> ## Documentation Index
> Fetch the complete documentation index at: https://docs.instructorphp.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Web to objects

## Overview

This example demonstrates how to extract structured data from a web page and get
it as PHP object.

## Example

In this example we will be extracting list of Laravel companies from The Manifest
website. The result will be a list of `Company` objects.

We use Webpage extractor to get the content of the page and specify 'none' scraper,
which means that we will be using built-in `file_get_contents` function to get the
content of the page.

In production environment you might want to use one of the supported scrapers:

* `browsershot`
* `scrapingbee`
* `scrapfly`
* `jinareader`

Commercial scrapers require API key, which can be set in the auxiliary web
configuration files (`/packages/auxiliary/config/web/default.yaml` and
`/packages/auxiliary/config/web/scrapers/*.yaml`).

```php theme={null}
<?php
require 'examples/boot.php';

use Cognesy\Auxiliary\Web\Webpage;
use Cognesy\Config\BasePath;
use Cognesy\Instructor\StructuredOutput;
use Cognesy\Instructor\StructuredOutputRuntime;
use Cognesy\Instructor\Enums\OutputMode;
use Cognesy\Polyglot\Inference\LLMProvider;
use Cognesy\Schema\Attributes\Instructions;
class Company {
    public string $name = '';
    public string $location = '';
    public string $description = '';
    public int $minProjectBudget = 0;
    public string $companySize = '';
    #[Instructions('Remove any tracking parameters from the URL')]
    public string $websiteUrl = '';
    /** @var string[] */
    public array $clients = [];
}

$companyGen = Webpage::withScraper('none')
    ->get(BasePath::get('examples/A05_Extras/WebToObjects/companies.html'))
    ->select('.directory-providers__list')
    ->selectMany(
        selector: '.provider-card',
        callback: fn($item) => $item->asMarkdown(),
        limit: 3
    );

$companies = [];
echo "Extracting company data from:\n\n";
foreach($companyGen as $companyDiv) {
    /** @var string $companyDiv */
    echo " > " . substr($companyDiv, 0, 32) . "...\n\n";
    $company = new StructuredOutput(
            StructuredOutputRuntime::fromProvider(LLMProvider::using('openai'))
                ->withOutputMode(OutputMode::Json)
        )
        ->with(
            messages: $companyDiv,
            responseModel: Company::class,
        )->get();
    $companies[] = $company;
    dump($company);
}

assert(count($companies) === 3);
?>
```