From dc8c77d72bb5404d0fcb3d25e510405769457443 Mon Sep 17 00:00:00 2001 From: "Mark E. Haase" Date: Thu, 5 May 2022 12:36:19 -0400 Subject: [PATCH 1/2] Fix JSON export for reports (#175) JSON exports were being rendered through an HTML template due to content negotiation in the browser (e.g. Accept:text/html). This commit forces the renderer to use JSON regardless of content negotiation and also cleans up the json and docx export code. --- src/tram/views.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/src/tram/views.py b/src/tram/views.py index 14352435f0..552f16e3cf 100644 --- a/src/tram/views.py +++ b/src/tram/views.py @@ -15,8 +15,8 @@ ) from django.shortcuts import render from django.views.decorators.http import require_POST -from rest_framework import viewsets -from rest_framework.decorators import api_view +from rest_framework import renderers, viewsets +from rest_framework.decorators import action, api_view from rest_framework.response import Response import tram.report.docx From d4e968e36f0ac70690afaac65e335cf3d226ac6d Mon Sep 17 00:00:00 2001 From: "Mark E. Haase" Date: Thu, 19 May 2022 15:56:29 -0400 Subject: [PATCH 2/2] Fix JSON exports for reports (#175) - take two Based on Nick's feedback, reverted to using two ViewSets, moved the docx code into a custom renderer, and use Django's built in format parameter instead of type. This makes for a much cleaner PR. --- README.md | 2 +- src/tram/renderers.py | 29 +++++++++++++ src/tram/templates/index.html | 6 +-- src/tram/urls.py | 4 +- src/tram/views.py | 76 +++++++++++++---------------------- tests/tram/test_views.py | 22 +++++----- 6 files changed, 73 insertions(+), 66 deletions(-) create mode 100644 src/tram/renderers.py diff --git a/README.md b/README.md index f2b570d6d1..3186ed622d 100644 --- a/README.md +++ b/README.md @@ -154,7 +154,7 @@ docker.errors.DockerException: Error while fetching server API version: ('Connec [97438] Failed to execute script docker-compose ``` -Then most likely Docker is not running and you need to start Docker. +Then most likely reason is that Docker is not running and you need to start it. ## Report Troubleshooting diff --git a/src/tram/renderers.py b/src/tram/renderers.py new file mode 100644 index 0000000000..7d4e325e44 --- /dev/null +++ b/src/tram/renderers.py @@ -0,0 +1,29 @@ +import io + +from rest_framework import renderers + +import tram.report.docx + + +class DocxReportRenderer(renderers.BaseRenderer): + """This custom renderer exports mappings into Word .docx format.""" + + media_type = ( + "application/vnd.openxmlformats-officedocument.wordprocessingml.document" + ) + format = "docx" + + def render(self, data, accepted_media_type=None, renderer_context=None): + """ + Export report mappings into Word .docx format. + + :param data: the report mappings dict + :param accepted_media_type: the content type negotiated by DRF + :param renderer_context: optional additional data + :returns bytes: .docx binary data + """ + document = tram.report.docx.build(data) + buffer = io.BytesIO() + document.save(buffer) + buffer.seek(0) + return buffer.read() diff --git a/src/tram/templates/index.html b/src/tram/templates/index.html index 2cab7722b5..eafe8832b4 100644 --- a/src/tram/templates/index.html +++ b/src/tram/templates/index.html @@ -71,10 +71,8 @@

Reports

Export
- JSON - DOCX + JSON + DOCX
{% if report.document_id is not None %} diff --git a/src/tram/urls.py b/src/tram/urls.py index 003db73e87..c326b0c188 100644 --- a/src/tram/urls.py +++ b/src/tram/urls.py @@ -29,7 +29,9 @@ router.register(r"jobs", views.DocumentProcessingJobViewSet) router.register(r"mappings", views.MappingViewSet) router.register(r"reports", views.ReportViewSet) -router.register(r"report-export", views.ReportExportViewSet) +router.register( + r"report-mappings", views.ReportMappingViewSet, basename="report-mapping" +) router.register(r"sentences", views.SentenceViewSet) diff --git a/src/tram/views.py b/src/tram/views.py index 552f16e3cf..5dbdb3d932 100644 --- a/src/tram/views.py +++ b/src/tram/views.py @@ -1,4 +1,3 @@ -import io import json import logging import time @@ -6,20 +5,13 @@ from constance import config from django.contrib.auth.decorators import login_required -from django.http import ( - Http404, - HttpResponse, - HttpResponseBadRequest, - JsonResponse, - StreamingHttpResponse, -) +from django.http import Http404, HttpResponse, HttpResponseBadRequest, JsonResponse from django.shortcuts import render from django.views.decorators.http import require_POST from rest_framework import renderers, viewsets -from rest_framework.decorators import action, api_view +from rest_framework.decorators import api_view from rest_framework.response import Response -import tram.report.docx from tram import serializers from tram.ml import base from tram.models import ( @@ -30,6 +22,7 @@ Report, Sentence, ) +from tram.renderers import DocxReportRenderer logger = logging.getLogger(__name__) @@ -62,56 +55,41 @@ class ReportViewSet(viewsets.ModelViewSet): serializer_class = serializers.ReportSerializer -class ReportExportViewSet(viewsets.ModelViewSet): - queryset = Report.objects.all() +class ReportMappingViewSet(viewsets.ModelViewSet): + """ + This viewset provides access to report mappings. + """ + serializer_class = serializers.ReportExportSerializer + renderer_classes = [renderers.JSONRenderer, DocxReportRenderer] def get_queryset(self): - queryset = ReportViewSet.queryset + """ + Override parent implementation to support lookup by document ID. + """ + queryset = Report.objects.all() document_id = self.request.query_params.get("doc-id", None) if document_id: queryset = queryset.filter(document__id=document_id) return queryset - def retrieve(self, request, *args, **kwargs): - - report_format = request.GET.get("type", "") - - # If an invalid report_format is given, just default to json - if report_format not in ["json", "docx"]: - report_format = "json" - logger.warning("Invalid File Type. Defaulting to JSON.") - - # Retrieve report data as json - response = super().retrieve(request, *args, **kwargs) - basename = quote(self.get_object().name, safe="") + def retrieve(self, request, pk=None): + """ + Get the mappings for a report. - if report_format == "json": - response["Content-Disposition"] = f'attachment; filename="{basename}.json"' - - elif report_format == "docx": - # Uses json dictionary to create formatted document - document = tram.report.docx.build(response.data) - - # save document info - buffer = io.BytesIO() - document.save(buffer) # save your memory stream - buffer.seek(0) # rewind the stream - - # put them to streaming content response within docx content_type - content_type = ( - "application/" - "vnd.openxmlformats-officedocument.wordprocessingml.document" - ) - response = StreamingHttpResponse( - streaming_content=buffer, # use the stream's content - content_type=content_type, - ) - - response["Content-Disposition"] = f'attachment; filename="{basename}.docx"' - response["Content-Encoding"] = "UTF-8" + Overrides the parent implementation to add a Content-Disposition header + so that the browser will download instead of displaying inline. + :param request: HTTP request + :param pk: primary key of a report + """ + response = super().retrieve(request, request, pk) + report = self.get_object() + filename = "{}.{}".format( + quote(report.name, safe=""), request.accepted_renderer.format + ) + response["Content-Disposition"] = f'attachment; filename="{filename}"' return response diff --git a/tests/tram/test_views.py b/tests/tram/test_views.py index 9e41208a94..58088093b8 100644 --- a/tests/tram/test_views.py +++ b/tests/tram/test_views.py @@ -254,28 +254,28 @@ def test_get_sentences_by_technique(self, logged_in_client): @pytest.mark.django_db -class TestReportExport: - def test_get_report_export_succeeds(self, logged_in_client, mapping): +class TestReportMappings: + def test_get_json(self, logged_in_client, mapping): # Act - response = logged_in_client.get("/api/report-export/1/") + response = logged_in_client.get("/api/report-mappings/1/?format=json") json_response = json.loads(response.content) # Assert - assert "sentences" in json_response + assert json_response["id"] == 1 assert len(json_response["sentences"][0]["mappings"]) == 1 - def test_export_docx_report(self, logged_in_client, mapping): + def test_get_docx(self, logged_in_client, mapping): """ Check that something that looks like a Word doc was returned. There are separate unit tests for the doc's content. """ # Act - response = logged_in_client.get("/api/report-export/1/?type=docx") - data = list(response.streaming_content) + response = logged_in_client.get("/api/report-mappings/1/?format=docx") + data = response.content # Assert - assert data[0].startswith(b"PK\x03\x04") + assert data.startswith(b"PK\x03\x04") def test_bootstrap_training_data_can_be_posted_as_json_report( self, logged_in_client @@ -286,7 +286,7 @@ def test_bootstrap_training_data_can_be_posted_as_json_report( # Act response = logged_in_client.post( - "/api/report-export/", json_string, content_type="application/json" + "/api/report-mappings/", json_string, content_type="application/json" ) # Assert @@ -295,7 +295,7 @@ def test_bootstrap_training_data_can_be_posted_as_json_report( def test_report_export_update_not_implemented(self, logged_in_client): # Act response = logged_in_client.post( - "/api/report-export/1/", "{}", content_type="application/json" + "/api/report-mappings/1/", "{}", content_type="application/json" ) # Assert @@ -311,7 +311,7 @@ def test_download_original_report(self, logged_in_client, document): def test_get_reports_by_doc_id(self, logged_in_client, report_with_document): # Act doc_id = report_with_document.document.id - response = logged_in_client.get(f"/api/report-export/?doc-id={doc_id}") + response = logged_in_client.get(f"/api/report-mappings/?doc-id={doc_id}") json_response = json.loads(response.content) # Assert