CU-8699gb53b Add API key based model download

webapp/Dockerfile

@@ -1,37 +1,69 @@
-FROM python:3.7
+# Stage 1: Build stage (dependencies and compilation)
+FROM python:3.12-slim as build
 
 # Create the required folders
 RUN mkdir -p /webapp/models
 
 # Copy everything
 COPY . /webapp
 
+# Install dependencies for building (git, etc.)
+RUN apt-get update && apt-get install -y --no-install-recommends \
+    git \
+    build-essential \
+    apt-utils \
+    cron \
+    sqlite3 \
+    libsqlite3-dev
+
+# Install Python dependencies
+ARG USE_CPU_TORCH=false
+# NOTE: Allow building without GPU so as to lower image size (disabled by default)
+RUN pip install -U pip && if [ "$USE_CPU_TORCH" = "true" ]; then \
+                              pip install -r /webapp/requirements.txt --extra-index-url https://download.pytorch.org/whl/cpu/; \
+                          else \
+                              pip install -r /webapp/requirements.txt; \
+                          fi
+
+# Get the spacy model (for later copy)
+RUN python -m spacy download en_core_web_md
+
+# Stage 2: Final (production) image
+FROM python:3.12-slim as final
+
+# Install runtime dependencies (you don’t need git in production)
+RUN apt-get update && apt-get install -y --no-install-recommends \
+    cron \
+    sqlite3 \
+    libsqlite3-dev && apt-get autoremove
+
+# Create the required folders (if not created already)
+RUN mkdir -p /webapp/models && mkdir -p /medcat_data
+
+# Copy only necessary files from build stage
+COPY --from=build /webapp /webapp
+# COPY --from=build /root/.cache /root/.cache  # Copy pip cache if needed
+
+# Copy Python site-packages (installed by pip) from build stage
+COPY --from=build /usr/local/lib/python3.12/site-packages /usr/local/lib/python3.12/site-packages
+
+# Set environment variables
 ENV VOCAB_URL=https://medcat.rosalind.kcl.ac.uk/media/vocab.dat
 ENV CDB_URL=https://medcat.rosalind.kcl.ac.uk/media/cdb-medmen-v1.dat
 
 ENV CDB_PATH=/webapp/models/cdb.dat
 ENV VOCAB_PATH=/webapp/models/vocab.dat
 
-# Create the data directory
-RUN mkdir -p /medcat_data
-
 # Set the pythonpath
 WORKDIR /webapp
 
-RUN pip install -r requirements.txt
-
-# Get the spacy model
-RUN python -m spacy download en_core_web_md
+# Create the db backup cron job (copied from your setup)
+COPY etc/cron.d/db-backup-cron /etc/cron.d/db-backup-cron
+RUN chmod 0644 /etc/cron.d/db-backup-cron && crontab /etc/cron.d/db-backup-cron
 
-# Build the db
+# Run migrations and collect static (could be in entrypoint script)
 RUN python manage.py makemigrations && \
     python manage.py makemigrations demo && \
     python manage.py migrate && \
     python manage.py migrate demo && \
     python manage.py collectstatic --noinput
-
-# Create the db backup cron job
-RUN apt-get update && apt-get install -y --no-install-recommends apt-utils cron sqlite3 libsqlite3-dev
-COPY etc/cron.d/db-backup-cron /etc/cron.d/db-backup-cron
-RUN chmod 0644 /etc/cron.d/db-backup-cron
-RUN crontab /etc/cron.d/db-backup-cron

webapp/demo/forms.py

@@ -46,3 +46,8 @@ class Meta:
             "funder": forms.TextInput(attrs={"size": 40}),
             "use_case": forms.Textarea(attrs={"rows": 5, "cols": 40}),
         }
+
+
+class UMLSApiKeyForm(forms.Form):
+    apikey = forms.CharField(label='UMLS API Key',
+                             widget=forms.TextInput(attrs={'size': 50}))

webapp/demo/templates/umls_api_key_entry.html

@@ -0,0 +1,49 @@
+{% extends 'base.html' %}
+{% load static %}
+
+{% block style %}
+<link rel="stylesheet" href="{% static 'css/home.css' %}">
+<link rel="stylesheet" href="{% static 'css/annotations.css' %}">
+<link rel="stylesheet" href="https://use.fontawesome.com/releases/v5.8.2/css/all.css" integrity="sha384-oS3vJWv+0UjzBfQzYUhtDYW+Pj2yciDJxpsK1OYPAYjqT085Qq/1cq5FLXAZQ7Ay" crossorigin="anonymous">
+<link rel="icon" href="{% static 'image/favicon.ico' %}">
+{% endblock %}
+
+{% block body %}
+<div class="container-fluid">
+    <h5>Please enter your UMLS API key to verify your license:</h5>
+
+    {% if message %}
+        <div class="alert alert-info">
+            {{ message }}
+        </div>
+    {% endif %}
+
+    {% if form.errors %}
+        <div class="alert alert-danger">
+            <ul>
+                {% for field in form %}
+                    {% for error in field.errors %}
+                        <li><strong>{{ field.label }}:</strong> {{ error }}</li>
+                    {% endfor %}
+                {% endfor %}
+                {% for error in form.non_field_errors %}
+                    <li>{{ error }}</li>
+                {% endfor %}
+            </ul>
+        </div>
+    {% endif %}
+
+    <form method="post">
+        {% csrf_token %}
+        <div class="form-group">
+            {{ form.apikey.label_tag }}
+            {{ form.apikey }}
+        </div>
+        <button type="submit" class="btn btn-primary">Verify API Key</button>
+    </form>
+
+    <p class="mt-3 text-muted">
+        You can find your API key by logging into your UMLS account and visiting your <a href="https://uts.nlm.nih.gov/uts/profile" target="_blank">UMLS Profile</a>.
+    </p>
+</div>
+{% endblock %}

webapp/demo/urls.py

@@ -5,5 +5,6 @@
 urlpatterns = [
     path('', show_annotations, name='train_annotations'),
     path('auth-callback', validate_umls_user, name='validate-umls-user'),
+    path('auth-callback-api', validate_umls_api_key, name='validate-umls-api-key'),
     path('download-model', download_model, name="download-model")
 ]

webapp/demo/views.py

@@ -2,6 +2,7 @@
 sys.path.insert(0, '/home/ubuntu/projects/MedCAT/')
 import os
 import json
+import requests
 from django.shortcuts import render
 from django.http import StreamingHttpResponse, HttpResponse
 from wsgiref.util import FileWrapper
@@ -13,12 +14,17 @@
 from urllib.error import HTTPError
 #from medcat.meta_cat import MetaCAT
 from .models import *
-from .forms import DownloaderForm
+from .forms import DownloaderForm, UMLSApiKeyForm
 
 AUTH_CALLBACK_SERVICE = 'https://medcat.rosalind.kcl.ac.uk/auth-callback'
 VALIDATION_BASE_URL = 'https://uts-ws.nlm.nih.gov/rest/isValidServiceValidate'
 VALIDATION_LOGIN_URL = f'https://uts.nlm.nih.gov/uts/login?service={AUTH_CALLBACK_SERVICE}'
 
+API_KEY_AUTH_URL = 'https://utslogin.nlm.nih.gov/cas/v1/api-key'
+UMLS_SERVICE = 'http://umlsks.nlm.nih.gov'  # required as 'service' parameter
+TEST_CUI = 'C0000005'  # harmless, public CUI for validation
+CONTENT_API_URL = f'https://uts-ws.nlm.nih.gov/rest/content/current/CUI/{TEST_CUI}'
+
 model_pack_path = os.getenv('MODEL_PACK_PATH', 'models/medmen_wstatus_2021_oct.zip')
 
 try:
@@ -101,6 +107,55 @@ def validate_umls_user(request):
         return render(request, 'umls_user_validation.html', context=context)
 
 
+def validate_umls_api_key(request):
+    if request.method == 'POST':
+        form = UMLSApiKeyForm(request.POST)
+        if form.is_valid():
+            apikey = form.cleaned_data['apikey']
+            try:
+                # Step 1: Get TGT
+                r = requests.post(API_KEY_AUTH_URL, data={'apikey': apikey}, timeout=10)
+                if r.status_code != 201:
+                    raise Exception('Invalid API key or auth server issue.')
+
+                tgt_url = r.headers['Location']
+
+                # Step 2: Get service ticket
+                r = requests.post(tgt_url, data={'service': UMLS_SERVICE}, timeout=10)
+                if r.status_code != 200:
+                    raise Exception('Could not get service ticket.')
+
+                service_ticket = r.text.strip()
+
+                # Step 3: Use ticket to call a known endpoint
+                params = {'ticket': service_ticket}
+                r = requests.get(CONTENT_API_URL, params=params, timeout=10)
+
+                if r.status_code == 200:
+                    context = {
+                        'is_valid': True,
+                        'message': 'License verified via API key!',
+                        'downloader_form': DownloaderForm(MedcatModel.objects.all())
+                    }
+                else:
+                    context = {
+                        'is_valid': False,
+                        'message': 'API key is not valid or user is not licensed for UMLS.'
+                    }
+
+            except Exception as e:
+                context = {
+                    'is_valid': False,
+                    'message': f'Error validating API key: {str(e)}'
+                }
+
+            return render(request, 'umls_user_validation.html', context=context)
+    else:
+        form = UMLSApiKeyForm()
+
+    return render(request, 'umls_api_key_entry.html', {'form': form})
+
+
 def download_model(request):
     if request.method == 'POST':
         downloader_form = DownloaderForm(MedcatModel.objects.all(), request.POST)

webapp/requirements.txt

@@ -2,5 +2,5 @@ Django==3.2.25
 django-dbbackup==4.0.0b0
 django-storages[boto3]==1.12.3
 django-cron==0.5.1
-medcat~=1.15.0
+medcat~=1.16.0
 urllib3==1.26.18