1
0
mirror of https://github.com/ilri/dspace-statistics-api.git synced 2024-11-16 19:37:04 +01:00
dspace-statistics-api/dspace_statistics_api/app.py
Alan Orth a35ecf2394
Add Swagger UI on /swagger
This includes a Swagger UI with an OpenAPI 3.0 JSON schema for easy
interactive demonstration and testing of the API. The JSON schema
was created with the standalone swagger-editor. Includes tests to
make sure that the /swagger and /docs/openapi.json paths are acce-
ssible.
2020-12-22 11:18:47 +02:00

205 lines
7.3 KiB
Python
Raw Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

import falcon
import psycopg2.extras
from falcon_swagger_ui import register_swaggerui_app
from .database import DatabaseManager
from .stats import get_downloads, get_views
from .util import set_statistics_scope, validate_post_parameters
class RootResource:
def on_get(self, req, resp):
resp.status = falcon.HTTP_200
resp.content_type = "text/html"
with open("dspace_statistics_api/docs/index.html", "r") as f:
resp.body = f.read()
class OpenAPIJSONResource:
def on_get(self, req, resp):
resp.status = falcon.HTTP_200
resp.content_type = "text/html"
with open("dspace_statistics_api/docs/openapi.json", "r") as f:
resp.body = f.read()
class AllStatisticsResource:
@falcon.before(set_statistics_scope)
def on_get(self, req, resp):
"""Handles GET requests"""
# Return HTTPBadRequest if id parameter is not present and valid
limit = req.get_param_as_int("limit", min_value=1, max_value=100) or 100
page = req.get_param_as_int("page", min_value=0) or 0
offset = limit * page
with DatabaseManager() as db:
db.set_session(readonly=True)
with db.cursor() as cursor:
# get total number of communities/collections/items so we can estimate the pages
cursor.execute(f"SELECT COUNT(id) FROM {req.context.statistics_scope}")
pages = round(cursor.fetchone()[0] / limit)
# get statistics and use limit and offset to page through results
cursor.execute(
f"SELECT id, views, downloads FROM {req.context.statistics_scope} ORDER BY id LIMIT %s OFFSET %s",
[limit, offset],
)
# create a list to hold dicts of stats
statistics = list()
# iterate over results and build statistics object
for result in cursor:
statistics.append(
{
"id": str(result["id"]),
"views": result["views"],
"downloads": result["downloads"],
}
)
message = {
"currentPage": page,
"totalPages": pages,
"limit": limit,
"statistics": statistics,
}
resp.media = message
@falcon.before(set_statistics_scope)
@falcon.before(validate_post_parameters)
def on_post(self, req, resp):
"""Handles POST requests.
Uses two `before` hooks to set the statistics "scope" and validate the
POST parameters. The "scope" is the type of statistics we want, which
will be items, communities, or collections, depending on the request.
"""
# Build the Solr date string, ie: [* TO *]
if req.context.dateFrom and req.context.dateTo:
solr_date_string = f"[{req.context.dateFrom} TO {req.context.dateTo}]"
elif not req.context.dateFrom and req.context.dateTo:
solr_date_string = f"[* TO {req.context.dateTo}]"
elif req.context.dateFrom and not req.context.dateTo:
solr_date_string = f"[{req.context.dateFrom} TO *]"
else:
solr_date_string = "[* TO *]"
# Helper variables to make working with pages/items/results easier and
# to make the code easier to understand
number_of_elements: int = len(req.context.elements)
pages: int = int(number_of_elements / req.context.limit)
first_element: int = req.context.page * req.context.limit
last_element: int = first_element + req.context.limit
# Get a subset of the POSTed items based on our limit. Note that Python
# list slicing and indexing are both zero based, but the first and last
# items in a slice can be confusing. See this ASCII diagram:
#
# +---+---+---+---+---+---+
# | P | y | t | h | o | n |
# +---+---+---+---+---+---+
# Slice position: 0 1 2 3 4 5 6
# Index position: 0 1 2 3 4 5
#
# So if we have a list of items with 240 items:
#
# 1st set: items[0:100] would give items at indexes 0 to 99
# 2nd set: items[100:200] would give items at indexes 100 to 199
# 3rd set: items[200:300] would give items at indexes 200 to 239
elements_subset: list = req.context.elements[first_element:last_element]
views: dict = get_views(
solr_date_string, elements_subset, req.context.views_facet_field
)
downloads: dict = get_downloads(
solr_date_string, elements_subset, req.context.downloads_facet_field
)
# create a list to hold dicts of stats
statistics = list()
# iterate over views dict to extract views and use the element id as an
# index to the downloads dict to extract downloads.
for k, v in views.items():
statistics.append({"id": k, "views": v, "downloads": downloads[k]})
message = {
"currentPage": req.context.page,
"totalPages": pages,
"limit": req.context.limit,
"statistics": statistics,
}
resp.status = falcon.HTTP_200
resp.media = message
class SingleStatisticsResource:
@falcon.before(set_statistics_scope)
def on_get(self, req, resp, id_):
"""Handles GET requests"""
# Adapt Pythons uuid.UUID type to PostgreSQLs uuid
# See: https://www.psycopg.org/docs/extras.html
psycopg2.extras.register_uuid()
with DatabaseManager() as db:
db.set_session(readonly=True)
with db.cursor() as cursor:
cursor = db.cursor()
cursor.execute(
f"SELECT views, downloads FROM {req.context.database} WHERE id=%s",
[str(id_)],
)
if cursor.rowcount == 0:
raise falcon.HTTPNotFound(
title=f"{req.context.statistics_scope} not found",
description=f'The {req.context.statistics_scope} with id "{str(id_)}" was not found.',
)
else:
results = cursor.fetchone()
statistics = {
"id": str(id_),
"views": results["views"],
"downloads": results["downloads"],
}
resp.media = statistics
api = application = falcon.API()
api.add_route("/", RootResource())
# Item routes
api.add_route("/items", AllStatisticsResource())
api.add_route("/item/{id_:uuid}", SingleStatisticsResource())
# Community routes
api.add_route("/communities", AllStatisticsResource())
api.add_route("/community/{id_:uuid}", SingleStatisticsResource())
# Collection routes
api.add_route("/collections", AllStatisticsResource())
api.add_route("/collection/{id_:uuid}", SingleStatisticsResource())
# Swagger configuration
SWAGGERUI_URL = "/swagger" # without trailing slash
SCHEMA_URL = "/docs/openapi.json"
api.add_route("/docs/openapi.json", OpenAPIJSONResource())
register_swaggerui_app(
api,
SWAGGERUI_URL,
SCHEMA_URL,
config={
"supportedSubmitMethods": ["get", "post"],
},
)
# vim: set sw=4 ts=4 expandtab: