如何通过PHP实现PDF在Solr中的自动索引与文件名检索？

Hey there! Let's get your PHP + Solr PDF indexing and search workflow sorted out. I’ve tackled this exact setup before, so here’s a step-by-step guide that should solve your issues:

Solr relies on Apache Tika to extract content from PDFs, so first we need to make sure that’s configured properly:

• Enable the Extract Request Handler: Open your Solr core’s solrconfig.xml and add (or verify) this configuration block:

  <requestHandler name="/update/extract" class="org.apache.solr.handler.extraction.ExtractingRequestHandler">
    <lst name="defaults">
      <str name="lowernames">true</str>
      <str name="uprefix">ignored_</str>
      <str name="fmap.content">content</str>
      <str name="fmap.filename">filename</str>
    </lst>
  </requestHandler>
  

  This maps the extracted PDF content to a content field and the filename to a filename field in your Solr schema.
• Configure Your Schema: In your core’s managed-schema (or schema.xml), add these fields if they don’t exist:

  <field name="id" type="string" indexed="true" stored="true" required="true" multiValued="false"/>
  <field name="filename" type="text_general" indexed="true" stored="true"/>
  <field name="content" type="text_general" indexed="true" stored="true"/>
  

  The text_general type lets you search both full content and filenames with partial matches; if you want exact filename searches only, use string instead.
• Check Tika JARs: Solr 7.x and above usually include Tika by default, but if you’re on an older version, make sure the Apache Tika JARs are in your Solr lib directory.

Next, we’ll handle the PHP upload and send the PDF directly to Solr for indexing. I recommend using the Solarium library (a popular Solr PHP client) for this—it’s way easier than raw cURL calls.

First, Install Solarium #

Use Composer to install it:

composer require solarium/solarium

PHP Upload & Index Code Example #

Here’s a complete snippet that handles file uploads, validates PDFs, and indexes them to Solr:

<?php
require 'vendor/autoload.php';

// Initialize Solarium client
$client = new Solarium\Client([
    'endpoint' => [
        'localhost' => [
            'host' => '127.0.0.1',
            'port' => 8983,
            'path' => '/solr/',
            'core' => 'your_core_name' // Replace with your Solr core name
        ]
    ]
]);

// Handle file upload
if ($_SERVER['REQUEST_METHOD'] === 'POST' && isset($_FILES['pdf_file'])) {
    $file = $_FILES['pdf_file'];
    
    // Validate it's a PDF
    if ($file['type'] !== 'application/pdf' || pathinfo($file['name'], PATHINFO_EXTENSION) !== 'pdf') {
        die('Please upload a valid PDF file.');
    }
    
    // Create an update query with extract handler
    $update = $client->createUpdate();
    $extractQuery = $update->createExtract();
    
    // Set the file to extract
    $extractQuery->setFile($file['tmp_name']);
    // Map filename to our Solr field
    $extractQuery->setParam('resource.name', $file['name']);
    // Don't just extract—index the content
    $extractQuery->setParam('extractOnly', 'false');
    // Auto-commit the index so changes are immediately available
    $extractQuery->setParam('commit', 'true');
    
    // Add the extract query to the update and execute
    $update->addExtract($extractQuery);
    $result = $client->update($update);
    
    echo "PDF uploaded and indexed successfully!";
}
?>

<!-- HTML Upload Form -->
<form method="POST" enctype="multipart/form-data">
    <input type="file" name="pdf_file" accept=".pdf" required>
    <button type="submit">Upload PDF</button>
</form>

If you don’t want to use Solarium, you can use raw cURL instead—here’s a quick snippet for that:

<?php
if ($_SERVER['REQUEST_METHOD'] === 'POST' && isset($_FILES['pdf_file'])) {
    $file = $_FILES['pdf_file'];
    $solrUrl = 'http://127.0.0.1:8983/solr/your_core_name/update/extract?commit=true&resource.name=' . urlencode($file['name']);
    
    $ch = curl_init($solrUrl);
    curl_setopt($ch, CURLOPT_POST, true);
    curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
    curl_setopt($ch, CURLOPT_POSTFIELDS, [
        'myfile' => new CURLFile($file['tmp_name'], $file['type'], $file['name'])
    ]);
    
    $response = curl_exec($ch);
    curl_close($ch);
    
    echo "PDF indexed: " . $response;
}
?>

Now let’s build the search functionality that lets users query text and get back matching PDF filenames. Again, using Solarium makes this straightforward:

<?php
require 'vendor/autoload.php';

// Initialize Solarium client (same as before)
$client = new Solarium\Client([
    'endpoint' => [
        'localhost' => [
            'host' => '127.0.0.1',
            'port' => 8983,
            'path' => '/solr/',
            'core' => 'your_core_name'
        ]
    ]
]);

// Handle search request
$searchQuery = $_GET['q'] ?? '';
if (!empty($searchQuery)) {
    // Create a select query
    $select = $client->createSelect();
    $query = $select->createQuery();
    
    // Search both content and filename fields
    $query->setQuery("content:$searchQuery OR filename:$searchQuery");
    // Only return the filename field (plus id if you need it)
    $select->setFields(['id', 'filename']);
    
    // Execute the query
    $resultset = $client->select($select);
    
    // Display results
    echo "Found " . $resultset->getNumFound() . " results:\n";
    foreach ($resultset as $document) {
        echo "- " . $document->filename . "\n";
    }
}
?>

<!-- HTML Search Form -->
<form method="GET">
    <input type="text" name="q" placeholder="Search PDF content or filename..." required>
    <button type="submit">Search</button>
</form>

• PDF content not indexing? Double-check that your Solr core’s solrconfig.xml has the Extract Request Handler enabled, and that Tika is properly set up. Also, make sure you’re setting extractOnly=false in your index request.
• Filename search not working? If you used string type for the filename field, you’ll need exact matches. Switch to text_general if you want partial matches (like searching for "report" and getting "annual_report.pdf").
• Changes not showing up in search? Always include commit=true in your index request, or set up auto-commit in Solr’s solrconfig.xml (not recommended for production, but fine for testing).

内容的提问来源于stack exchange，提问作者amit rawat