39,6 +56,15 @@ class App extends Component {
     this.checkAuthStatus();
   }
 
+  componentDidUpdate(prevState) {
+    // Clear targetTab after navigation is complete
+    if (this.state.targetTab && prevState.currentView !== this.state.currentView) {
+      setTimeout(() => {
+        this.setState({ targetTab: null });
+      }, 100); // Small delay to ensure navigation completes
+    }
+  }
+
   checkAuthStatus = async () => {
     try {
       const token = localStorage.getItem('token');
@@ -46,6 +72,7 @@ class App extends Component {
         const user = await this.authService.verifyToken(token);
         if (user) {
           this.setState({ isAuthenticated: true, user, loading: false });
+          this.fetchDocumentStatus();
           return;
         }
       }
@@ -62,6 +89,7 @@ class App extends Component {
       if (result.success) {
         localStorage.setItem('token', result.token);
         this.setState({ isAuthenticated: true, user: result.user });
+        this.fetchDocumentStatus();
       }
     } catch (error) {
       console.error('Login failed:', error);
@@ -83,8 +111,131 @@ class App extends Component {
     this.setState({ currentView: newValue });
   };
 
+  showSnackbar = (message, severity = 'info') => {
+    this.setState({
+      snackbar: {
+        open: true,
+        message,
+        severity
+      }
+    });
+  };
+
+  handleSnackbarClose = (event, reason) => {
+    if (reason === 'clickaway') {
+      return;
+    }
+    this.setState({
+      snackbar: {
+        ...this.state.snackbar,
+        open: false
+      }
+    });
+  };
+
+  fetchDocumentStatus = async () => {
+    try {
+      const token = localStorage.getItem('token');
+      if (!token) {
+        console.log('No token found for document status');
+        return;
+      }
+
+      console.log('Fetching document status...');
+      const response = await fetch('/api/data/document-status', {
+        headers: {
+          'Authorization': `Bearer ${token}`,
+          'Content-Type': 'application/json',
+        },
+      });
+
+      if (response.ok) {
+        const status = await response.json();
+        console.log('Document status received:', status);
+        this.setState({ documentStatus: status });
+      } else {
+        console.error('Failed to fetch document status:', response.status, await response.text());
+      }
+    } catch (error) {
+      console.error('Error fetching document status:', error);
+    }
+  };
+
+  handleProcessing = async (processType) => {
+    if (this.state.processingStatus[processType]) {
+      return; // Already processing
+    }
+
+    // Handle datev upload navigation
+    if (processType === 'datev-upload') {
+      this.setState({ 
+        currentView: 'tables',
+        targetTab: { 
+          level1: 3, // CSV Import tab
+          level2: 'DATEV_LINKS' // DATEV Beleglinks tab
+        }
+      });
+      return;
+    }
+
+    // Check if there are documents to process
+    const statusKey = processType === 'datev-sync' ? 'needDatevSync' : 
+                     processType === 'extraction' ? 'needExtraction' : 'needMarkdown';
+    
+    if (!this.state.documentStatus || this.state.documentStatus[statusKey] === 0) {
+      this.showSnackbar(`No documents need ${processType} processing at this time.`, 'info');
+      return;
+    }
+
+    this.setState(prevState => ({
+      processingStatus: {
+        ...prevState.processingStatus,
+        [processType]: true
+      }
+    }));
+
+    try {
+      const token = localStorage.getItem('token');
+      if (!token) return;
+
+      const response = await fetch(`/api/data/process-${processType}`, {
+        method: 'POST',
+        headers: {
+          'Authorization': `Bearer ${token}`,
+          'Content-Type': 'application/json',
+        },
+      });
+
+      if (response.ok) {
+        const result = await response.json();
+        console.log(`${processType} processing result:`, result);
+        this.showSnackbar(`${processType} processing completed successfully!`, 'success');
+        // Refresh document status after successful processing
+        await this.fetchDocumentStatus();
+      } else {
+        const error = await response.json();
+        console.error(`Failed to process ${processType}:`, error);
+        this.showSnackbar(`Failed to process ${processType}: ${error.error || response.status}`, 'error');
+      }
+    } catch (error) {
+      console.error(`Error processing ${processType}:`, error);
+      this.showSnackbar(`Error processing ${processType}: ${error.message}`, 'error');
+    } finally {
+      this.setState(prevState => ({
+        processingStatus: {
+          ...prevState.processingStatus,
+          [processType]: false
+        }
+      }));
+    }
+  };
+
   render() {
-    const { isAuthenticated, user, loading, currentView } = this.state;
+    const { isAuthenticated, user, loading, currentView, documentStatus, processingStatus, snackbar } = this.state;
+    
+    // Debug logging
+    console.log('App render - documentStatus:', documentStatus);
+    console.log('App render - isAuthenticated:', isAuthenticated);
 
     if (loading) {
       return (
@@ -149,6 +300,97 @@ class App extends Component {
                     sx={{ minHeight: 48 }}
                   />
                 </Tabs>
+                <Divider orientation="vertical" flexItem sx={{ mx: 2, backgroundColor: 'rgba(255, 255, 255, 0.3)' }} />
+                <Box sx={{ display: 'flex', alignItems: 'center', gap: 1 }}>
+                  <Button
+                    color="inherit"
+                    size="small"
+                    onClick={() => this.handleProcessing('markdown')}
+                    disabled={processingStatus.markdown || !documentStatus}
+                    sx={{ 
+                      minWidth: 'auto',
+                      px: 1,
+                      '&:hover': { backgroundColor: 'rgba(255, 255, 255, 0.1)' }
+                    }}
+                    title="Process markdown conversion"
+                  >
+                    <Badge 
+                      badgeContent={documentStatus?.needMarkdown || 0} 
+                      color={documentStatus?.needMarkdown > 0 ? "error" : "default"}
+                      max={999999}
+                      sx={{ mr: 0.5 }}
+                    >
+                      <DocumentScannerIcon fontSize="small" />
+                    </Badge>
+                    {processingStatus.markdown && <PlayArrowIcon fontSize="small" />}
+                  </Button>
+                  <Button
+                    color="inherit"
+                    size="small"
+                    onClick={() => this.handleProcessing('extraction')}
+                    disabled={processingStatus.extraction || !documentStatus}
+                    sx={{ 
+                      minWidth: 'auto',
+                      px: 1,
+                      '&:hover': { backgroundColor: 'rgba(255, 255, 255, 0.1)' }
+                    }}
+                    title="Process data extraction"
+                  >
+                    <Badge 
+                      badgeContent={documentStatus?.needExtraction || 0} 
+                      color={documentStatus?.needExtraction > 0 ? "warning" : "default"}
+                      max={999999}
+                      sx={{ mr: 0.5 }}
+                    >
+                      <ExtractIcon fontSize="small" />
+                    </Badge>
+                    {processingStatus.extraction && <PlayArrowIcon fontSize="small" />}
+                  </Button>
+                  <Button
+                    color="inherit"
+                    size="small"
+                    onClick={() => this.handleProcessing('datev-sync')}
+                    disabled={processingStatus.datevSync || !documentStatus}
+                    sx={{ 
+                      minWidth: 'auto',
+                      px: 1,
+                      '&:hover': { backgroundColor: 'rgba(255, 255, 255, 0.1)' }
+                    }}
+                    title="Process Datev sync"
+                  >
+                    <Badge 
+                      badgeContent={documentStatus?.needDatevSync || 0} 
+                      color={documentStatus?.needDatevSync > 0 ? "info" : "default"}
+                      max={999999}
+                      sx={{ mr: 0.5 }}
+                    >
+                      <EmailIcon fontSize="small" />
+                    </Badge>
+                    {processingStatus.datevSync && <PlayArrowIcon fontSize="small" />}
+                  </Button>
+                  <Button
+                    color="inherit"
+                    size="small"
+                    onClick={() => this.handleProcessing('datev-upload')}
+                    disabled={processingStatus.datevUpload || !documentStatus}
+                    sx={{ 
+                      minWidth: 'auto',
+                      px: 1,
+                      '&:hover': { backgroundColor: 'rgba(255, 255, 255, 0.1)' }
+                    }}
+                    title="Process Datev CSV upload"
+                  >
+                    <Badge 
+                      badgeContent={documentStatus?.needDatevUpload || 0} 
+                      color={documentStatus?.needDatevUpload > 0 ? "secondary" : "default"}
+                      max={999999}
+                      sx={{ mr: 0.5 }}
+                    >
+                      <UploadIcon fontSize="small" />
+                    </Badge>
+                    {processingStatus.datevUpload && <PlayArrowIcon fontSize="small" />}
+                  </Button>
+                </Box>
                 {this.state.exportData && (
                   <Button 
                     color="inherit" 
@@ -194,12 +436,29 @@ class App extends Component {
                 onUpdateExportData={this.updateExportData}
                 currentView={currentView}
                 onViewChange={this.handleViewChange}
+                targetTab={this.state.targetTab}
               />
             ) : (
               <Login onLogin={this.handleLogin} />
             )}
           </Container>
         </Box>
+
+        <Snackbar
+          open={snackbar.open}
+          autoHideDuration={6000}
+          onClose={this.handleSnackbarClose}
+          anchorOrigin={{ vertical: 'bottom', horizontal: 'right' }}
+        >
+          <Alert
+            onClose={this.handleSnackbarClose}
+            severity={snackbar.severity}
+            variant="filled"
+            sx={{ width: '100%' }}
+          >
+            {snackbar.message}
+          </Alert>
+        </Snackbar>
       </ThemeProvider>
     );
   }

client/src/components/CSVImportDialog.js

@@ -56,6 +56,20 @@ class CSVImportPanel extends Component {
     this.datevFileInputRef = React.createRef();
   }
 
+  componentDidMount() {
+    // Check if we should navigate to a specific tab
+    if (this.props.targetTab) {
+      this.setState({ activeTab: this.props.targetTab });
+    }
+  }
+
+  componentDidUpdate(prevProps) {
+    // Handle targetTab changes
+    if (this.props.targetTab !== prevProps.targetTab && this.props.targetTab) {
+      this.setState({ activeTab: this.props.targetTab });
+    }
+  }
+
   // Tab switch resets type-specific state but keeps success state as-is
   handleTabChange = (_e, value) => {
     this.setState({
@@ -344,6 +358,7 @@ class CSVImportPanel extends Component {
         )}
 
         {currentHeaders && (
+          
           <Box sx={{ mb: 2 }}>
             <Typography variant="subtitle2" gutterBottom>
               Erkannte Spalten ({currentHeaders.length}):

client/src/components/DataViewer.js

@@ -165,7 +165,7 @@ class DataViewer extends Component {
           </>
         ) : (
           <Box sx={{ flex: 1, minHeight: 0, overflow: 'auto', p: 2 }}>
-            <TableManagement user={user} />
+            <TableManagement user={user} targetTab={this.props.targetTab} />
           </Box>
         )}
       </Box>

client/src/components/TableManagement.js

@@ -26,6 +26,21 @@ class TableManagement extends Component {
     };
   }
 
+  componentDidMount() {
+    // Check if we should navigate to a specific tab
+    if (this.props.targetTab?.level1 !== undefined) {
+      this.setState({ activeTab: this.props.targetTab.level1 });
+    }
+  }
+
+  componentDidUpdate(prevProps) {
+    // Handle targetTab changes
+    if (this.props.targetTab?.level1 !== prevProps.targetTab?.level1 && 
+        this.props.targetTab?.level1 !== undefined) {
+      this.setState({ activeTab: this.props.targetTab.level1 });
+    }
+  }
+
   handleTabChange = (event, newValue) => {
     this.setState({ activeTab: newValue });
   };
@@ -92,6 +107,7 @@ class TableManagement extends Component {
                 </Typography>
                 <CSVImportPanel 
                   user={user}
+                  targetTab={this.props.targetTab?.level2}
                 />
               </Box>
             )}

package-lock.json

@@ -21,6 +21,8 @@
         "google-auth-library": "^9.0.0",
         "jsonwebtoken": "^9.0.0",
         "mssql": "^9.1.0",
+        "nodemailer": "^7.0.5",
+        "openai": "^5.12.0",
         "react": "^18.2.0",
         "react-dom": "^18.2.0"
       },
@@ -7142,6 +7144,15 @@
       "dev": true,
       "license": "MIT"
     },
+    "node_modules/nodemailer": {
+      "version": "7.0.5",
+      "resolved": "https://registry.npmjs.org/nodemailer/-/nodemailer-7.0.5.tgz",
+      "integrity": "sha512-nsrh2lO3j4GkLLXoeEksAMgAOqxOv6QumNRVQTJwKH4nuiww6iC2y7GyANs9kRAxCexg3+lTWM3PZ91iLlVjfg==",
+      "license": "MIT-0",
+      "engines": {
+        "node": ">=6.0.0"
+      }
+    },
     "node_modules/nodemon": {
       "version": "3.1.10",
       "resolved": "https://registry.npmjs.org/nodemon/-/nodemon-3.1.10.tgz",
@@ -7365,6 +7376,27 @@
         "url": "https://github.com/sponsors/sindresorhus"
       }
     },
+    "node_modules/openai": {
+      "version": "5.12.0",
+      "resolved": "https://registry.npmjs.org/openai/-/openai-5.12.0.tgz",
+      "integrity": "sha512-vUdt02xiWgOHiYUmW0Hj1Qu9OKAiVQu5Bd547ktVCiMKC1BkB5L3ImeEnCyq3WpRKR6ZTaPgekzqdozwdPs7Lg==",
+      "license": "Apache-2.0",
+      "bin": {
+        "openai": "bin/cli"
+      },
+      "peerDependencies": {
+        "ws": "^8.18.0",
+        "zod": "^3.23.8"
+      },
+      "peerDependenciesMeta": {
+        "ws": {
+          "optional": true
+        },
+        "zod": {
+          "optional": true
+        }
+      }
+    },
     "node_modules/own-keys": {
       "version": "1.0.1",
       "resolved": "https://registry.npmjs.org/own-keys/-/own-keys-1.0.1.tgz",
@@ -9827,7 +9859,7 @@
       "version": "8.18.3",
       "resolved": "https://registry.npmjs.org/ws/-/ws-8.18.3.tgz",
       "integrity": "sha512-PEIGCY5tSlUt50cqyMXfCzX+oOPqN0vuGqWzbcJ2xvnkzkq46oOpz7dQaTDBdfICb4N14+GARUDw2XV2N4tvzg==",
-      "dev": true,
+      "devOptional": true,
       "license": "MIT",
       "engines": {
         "node": ">=10.0.0"

package.json

@@ -32,6 +32,8 @@
     "google-auth-library": "^9.0.0",
     "jsonwebtoken": "^9.0.0",
     "mssql": "^9.1.0",
+    "nodemailer": "^7.0.5",
+    "openai": "^5.12.0",
     "react": "^18.2.0",
     "react-dom": "^18.2.0"
   },

src/config/database.js

@@ -73,10 +73,42 @@ const executeQuery = async (query, params = {}) => {
   }
 };
 
+const executeTransaction = async (callback) => {
+  if (!process.env.DB_SERVER) {
+    throw new Error('Database not configured');
+  }
+  
+  let pool;
+  let transaction;
+  
+  try {
+    pool = await getPool();
+    transaction = new sql.Transaction(pool);
+    
+    await transaction.begin();
+    
+    const result = await callback(transaction);
+    
+    await transaction.commit();
+    return result;
+  } catch (error) {
+    if (transaction) {
+      try {
+        await transaction.rollback();
+      } catch (rollbackError) {
+        console.error('Transaction rollback failed:', rollbackError);
+      }
+    }
+    console.error('Transaction error:', error);
+    throw error;
+  }
+};
+
 module.exports = {
   config,
   getPool,
   testConnection,
   executeQuery,
+  executeTransaction,
   sql,
 }; 

src/routes/data/documentProcessing.js

@@ -0,0 +1,418 @@
+const express = require('express');
+const { authenticateToken } = require('../../middleware/auth');
+const { executeQuery, executeTransaction } = require('../../config/database');
+const sql = require('mssql');
+const nodemailer = require('nodemailer');
+
+const router = express.Router();
+
+// Get document processing status
+router.get('/document-status', authenticateToken, async (req, res) => {
+  try {
+    console.log('Document status endpoint called');
+    const queries = {
+      needMarkdownUmsatz: "SELECT COUNT(*) as count FROM tUmsatzBeleg WHERE markDown is null",
+      needMarkdownPdf: "SELECT COUNT(*) as count FROM tPdfObjekt WHERE markDown is null",
+      needExtractionUmsatz: "SELECT COUNT(*) as count FROM tUmsatzBeleg WHERE markDown is not null and extraction is null",
+      needExtractionPdf: "SELECT COUNT(*) as count FROM tPdfObjekt WHERE markDown is not null and extraction is null",
+      needDatevSyncUmsatz: "SELECT COUNT(*) as count FROM tUmsatzBeleg WHERE markDown is not null and datevlink is null",
+      needDatevSyncPdf: "SELECT COUNT(*) as count FROM tPdfObjekt WHERE markDown is not null and datevlink is null",
+      needDatevUploadUmsatz: "SELECT COUNT(*) as count FROM tUmsatzBeleg WHERE datevlink = 'pending'",
+      needDatevUploadPdf: "SELECT COUNT(*) as count FROM tPdfObjekt WHERE datevlink = 'pending'"
+    };
+
+    const results = {};
+    for (const [key, query] of Object.entries(queries)) {
+      const result = await executeQuery(query);
+      results[key] = result.recordset[0].count;
+    }
+
+    const status = {
+      needMarkdown: results.needMarkdownUmsatz + results.needMarkdownPdf,
+      needExtraction: results.needExtractionUmsatz + results.needExtractionPdf,
+      needDatevSync: results.needDatevSyncUmsatz + results.needDatevSyncPdf,
+      needDatevUpload: results.needDatevUploadUmsatz + results.needDatevUploadPdf,
+      details: {
+        markdown: {
+          umsatzBeleg: results.needMarkdownUmsatz,
+          pdfObjekt: results.needMarkdownPdf
+        },
+        extraction: {
+          umsatzBeleg: results.needExtractionUmsatz,
+          pdfObjekt: results.needExtractionPdf
+        },
+        datevSync: {
+          umsatzBeleg: results.needDatevSyncUmsatz,
+          pdfObjekt: results.needDatevSyncPdf
+        },
+        datevUpload: {
+          umsatzBeleg: results.needDatevUploadUmsatz,
+          pdfObjekt: results.needDatevUploadPdf
+        }
+      }
+    };
+
+    console.log('Document status computed:', status);
+    res.json(status);
+  } catch (error) {
+    console.error('Error fetching document processing status:', error);
+    res.status(500).json({ error: 'Failed to fetch document processing status' });
+  }
+});
+
+// Process markdown conversion
+router.post('/process-markdown', authenticateToken, async (req, res) => {
+  try {
+    const { OpenAI } = require('openai');
+    
+    // Check environment for OpenAI API key
+    if (!process.env.OPENAI_API_KEY) {
+      return res.status(500).json({ error: 'OpenAI API key not configured' });
+    }
+
+    const openai = new OpenAI({
+      apiKey: process.env.OPENAI_API_KEY,
+    });
+
+    await executeTransaction(async (transaction) => {
+      // Process UmsatzBeleg documents
+      const umsatzResult = await new sql.Request(transaction).query(
+        "SELECT TOP 1 kUmsatzBeleg, content FROM tUmsatzBeleg WHERE markDown is null"
+      );
+
+      if (umsatzResult.recordset.length > 0) {
+        const { kUmsatzBeleg, content } = umsatzResult.recordset[0];
+        
+        const response = await openai.responses.create({
+          model: "gpt-4o",
+          input: [
+            { "role": "developer", "content": [{ "type": "input_text", "text": "Convert to Markdown" }] },
+            { "role": "user", "content": [{ "type": "input_file", "filename": "invoice.pdf", "file_data": "data:application/pdf;base64," + content.toString('base64') }] }
+          ],
+          text: {
+            "format": {
+              "type": "json_schema", "name": "markdown", "strict": true, "schema": { "type": "object", "properties": {
+                "output": { "type": "string", "description": "Input converted to Markdown" }
+              }, "required": ["output"], "additionalProperties": false }
+            }
+          },
+          tools: [],
+          store: false
+        });
+
+        const markdown = JSON.parse(response.output_text);
+        
+        await new sql.Request(transaction)
+          .input('kUmsatzBeleg', kUmsatzBeleg)
+          .input('markDown', markdown.output)
+          .query("UPDATE tUmsatzBeleg SET markDown = @markDown WHERE kUmsatzBeleg = @kUmsatzBeleg");
+      }
+
+      // Process PdfObjekt documents
+      const pdfResult = await new sql.Request(transaction).query(
+        "SELECT TOP 1 kPdfObjekt, content FROM tPdfObjekt WHERE markDown is null"
+      );
+
+      if (pdfResult.recordset.length > 0) {
+        const { kPdfObjekt, content } = pdfResult.recordset[0];
+        
+        const response = await openai.responses.create({
+          model: "gpt-4o",
+          input: [
+            { "role": "developer", "content": [{ "type": "input_text", "text": "Convert to Markdown" }] },
+            { "role": "user", "content": [{ "type": "input_file", "filename": "invoice.pdf", "file_data": "data:application/pdf;base64," + content.toString('base64') }] }
+          ],
+          text: {
+            "format": {
+              "type": "json_schema", "name": "markdown", "strict": true, "schema": { "type": "object", "properties": {
+                "output": { "type": "string", "description": "Input converted to Markdown" }
+              }, "required": ["output"], "additionalProperties": false }
+            }
+          },
+          tools: [],
+          store: false
+        });
+
+        const markdown = JSON.parse(response.output_text);
+        
+        await new sql.Request(transaction)
+          .input('kPdfObjekt', kPdfObjekt)
+          .input('markDown', markdown.output)
+          .query("UPDATE tPdfObjekt SET markDown = @markDown WHERE kPdfObjekt = @kPdfObjekt");
+      }
+    });
+
+    res.json({ success: true, message: 'Markdown processing completed' });
+  } catch (error) {
+    console.error('Error processing markdown:', error);
+    res.status(500).json({ error: 'Failed to process markdown: ' + error.message });
+  }
+});
+
+// Process data extraction
+router.post('/process-extraction', authenticateToken, async (req, res) => {
+  try {
+    const { OpenAI } = require('openai');
+    
+    if (!process.env.OPENAI_API_KEY) {
+      return res.status(500).json({ error: 'OpenAI API key not configured' });
+    }
+
+    const openai = new OpenAI({
+      apiKey: process.env.OPENAI_API_KEY,
+    });
+
+    await executeTransaction(async (transaction) => {
+      // Get creditor IDs for extraction
+      const creditorResult = await new sql.Request(transaction).query(
+        "SELECT kreditorId FROM fibdash.Kreditor ORDER BY kreditorId"
+      );
+      const creditorIDs = creditorResult.recordset.map(r => r.kreditorId).join(', ');
+
+      // Process UmsatzBeleg documents
+      const umsatzResult = await new sql.Request(transaction).query(
+        "SELECT TOP 1 kUmsatzBeleg, markDown FROM tUmsatzBeleg WHERE markDown is not null and extraction is null"
+      );
+
+      if (umsatzResult.recordset.length > 0) {
+        const { kUmsatzBeleg, markDown } = umsatzResult.recordset[0];
+        
+        const response = await openai.responses.create({
+          model: "gpt-4o-mini",
+          input: [
+            { "role": "developer", "content": [{ "type": "input_text", "text": `Extract specified information from provided input and structure it in a JSON format. 
+
+              The aim is to accurately identify and capture the following elements: 
+              - Rechnungsdatum/Belegdatum (Invoice Date/Document Date), 
+              - Rechnungsnummer/Belegnummer (Invoice Number/Document Number), 
+              - Netto Betrag (Net Amount), 
+              - Brutto Betrag (Gross Amount), 
+              - and Absender (Sender).
+
+              # Steps
+
+              1. **Identify Dates**: Find and extract the invoice or document date (Rechnungsdatum/Belegdatum) from the input text.
+              2. **Extract Numbers**: Locate and pull out the invoice or document number (Rechnungsnummer/Belegnummer).
+              3. **Determine Amounts**: Identify the net amount (Netto Betrag) and the gross amount (Brutto Betrag) and the currency in the text.
+              4. **Source the Sender**: Extract the sender's information (Absender, Country).
+              5. **Structure Data**: Organize the extracted information into a JSON format following the specified schema.
+
+              # Notes
+
+              - Ensure that dates are formatted consistently.
+              - Be mindful of various numerical representations (e.g., with commas or periods).
+              - The sender's information might include company names, so recognize various formats.
+              - Prioritize accuracy in identifying the correct fields, as there can be similar text elements present.
+
+              Also select the CreditorID, from that List: ${creditorIDs}` }] },
+            { "role": "user", "content": [{ "type": "input_text", "text": markDown }] }
+          ],
+          text: {
+            "format": {
+              "type": "json_schema", "name": "invoice", "strict": true, "schema": { "type": "object", "properties": {
+                "date": { "type": "string", "description": "Rechungsdatum / Belegdatum in ISO 8601" },
+                "invoice_number": { "type": "string", "description": "Rechnungsnummer / Belegnummer / Invoicenr" },
+                "net_amounts_and_tax": {
+                  "type": "array", "description": "Liste von Nettobeträgen mit jeweiligem Steuersatz und Steuerbetrag, ein Listeneintrag pro Steuersatz",
+                  "items": { "type": "object", "properties": {
+                    "net_amount": { "type": "number", "description": "Netto Betrag" },
+                    "tax_rate": { "type": "number", "description": "Steuersatz in Prozent" },
+                    "tax_amount": { "type": "number", "description": "Steuerbetrag" }
+                  }, "required": ["net_amount", "tax_rate", "tax_amount"], "additionalProperties": false }
+                },
+                "gross_amount": { "type": "number", "description": "Brutto Betrag (muss der Summe aller net_amount + tax_amount entsprechen)" },
+                "currency": { "type": "string", "description": "currency code in ISO 4217" },
+                "country": { "type": "string", "description": "country of origin in ISO 3166" },
+                "sender": { "type": "string", "description": "Absender" },
+                "creditorID": { "type": "string", "description": "CreditorID or empty if unknown" }
+              }, "required": ["date", "invoice_number", "net_amounts_and_tax", "gross_amount", "currency", "country", "sender", "creditorID"], "additionalProperties": false }
+            }
+          },
+          reasoning: { "effort": "medium", "summary": "auto" },
+          tools: [],
+          store: false
+        });
+
+        const extraction = JSON.parse(response.output_text);
+        
+        await new sql.Request(transaction)
+          .input('kUmsatzBeleg', kUmsatzBeleg)
+          .input('extraction', JSON.stringify(extraction))
+          .query("UPDATE tUmsatzBeleg SET extraction = @extraction WHERE kUmsatzBeleg = @kUmsatzBeleg");
+      }
+
+      // Process PdfObjekt documents
+      const pdfResult = await new sql.Request(transaction).query(
+        "SELECT TOP 1 kPdfObjekt, markDown FROM tPdfObjekt WHERE markDown is not null and extraction is null"
+      );
+
+      if (pdfResult.recordset.length > 0) {
+        const { kPdfObjekt, markDown } = pdfResult.recordset[0];
+        
+        const response = await openai.responses.create({
+          model: "gpt-4o-mini",
+          input: [
+            { "role": "developer", "content": [{ "type": "input_text", "text": `Extract specified information from provided input and structure it in a JSON format. 
+
+              The aim is to accurately identify and capture the following elements: 
+              - Rechnungsdatum/Belegdatum (Invoice Date/Document Date), 
+              - Rechnungsnummer/Belegnummer (Invoice Number/Document Number), 
+              - Netto Betrag (Net Amount), 
+              - Brutto Betrag (Gross Amount), 
+              - and Absender (Sender).
+
+              # Steps
+
+              1. **Identify Dates**: Find and extract the invoice or document date (Rechnungsdatum/Belegdatum) from the input text.
+              2. **Extract Numbers**: Locate and pull out the invoice or document number (Rechnungsnummer/Belegnummer).
+              3. **Determine Amounts**: Identify the net amount (Netto Betrag) and the gross amount (Brutto Betrag) and the currency in the text.
+              4. **Source the Sender**: Extract the sender's information (Absender, Country).
+              5. **Structure Data**: Organize the extracted information into a JSON format following the specified schema.
+
+              # Notes
+
+              - Ensure that dates are formatted consistently.
+              - Be mindful of various numerical representations (e.g., with commas or periods).
+              - The sender's information might include company names, so recognize various formats.
+              - Prioritize accuracy in identifying the correct fields, as there can be similar text elements present.
+
+              Also select the CreditorID, from that List: ${creditorIDs}` }] },
+            { "role": "user", "content": [{ "type": "input_text", "text": markDown }] }
+          ],
+          text: {
+            "format": {
+              "type": "json_schema", "name": "invoice", "strict": true, "schema": { "type": "object", "properties": {
+                "date": { "type": "string", "description": "Rechungsdatum / Belegdatum in ISO 8601" },
+                "invoice_number": { "type": "string", "description": "Rechnungsnummer / Belegnummer / Invoicenr" },
+                "net_amounts_and_tax": {
+                  "type": "array", "description": "Liste von Nettobeträgen mit jeweiligem Steuersatz und Steuerbetrag, ein Listeneintrag pro Steuersatz",
+                  "items": { "type": "object", "properties": {
+                    "net_amount": { "type": "number", "description": "Netto Betrag" },
+                    "tax_rate": { "type": "number", "description": "Steuersatz in Prozent" },
+                    "tax_amount": { "type": "number", "description": "Steuerbetrag" }
+                  }, "required": ["net_amount", "tax_rate", "tax_amount"], "additionalProperties": false }
+                },
+                "gross_amount": { "type": "number", "description": "Brutto Betrag (muss der Summe aller net_amount + tax_amount entsprechen)" },
+                "currency": { "type": "string", "description": "currency code in ISO 4217" },
+                "country": { "type": "string", "description": "country of origin in ISO 3166" },
+                "sender": { "type": "string", "description": "Absender" },
+                "creditorID": { "type": "string", "description": "CreditorID or empty if unknown" }
+              }, "required": ["date", "invoice_number", "net_amounts_and_tax", "gross_amount", "currency", "country", "sender", "creditorID"], "additionalProperties": false }
+            }
+          },
+          reasoning: { "effort": "medium", "summary": "auto" },
+          tools: [],
+          store: false
+        });
+
+        const extraction = JSON.parse(response.output_text);
+        
+        await new sql.Request(transaction)
+          .input('kPdfObjekt', kPdfObjekt)
+          .input('extraction', JSON.stringify(extraction))
+          .query("UPDATE tPdfObjekt SET extraction = @extraction WHERE kPdfObjekt = @kPdfObjekt");
+      }
+    });
+
+    res.json({ success: true, message: 'Extraction processing completed' });
+  } catch (error) {
+    console.error('Error processing extraction:', error);
+    res.status(500).json({ error: 'Failed to process extraction: ' + error.message });
+  }
+});
+
+// Process Datev sync
+router.post('/process-datev-sync', authenticateToken, async (req, res) => {
+  try {
+    const transporter = nodemailer.createTransport({
+      host: "smtp.gmail.com",
+      port: 587,
+      secure: false, // true for 465, false for other ports
+      auth: {
+        user: "sebgreenbus@gmail.com",
+        pass: "abrp idub thbi kdws", // For Gmail, you might need an app-specific password
+      },
+    });
+
+    await executeTransaction(async (transaction) => {
+      // Process UmsatzBeleg documents
+      const umsatzResult = await new sql.Request(transaction).query(
+        "SELECT TOP 1 kUmsatzBeleg, content FROM tUmsatzBeleg WHERE markDown is not null and datevlink is null"
+      );
+
+      if (umsatzResult.recordset.length > 0) {
+        const { kUmsatzBeleg, content } = umsatzResult.recordset[0];
+
+        const mailOptions = {
+          from: '"Growheads" <sebgreenbus@gmail.com>',
+          to: "97bfd9eb-770f-481a-accb-e69649d36a9e@uploadmail.datev.de",
+          subject: `Beleg ${kUmsatzBeleg} für Datev`,
+          text: "", // No body text as requested
+          attachments: [
+            {
+              filename: `UmsatzBeleg${kUmsatzBeleg}.pdf`,
+              content: content,
+              contentType: "application/pdf",
+            },
+          ],
+        };
+
+        try {
+          let info = await transporter.sendMail(mailOptions);
+          console.log("Message sent: %s", info.messageId);
+
+          await new sql.Request(transaction)
+            .input('kUmsatzBeleg', kUmsatzBeleg)
+            .input('datevlink', 'pending')
+            .query("UPDATE tUmsatzBeleg SET datevlink = @datevlink WHERE kUmsatzBeleg = @kUmsatzBeleg");
+        } catch (emailError) {
+          console.error("Error sending email:", emailError);
+          throw emailError;
+        }
+      }
+
+      // Process PdfObjekt documents
+      const pdfResult = await new sql.Request(transaction).query(
+        "SELECT TOP 1 kPdfObjekt, content FROM tPdfObjekt WHERE markDown is not null and datevlink is null"
+      );
+
+      if (pdfResult.recordset.length > 0) {
+        const { kPdfObjekt, content } = pdfResult.recordset[0];
+
+        const mailOptions = {
+          from: '"Growheads" <sebgreenbus@gmail.com>',
+          to: "97bfd9eb-770f-481a-accb-e69649d36a9e@uploadmail.datev.de",
+          subject: `Rechnung ${kPdfObjekt} für Datev`,
+          text: "", // No body text as requested
+          attachments: [
+            {
+              filename: `Rechnung${kPdfObjekt}.pdf`,
+              content: content,
+              contentType: "application/pdf",
+            },
+          ],
+        };
+
+        try {
+          let info = await transporter.sendMail(mailOptions);
+          console.log("Message sent: %s", info.messageId);
+
+          await new sql.Request(transaction)
+            .input('kPdfObjekt', kPdfObjekt)
+            .input('datevlink', 'pending')
+            .query("UPDATE tPdfObjekt SET datevlink = @datevlink WHERE kPdfObjekt = @kPdfObjekt");
+        } catch (emailError) {
+          console.error("Error sending email:", emailError);
+          throw emailError;
+        }
+      }
+    });
+
+    res.json({ success: true, message: 'Datev sync processing completed' });
+  } catch (error) {
+    console.error('Error processing Datev sync:', error);
+    res.status(500).json({ error: 'Failed to process Datev sync: ' + error.message });
+  }
+});
+
+module.exports = router;

src/routes/data/index.js

@@ -8,6 +8,7 @@ const kreditors = require('./kreditors');
 const bankingTransactions = require('./bankingTransactions');
 const accountingItems = require('./accountingItems');
 const csvImport = require('./csvImport');
+const documentProcessing = require('./documentProcessing');
 
 const router = express.Router();
 
@@ -20,5 +21,6 @@ router.use(kreditors);
 router.use(bankingTransactions);
 router.use(accountingItems);
 router.use(csvImport);
+router.use(documentProcessing);
 
 module.exports = router;