lvl0/fedi-feed-router
1
0
Fork 0
Code Issues 27 Pull requests Projects Releases 13 Packages Wiki Activity Actions

115 - Add Belga JSON discovery parser for press-release API

Browse source
This commit is contained in:
myrmidex 2026-08-01 13:43:50 +02:00
parent 00366ffeaa
commit 2a4d4ad84d
5 changed files with 165 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

1
.gitignore vendored
View file

@ -24,3 +24,4 @@ yarn-error.log
/coverage-report*
/coverage.xml
/.php-cs-fixer.dist.php
/.php-cs-fixer.cache

25
app/Services/Parsers/BelgaHomepageParser.php Normal file
View file

@ -0,0 +1,25 @@
<?php
namespace App\Services\Parsers;
class BelgaHomepageParser
{
private const ARTICLE_URL_TEMPLATE = 'https://www.belganewsagency.eu/press-releases/%s/';
public static function extractArticleUrls(string $json): array
{
$decoded = json_decode($json, true);
if (! is_array($decoded) || ! isset($decoded['data']) || ! is_array($decoded['data'])) {
return [];
}
return collect($decoded['data'])
->pluck('id')
->filter(fn($id) => is_int($id) || (is_string($id) && ctype_digit($id)))
->map(fn($id) => sprintf(self::ARTICLE_URL_TEMPLATE, $id))
->unique()
->values()
->toArray();
}
}

34
app/Services/Parsers/BelgaHomepageParserAdapter.php Normal file
View file

@ -0,0 +1,34 @@
<?php
namespace App\Services\Parsers;
use App\Contracts\HomepageParserInterface;
class BelgaHomepageParserAdapter implements HomepageParserInterface
{
private const API_URL = 'https://capi.belga.press/belgapress/api/public/pressreleases?offset=1&count=6&search=&start=&end=&newsroomId=70&language=EN';
public function __construct(
private string $language = 'en',
) {}
public function canParse(string $url): bool
{
return str_contains($url, 'belganewsagency.eu') || str_contains($url, 'capi.belga.press');
}
public function extractArticleUrls(string $html): array
{
return BelgaHomepageParser::extractArticleUrls($html);
}
public function getHomepageUrl(): string
{
return str_replace('language=EN', 'language=' . strtoupper($this->language), self::API_URL);
}
public function getSourceName(): string
{
return 'Belga News Agency';
}
}

1
tests/Fixtures/belga-pressreleases.json Normal file
View file

File diff suppressed because one or more lines are too long

104
tests/Unit/Services/Parsers/BelgaHomepageParserTest.php Normal file
View file

@ -0,0 +1,104 @@
<?php
namespace Tests\Unit\Services\Parsers;
use App\Services\Parsers\BelgaHomepageParser;
use App\Services\Parsers\BelgaHomepageParserAdapter;
use PHPUnit\Framework\TestCase;
class BelgaHomepageParserTest extends TestCase
{
private function fixture(): string
{
return file_get_contents(__DIR__ . '/../../../Fixtures/belga-pressreleases.json');
}
public function test_extracts_article_urls_from_api_json(): void
{
$urls = BelgaHomepageParser::extractArticleUrls($this->fixture());
$this->assertCount(6, $urls);
$this->assertSame('https://www.belganewsagency.eu/press-releases/35285/', $urls[0]);
$this->assertContains('https://www.belganewsagency.eu/press-releases/35269/', $urls);
}
public function test_deduplicates_urls(): void
{
$json = json_encode(['data' => [
['id' => 100],
['id' => 100],
['id' => '100'],
['id' => 101],
]]);
$urls = BelgaHomepageParser::extractArticleUrls($json);
$this->assertSame([
'https://www.belganewsagency.eu/press-releases/100/',
'https://www.belganewsagency.eu/press-releases/101/',
], $urls);
}
public function test_returns_empty_array_for_malformed_json(): void
{
$this->assertSame([], BelgaHomepageParser::extractArticleUrls('<html>404 not found</html>'));
$this->assertSame([], BelgaHomepageParser::extractArticleUrls('{"data": ['));
$this->assertSame([], BelgaHomepageParser::extractArticleUrls(''));
}
public function test_returns_empty_array_when_data_key_missing_or_empty(): void
{
$this->assertSame([], BelgaHomepageParser::extractArticleUrls('{}'));
$this->assertSame([], BelgaHomepageParser::extractArticleUrls('{"data": []}'));
$this->assertSame([], BelgaHomepageParser::extractArticleUrls('{"data": "not-an-array"}'));
}
public function test_skips_entries_without_usable_id(): void
{
$json = json_encode(['data' => [
['id' => 200],
['id' => null],
['publishDate' => '2026-08-01T06:24:19'],
['id' => 'not-numeric'],
['id' => 201],
]]);
$urls = BelgaHomepageParser::extractArticleUrls($json);
$this->assertSame([
'https://www.belganewsagency.eu/press-releases/200/',
'https://www.belganewsagency.eu/press-releases/201/',
], $urls);
}
public function test_adapter_delegates_and_reports_source(): void
{
$adapter = new BelgaHomepageParserAdapter('en');
$this->assertSame(
BelgaHomepageParser::extractArticleUrls($this->fixture()),
$adapter->extractArticleUrls($this->fixture())
);
$this->assertTrue($adapter->canParse('https://www.belganewsagency.eu/press-releases/35285/'));
$this->assertTrue($adapter->canParse('https://capi.belga.press/belgapress/api/public/pressreleases'));
$this->assertFalse($adapter->canParse('https://www.vrt.be/vrtnws/en/'));
$this->assertSame('Belga News Agency', $adapter->getSourceName());
}
public function test_adapter_uppercases_language_in_api_url(): void
{
$this->assertStringContainsString('language=EN', (new BelgaHomepageParserAdapter('en'))->getHomepageUrl());
$this->assertStringContainsString('language=NL', (new BelgaHomepageParserAdapter('nl'))->getHomepageUrl());
}
public function test_adapter_api_url_pins_load_bearing_query_params(): void
{
$url = (new BelgaHomepageParserAdapter('en'))->getHomepageUrl();
$this->assertStringStartsWith('https://capi.belga.press/belgapress/api/public/pressreleases?', $url);
$this->assertStringContainsString('newsroomId=70', $url);
$this->assertStringContainsString('offset=1', $url);
$this->assertStringContainsString('count=6', $url);
}
}