def test_iter(self):
"""
the iteration method should handle pagination automatically
based on rows and max_rows; use the mock solr server to query
data row by row and ensure that the data return as expected
"""
sq = SearchQuery(q="unittest", rows=1, max_pages=20)
with MockSolrResponse(sq.HTTP_ENDPOINT):
self.assertEqual(sq._query['start'], 0)
self.assertEqual(next(sq).bibcode, '1971Sci...174..142S')
self.assertEqual(len(sq.articles), 1)
self.assertEqual(sq._query['start'], 1)
self.assertEqual(next(sq).bibcode, '2012GCN..13229...1S')
self.assertEqual(len(list(sq)), 19) # 2 already returned
with self.assertRaisesRegexp(
StopIteration,
"Maximum number of pages queried"):
next(sq)
sq.max_pages = 500
self.assertEqual(len(list(sq)), 28-19-2)
with self.assertRaisesRegexp(
StopIteration,
"All records found"):
next(sq)
def test_init(self):
"""
init should result in a properly formatted query attribute
"""
with self.assertRaisesRegexp(AssertionError, "q must not be empty"):
SearchQuery()
sq = SearchQuery(q="star")
self.assertIn("star", sq.query['q'])
sq = SearchQuery(title="t", author="ln, fn")
# In Python 3, assertItemsEqual is named assertCountEqual
assertItemsEqual = self.assertItemsEqual if sys.version_info[0] == 2 \
else self.assertCountEqual
assertItemsEqual(
sq.query['q'].split(),
'title:"t" author:"ln, fn"'.split(),
)
sq = SearchQuery(q="star", aff="institute")
assertItemsEqual(
sq.query['q'].split(),
'aff:"institute" star'.split(),
)
def test_no_highlights(self):
"""
Test when there are no highlights
"""
sq = SearchQuery(q='star', fl=['bibcode'], hl=['abstract'])
with MockSolrResponse(SEARCH_URL):
p = list(sq)[1]
highlights = sq.highlights(p)
self.assertEqual(highlights, {})
def test_no_highlights(self):
"""
Test when there are no highlights
"""
sq = SearchQuery(q='star', fl=['bibcode'], hl=['abstract'])
with MockSolrResponse(SEARCH_URL):
p = list(sq)[1]
highlights = sq.highlights(p)
self.assertEqual(highlights, {})
def test_get_highlight(self):
"""
Test can retrieve a highlight for a given bibcode for a given query
"""
sq = SearchQuery(q='star')
self.assertNotIn('hl', sq.query)
self.assertNotIn('hl.fl', sq.query)
sq = SearchQuery(q='star', fl=['bibcode'], hl=['abstract'])
self.assertEqual(sq.query['hl'], 'true')
self.assertEqual(sq.query['hl.fl'], ['abstract'])
with MockSolrResponse(SEARCH_URL):
p = list(sq)[0]
highlights = sq.highlights(p)
self.assertEqual(highlights, {'abstract': 'astronomy abstract'})
def test_print_methods(self):
"""
the class should return a user-friendly formatted identified when the
__str__ or __unicode__ methods are called
"""
self.assertEqual(
'<Sudilovsky, V. et al. 2013, 2013A&A...552A.143S>',
self.article.__str__()
)
self.assertEqual(self.article.__unicode__(), self.article.__str__())
self.article.first_author = None
self.article.author = None
self.article.bibcode = None
self.article.year = None
self.assertEqual(
self.article.__str__(),
"<Unknown author Unknown year, Unknown bibcode>"
)
# accessing print methods should use lazy loaded attributes: issue#98
sq = SearchQuery(q="unittest", rows=1, fl=["bibcode"])
with MockSolrResponse(sq.HTTP_ENDPOINT):
article = list(sq)[0]
self.assertEqual(article.__unicode__(), '<Sudilovsky, Oscar et al. 1971, 1971Sci...174..142S>')
def test_get_highlight(self):
"""
Test can retrieve a highlight for a given bibcode for a given query
"""
sq = SearchQuery(q='star')
self.assertNotIn('hl', sq.query)
self.assertNotIn('hl.fl', sq.query)
sq = SearchQuery(q='star', fl=['bibcode'], hl=['abstract'])
self.assertEqual(sq.query['hl'], 'true')
self.assertEqual(sq.query['hl.fl'], ['abstract'])
with MockSolrResponse(SEARCH_URL):
p = list(sq)[0]
highlights = sq.highlights(p)
self.assertEqual(highlights, {'abstract': 'astronomy abstract'})
def test_incorrect_highlight_fl(self):
"""
Test when user passes incorrect fields it raises an exception
"""
with self.assertRaises(Exception):
SearchQuery(q='star',
fl=['bibcode'],
hl=['foo', 'bar', 'abstract', 'abstract'])
def test_iter(self):
"""
the iteration method should handle pagination automatically
based on rows and max_rows; use the mock solr server to query
data row by row and ensure that the data return as expected
"""
sq = SearchQuery(q="unittest", rows=1, max_pages=20, start=0)
with MockSolrResponse(sq.HTTP_ENDPOINT):
self.assertEqual(sq._query['start'], 0)
self.assertEqual(next(sq).bibcode, '1971Sci...174..142S')
self.assertEqual(len(sq.articles), 1)
self.assertEqual(sq._query['start'], 1)
self.assertEqual(next(sq).bibcode, '2012GCN..13229...1S')
self.assertEqual(len(list(sq)), 18) # 2 already returned
with six.assertRaisesRegex(self, StopIteration,
"Maximum number of pages queried"):
next(sq)
sq.max_pages = 500
self.assertEqual(len(list(sq)), 28 - 18 - 2)
with six.assertRaisesRegex(self, StopIteration,
"All records found"):
next(sq)
# not setting max_pages should return the exact number of rows requests
sq = SearchQuery(q="unittest", rows=3)
with MockSolrResponse(sq.HTTP_ENDPOINT):
self.assertEqual(len(list(sq)), 3)
# Default should use cursorMark
sq = SearchQuery(q="unittest", sort="date")
with MockSolrResponse(sq.HTTP_ENDPOINT):
self.assertEqual(next(sq).bibcode, '1971Sci...174..142S')
self.assertEqual(sq.query['cursorMark'],
sq.response.json['nextCursorMark'])
def test_iter(self):
"""
the iteration method should handle pagination automatically
based on rows and max_rows; use the mock solr server to query
data row by row and ensure that the data return as expected
"""
sq = SearchQuery(q="unittest", rows=1, max_pages=20, start=0)
with MockSolrResponse(sq.HTTP_ENDPOINT):
self.assertEqual(sq._query['start'], 0)
self.assertEqual(next(sq).bibcode, '1971Sci...174..142S')
self.assertEqual(len(sq.articles), 1)
self.assertEqual(sq._query['start'], 1)
self.assertEqual(next(sq).bibcode, '2012GCN..13229...1S')
self.assertEqual(len(list(sq)), 18) # 2 already returned
with self.assertRaisesRegexp(
StopIteration,
"Maximum number of pages queried"):
next(sq)
sq.max_pages = 500
self.assertEqual(len(list(sq)), 28-18-2)
with self.assertRaisesRegexp(
StopIteration,
"All records found"):
next(sq)
# not setting max_pages should return the exact number of rows requests
sq = SearchQuery(q="unittest", rows=3)
with MockSolrResponse(sq.HTTP_ENDPOINT):
self.assertEqual(len(list(sq)), 3)
# Default should use cursorMark
sq = SearchQuery(q="unittest", sort="date")
with MockSolrResponse(sq.HTTP_ENDPOINT):
self.assertEqual(next(sq).bibcode, '1971Sci...174..142S')
self.assertEqual(
sq.query['cursorMark'], sq.response.json['nextCursorMark']
)
def test_get_field_bibtex(self):
"""
should emit a warning when calling _get_field with bibtex but otherwise work
as expected both in the case of fl=bibtex and fl=None
"""
for fl in [None, "bibtex"]:
sq = SearchQuery(q="*", fl=fl)
with warnings.catch_warnings(record=True) as w:
with MockSolrResponse(SEARCH_URL):
with MockExportResponse("{base}bibtex".format(base=EXPORT_URL)):
article = next(sq)
article.bibtex
if six.PY3:
msg = w[1].message.args[0]
self.assertEqual(msg, "bibtex should be queried with ads.ExportQuery(); "
"You will hit API ratelimits very quickly otherwise.")
self.assertTrue(article.bibtex.startswith("@ARTICLE{2013A&A...552A.143S"))
def test_rows_rewrite(self):
"""
if the responseHeader "rows" is not the same as the query's "rows",
the query's "rows" should be re-written and a warning should be emitted
"""
sq = SearchQuery(q="unittest", rows=10e6)
with MockSolrResponse(sq.HTTP_ENDPOINT):
self.assertEqual(sq.query['rows'], 10e6)
with warnings.catch_warnings(record=True) as w:
next(sq)
if six.PY3:
msg = w[-1].message.args[0]
elif six.PY2:
msg = w[-1].message.message
self.assertEqual(
msg,
"Response rows did not match input rows. Setting this query's rows to 300"
)
self.assertEqual(sq.query['rows'], 300)
def test_init(self):
"""
init should result in a properly formatted query attribute
"""
with six.assertRaisesRegex(self, AssertionError,
"q must not be empty"):
SearchQuery()
sq = SearchQuery(q="star")
self.assertIn("star", sq.query['q'])
sq = SearchQuery(title="t", author="ln, fn")
# In Python 3, assertItemsEqual is named assertCountEqual
assertItemsEqual = self.assertItemsEqual if sys.version_info[0] == 2 \
else self.assertCountEqual
assertItemsEqual(
sq.query['q'].split(),
'title:"t" author:"ln, fn"'.split(),
)
sq = SearchQuery(q="star", aff="institute")
assertItemsEqual(
sq.query['q'].split(),
'aff:"institute" star'.split(),
)
sq = SearchQuery(q="star", token="test-token")
self.assertEqual(sq.token, "test-token")
self.assertEqual(sq._token, "test-token")
sq = SearchQuery(q="star", sort="date")
self.assertEqual(sq.query['cursorMark'], '*')
self.assertNotIn('start', sq.query)
self.assertEqual(sq.query['sort'], 'date desc,id desc')
sq = SearchQuery(q="star", start=0)
self.assertEqual(sq.query['start'], 0)
self.assertNotIn('cursorMark', sq.query)
self.assertEqual(sq.query['sort'], "score desc")
sq = SearchQuery(q="star", start=0, sort="date")
self.assertEqual(sq.query['start'], 0)
self.assertEqual(sq.query['sort'], "date desc")
sq = SearchQuery({"q": "star"})
self.assertEqual(sq.query['cursorMark'], "*")
self.assertEqual(sq.query['sort'], "score desc,id desc")
self.assertNotIn('start', sq.query)
# test unicode
sq = SearchQuery(author=u'é')
self.assertIn(u'é', sq.query['q'])
with six.assertRaisesRegex(self, AssertionError,
".+mutually exclusive.+"):
SearchQuery(q="start", start=0, cursorMark="*")
# test that bibtex/metrics is excluded from sq.query['fl']
sq = SearchQuery(q="star", fl=["f1", "bibtex", "f2", "metrics", "f3"])
self.assertEqual(sq.query['fl'], ["id", "f1", "f2", "f3"])
