deepseek_chatbot/api/chatbot.php

<?php
require_once 'CsvReader.php';
set_time_limit(2500); // 2 Minuten erlauben

header('Content-Type: application/json');

// only accept POST requests
if ($_SERVER['REQUEST_METHOD'] !== 'POST') {
    http_response_code(405);
    echo json_encode(['error' => 'Only POST allowed']);
    exit;
}


/** Extract text from a PDF file using pdftotext command line tool.
 *
 * @param string $file Path to the PDF file.
 * @return string Extracted text from the PDF.
 */
function extractPdfText($file) {
    $tmp = tempnam(sys_get_temp_dir(), 'pdftext');
    exec("pdftotext " . escapeshellarg($file) . " " . escapeshellarg($tmp));
    $text = file_get_contents($tmp);
    unlink($tmp);
    return $text ?: '';
}

// Build prompt for deepseek
$input = json_decode(file_get_contents('php://input'), true);
$question = isset($input['question']) ? $input['question'] : '';
$roleFile = __DIR__ . '/data/role.txt';
$role = file_exists($roleFile) ? file_get_contents($roleFile) : '';

$reader = new CsvReader(__DIR__ . '/data/csv');
$rows = $reader->readAll();
$contextText = "";
foreach ($rows as $row) {
    $contextText .= "- " . implode(" | ", $row) . "\n";
}

$pdf1 = extractPdfText(__DIR__ . '/data/pdf/onboarding.pdf');
$pdf2 = extractPdfText(__DIR__ . '/data/pdf/urlaub.pdf');

$pdfContext = "\n\nAus Onboarding-Dokument:\n" . $pdf1 . "\n\nAus Urlaubsdokument:\n" . $pdf2;

$htmlDir = __DIR__ . '/data/html';
$htmlContext = '';

$prompt = "";
$prompt = "Deine Rolle ist: ". $role;
$prompt .= "\nDeine Frage ist: " . $question;
$prompt .= "\nCSV-Daten:\n" . $contextText;
$prompt .= "\nPDF-Kontext:\n" . $pdfContext;
$prompt .= "\nHTML-Wissensdatenbank:\n" . $htmlContext;

// ollama stuff
//$payload = json_encode([
//    "model" => "deepseek-r1",
////    "model" => "deepseek-llm",
////    "model" => "deepseek-coder:6.7b",
//    "prompt" => $prompt,
//    "stream" => false,
//    "stop" => ["</think>"]
//]);
//$promptContext = "Kontextdaten:\n\nCSV:\n$contextText\n\nPDF:\n$pdfContext\nHTML:\n$htmlContext";
$promptContext = "Kontextdaten:\n\nCSV:\n$contextText\n\nPDF:\n$pdfContext";
error_log("System prompt length: " . strlen($promptContext));

$payload = json_encode([
    "model" => "deepseek-llm-7b-chat",
    "messages" => [
//        ["role" => "system", "content" => "Rolle:\n$role\n\n$promptContext"],
        ["role" => "system", "content" => "Rolle:\n$role"],
        ["role" => "user", "content" => $question]
//        ["role" => "context", "content" => $promptContext]
    ],
    "temperature" => 0.5,
    "max_tokens" => -1,
    "stream" => false
]);
$start = microtime(true);
// ### Send Request
//$ch = curl_init('http://localhost:11434/api/generate');
$ch = curl_init('http://localhost:1234/api/v0/chat/completions');
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
curl_setopt($ch, CURLOPT_POST, true);
curl_setopt($ch, CURLOPT_HTTPHEADER, ['Content-Type: application/json']);
curl_setopt($ch, CURLOPT_POSTFIELDS, $payload);
$response = curl_exec($ch);
curl_close($ch);
$data = json_decode($response, true);
$end = microtime(true);

// ### Prepare Response
//$responseText = isset($data['response']) ? $data['response'] : 'No answer from Deepseek!';
//$responseText = $data['choices'][0]['message']['content'] ?? 'No answer from Deepseek!';
$responseText = isset($data['choices'][0]['message']['content']) ? $data['choices'][0]['message']['content'] : '';
//$responseText = preg_replace('/<think>.*?<\/think>/s', ' ', $responseText);
$responseText = trim($responseText);

echo json_encode([
    'reply' => $responseText,
//    'misc' => $htmlContext,
    'duration_seconds' => round($end - $start, 3)
]);