Merge pull request #1039 from cmu-delphi/release/delphi-epidata-0.4.3

Release Delphi Epidata 0.4.3

Author: krivard

.bumpversion.cfg

@@ -1,5 +1,5 @@
 [bumpversion]
-current_version = 0.4.2
+current_version = 0.4.3
 commit = False
 tag = False
 

dev/local/Makefile

@@ -83,7 +83,7 @@ web:
 	@# Run the web server
 	@docker run --rm -p 127.0.0.1:10080:80 \
 		--env "SQLALCHEMY_DATABASE_URI=mysql+mysqldb://user:pass@delphi_database_epidata:3306/epidata" \
-		--env "FLASK_SECRET=abc" --env "FLASK_PREFIX=/epidata" \
+		--env "FLASK_SECRET=abc" --env "FLASK_PREFIX=/epidata" --env "LOG_DEBUG" \
 		--network delphi-net --name delphi_web_epidata \
 		delphi_web_epidata >$(LOG_WEB) 2>&1 &
 
@@ -139,6 +139,20 @@ test:
 		--env "FLASK_SECRET=abc" \
 		delphi_web_python python -m pytest --import-mode importlib $(pdb) $(test) | tee test_output_$(NOW).log
 
+.PHONY=bash
+bash:
+	@docker run -it --rm --network delphi-net \
+		--mount type=bind,source=$(CWD)repos/delphi/delphi-epidata,target=/usr/src/app/repos/delphi/delphi-epidata,readonly \
+		--mount type=bind,source=$(CWD)repos/delphi/delphi-epidata/src,target=/usr/src/app/delphi/epidata,readonly \
+		--env "SQLALCHEMY_DATABASE_URI=mysql+mysqldb://user:pass@delphi_database_epidata:3306/epidata" \
+		--env "FLASK_SECRET=abc" \
+		delphi_web_python bash
+
+.PHONY=sql
+sql:
+	@docker run --rm -it --network delphi-net --cap-add=sys_nice \
+		percona mysql --user=user --password=pass --port 3306 --host delphi_database_epidata epidata
+
 .PHONY=clean
 clean:
 	@docker images -f "dangling=true" -q | xargs docker rmi >/dev/null 2>&1

requirements.txt

@@ -12,3 +12,4 @@ tenacity==7.0.0
 newrelic
 epiweeks==2.1.2
 typing-extensions
+structlog==22.1.0

src/acquisition/covidcast/logger.py

@@ -1,10 +1,10 @@
 """Structured logger utility for creating JSON logs in Delphi pipelines."""
 import logging
+import os
 import sys
 import threading
 import structlog
 
-
 def handle_exceptions(logger):
     """Handle exceptions using the provided logger."""
     def exception_handler(etype, value, traceback):
@@ -45,12 +45,24 @@ def get_structured_logger(name=__name__,
     if filename:
         handlers.append(logging.FileHandler(filename))
 
+    if "LOG_DEBUG" in os.environ:
+        log_level = logging.DEBUG
+    else:
+        log_level = logging.INFO
+
     logging.basicConfig(
         format="%(message)s",
-        level=logging.INFO,
+        level=log_level,
         handlers=handlers
         )
 
+    def add_pid(logger, method_name, event_dict):
+        """
+        Add current PID to the event dict.
+        """
+        event_dict["pid"] = os.getpid()
+        return event_dict
+
     # Configure structlog. This uses many of the standard suggestions from
     # the structlog documentation.
     structlog.configure(
@@ -61,6 +73,8 @@ def get_structured_logger(name=__name__,
             structlog.stdlib.add_logger_name,
             # Include log level in output.
             structlog.stdlib.add_log_level,
+            # Include PID in output.
+            add_pid,
             # Allow formatting into arguments e.g., logger.info("Hello, %s",
             # name)
             structlog.stdlib.PositionalArgumentsFormatter(),

src/client/delphi_epidata.R

@@ -15,7 +15,7 @@ Epidata <- (function() {
   # API base url
   BASE_URL <- 'https://delphi.cmu.edu/epidata/api.php'
 
-  client_version <- '0.4.2'
+  client_version <- '0.4.3'
 
   # Helper function to cast values and/or ranges to strings
   .listitem <- function(value) {

src/client/delphi_epidata.js

@@ -22,7 +22,7 @@
   }
 })(this, function (exports, fetchImpl, jQuery) {
   const BASE_URL = "https://delphi.cmu.edu/epidata/";
-  const client_version = "0.4.2";
+  const client_version = "0.4.3";
 
   // Helper function to cast values and/or ranges to strings
   function _listitem(value) {

src/client/packaging/npm/package.json

@@ -2,7 +2,7 @@
   "name": "delphi_epidata",
   "description": "Delphi Epidata API Client",
   "authors": "Delphi Group",
-  "version": "0.4.2",
+  "version": "0.4.3",
   "license": "MIT",
   "homepage": "https://github.com/cmu-delphi/delphi-epidata",
   "bugs": {

src/client/packaging/pypi/delphi_epidata/__init__.py

@@ -1,4 +1,4 @@
 from .delphi_epidata import Epidata
 
 name = 'delphi_epidata'
-__version__ = '0.4.2'
+__version__ = '0.4.3'

src/client/packaging/pypi/setup.py

@@ -5,7 +5,7 @@
 
 setuptools.setup(
     name="delphi_epidata",
-    version="0.4.2",
+    version="0.4.3",
     author="David Farrow",
     author_email="dfarrow0@gmail.com",
     description="A programmatic interface to Delphi's Epidata API.",

src/server/_common.py

@@ -1,12 +1,15 @@
 from typing import cast
+import time
 
 from flask import Flask, g, request
+from sqlalchemy import event
 from sqlalchemy.engine import Connection
 from werkzeug.local import LocalProxy
 
+from .utils.logger import get_structured_logger
 from ._config import SECRET
 from ._db import engine
-from ._exceptions import DatabaseErrorException
+from ._exceptions import DatabaseErrorException, EpiDataException
 
 app = Flask("EpiData", static_url_path="")
 app.config["SECRET"] = SECRET
@@ -24,19 +27,53 @@ def _get_db() -> Connection:
 """
 db: Connection = cast(Connection, LocalProxy(_get_db))
 
+@event.listens_for(engine, "before_cursor_execute")
+def before_cursor_execute(conn, cursor, statement, parameters, context, executemany):
+    context._query_start_time = time.time()
+
+
+@event.listens_for(engine, "after_cursor_execute")
+def after_cursor_execute(conn, cursor, statement, parameters, context, executemany):
+    # this timing info may be suspect, at least in terms of dbms cpu time...
+    # it is likely that it includes that time as well as any overhead that
+    # comes from throttling or flow control on the streamed data, as well as
+    # any row transform/processing time
+    total_time = time.time() - context._query_start_time
+
+    # Convert to milliseconds
+    total_time *= 1000
+    get_structured_logger('server_api').info("Executed SQL", statement=statement, params=parameters, elapsed_time_ms=total_time)
+
 
 @app.before_request
-def connect_db():
+def before_request_execute():
+    # Set timer for statement
+    g._request_start_time = time.time()
+
+    # Log statement
+    get_structured_logger('server_api').info("Received API request", method=request.method, url=request.url, form_args=request.form, req_length=request.content_length, remote_addr=request.remote_addr, user_agent=request.user_agent.string)
+
     if request.path.startswith('/lib'):
         return
     # try to get the db
     try:
         _get_db()
-    except:
-        app.logger.error('database connection error', exc_info=True)
+    except Exception as e:
+        get_structured_logger('server_error').error('database connection error', exception=e)
         raise DatabaseErrorException()
 
 
+@app.after_request
+def after_request_execute(response):
+    total_time = time.time() - g._request_start_time
+    # Convert to milliseconds
+    total_time *= 1000
+    get_structured_logger('server_api').info('Served API request', method=request.method, url=request.url, form_args=request.form, req_length=request.content_length, remote_addr=request.remote_addr, user_agent=request.user_agent.string,
+                                             values=request.values.to_dict(flat=False), blueprint=request.blueprint, endpoint=request.endpoint,
+                                             response_status=response.status, content_length=response.calculate_content_length(), elapsed_time_ms=total_time)
+    return response
+
+
 @app.teardown_appcontext
 def teardown_db(exception=None):
     # close the db connection
@@ -46,6 +83,16 @@ def teardown_db(exception=None):
         db.close()
 
 
+@app.errorhandler(EpiDataException)
+def handle_exception(e):
+    # Log error and pass through; EpiDataExceptions are HTTPExceptions which are valid WSGI responses (see https://werkzeug.palletsprojects.com/en/2.2.x/exceptions/ )
+    if isinstance(e, DatabaseErrorException):
+        get_structured_logger('server_error').error('Received DatabaseErrorException', exception=str(e), exc_info=True)
+    else:
+        get_structured_logger('server_error').warn('Encountered user-side error', exception=str(e))
+    return e
+
+
 def is_compatibility_mode() -> bool:
     """
     checks whether this request is in compatibility mode