Merge pull request #10 from snowplow-incubator/enrichments

Add support for self-service enrichments

app/Dockerfile → control-plane/Dockerfile

@@ -1,5 +1,5 @@
-FROM python:3.12
+FROM python:3.12-slim
 WORKDIR /app
 COPY requirements.txt .
 RUN pip install -r requirements.txt
-COPY . .
+COPY app.py app.py

control-plane/app.py

@@ -0,0 +1,98 @@
+from base64 import b64decode, b64encode
+from glob import glob
+import json
+import logging
+
+import docker
+import requests
+from flask import Flask, request
+from flask_cors import CORS
+
+
+app = Flask(__name__)
+CORS(app)
+logging.getLogger('flask_cors').level = logging.DEBUG
+
+valid_containers = ['snowplow-stream-collector', 'snowplow-enrich', 'snowplow-iglu-server',
+                    'connect', 'snowflake-streaming-loader', 'snowflake-streaming-loader-incomplete',
+                    'lake-loader', 'bigquery-loader', 'snowbridge', 'ngrok-tunnel']
+
+client = docker.from_env()
+
+PATH = '../enrich/enrichments'
+## Container API
+
+def restart(name: str):
+    try:
+        if name in valid_containers:
+            container = client.containers.get(name)
+            container.restart()
+            return {'restart': True}
+        else:
+            return {'restart': False, 'error': 'Invalid container name'}
+    except Exception as e:
+        return {'restarted': False, 'exception': str(e)}
+
+@app.route('/containers/restart')
+def restart_container():
+    name = request.args.get('container')
+    return restart(name)
+
+@app.route('/containers/start')
+def start_container():
+    name = request.args.get('container')
+    try:
+        if name in valid_containers:
+            container = client.containers.get(name)
+            container.start()
+            return {'start': True}
+        else:
+            return {'start': False, 'error': 'Invalid container name'}
+    except Exception as e:
+        return {'start': False, 'exception': str(e)}
+
+@app.route('/containers/list')
+def list_containers():
+    containers = client.containers.list()
+    output = [{'name': container.name, 'status': container.status} for container in containers if container.name in valid_containers]
+    return output
+
+@app.route('/containers/status')
+def container_status():
+    name = request.args.get('container')
+    try:
+        container = client.containers.get(name)
+        return {'status': container.status}
+    except Exception as e:
+        return {'status': 'error', 'exception': str(e)}
+
+
+## Enrichment API
+
+## List enrichments
+@app.route('/pipelines/enrichments/', methods=['GET'])
+def read_enrichments():
+    enrichments = []
+    for filename in glob(f'{PATH}/*.json'):
+        print(filename)
+        with open(filename, 'r') as file:
+            enrichments.append({
+                **json.loads(file.read())
+            })
+    return enrichments
+
+## Write single enrichment
+@app.route('/pipelines/enrichments/', methods=['POST'])
+def write_enrichment():
+    data = request.json
+    enrichment_name = data['data']['name']
+    if enrichment_name == 'javascript_script_config':
+        script = data['data']['parameters']['script'].encode()
+        data['data']['parameters']['script'] = b64encode(script).decode('utf-8')
+
+    with open(f'{PATH}/{enrichment_name}.json', 'w') as file:
+        file.write(json.dumps(data, indent=4))
+    restart('snowplow-enrich')
+    return {'success': True}
+
+

control-plane/requirements.txt

@@ -0,0 +1,3 @@
+flask==3.0.3
+flask-cors==5.0.0
+docker==7.1.0

docker-compose.yml

@@ -287,7 +287,7 @@ services:
 
   local-failed-ui:
     container_name: "local-failed-ui"
-    image: "snowplow/failed-events-ui:0.0.1"
+    image: "snowplow/failed-events-ui:0.0.2"
     ports:
       - "3001:3000"
 
@@ -305,4 +305,17 @@ services:
       - "./tunnel/ngrok.yml:/etc/ngrok.yml"
       - "./tunnel/policy.yml:/etc/policy.yml"
     ports:
-      - "4040:4040"
+      - "4040:4040"
+
+  control-plane:
+    container_name: "snowplow-control-plane"
+    build: "control-plane/"
+    command: "flask run --host=0.0.0.0 --debug"
+    environment:
+      - "FLASK_APP=app"
+    volumes:
+      - "./enrich:/enrich"
+      - "/var/run/docker.sock:/var/run/docker.sock"
+    ports:
+      - "8083:5000"
+

enrich/enrichments/anon_ip.json

@@ -0,0 +1,11 @@
+{
+  "schema": "iglu:com.snowplowanalytics.snowplow/anon_ip/jsonschema/1-0-0",
+  "data": {
+    "name": "anon_ip",
+    "vendor": "com.snowplowanalytics.snowplow",
+    "enabled": false,
+    "parameters": {
+      "anonOctets": 1
+    }
+  }
+}

enrich/enrichments/api_request_enrichment_config.json

@@ -0,0 +1,43 @@
+{
+  "schema": "iglu:com.snowplowanalytics.snowplow.enrichments/api_request_enrichment_config/jsonschema/1-0-0",
+  "data": {
+    "name": "api_request_enrichment_config",
+    "vendor": "com.snowplowanalytics.snowplow.enrichments",
+    "enabled": false,
+    "parameters": {
+      "inputs": [
+        {
+          "key": "user",
+          "pojo": {
+            "field": "user_id"
+          }
+        }
+      ],
+      "api": {
+        "http": {
+          "method": "GET",
+          "uri": "http://api.acme.com/users/{{client}}/{{user}}?format=json",
+          "timeout": 5000,
+          "authentication": {
+            "httpBasic": {
+              "username": "NA",
+              "password": "NA"
+            }
+          }
+        }
+      },
+      "outputs": [
+        {
+          "schema": "iglu:com.acme/user/jsonschema/1-0-0",
+          "json": {
+            "jsonPath": "$.record"
+          }
+        }
+      ],
+      "cache": {
+        "size": 3000,
+        "ttl": 60
+      }
+    }
+  }
+}

enrich/enrichments/campaign_attribution.json

@@ -0,0 +1,18 @@
+{
+  "schema": "iglu:com.snowplowanalytics.snowplow/campaign_attribution/jsonschema/1-0-0",
+  "data": {
+    "name": "campaign_attribution",
+    "vendor": "com.snowplowanalytics.snowplow",
+    "enabled": true,
+    "parameters": {
+      "mapping": "static",
+      "fields": {
+        "mktMedium": ["utm_medium"],
+        "mktSource": ["utm_source"],
+        "mktTerm": ["utm_term"],
+        "mktContent": ["utm_content"],
+        "mktCampaign": ["utm_campaign"]
+      }
+    }
+  }
+}

enrich/enrichments/cookie_extractor_config.json

@@ -0,0 +1,11 @@
+{
+  "schema": "iglu:com.snowplowanalytics.snowplow/cookie_extractor_config/jsonschema/1-0-0",
+  "data": {
+    "name": "cookie_extractor_config",
+    "vendor": "com.snowplowanalytics.snowplow",
+    "enabled": false,
+    "parameters": {
+      "cookies": []
+    }
+  }
+}

enrich/enrichments/cross_navigation_enrichment.json

@@ -0,0 +1,8 @@
+{
+  "schema": "iglu:com.snowplowanalytics.snowplow.enrichments/cross_navigation_config/jsonschema/1-0-0",
+  "data": {
+    "enabled": false,
+    "vendor": "com.snowplowanalytics.snowplow.enrichments",
+    "name": "cross_navigation_config"
+  }
+}

enrich/enrichments/currency_conversion_config.json

@@ -0,0 +1,14 @@
+{
+  "schema": "iglu:com.snowplowanalytics.snowplow/currency_conversion_config/jsonschema/1-0-0",
+  "data": {
+    "enabled": false,
+    "vendor": "com.snowplowanalytics.snowplow",
+    "name": "currency_conversion_config",
+    "parameters": {
+      "accountType": "DEVELOPER",
+      "apiKey": "PLACEHOLDER",
+      "baseCurrency": "EUR",
+      "rateAt": "EOD_PRIOR"
+    }
+  }
+}

enrich/enrichments/event_fingerprint_config.json

@@ -0,0 +1,17 @@
+{
+    "data": {
+        "enabled": true,
+        "name": "event_fingerprint_config",
+        "parameters": {
+            "excludeParameters": [
+                "cv",
+                "eid",
+                "nuid",
+                "stm"
+            ],
+            "hashAlgorithm": "SHA1"
+        },
+        "vendor": "com.snowplowanalytics.snowplow"
+    },
+    "schema": "iglu:com.snowplowanalytics.snowplow/event_fingerprint_config/jsonschema/1-0-1"
+}

enrich/enrichments/http_header_extractor_config.json

@@ -0,0 +1,11 @@
+{
+  "schema": "iglu:com.snowplowanalytics.snowplow.enrichments/http_header_extractor_config/jsonschema/1-0-0",
+  "data": {
+    "name": "http_header_extractor_config",
+    "vendor": "com.snowplowanalytics.snowplow.enrichments",
+    "enabled": false,
+    "parameters": {
+      "headersPattern": ".*"
+    }
+  }
+}

enrich/enrichments/iab_spiders_and_robots_enrichment.json

@@ -0,0 +1,22 @@
+{
+    "schema": "iglu:com.snowplowanalytics.snowplow.enrichments/iab_spiders_and_robots_enrichment/jsonschema/1-0-0",
+    "data": {
+        "name": "iab_spiders_and_robots_enrichment",
+        "vendor": "com.snowplowanalytics.snowplow.enrichments",
+        "enabled": false,
+        "parameters": {
+            "ipFile": {
+                "database": "ip_exclude_current_cidr.txt",
+                "uri": "gs://bucket-name"
+            },
+            "excludeUseragentFile": {
+                "database": "exclude_current.txt",
+                "uri": "gs://bucket-name"
+            },
+            "includeUseragentFile": {
+                "database": "include_current.txt",
+                "uri": "gs://bucket-name"
+            }
+        }
+    }
+}

enrich/enrichments/ip_lookups.json

@@ -0,0 +1,14 @@
+{
+  "schema": "iglu:com.snowplowanalytics.snowplow/ip_lookups/jsonschema/2-0-0",
+  "data": {
+    "name": "ip_lookups",
+    "vendor": "com.snowplowanalytics.snowplow",
+    "enabled": false,
+    "parameters": {
+      "geo": {
+        "database": "GeoLite2-City.mmdb",
+        "uri": "gs://max-mind-geolite2/"
+      }
+    }
+  }
+}

enrich/enrichments/javascript_script_config.json

@@ -0,0 +1,11 @@
+{
+    "schema": "iglu:com.snowplowanalytics.snowplow/javascript_script_config/jsonschema/1-0-1",
+    "data": {
+        "vendor": "com.snowplowanalytics.snowplow",
+        "name": "javascript_script_config",
+        "parameters": {
+            "script": "="
+        },
+        "enabled": false
+    }
+}