Skip to content

Commit 742e57c

Browse files
jbtronicsjbtronics
committed
Added basic provider for pollin
1 parent a596166 commit 742e57c

File tree

2 files changed

+239
-1
lines changed

2 files changed

+239
-1
lines changed

‎.env‎

Lines changed: 6 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -220,7 +220,6 @@ PROVIDER_OEMSECRETS_SORT_CRITERIA=C
220220
# Reichelt provider:
221221
# Reichelt.com offers no official API, so this info provider webscrapes the website to extract info
222222
# It could break at any time, use it at your own risk
223-
224223
# We dont require an API key for Reichelt, just set this to 1 to enable Reichelt support
225224
PROVIDER_REICHELT_ENABLED=0
226225
# The country to get prices for
@@ -232,6 +231,12 @@ PROVIDER_REICHELT_INCLUDE_VAT=1
232231
# The currency to get prices in (only for countries with countries other than EUR)
233232
PROVIDER_REICHELT_CURRENCY=EUR
234233

234+
# Pollin provider:
235+
# Pollin.de offers no official API, so this info provider webscrapes the website to extract info
236+
# It could break at any time, use it at your own risk
237+
# We dont require an API key for Pollin, just set this to 1 to enable Pollin support
238+
PROVIDER_POLLIN_ENABLED=0
239+
235240
##################################################################################
236241
# EDA integration related settings
237242
##################################################################################

‎src/Services/InfoProviderSystem/Providers/PollinProvider.php‎

Lines changed: 233 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,233 @@
1+
<?php
2+
/*
3+
* This file is part of Part-DB (https://github.com/Part-DB/Part-DB-symfony).
4+
*
5+
* Copyright (C) 2019 - 2025 Jan Böhmer (https://github.com/jbtronics)
6+
*
7+
* This program is free software: you can redistribute it and/or modify
8+
* it under the terms of the GNU Affero General Public License as published
9+
* by the Free Software Foundation, either version 3 of the License, or
10+
* (at your option) any later version.
11+
*
12+
* This program is distributed in the hope that it will be useful,
13+
* but WITHOUT ANY WARRANTY; without even the implied warranty of
14+
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
15+
* GNU Affero General Public License for more details.
16+
*
17+
* You should have received a copy of the GNU Affero General Public License
18+
* along with this program. If not, see <https://www.gnu.org/licenses/>.
19+
*/
20+
21+
declare(strict_types=1);
22+
23+
24+
namespace App\Services\InfoProviderSystem\Providers;
25+
26+
use App\Entity\Parts\ManufacturingStatus;
27+
use App\Entity\Parts\Part;
28+
use App\Services\InfoProviderSystem\DTOs\FileDTO;
29+
use App\Services\InfoProviderSystem\DTOs\ParameterDTO;
30+
use App\Services\InfoProviderSystem\DTOs\PartDetailDTO;
31+
use App\Services\InfoProviderSystem\DTOs\PriceDTO;
32+
use App\Services\InfoProviderSystem\DTOs\PurchaseInfoDTO;
33+
use App\Services\InfoProviderSystem\DTOs\SearchResultDTO;
34+
use Symfony\Component\DependencyInjection\Attribute\Autowire;
35+
use Symfony\Component\DomCrawler\Crawler;
36+
use Symfony\Contracts\HttpClient\HttpClientInterface;
37+
38+
class PollinProvider implements InfoProviderInterface
39+
{
40+
41+
public function __construct(private readonly HttpClientInterface $client,
42+
#[Autowire(env: 'bool:PROVIDER_POLLIN_ENABLED')]
43+
private readonly bool $enabled = true,
44+
)
45+
{
46+
}
47+
48+
public function getProviderInfo(): array
49+
{
50+
return [
51+
'name' => 'Pollin',
52+
'description' => 'Webscrapping from pollin.de to get part information',
53+
'url' => 'https://www.reichelt.de/',
54+
'disabled_help' => 'Set PROVIDER_POLLIN_ENABLED env to 1'
55+
];
56+
}
57+
58+
public function getProviderKey(): string
59+
{
60+
return 'pollin';
61+
}
62+
63+
public function isActive(): bool
64+
{
65+
return $this->enabled;
66+
}
67+
68+
public function searchByKeyword(string $keyword): array
69+
{
70+
$response = $this->client->request('GET', 'https://www.pollin.de/search', [
71+
'query' => [
72+
'search' => $keyword
73+
]
74+
]);
75+
76+
$content = $response->getContent();
77+
$dom = new Crawler($content);
78+
79+
$results = [];
80+
81+
//Iterate over each div.product-box
82+
$dom->filter('div.product-box')->each(function (Crawler $node) use (&$results) {
83+
$results[] = new SearchResultDTO(
84+
provider_key: $this->getProviderKey(),
85+
provider_id: $node->filter('meta[itemprop="productID"]')->attr('content'),
86+
name: $node->filter('a.product-name')->text(),
87+
description: '',
88+
preview_image_url: $node->filter('img.product-image')->attr('src'),
89+
manufacturing_status: $this->mapAvailability($node->filter('link[itemprop="availability"]')->attr('href')),
90+
provider_url: $node->filter('a.product-name')->attr('href')
91+
);
92+
});
93+
94+
return $results;
95+
}
96+
97+
private function mapAvailability(string $availabilityURI): ManufacturingStatus
98+
{
99+
return match( $availabilityURI) {
100+
'http://schema.org/InStock' => ManufacturingStatus::ACTIVE,
101+
'http://schema.org/OutOfStock' => ManufacturingStatus::DISCONTINUED,
102+
default => ManufacturingStatus::NOT_SET
103+
};
104+
}
105+
106+
public function getDetails(string $id): PartDetailDTO
107+
{
108+
//Ensure that $id is numeric
109+
if (!is_numeric($id)) {
110+
throw new \InvalidArgumentException("The id must be numeric!");
111+
}
112+
113+
$response = $this->client->request('GET', 'https://www.pollin.de/search', [
114+
'query' => [
115+
'search' => $id
116+
]
117+
]);
118+
119+
//The response must have us redirected to the product page
120+
if ($response->getInfo('redirect_count') > 0) {
121+
throw new \RuntimeException("Could not resolve the product page for the given id!");
122+
}
123+
124+
$content = $response->getContent();
125+
126+
return $this->parseProductPage($content);
127+
}
128+
129+
private function parseProductPage(string $content): PartDetailDTO
130+
{
131+
$dom = new Crawler($content);
132+
133+
$productPageUrl = $dom->filter('meta[property="product:product_link"]')->attr('content');
134+
$orderId = trim($dom->filter('span[itemprop="sku"]')->text()); //Text is important here
135+
136+
//Calculate the mass
137+
$massStr = $dom->filter('meta[itemprop="weight"]')->attr('content');
138+
//Remove the unit
139+
$massStr = str_replace('kg', '', $massStr);
140+
//Convert to float and convert to grams
141+
$mass = (float) $massStr * 1000;
142+
143+
//Parse purchase info
144+
$purchaseInfo = new PurchaseInfoDTO(
145+
'Pollin',
146+
$orderId,
147+
[
148+
new PriceDTO(1, $dom->filter('meta[property="product:price:amount"]')->attr('content'), $dom->filter('meta[property="product:price:currency"]')->attr('content'))
149+
],
150+
$productPageUrl
151+
);
152+
153+
154+
155+
return new PartDetailDTO(
156+
provider_key: $this->getProviderKey(),
157+
provider_id: $orderId,
158+
name: trim($dom->filter('meta[property="og:title"]')->attr('content')),
159+
description: $dom->filter('meta[property="og:description"]')->attr('content'),
160+
category: $this->parseCategory($dom),
161+
manufacturer: $dom->filter('meta[property="product:brand"]')->count() > 0 ? $dom->filter('meta[property="product:brand"]')->attr('content') : null,
162+
preview_image_url: $dom->filter('meta[property="og:image"]')->attr('content'),
163+
manufacturing_status: $this->mapAvailability($dom->filter('link[itemprop="availability"]')->attr('href')),
164+
provider_url: $productPageUrl,
165+
notes: $this->parseNotes($dom),
166+
datasheets: $this->parseDatasheets($dom),
167+
parameters: $this->parseParameters($dom),
168+
vendor_infos: [$purchaseInfo],
169+
mass: $mass,
170+
);
171+
}
172+
173+
private function parseDatasheets(Crawler $dom): array
174+
{
175+
//Iterate over each a element withing div.pol-product-detail-download-files
176+
$datasheets = [];
177+
$dom->filter('div.pol-product-detail-download-files a')->each(function (Crawler $node) use (&$datasheets) {
178+
$datasheets[] = new FileDTO($node->attr('href'), $node->text());
179+
});
180+
181+
return $datasheets;
182+
}
183+
184+
private function parseParameters(Crawler $dom): array
185+
{
186+
$parameters = [];
187+
188+
//Iterate over each tr.properties-row inside table.product-detail-properties-table
189+
$dom->filter('table.product-detail-properties-table tr.properties-row')->each(function (Crawler $node) use (&$parameters) {
190+
$parameters[] = ParameterDTO::parseValueField(
191+
name: rtrim($node->filter('th.properties-label')->text(), ':'),
192+
value: $node->filter('td.properties-value')->text()
193+
);
194+
});
195+
196+
return $parameters;
197+
}
198+
199+
private function parseCategory(Crawler $dom): string
200+
{
201+
$category = '';
202+
203+
//Iterate over each li.breadcrumb-item inside ol.breadcrumb
204+
$dom->filter('ol.breadcrumb li.breadcrumb-item')->each(function (Crawler $node) use (&$category) {
205+
//Skip if it has breadcrumb-item-home class
206+
if (str_contains($node->attr('class'), 'breadcrumb-item-home')) {
207+
return;
208+
}
209+
210+
211+
$category .= $node->text() . ' -> ';
212+
});
213+
214+
//Remove the last ' -> '
215+
return substr($category, 0, -4);
216+
}
217+
218+
private function parseNotes(Crawler $dom): string
219+
{
220+
//Concat product highlights and product description
221+
return $dom->filter('div.product-detail-top-features')->html() . '<br><br>' . $dom->filter('div.product-detail-description-text')->html();
222+
}
223+
224+
public function getCapabilities(): array
225+
{
226+
return [
227+
ProviderCapabilities::BASIC,
228+
ProviderCapabilities::PICTURE,
229+
ProviderCapabilities::PRICE,
230+
ProviderCapabilities::DATASHEET
231+
];
232+
}
233+
}

0 commit comments

Comments
 (0)