def test_shelfmark_partialmatch(self, empty_solr, multifragment):
# integration test for shelfmark indexing with partial matching
# - using empty solr fixture to ensure solr is empty when this test starts
# multifragment shelfmark can test for this problem: T-S 16.377
doc1 = Document.objects.create()
TextBlock.objects.create(document=doc1, fragment=multifragment)
# create an arbitrary fragment with similar numeric shelfmark
folder_fragment = Fragment.objects.create(shelfmark="T-S 16.378")
doc2 = Document.objects.create()
TextBlock.objects.create(document=doc2, fragment=folder_fragment)
# ensure solr index is updated with the two test documents
SolrClient().update.index(
[
doc1.index_data(),
doc2.index_data(),
],
commit=True,
)
docsearch_view = DocumentSearchView()
docsearch_view.request = Mock()
# sort doesn't matter in this case
docsearch_view.request.GET = {"q": "T-S 16"}
qs = docsearch_view.get_queryset()
# should return both documents
assert qs.count() == 2
resulting_ids = [result["pgpid"] for result in qs]
assert doc1.id in resulting_ids
assert doc2.id in resulting_ids
def test_get_paginate_by(self):
"""Should set pagination to 2 per page"""
docsearch_view = DocumentSearchView(kwargs={})
docsearch_view.request = Mock()
docsearch_view.request.GET = {"per_page": "2"}
qs = docsearch_view.get_queryset()
assert docsearch_view.get_paginate_by(qs) == 2
def test_get_context_data(self, mock_get_queryset, rf, mock_solr_queryset):
with patch(
"geniza.corpus.views.DocumentSolrQuerySet",
new=mock_solr_queryset(
DocumentSolrQuerySet, extra_methods=["admin_search", "keyword_search"]
),
) as mock_queryset_cls:
mock_qs = mock_queryset_cls.return_value
mock_qs.count.return_value = 22
mock_qs.get_facets.return_value.facet_fields = {}
# mock_qs.__getitem__.return_value = docsearch_view.queryset
mock_get_queryset.return_value = mock_qs
docsearch_view = DocumentSearchView(kwargs={})
docsearch_view.queryset = mock_qs
docsearch_view.object_list = mock_qs
docsearch_view.request = rf.get("/documents/")
context_data = docsearch_view.get_context_data()
assert (
context_data["highlighting"]
== context_data["page_obj"].object_list.get_highlighting.return_value
)
assert context_data["page_obj"].start_index() == 0
def test_get_form_kwargs(self):
docsearch_view = DocumentSearchView()
docsearch_view.request = Mock()
# no params
docsearch_view.request.GET = {}
assert docsearch_view.get_form_kwargs() == {
"initial": {
"sort": "scholarship_desc",
},
"prefix": None,
"data": {"sort": "scholarship_desc"},
}
# keyword search param
docsearch_view.request.GET = {"q": "contract"}
assert docsearch_view.get_form_kwargs() == {
"initial": {"sort": "scholarship_desc"},
"prefix": None,
"data": {
"q": "contract",
"sort": "relevance",
},
}
# sort search param
docsearch_view.request.GET = {"sort": "scholarship_desc"}
assert docsearch_view.get_form_kwargs() == {
"initial": {
"sort": "scholarship_desc",
},
"prefix": None,
"data": {
"sort": "scholarship_desc",
},
}
# keyword and sort search params
docsearch_view.request.GET = {"q": "contract", "sort": "scholarship_desc"}
assert docsearch_view.get_form_kwargs() == {
"initial": {
"sort": "scholarship_desc",
},
"prefix": None,
"data": {
"q": "contract",
"sort": "scholarship_desc",
},
}
def test_shelfmark_boost(self, empty_solr, document, multifragment):
# integration test for shelfmark field boosting
# in solr configuration
# - using empty solr fixture to ensure solr is empty when this test starts
# create a second document with a different shelfmark
# that references the shelfmark of the first in the description
related_doc = Document.objects.create(
description="See also %s" % document.shelfmark
)
TextBlock.objects.create(document=related_doc, fragment=multifragment)
# third document with similar shelfmark
frag = Fragment.objects.create(
shelfmark="CUL Add.300", # fixture has shelfmark CUL Add.2586
)
neighbor_doc = Document.objects.create()
TextBlock.objects.create(document=neighbor_doc, fragment=frag)
# ensure solr index is updated with all three test documents
SolrClient().update.index(
[
document.index_data(),
neighbor_doc.index_data(),
related_doc.index_data(),
],
commit=True,
)
docsearch_view = DocumentSearchView()
docsearch_view.request = Mock()
# assuming relevance sort is default; update if that changes
docsearch_view.request.GET = {"q": document.shelfmark, "sort": "relevance"}
qs = docsearch_view.get_queryset()
# should return all three documents
assert qs.count() == 3
# document with exact match on shelfmark should be returned first
assert (
qs[0]["pgpid"] == document.id
), "document with matching shelfmark returned first"
# document with full shelfmark should in description should be second
assert (
qs[1]["pgpid"] == related_doc.id
), "document with shelfmark in description returned second"
def test_ignore_suppressed_documents(self, document, empty_solr):
suppressed_document = Document.objects.create(status=Document.SUPPRESSED)
Document.index_items([document, suppressed_document])
SolrClient().update.index([], commit=True)
# [d.index_data() for d in [document, suppressed_document]], commit=True
# )
print(suppressed_document.index_data())
docsearch_view = DocumentSearchView()
# mock request with empty keyword search
docsearch_view.request = Mock()
docsearch_view.request.GET = {"q": ""}
qs = docsearch_view.get_queryset()
result_pgpids = [obj["pgpid"] for obj in qs]
print(result_pgpids)
print(qs)
assert qs.count() == 1
assert document.id in result_pgpids
assert suppressed_document.id not in result_pgpids
def test_doctype_filter(self, document, join, empty_solr):
"""Integration test for document type filter"""
SolrClient().update.index(
[
document.index_data(), # type = Legal document
join.index_data(), # type = Letter
],
commit=True,
)
docsearch_view = DocumentSearchView()
docsearch_view.request = Mock()
# no filter
docsearch_view.request.GET = {}
qs = docsearch_view.get_queryset()
assert qs.count() == 2
# filter by doctype "Legal document"
docsearch_view.request.GET = {"doctype": ["Legal document"]}
qs = docsearch_view.get_queryset()
assert qs.count() == 1
assert qs[0]["pgpid"] == document.id, "Only legal document returned"
def test_scholarship_sort(
self,
document,
join,
empty_solr,
source,
twoauthor_source,
multiauthor_untitledsource,
):
"""integration test for sorting by scholarship asc and desc"""
Footnote.objects.create(
content_object=join,
source=source,
doc_relation=Footnote.EDITION,
)
doc_three_records = Document.objects.create(
description="testing description",
)
for src in [source, twoauthor_source, multiauthor_untitledsource]:
Footnote.objects.create(
content_object=doc_three_records,
source=src,
doc_relation=Footnote.EDITION,
)
# ensure solr index is updated with all three test documents
SolrClient().update.index(
[
document.index_data(), # no scholarship records
join.index_data(), # one scholarship record
doc_three_records.index_data(), # 3 scholarship records
],
commit=True,
)
docsearch_view = DocumentSearchView()
docsearch_view.request = Mock()
# no sort, no query
docsearch_view.request.GET = {}
qs = docsearch_view.get_queryset()
# should return all three documents
assert qs.count() == 3
# by default, should return document with most records first
assert (
qs[0]["pgpid"] == doc_three_records.id
), "document with most scholarship records returned first"
# sort by scholarship desc
docsearch_view.request.GET = {"sort": "scholarship_desc"}
qs = docsearch_view.get_queryset()
# should return document with most records first
assert (
qs[0]["pgpid"] == doc_three_records.id
), "document with most scholarship records returned first"
# sort by scholarship asc
docsearch_view.request.GET = {"sort": "scholarship_asc"}
qs = docsearch_view.get_queryset()
# should return document with fewest records first
assert (
qs[0]["pgpid"] == document.id
), "document with fewest scholarship records returned first"
# sort by scholarship asc with query
docsearch_view.request.GET = {"sort": "scholarship_asc", "q": "testing"}
qs = docsearch_view.get_queryset()
# should return 2 documents
assert qs.count() == 2
# should return document with fewest records first
assert (
qs[0]["pgpid"] == join.id
), "document with matching description and fewest scholarship records returned first"
def test_get_queryset(self, mock_solr_queryset):
with patch(
"geniza.corpus.views.DocumentSolrQuerySet",
new=self.mock_solr_queryset(
DocumentSolrQuerySet, extra_methods=["admin_search", "keyword_search"]
),
) as mock_queryset_cls:
docsearch_view = DocumentSearchView()
docsearch_view.request = Mock()
# keyword search param
docsearch_view.request.GET = {"q": "six apartments"}
qs = docsearch_view.get_queryset()
mock_queryset_cls.assert_called_with()
mock_sqs = mock_queryset_cls.return_value
mock_sqs.keyword_search.assert_called_with("six apartments")
mock_sqs.keyword_search.return_value.highlight.assert_any_call(
"description", snippets=3, method="unified", requireFieldMatch=True
)
mock_sqs.also.assert_called_with("score")
mock_sqs.also.return_value.order_by.assert_called_with("-score")
# sort search param
mock_sqs.reset_mock()
docsearch_view.request = Mock()
docsearch_view.request.GET = {"sort": "relevance"}
qs = docsearch_view.get_queryset()
mock_sqs = mock_queryset_cls.return_value
mock_sqs.keyword_search.assert_not_called()
# filter called once to limit by status
assert mock_sqs.filter.call_count == 1
mock_sqs.filter.assert_called_with(status=Document.STATUS_PUBLIC)
# order_by should not be called when there is no search query
mock_sqs.order_by.assert_not_called()
# sort and keyword search params
mock_sqs.reset_mock()
docsearch_view.request = Mock()
docsearch_view.request.GET = {"q": "six apartments", "sort": "relevance"}
qs = docsearch_view.get_queryset()
mock_sqs = mock_queryset_cls.return_value
mock_sqs.keyword_search.assert_called_with("six apartments")
mock_sqs.keyword_search.return_value.also.return_value.order_by.return_value.filter.assert_called_with(
status=Document.STATUS_PUBLIC
)
mock_sqs.keyword_search.return_value.also.return_value.order_by.assert_called_with(
"-score"
)
# keyword, sort, and doctype filter search params
mock_sqs.reset_mock()
docsearch_view.request = Mock()
docsearch_view.request.GET = {
"q": "six apartments",
"sort": "scholarship_desc",
"doctype": ["Legal"],
}
qs = docsearch_view.get_queryset()
mock_sqs = mock_queryset_cls.return_value
mock_sqs.keyword_search.assert_called_with("six apartments")
mock_sqs.keyword_search.return_value.also.return_value.order_by.return_value.filter.assert_called()
mock_sqs.keyword_search.return_value.also.return_value.order_by.assert_called_with(
"-scholarship_count_i"
)
# empty params
mock_sqs.reset_mock()
docsearch_view.request = Mock()
docsearch_view.request.GET = {"q": "", "sort": ""}
qs = docsearch_view.get_queryset()
mock_sqs = mock_queryset_cls.return_value
mock_sqs.keyword_search.assert_not_called()
mock_sqs.order_by.assert_called_with("-scholarship_count_i")
# no params
mock_sqs.reset_mock()
docsearch_view.request = Mock()
docsearch_view.request.GET = {}
qs = docsearch_view.get_queryset()
mock_sqs = mock_queryset_cls.return_value
mock_sqs.keyword_search.assert_not_called()
mock_sqs.order_by.assert_called_with("-scholarship_count_i")