MetDataAnalysis / met_api.py
yash-b18
Update to CLI to prevent crashing on invalid inputs
f78d74e
import time
import requests
import random
import pandas as pd
URL = "https://collectionapi.metmuseum.org/public/collection/v1/"
'''
Need to add error handling
'''
# gets information on a single object
def get_object(objectID):
try:
response = requests.get(f"{URL}/objects/{objectID}")
response.raise_for_status()
return response.json()
except Exception as e:
print(f"Error fetching object {objectID}: {e}")
return None
# gets all the objects that have some sort of image
def get_objectsWithImages():
try:
response = requests.get(f"{URL}search?hasImages=true&q=*")
response.raise_for_status()
data = response.json()
total = data.get("total", 0)
objectIDs = data.get("objectIDs", [])
return total, objectIDs
except Exception as e:
print(f"Error fetching objects with images: {e}")
return 0, []
# gets the urls for random objects with images
def get_images(totalObjects, objectIDs, limit):
try:
images = []
# grabbing extra in case a primary image is blank (Works best on small limits)
rand_indexes = random.sample(range(totalObjects), limit + 20)
for i in rand_indexes:
obj = get_object(objectIDs[i])
if obj and obj.get("primaryImage"):
images.append((obj["primaryImage"], obj.get("title", "Untitled")))
if len(images) == limit:
break
return images
except Exception as e:
print(f"Error in get_images: {e}")
return []
def department_counts(q="*", max_ids=200):
"""
Analytic: return a list of (department, count) for search results.
- q: search query (default '*' = anything)
- max_ids: cap how many object IDs to inspect (keeps it fast)
Uses the Met search endpoint (images only), then tallies the 'department'
field from each object's metadata.
"""
try:
resp = requests.get(f"{URL}search", params={"q": q, "hasImages": True}, timeout=15)
resp.raise_for_status()
ids = (resp.json().get("objectIDs") or [])[:max_ids]
except Exception:
return []
counts = {}
for oid in ids:
try:
obj = get_object(oid) # uses your existing helper
dep = obj.get("department") or "(unknown)"
counts[dep] = counts.get(dep, 0) + 1
except Exception:
continue
# return sorted (department, count) pairs, highest first
return sorted(counts.items(), key=lambda x: x[1], reverse=True)
"""
Get a list of departments
"""
def list_met_departments():
"""
Return a DataFrame of all Met departments (id + name) to help you choose.
"""
try:
r = requests.get(f"{URL}/departments")
r.raise_for_status()
depts = r.json().get("departments", [])
return pd.DataFrame(depts)[["departmentId", "displayName"]]
except Exception as e:
print(f"Error fetching departments: {e}")
return pd.DataFrame(columns=["departmentId", "displayName"])
"""
returns a dataframe of list of objects with images and metadata that matches search term
To be polite, will only get default max 5 random results from each department,
but can be specified as parameter. Also, can choose the departments to search from.
"""
def search_for_images(query,
max_per_department=5,
departments=None):
if not query or not query.strip():
raise ValueError("Please provide a query.")
# 1) Departments
if departments is None:
resp = requests.get(f"{URL}/departments")
resp.raise_for_status()
dept_list = resp.json().get("departments", [])
dept_ids = [d["departmentId"] for d in dept_list]
dept_id_to_name = {d["departmentId"]: d["displayName"] for d in dept_list}
else:
dept_ids = list(departments)
# Names will be filled from object details; provide a generic fallback
dept_id_to_name = {d: f"Department {d}" for d in dept_ids}
# 2) Per-department search → random sample of IDs → fetch details
rows = []
#session = requests.Session()
print("Searching for", query)
print("Across departments:", dept_ids)
for dept_id in dept_ids:
# limit to only those with images and is highlighted
params = {
"q": query,
"hasImages": "true",
#"isHighlight": "true",
"departmentId": dept_id,
}
try:
r = requests.get(f"{URL}/search", params=params)
#print(r.url)
r.raise_for_status()
except requests.RequestException:
continue
object_ids = (r.json() or {}).get("objectIDs") or []
if not object_ids:
continue
sample_ids = random.sample(object_ids, k=min(max_per_department, len(object_ids)))
for oid in sample_ids:
#print(f"Found: {oid}")
try:
obj = requests.get(f"{URL}/objects/{oid}").json()
except requests.RequestException:
continue
# skip if no image
if not obj["primaryImage"]:
continue
rows.append({
"objectID": obj.get("objectID"),
"title": obj.get("title"),
"artistDisplayName": obj.get("artistDisplayName"),
"objectDate": obj.get("objectDate"),
"culture": obj.get("culture"),
"medium": obj.get("medium"),
"department": obj.get("department") or dept_id_to_name.get(dept_id),
"objectName": obj.get("objectName"),
"classification": obj.get("classification"),
"primaryImageSmall": obj.get("primaryImageSmall"),
"primaryImage": obj.get("primaryImage"),
"objectURL": obj.get("objectURL"),
"isPublicDomain": obj.get("isPublicDomain"),
})
df = pd.DataFrame(rows)
if not df.empty:
df = df.sort_values(["department", "title"]).reset_index(drop=True)
return df
"""
CLI when run from command line
Displays the departments to allow user to input a department
Then searches the department based on input, and displays the results with images.
"""
def main():
print("Welcome to Met Search.")
departments = list_met_departments()
print(departments.to_string(index=False))
random.seed(time.time())
while True:
dept_no = input("Choose a departmentId #: (Type 'q', 'quit' to stop the program, or enter for all) ").strip()
if dept_no.lower() in ['q', 'quit']:
break
elif dept_no == '':
dept = None
else:
if not dept_no.isdigit():
print("Invalid input. Please try again.")
continue
dept = [int(dept_no)]
query = input("Search the Met for: ").strip()
if query == '':
print("Please enter a query.")
continue
results = search_for_images(query,2, departments=dept)
if results.empty:
print("No results found.")
continue
print(results.to_string(index=False))
if __name__ == '__main__':
main()