Skip to content

API_EXAMPLES.md

Latest commit

 

History

History
149 lines (109 loc) · 3.45 KB

API_EXAMPLES.md

File metadata and controls

149 lines (109 loc) · 3.45 KB

API Examples

Requires a running DECIMER server.

For full request/response/error contract, see ../API_ENDPOINTS.md. For script selection by experience level, see README.md.

Using the decimerapi Python Package

See also: decimer_server_usage_example.py

from pathlib import Path
from decimerapi.decimerapi import DecimerAPI

decimer_api = DecimerAPI()  # default host localhost, port 8099
# decimer_api = DecimerAPI("192.168.50.170", 8888)  # custom host/port
print(decimer_api.server_status())

current_file_path = Path(__file__).resolve()
input_image = current_file_path.parents[0] / "structure.png"
smiles = decimer_api.call_image2smiles(input_image, hand_drawn=False)
print(smiles)

Using Direct HTTP Requests (No Python Wrapper)

See also: decimer_naked_api_eg_docker_only_example.py

Python

Check server status:

import requests

url = "http://localhost:8099"

try:
    response = requests.get(url)
    response.raise_for_status()
    print("Server is running.")
except requests.exceptions.ConnectionError:
    print("Error: Could not connect to server. Is the server/container running?")
    raise SystemExit(1)
except requests.exceptions.HTTPError as e:
    print(f"Error: Server returned {e.response.status_code}")
    raise SystemExit(1)

Convert image to SMILES:

import base64
from pathlib import Path

import requests

url_i2s = "http://localhost:8099/image2smiles/"
input_image = "structure.png"
encoded_image = base64.b64encode(Path(input_image).read_bytes()).decode("utf-8")

data = {
    "encoded_image": encoded_image,  # required
    "is_hand_drawn": "false",      # optional
    "classify_image": "false",     # optional
}

response = requests.post(url_i2s, data=data)
if response.status_code != 200:
    print(f"Error: {response.status_code} -> {response.text}")
else:
    response_json = response.json()
    print(response_json["smiles"])

Curl

Health check:

curl -X GET "http://localhost:8099/"

Checking System Status & Hardware Acceleration

Query system information including hardware accelerator type (CPU/CUDA/Metal).

Using the decimerapi Python Package

from decimerapi import DecimerAPI

api = DecimerAPI()  # default localhost:8099
status = api.get_system_status()

if status:
    print(f"Server Status: {status['status']}")
    print(f"Accelerator: {status['accelerator_type']}")
    print(f"TensorFlow: {status['tensorflow_version']}")
else:
    print("Could not retrieve system status")

# Quick check for GPU availability
if status and status['accelerator_type'] != 'cpu':
    print(f"GPU acceleration available: {status['accelerator_type']}")
else:
    print("Running on CPU")

Using Direct HTTP Requests

Python

import requests

url = "http://localhost:8099/system/status"

try:
    response = requests.get(url)
    response.raise_for_status()
    status = response.json()
    print(f"Accelerator Type: {status['accelerator_type']}")
    print(f"TensorFlow Version: {status['tensorflow_version']}")
except requests.exceptions.RequestException as e:
    print(f"Error: {e}")

Curl

curl -X GET "http://localhost:8099/system/status"

Expected response:

{
  "status": "ready",
  "accelerator_type": "cuda",
  "tensorflow_version": "2.15.0"
}

Possible accelerator_type values:

  • cpu: CPU-only processing
  • cuda: NVIDIA CUDA GPU (Linux/Windows)
  • metal: Apple Metal (macOS with M1-M4)