reedgraff/hingesdkSDK de l'application de rencontres Hinge pour python pour les ingénieurs logiciels solitaires et ennuyés.
SDK Python non officiel pour l'API Hinge. Interagissez par programmation avec Hinge pour récupérer des recommandations, envoyer des messages, télécharger des médias et automatiser des interactions.
Pour utiliser ce SDK, vous avez besoin d'informations d'authentification valides. Vous pouvez les stocker dans un fichier .env ou les transmettre directement au client.
BEARER_TOKEN : Votre jeton d'authentification JWT (par ex., L2euNWN...).SESSION_ID : L'UUID de la session actuelle.USER_ID : Votre identifiant unique d'utilisateur/joueur.Si vous souhaitez utiliser votre compte Hinge existant sans risquer d'être signalé par les connexions depuis un nouvel appareil, vous devez extraire ces valeurs du trafic réseau de l'application.
Remarque : Vous n'avez besoin de le faire qu'une seule fois pour obtenir les bonnes variables. Une fois celles-ci capturées, vous pouvez utiliser la méthode de connexion par SMS pour les sessions futures.
prod-api.hingeaws.net ou des points de terminaison Hinge similaires.authorization (jeton Bearer)x-session-idx-device-idx-install-idUser ID (souvent dans le corps de la réponse sous subjectId ou similaire)Vous pouvez générer de nouveaux identifiants en vous connectant via SMS par l'intermédiaire du SDK. Cela effectue un nouveau flux de connexion.
from hingesdk.client import HingeClient
# These IDs must match the actual values associated with your account.
# You cannot use random UUIDs; they must align with the account's registered device.
client = HingeClient.login_with_sms(
phone_number="+15551234567",
device_id="your_device_id_uuid",
install_id="your_install_id_uuid"
)
print(f"BEARER_TOKEN={client.auth_token}")
print(f"SESSION_ID={client.session_id}")
print(f"USER_ID={client.user_id}")
Clonez le dépôt et installez le package :
git clone https://github.com/reedgraff/hingesdk
cd hingesdk
pip install .
Voici un exemple minimal pour s'authentifier et récupérer des recommandations d'utilisateurs.
import os
from hingesdk.api import HingeAPIClient
# Initialize with credentials
client = HingeAPIClient(
auth_token=os.getenv("BEARER_TOKEN"),
session_id=os.getenv("SESSION_ID"),
user_id=os.getenv("USER_ID")
)
# Fetch recommendations
recs = client.get_recommendations()
print(f"Successfully fetched recommendations request.")
from hingesdk.client import HingeAPIClient
auth_token = 'your_auth_token'
user_id = 'your_user_id'
client = HingeAPIClient(auth_token=auth_token, user_id=user_id)
response = client.send_message(
subject_id='receiver_id',
message='Hello, this is a test message!'
)
print(response)
from hingesdk.api import HingeAPIClient
client = HingeAPIClient(auth_token=auth_token, user_id=user_id)
recommendations = client.get_recommendations()
print(recommendations)
from hingesdk.tools import HingeTools
from hingesdk.api import HingeAPIClient
from hingesdk.media import HingeMediaClient
api_client = HingeAPIClient(auth_token=auth_token, user_id=user_id)
media_client = HingeMediaClient(auth_token=auth_token)
tools = HingeTools(api_client, media_client)
tools.download_recommendation_content(output_path='path_to_save_images')
import os
import json
from hingesdk.tools import HingeTools
from hingesdk.api import HingeAPIClient
from hingesdk.media import HingeMediaClient
auth_token = os.getenv("BEARER_TOKEN")
session_id = os.getenv("SESSION_ID")
user_id = os.getenv("USER_ID")
api_client = HingeAPIClient(
auth_token=auth_token,
session_id=session_id,
user_id=user_id
)
media_client = HingeMediaClient(auth_token=auth_token)
tools = HingeTools(api_client, media_client)
# Example: Get user info
tools.create_profile_json(
source=ProfileSource.STANDOUTS,
output_file="standouts.json"
)
# Example: Like a user
with open("standouts.json", "r") as f:
profiles = json.load(f)
personData = profiles["35582109789..."]
# Like a user's question
questionIDToLike = "5c4a346828fd883a24..."
response = api_client.like_profile(
subject_id=personData["interaction_data"]["subject_id"],
rating_token=personData["interaction_data"]["rating_token"],
prompt={
"questionId": questionIDToLike,
"response": "I've been in Miami for a year and still haven't gone (╥﹏╥)"
}
)
print(response)
# Like a user's photo
# photoIDToLike = "2c6411ac-66e4-4194-..."
# response = api_client.like_profile(
# subject_id=personData["interaction_data"]["subject_id"],
# rating_token=personData["interaction_data"]["rating_token"],
# photo={
# "contentId": photoIDToLike,
# "comment": "So how many people have commented saying they've been here before?"
# }
# )
# print(response)
Cet exemple montre comment extraire des recommandations et filtrer les profils en fonction de critères spécifiques (par exemple, les meilleures universités).
def find_top_50_university_students(json_file_path, age_min=18, age_max=24):
# List to store matching profiles
matching_profiles = []
# Dictionary of patterns for the top 50 universities (case insensitive)
top_50_patterns = {
"Princeton University": [r"princeton", r"\bpu\b", r"princeton\s+university"],
"Massachusetts Institute of Technology": [r"mit", r"massachusetts\s+institute\s+of\s+technology", r"mass\s+tech"],
"Harvard University": [r"harvard", r"\bhu\b", r"harvard\s+university"],
"Stanford University": [r"stanford", r"\bsu\b", r"stanford\s+university"],
"Yale University": [r"yale", r"\byu\b", r"yale\s+university"],
"California Institute of Technology": [r"caltech", r"california\s+institute\s+of\s+technology"],
"Duke University": [r"duke", r"\bdu\b", r"duke\s+university"],
"Johns Hopkins University": [r"johns\s+hopkins", r"\bjhu\b", r"hopkins"],
"Northwestern University": [r"northwestern", r"\bnu\b", r"northwestern\s+university"],
"University of Pennsylvania": [r"upenn", r"penn", r"university\s+of\s+pennsylvania"],
"Cornell University": [r"cornell", r"\bcu\b", r"cornell\s+university"],
"University of Chicago": [r"uchicago", r"university\s+of\s+chicago", r"u\s+chicago"],
"Brown University": [r"brown", r"\bbu\b", r"brown\s+university"],
"Columbia University": [r"columbia", r"\bcu\b", r"columbia\s+university"],
"Dartmouth College": [r"dartmouth", r"\bdc\b", r"dartmouth\s+college"],
"University of California--Los Angeles": [r"ucla", r"university\s+of\s+california\s+los\s+angeles", r"uc\s+la"],
"University of California, Berkeley": [r"uc\s+berkeley", r"berkeley", r"university\s+of\s+california\s+berkeley"],
"Rice University": [r"rice", r"\bru\b", r"rice\s+university"],
"University of Notre Dame": [r"notre\s+dame", r"\bnd\b", r"university\s+of\s+notre\s+dame"],
"Vanderbilt University": [r"vanderbilt", r"\bvu\b", r"vandy"],
"Carnegie Mellon University": [r"carnegie\s+mellon", r"\bcmu\b", r"cmu"],
"University of Michigan--Ann Arbor": [r"umich", r"michigan", r"university\s+of\s+michigan"],
"Washington University in St. Louis": [r"washu", r"washington\s+university", r"wu\s+stl"],
"Emory University": [r"emory", r"\beu\b", r"emory\s+university"],
"Georgetown University": [r"georgetown", r"\bgu\b", r"georgetown\s+university"],
"University of Virginia": [r"uva", r"virginia", r"university\s+of\s+virginia"],
"University of North Carolina--Chapel Hill": [r"unc", r"chapel\s+hill", r"university\s+of\s+north\s+carolina"],
"University of Southern California": [r"usc", r"southern\s+california", r"university\s+of\s+southern\s+california"],
"University of California, San Diego": [r"ucsd", r"uc\s+san\s+diego", r"university\s+of\s+california\s+san\s+diego"],
"New York University": [r"nyu", r"new\s+york\s+university"],
"University of Florida": [r"uf", r"florida", r"university\s+of\s+florida"],
"The University of Texas--Austin": [r"ut\s+austin", r"utexas", r"university\s+of\s+texas"],
"Georgia Institute of Technology": [r"gatech", r"georgia\s+tech", r"georgia\s+institute\s+of\s+technology"],
"University of California, Davis": [r"uc\s+davis", r"ucd", r"university\s+of\s+california\s+davis"],
"University of California--Irvine": [r"uci", r"uc\s+irvine", r"university\s+of\s+california\s+irvine"],
"University of Illinois Urbana-Champaign": [r"uiuc", r"illinois", r"university\s+of\s+illinois"],
"Boston College": [r"bc", r"boston\s+college"],
"Tufts University": [r"tufts", r"\btu\b", r"tufts\s+university"],
"University of California, Santa Barbara": [r"ucsb", r"uc\s+santa\s+barbara", r"university\s+of\s+california\s+santa\s+barbara"],
"University of Wisconsin--Madison": [r"uw\s+madison", r"wisconsin", r"university\s+of\s+wisconsin"],
"Boston University": [r"bu", r"boston\s+university"],
"The Ohio State University": [r"ohio\s+state", r"osu", r"the\s+ohio\s+state\s+university"],
"Rutgers University--New Brunswick": [r"rutgers", r"ru", r"rutgers\s+university"],
"University of Maryland, College Park": [r"umd", r"maryland", r"university\s+of\s+maryland"],
"University of Rochester": [r"rochester", r"\bur\b", r"university\s+of\s+rochester"],
"Lehigh University": [r"lehigh", r"\blu\b", r"lehigh\s+university"],
"Purdue University--Main Campus": [r"purdue", r"\bpu\b", r"purdue\s+university"],
"University of Georgia": [r"uga", r"georgia", r"university\s+of\s+georgia"],
"University of Washington": [r"uw", r"washington", r"university\s+of\s+washington"],
"Wake Forest University": [r"wake\s+forest", r"\bwfu\b", r"wake"],
"Case Western Reserve University": [r"case\s+western", r"\bcwru\b", r"case"],
"Texas A&M University": [r"texas\s+a&m", r"tamu", r"a&m"],
"Virginia Tech": [r"virginia\s+tech", r"vt", r"vtech"],
"Florida State University": [r"fsu", r"florida\s+state", r"florida\s+state\s+university"],
"University of Miami": [r'umiami', r'\bum\b', r'university\s+of\s+miami']
}
try:
# Read the JSON file
with open(json_file_path, 'r') as file:
data = json.load(file)
# Iterate through each user profile
for user_id, profile in data.items():
profile_info = profile.get('profile_info', {})
# Check age range (default to configurable min/max)
age = profile_info.get('age', 0)
if not (age_min <= age <= age_max):
continue
# Get education list (default to empty list if not present)
educations = profile_info.get('educations', [])
# Check each education string for top 50 university references
found_match = False
matched_university = None
for edu in educations:
if not isinstance(edu, str):
continue
# Convert to lowercase for case-insensitive matching
edu_lower = edu.lower()
# Check each university's patterns
for university, patterns in top_50_patterns.items():
for pattern in patterns:
if re.search(pattern, edu_lower):
found_match = True
matched_university = university
break
if found_match:
break
if found_match:
break
# If we found a match, add the profile to our results
if found_match:
# Get images list (default to empty list if not present)
images = profile.get('images', [])
image_urls = [img.get('url', '') for img in images if img.get('url')]
matching_profiles.append({
'user_id': user_id,
'age': age,
'firstName': profile_info.get('firstName', ''),
'educations': educations,
'matched_university': matched_university,
'location': profile_info.get('location', {}).get('name', ''),
'image_urls': image_urls
})
return matching_profiles
except FileNotFoundError:
print(f"Error: File {json_file_path} not found.")
return []
except json.JSONDecodeError:
print(f"Error: Invalid JSON format in {json_file_path}.")
return []
except Exception as e:
print(f"Unexpected error: {str(e)}")
return []
def main():
# Initialize clients with your auth token
auth_token = os.getenv("BEARER_TOKEN")
session_id = os.getenv("SESSION_ID")
user_id = os.getenv("USER_ID")
api_client = HingeAPIClient(
auth_token=auth_token,
session_id=session_id,
user_id=user_id
)
media_client = HingeMediaClient(auth_token=auth_token)
tools = HingeTools(api_client, media_client)
# Example: Mass Scraping
# tools.scrape_recommendations_multiple(
# iterations=40, # 40 seems like the max before needing to skip people...
# min_sleep = 20,
# max_sleep = 60,
# )
json_file_path = 'all_recommendations.json'
csv_file_path = 'university_matches.csv'
results = find_top_50_university_students(json_file_path)
print(f"Found {len(results)} matching profiles.")
# Define CSV headers with separate columns for each image
headers = ['timestamp', 'user_id', 'name', 'age', 'location', 'education',
'image1', 'image2', 'image3', 'image4', 'image5', 'image6']
# Open CSV file in append mode
with open(csv_file_path, 'a', newline='', encoding='utf-8') as csvfile:
writer = csv.DictWriter(csvfile, fieldnames=headers)
# Write headers if file is empty
if csvfile.tell() == 0:
writer.writeheader()
# Write each profile as a row
for profile in results:
# Create a dictionary for the row
row_data = {
'timestamp': datetime.now().strftime('%Y-%m-%d %H:%M:%S'),
'user_id': profile['user_id'],
'name': profile['firstName'],
'age': profile['age'],
'location': profile['location'],
'education': ', '.join(profile['educations'])
}
# Add image URLs to separate columns
for i in range(6):
image_key = f'image{i+1}'
if profile['image_urls'] and i < len(profile['image_urls']):
row_data[image_key] = profile['image_urls'][i]
else:
row_data[image_key] = ''
writer.writerow(row_data)
print(f"Results have been appended to {csv_file_path}")
Le SDK est organisé en modules logiques pour séparer les préoccupations.
hingesdk/
├── __init__.py
├── client.py # Base HingeClient: Handles HTTP requests, headers, and Auth
├── api.py # HingeAPIClient: Core API methods (like, message, get_recs)
├── media.py # HingeMediaClient: Helpers for downloading/processing images
├── tools.py # HingeTools: High-level workflows (batch scraping, exports)
├── models.py # Pydantic models and data structures (if applicable)
├── exceptions.py # Custom exception classes (HingeAPIError, HingeAuthError)
└── assets/ # Static resources (e.g., prompt definitions)
Ce projet est sous licence MIT.