def test_group_merge_candidates_same_desc():
# merge based on same description
command = merge_documents.Command()
doc1 = Document.objects.create(description="a marriage contract")
doc2 = Document.objects.create(description=doc1.description)
shelfmark_id = "shelfmark / letter"
report_rows = command.group_merge_candidates(
{
shelfmark_id: [doc1, doc2],
}
)
assert report_rows[0] == [
shelfmark_id,
1,
"MERGE",
"all descriptions match",
"primary",
doc1.pk,
doc1.description,
]
assert report_rows[1] == [
shelfmark_id,
1,
"MERGE",
"all descriptions match",
"merge",
doc2.pk,
doc2.description,
]
def test_merge_group_error_handling(mock_merge_with, document, join):
stderr = StringIO()
# no primary
command = merge_documents.Command(stderr=stderr)
rval = command.merge_group("a", [])
assert "Could not identify primary document" in stderr.getvalue()
assert rval == 0
# no merge docs
command.merge_group("a", [{"pgpid": "1", "role": "primary"}])
assert "No merge documents" in stderr.getvalue()
# primary does not exist in the database
command.merge_group(
"a",
[
{"pgpid": 14151, "role": "primary", "status": "test merge"},
{"pgpid": 23426, "role": ""},
],
)
assert "Primary document 14151 not found" in stderr.getvalue()
# not all merge documents exist in the database
command.merge_group(
"a",
[
{"pgpid": document.id, "role": "primary", "status": "test merge"},
{"pgpid": 23043, "role": ""},
{"pgpid": join.id, "role": ""},
],
)
assert "Not all merge documents found" in stderr.getvalue()
def test_handle_report(mock_generate_report, mock_group_merge, mock_get_merge):
cmd = merge_documents.Command()
cmd.handle(mode="report", path="test-report.csv")
# confirm that the correct sequence of methods is called
mock_get_merge.assert_any_call()
mock_group_merge.assert_called_with(mock_get_merge.return_value)
mock_generate_report.assert_called_with(
mock_group_merge.return_value, "test-report.csv"
)
def test_generate_report(tmpdir):
report_path = tmpdir.join("report.csv")
command = merge_documents.Command()
test_rows = [["a", "b", "c", "d", "e", "f"], ["g", "h", "j", "k", "l", "m"]]
command.generate_report(test_rows, report_path)
report = report_path.read()
assert "merge group,group id,action,status,role,pgpid,description" in report
for row in test_rows:
assert ",".join(row) in report
def test_handle_file_not_found(tmpdir):
command = merge_documents.Command()
with pytest.raises(CommandError) as err:
command.handle(mode="merge", path="/tmp/example/not-here.csv")
assert "Report file not found: /tmp/example/not-here.csv" in str(err)
# test with default path
# change working directory to tmpdir to ensure file is not found
os.chdir(tmpdir)
stderr = StringIO()
with pytest.raises(CommandError) as err:
call_command("merge_documents", "merge", stderr=stderr)
assert "Report file not found: merge-report.csv" in str(err)
def test_group_merge_candidates_see_pgpid():
# merge based on "see pgpid" text in secondary documents
command = merge_documents.Command()
doc1 = Document.objects.create(description="a marriage contract", pk=134552)
doc2 = Document.objects.create(description="See PGPID %d" % doc1.pk)
shelfmark_id = "shelfmark / unknown"
report_rows = command.group_merge_candidates(
{
shelfmark_id: [doc1, doc2],
}
)
# status should be merge; rationale should be "see join"
assert report_rows[0][2] == "MERGE"
assert report_rows[0][3] == "see PGPID"
def test_group_merge_candidates_empty_description():
# merge based on empty description text in secondary documents
command = merge_documents.Command()
doc1 = Document.objects.create(description="a marriage contract")
doc2 = Document.objects.create()
shelfmark_id = "shelfmark / letter"
report_rows = command.group_merge_candidates(
{
shelfmark_id: [doc1, doc2],
}
)
# status should be merge; rationale from empty description
assert report_rows[0][2] == "MERGE"
assert report_rows[0][3] == "one description, others empty"
def test_merge_group(mock_merge_with, document, join):
stderr = StringIO()
command = merge_documents.Command(stderr=stderr)
rval = command.merge_group(
"a",
[
{"pgpid": document.id, "role": "primary", "status": "test merge"},
{"pgpid": join.id, "role": ""},
],
)
assert rval == 1
# inspect the call (can't use assert because document eqaulity check fails)
args, kwargs = mock_merge_with.call_args
assert args[0][0].pk == join.pk
assert args[1] == "test merge"
def test_group_merge_candidates_see_join():
# merge based on "see join" text in secondary documents
command = merge_documents.Command()
doc1 = Document.objects.create(description="a marriage contract")
doc2 = Document.objects.create(description="See join.")
shelfmark_id = "shelfmark / letter"
report_rows = command.group_merge_candidates(
{
shelfmark_id: [doc1, doc2],
}
)
# status should be merge; rationale should be "see join"
assert report_rows[0][2] == "MERGE"
assert report_rows[0][3] == "see join"
assert report_rows[0][4] == "primary"
assert report_rows[0][5] == doc1.pk
def test_handle_merge(mock_load_report, mock_merge_group):
stdout = StringIO()
path = "merge-groups.csv"
cmd = merge_documents.Command(stdout=stdout)
# return some mock groups from load method
mock_load_report.return_value = [
# group id, group
("a", [1, 2, 3]),
("b", [4, 5, 5]),
]
# report success, then failure from merge method
mock_merge_group.side_effect = (1, 0)
cmd.handle(mode="merge", path=path)
# confirm that the correct sequence of methods is called
mock_load_report.assert_called_with(path)
for group_id, docs in mock_load_report.return_value:
mock_merge_group.assert_any_call(group_id, docs)
output = stdout.getvalue()
assert "Successfully merged 1 group" in output
assert "skipped 1" in output
def test_get_merge_candidates(fragment, multifragment, join):
# looks for more than one document on the same set of fragments
# join is a document associated with fragment & multifragment
# create two other documents to be merged
doc2 = Document.objects.create(
description="see other fragment", doctype=join.doctype
)
# associate in same order as join
TextBlock.objects.create(document=doc2, fragment=fragment, order=1)
TextBlock.objects.create(document=doc2, fragment=multifragment, order=2)
doc3 = Document.objects.create(
description="see other fragment", doctype=join.doctype
)
# associate in different order as join doc
TextBlock.objects.create(document=doc3, fragment=fragment, order=2)
TextBlock.objects.create(document=doc3, fragment=multifragment, order=1)
# doc on the same fragments with different type (unknown)
unknown_doc = Document.objects.create(
description="something else",
)
# associate in same order as join doc
TextBlock.objects.create(document=unknown_doc, fragment=fragment, order=1)
TextBlock.objects.create(document=unknown_doc, fragment=multifragment, order=3)
command = merge_documents.Command()
candidates = command.get_merge_candidates()
# should result in one candidate group
assert len(candidates) == 1
# key should be shelfmark + type
shelfmark_type = "%s / %s" % (join.shelfmark, join.doctype.name)
assert shelfmark_type in candidates
# should not include document with different type
assert len(candidates[shelfmark_type]) == 3
for doc in [join, doc2, doc3]:
assert doc in candidates[shelfmark_type]