Skip to content

Commit

Permalink
Use fixtures for functional tests
Browse files Browse the repository at this point in the history
  • Loading branch information
@jtojnar
jtojnar committed Jan 8, 2022
1 parent 5928fee commit a2d0269
Show file tree
Hide file tree
Showing 10 changed files with 8,321 additions and 8 deletions.
35 changes: 27 additions & 8 deletions tests/GrabyFunctionalTest.php
View file
Original file line number Diff line number Diff line change
Expand Up @@ -3,6 +3,8 @@
namespace Tests\Graby;

use Graby\Graby;
use GuzzleHttp\Psr7\Response;
use Http\Mock\Client as HttpMockClient;
use Monolog\Handler\TestHandler;
use Monolog\Logger;
use PHPUnit\Framework\TestCase;
Expand All @@ -19,8 +21,10 @@ public function testRealFetchContent(): void
$logger = new Logger('foo');
$handler = new TestHandler($level = Logger::INFO);
$logger->pushHandler($handler);
$httpMockClient = new HttpMockClient();
$httpMockClient->addResponse(new Response(200, ['Connection' => ['keep-alive'], 'Content-Type' => ['text/html; charset=UTF-8'], 'X-Protected-By' => ['Sqreen'], 'Set-Cookie' => ['critical-article-free-desktop=19e41f325845da04f3af0e02a715e61a; expires=Mon, 07-Feb-2022 05:02:22 GMT; Max-Age=2592000; path=/; secure'], 'X-Frame-Options' => ['SAMEORIGIN'], 'X-XSS-Protection' => ['1; mode=block'], 'Via' => ['1.1 google, 1.1 varnish, 1.1 varnish'], 'Cache-Control' => ['private, max-age=0'], 'Accept-Ranges' => ['bytes'], 'Date' => ['Sat, 08 Jan 2022 05:02:22 GMT'], 'X-Served-By' => ['cache-cdg20729-CDG, cache-fra19154-FRA'], 'X-Cache' => ['MISS, MISS'], 'X-Cache-Hits' => ['0, 0'], 'X-Timer' => ['S1641618142.345590,VS0,VE99'], 'Vary' => ['Accept-Encoding'], 'Strict-Transport-Security' => ['max-age=31557600'], 'transfer-encoding' => ['chunked']], (string) file_get_contents(__DIR__ . '/fixtures/content/https___www.lemonde.fr_actualite-medias_article_2015_04_12_radio-france-vers-une-sortie-du-conflit_4614610_3236.html')));

$graby = new Graby(['debug' => true]);
$graby = new Graby(['debug' => true], $httpMockClient);
$graby->setLogger($logger);

$res = $graby->fetchContent('https://www.lemonde.fr/actualite-medias/article/2015/04/12/radio-france-vers-une-sortie-du-conflit_4614610_3236.html');
Expand Down Expand Up @@ -71,7 +75,9 @@ public function testRealFetchContent(): void

public function testRealFetchContent2(): void
{
$graby = new Graby(['debug' => true]);
$httpMockClient = new HttpMockClient();
$httpMockClient->addResponse(new Response(200, ['Content-Type' => ['text/html; charset=UTF-8'], 'Expires' => ['Sat, 08 Jan 2022 05:02:22 GMT'], 'Date' => ['Sat, 08 Jan 2022 05:02:22 GMT'], 'Cache-Control' => ['private, max-age=0'], 'Last-Modified' => ['Mon, 06 Dec 2021 04:36:10 GMT'], 'X-Content-Type-Options' => ['nosniff'], 'X-XSS-Protection' => ['1; mode=block'], 'Server' => ['GSE'], 'Alt-Svc' => ['h3=":443"; ma=2592000,h3-29=":443"; ma=2592000,h3-Q050=":443"; ma=2592000,h3-Q046=":443"; ma=2592000,h3-Q043=":443"; ma=2592000,quic=":443"; ma=2592000; v="46,43"'], 'Accept-Ranges' => ['none'], 'Vary' => ['Accept-Encoding'], 'Transfer-Encoding' => ['chunked']], (string) file_get_contents(__DIR__ . '/fixtures/content/https___bjori.blogspot.com_2015_04_next-gen-mongodb-driver.html')));
$graby = new Graby(['debug' => true], $httpMockClient);
$res = $graby->fetchContent('https://bjori.blogspot.com/2015/04/next-gen-mongodb-driver.html');

$this->assertCount(11, $res);
Expand Down Expand Up @@ -99,7 +105,9 @@ public function testRealFetchContent2(): void

public function testPdfFile(): void
{
$graby = new Graby(['debug' => true]);
$httpMockClient = new HttpMockClient();
$httpMockClient->addResponse(new Response(200, ['Server' => ['nginx/1.14.2'], 'Date' => ['Sat, 08 Jan 2022 05:02:23 GMT'], 'Content-Type' => ['application/pdf'], 'Content-Length' => ['414203'], 'Last-Modified' => ['Thu, 06 Mar 2008 11:04:29 GMT'], 'Connection' => ['keep-alive'], 'ETag' => ['"47cfcfbd-651fb"'], 'Accept-Ranges' => ['bytes']], (string) file_get_contents(__DIR__ . '/fixtures/content/http___img3.free.fr_im_tv_telesites_documentation.pdf')));
$graby = new Graby(['debug' => true], $httpMockClient);
$res = $graby->fetchContent('http://img3.free.fr/im_tv/telesites/documentation.pdf');

$this->assertCount(11, $res);
Expand Down Expand Up @@ -130,7 +138,9 @@ public function testPdfFile(): void

public function testImageFile(): void
{
$graby = new Graby(['debug' => true]);
$httpMockClient = new HttpMockClient();
$httpMockClient->addResponse(new Response(200, ['Connection' => ['keep-alive'], 'Content-Length' => ['158175'], 'Last-Modified' => ['Thu, 17 May 2018 20:44:59 GMT'], 'ETag' => ['"c2fb1185cf396855391e593d56c135fb"'], 'x-amz-storage-class' => ['STANDARD_IA'], 'Content-Type' => ['image/jpeg'], 'cache-control' => ['public, max-age=31536000'], 'Accept-Ranges' => ['bytes'], 'Date' => ['Sat, 08 Jan 2022 05:02:23 GMT'], 'Age' => ['4746941'], 'X-Served-By' => ['cache-bwi5144-BWI, cache-iad-kcgs7200076-IAD, cache-fra19158-FRA'], 'X-Cache' => ['HIT, HIT, HIT'], 'X-Cache-Hits' => ['1, 1, 1'], 'X-Timer' => ['S1641618144.910034,VS0,VE1'], 'Strict-Transport-Security' => ['max-age=300'], 'Access-Control-Allow-Methods' => ['GET, OPTIONS'], 'Access-Control-Allow-Origin' => ['*'], 'Server' => ['cat factory 1.0'], 'X-Content-Type-Options' => ['nosniff']], (string) file_get_contents(__DIR__ . '/fixtures/content/https___i.imgur.com_KQQ7D9z.jpg')));
$graby = new Graby(['debug' => true], $httpMockClient);
$res = $graby->fetchContent('https://i.imgur.com/KQQ7D9z.jpg');

$this->assertCount(11, $res);
Expand Down Expand Up @@ -201,7 +211,9 @@ public function testAccentuedUrls(string $url): void
*/
public function testYoutubeOembed(): void
{
$graby = new Graby(['debug' => true]);
$httpMockClient = new HttpMockClient();
$httpMockClient->addResponse(new Response(200, ['Content-Type' => ['application/xml'], 'Vary' => ['X-Origin', 'Referer', 'Origin,Accept-Encoding'], 'Date' => ['Sat, 08 Jan 2022 05:02:24 GMT'], 'Server' => ['scaffolding on HTTPServer2'], 'Cache-Control' => ['private'], 'X-XSS-Protection' => ['0'], 'X-Frame-Options' => ['SAMEORIGIN'], 'X-Content-Type-Options' => ['nosniff'], 'Alt-Svc' => ['h3=":443"; ma=2592000,h3-29=":443"; ma=2592000,h3-Q050=":443"; ma=2592000,h3-Q046=":443"; ma=2592000,h3-Q043=":443"; ma=2592000,quic=":443"; ma=2592000; v="46,43"'], 'Accept-Ranges' => ['none'], 'Transfer-Encoding' => ['chunked']], (string) file_get_contents(__DIR__ . '/fixtures/content/https___www.youtube.com_oembed_url_https___www.youtube.com_watch_v_td0P8qrS8iI_format_xml.html')));
$graby = new Graby(['debug' => true], $httpMockClient);
$res = $graby->fetchContent('https://www.youtube.com/oembed?url=https://www.youtube.com/watch?v=td0P8qrS8iI&format=xml');

$this->assertCount(11, $res);
Expand Down Expand Up @@ -254,7 +266,9 @@ public function testEncodedUrl(): void

public function testKoreanPage(): void
{
$graby = new Graby(['debug' => true]);
$httpMockClient = new HttpMockClient();
$httpMockClient->addResponse(new Response(200, ['Date' => ['Sat, 08 Jan 2022 05:02:26 GMT'], 'Server' => ['Apache/2.4.34 (Unix) OpenSSL/1.0.1e-fips PHP/5.6.30'], 'X-Powered-By' => ['PHP/5.6.30'], 'Set-Cookie' => ['PHPSESSID=atrh7ouhppo4ofdhbe8pd4dbf2; path=/'], 'Vary' => ['Accept-Encoding'], 'Connection' => ['close'], 'Transfer-Encoding' => ['chunked'], 'Content-Type' => ['text/html; charset=UTF-8']], (string) file_get_contents(__DIR__ . '/fixtures/content/http___www.newstown.co.kr_news_articleView.html_idxno_243722.html')));
$graby = new Graby(['debug' => true], $httpMockClient);
$res = $graby->fetchContent('http://www.newstown.co.kr/news/articleView.html?idxno=243722');

$this->assertCount(11, $res);
Expand All @@ -279,14 +293,16 @@ public function testKoreanPage(): void

public function testMultipage(): void
{
$httpMockClient = new HttpMockClient();
$httpMockClient->addResponse(new Response(200, ['content-type' => ['text/html; charset=UTF-8'], 'vary' => ['Accept-Encoding'], 'cache-control' => ['max-age=3600, public'], 'date' => ['Sat, 08 Jan 2022 04:20:20 GMT'], 'x-frame-options' => ['deny'], 'x-content-type-options' => ['nosniff'], 'referrer-policy' => ['origin'], 'etag' => ['W/"59797fc71df1c3e896ee2231e6ea263b"'], 'x-via-popn' => ['front01'], 'strict-transport-security' => ['max-age=63072000; includeSubDomains; preload'], 'x-cacheable' => ['yes'], 'x-varnish' => ['594842719 593544597'], 'age' => ['2528'], 'x-via-popv' => ['front02'], 'x-cache' => ['HIT'], 'accept-ranges' => ['bytes'], 'content-length' => ['252912'], 'x-via-poph' => ['front03']], (string) file_get_contents(__DIR__ . '/fixtures/content/https___www.clubic.com_carte-graphique_carte-graphique-amd_article-478936-1-radeon-hd-7750-7770.html')));
$graby = new Graby([
'debug' => true,
'extractor' => [
'config_builder' => [
'site_config' => [__DIR__ . '/fixtures/site_config'],
],
],
]);
], $httpMockClient);
$res = $graby->fetchContent('https://www.clubic.com/carte-graphique/carte-graphique-amd/article-478936-1-radeon-hd-7750-7770.html');

$this->assertCount(11, $res);
Expand All @@ -312,14 +328,17 @@ public function testMultipage(): void

public function testCookie(): void
{
$httpMockClient = new HttpMockClient();
$httpMockClient->addResponse(new Response(301, ['Server' => ['AkamaiGHost'], 'Content-Length' => ['0'], 'Location' => ['https://www.npr.org/sections/parallels/2017/05/19/529148729/michael-flynns-contradictory-line-on-russia'], 'Date' => ['Sat, 08 Jan 2022 05:02:29 GMT'], 'Connection' => ['keep-alive']], (string) file_get_contents(__DIR__ . '/fixtures/content/http___www.npr.org_sections_parallels_2017_05_19_529148729_michael-flynns-contradictory-line-on-russia.html')));
$httpMockClient->addResponse(new Response(200, ['Content-Type' => ['text/html;charset=UTF-8'], 'X-Cache-NPR' => ['MISS'], 'Access-Control-Allow-Origin' => ['http://www.google.co.uk'], 'Access-Control-Allow-Credentials' => ['true'], 'X-NPR-Trace-ID' => ['renIaLEEV1Q'], 'X-Content-Type-Options' => ['nosniff'], 'X-XSS-Protection' => ['1; mode=block'], 'X-Served-By' => ['pod-www-render-nginx-776fbf4c8c-7sjbx'], 'Referrer-Policy' => ['no-referrer-when-downgrade'], 'Strict-Transport-Security' => ['max-age=15724800; includeSubDomains'], 'Cache-Control' => ['no-cache'], 'Expires' => ['Sat, 08 Jan 2022 05:02:30 GMT'], 'Date' => ['Sat, 08 Jan 2022 05:02:30 GMT'], 'Transfer-Encoding' => ['chunked'], 'Connection' => ['keep-alive', 'Transfer-Encoding']], (string) file_get_contents(__DIR__ . '/fixtures/content/http___www.npr.org_sections_parallels_2017_05_19_529148729_michael-flynns-contradictory-line-on-russia')));
$graby = new Graby([
'debug' => true,
'extractor' => [
'config_builder' => [
'site_config' => [__DIR__ . '/fixtures/site_config'],
],
],
]);
], $httpMockClient);
$res = $graby->fetchContent('http://www.npr.org/sections/parallels/2017/05/19/529148729/michael-flynns-contradictory-line-on-russia');

$this->assertCount(11, $res);
Expand Down
Binary file added tests/fixtures/content/http___img3.free.fr_im_tv_telesites_documentation.pdf
View file
Binary file not shown.

0 comments on commit a2d0269

Please sign in to comment.