safety-screen/Nominatim
1
0
Fork 0
forked from hans/Nominatim
Code Pull Requests Activity

replace TokenType enum with simple char constants

Browse Source
This commit is contained in:
Sarah Hoffmann
2025-02-21 10:23:41 +01:00
parent 4577669213
commit 31412e0674
8 changed files with 281 additions and 281 deletions
Download Patch File Download Diff File Expand all files Collapse all files

24
src/nominatim_api/search/db_search_builder.py
View File

@@ -97,7 +97,7 @@ class SearchBuilder:
builder = self.build_poi_search(sdata)
elif assignment.housenumber:
hnr_tokens = self.query.get_tokens(assignment.housenumber,
qmod.TokenType.HOUSENUMBER)
qmod.TOKEN_HOUSENUMBER)
builder = self.build_housenumber_search(sdata, hnr_tokens, assignment.address)
else:
builder = self.build_special_search(sdata, assignment.address,
@@ -173,7 +173,7 @@ class SearchBuilder:
list(partials), lookups.LookupAll))
else:
addr_fulls = [t.token for t
in self.query.get_tokens(address[0], qmod.TokenType.WORD)]
in self.query.get_tokens(address[0], qmod.TOKEN_WORD)]
if len(addr_fulls) > 5:
return
sdata.lookups.append(
@@ -216,7 +216,7 @@ class SearchBuilder:
addr_count = min(t.addr_count for t in addr_partials) if addr_partials else 30000
# Partial term to frequent. Try looking up by rare full names first.
name_fulls = self.query.get_tokens(name, qmod.TokenType.WORD)
name_fulls = self.query.get_tokens(name, qmod.TOKEN_WORD)
if name_fulls:
fulls_count = sum(t.count for t in name_fulls)
@@ -285,7 +285,7 @@ class SearchBuilder:
db_field: str = 'name_vector') -> dbf.FieldRanking:
""" Create a ranking expression for a name term in the given range.
"""
name_fulls = self.query.get_tokens(trange, qmod.TokenType.WORD)
name_fulls = self.query.get_tokens(trange, qmod.TOKEN_WORD)
ranks = [dbf.RankedTokens(t.penalty, [t.token]) for t in name_fulls]
ranks.sort(key=lambda r: r.penalty)
# Fallback, sum of penalty for partials
@@ -304,10 +304,10 @@ class SearchBuilder:
while todo:
neglen, pos, rank = heapq.heappop(todo)
for tlist in self.query.nodes[pos].starting:
if tlist.ttype in (qmod.TokenType.PARTIAL, qmod.TokenType.WORD):
if tlist.ttype in (qmod.TOKEN_PARTIAL, qmod.TOKEN_WORD):
if tlist.end < trange.end:
chgpenalty = PENALTY_WORDCHANGE[self.query.nodes[tlist.end].btype]
if tlist.ttype == qmod.TokenType.PARTIAL:
if tlist.ttype == qmod.TOKEN_PARTIAL:
penalty = rank.penalty + chgpenalty \
+ max(t.penalty for t in tlist.tokens)
heapq.heappush(todo, (neglen - 1, tlist.end,
@@ -317,7 +317,7 @@ class SearchBuilder:
heapq.heappush(todo, (neglen - 1, tlist.end,
rank.with_token(t, chgpenalty)))
elif tlist.end == trange.end:
if tlist.ttype == qmod.TokenType.PARTIAL:
if tlist.ttype == qmod.TOKEN_PARTIAL:
ranks.append(dbf.RankedTokens(rank.penalty
+ max(t.penalty for t in tlist.tokens),
rank.tokens))
@@ -357,11 +357,11 @@ class SearchBuilder:
if assignment.housenumber:
sdata.set_strings('housenumbers',
self.query.get_tokens(assignment.housenumber,
qmod.TokenType.HOUSENUMBER))
qmod.TOKEN_HOUSENUMBER))
if assignment.postcode:
sdata.set_strings('postcodes',
self.query.get_tokens(assignment.postcode,
qmod.TokenType.POSTCODE))
qmod.TOKEN_POSTCODE))
if assignment.qualifier:
tokens = self.get_qualifier_tokens(assignment.qualifier)
if not tokens:
@@ -391,7 +391,7 @@ class SearchBuilder:
optionally filtered by the country list from the details
parameters.
"""
tokens = self.query.get_tokens(trange, qmod.TokenType.COUNTRY)
tokens = self.query.get_tokens(trange, qmod.TOKEN_COUNTRY)
if self.details.countries:
tokens = [t for t in tokens if t.lookup_word in self.details.countries]
@@ -402,7 +402,7 @@ class SearchBuilder:
optionally filtered by the qualifier list from the details
parameters.
"""
tokens = self.query.get_tokens(trange, qmod.TokenType.QUALIFIER)
tokens = self.query.get_tokens(trange, qmod.TOKEN_QUALIFIER)
if self.details.categories:
tokens = [t for t in tokens if t.get_category() in self.details.categories]
@@ -415,7 +415,7 @@ class SearchBuilder:
"""
if assignment.near_item:
tokens: Dict[Tuple[str, str], float] = {}
for t in self.query.get_tokens(assignment.near_item, qmod.TokenType.NEAR_ITEM):
for t in self.query.get_tokens(assignment.near_item, qmod.TOKEN_NEAR_ITEM):
cat = t.get_category()
# The category of a near search will be that of near_item.
# Thus, if search is restricted to a category parameter,

34
src/nominatim_api/search/icu_tokenizer.py
View File

@@ -29,11 +29,11 @@ from .query_analyzer_factory import AbstractQueryAnalyzer
DB_TO_TOKEN_TYPE = {
'W': qmod.TokenType.WORD,
'w': qmod.TokenType.PARTIAL,
'H': qmod.TokenType.HOUSENUMBER,
'P': qmod.TokenType.POSTCODE,
'C': qmod.TokenType.COUNTRY
'W': qmod.TOKEN_WORD,
'w': qmod.TOKEN_PARTIAL,
'H': qmod.TOKEN_HOUSENUMBER,
'P': qmod.TOKEN_POSTCODE,
'C': qmod.TOKEN_COUNTRY
}
PENALTY_IN_TOKEN_BREAK = {
@@ -225,12 +225,12 @@ class ICUQueryAnalyzer(AbstractQueryAnalyzer):
if row.type == 'S':
if row.info['op'] in ('in', 'near'):
if trange.start == 0:
query.add_token(trange, qmod.TokenType.NEAR_ITEM, token)
query.add_token(trange, qmod.TOKEN_NEAR_ITEM, token)
else:
if trange.start == 0 and trange.end == query.num_token_slots():
query.add_token(trange, qmod.TokenType.NEAR_ITEM, token)
query.add_token(trange, qmod.TOKEN_NEAR_ITEM, token)
else:
query.add_token(trange, qmod.TokenType.QUALIFIER, token)
query.add_token(trange, qmod.TOKEN_QUALIFIER, token)
else:
query.add_token(trange, DB_TO_TOKEN_TYPE[row.type], token)
@@ -297,8 +297,8 @@ class ICUQueryAnalyzer(AbstractQueryAnalyzer):
"""
for part, node, i in zip(parts, query.nodes, range(1000)):
if len(part.token) <= 4 and part.token.isdigit()\
and not node.has_tokens(i+1, qmod.TokenType.HOUSENUMBER):
query.add_token(qmod.TokenRange(i, i+1), qmod.TokenType.HOUSENUMBER,
and not node.has_tokens(i+1, qmod.TOKEN_HOUSENUMBER):
query.add_token(qmod.TokenRange(i, i+1), qmod.TOKEN_HOUSENUMBER,
ICUToken(penalty=0.5, token=0,
count=1, addr_count=1, lookup_word=part.token,
word_token=part.token, info=None))
@@ -307,19 +307,19 @@ class ICUQueryAnalyzer(AbstractQueryAnalyzer):
""" Add penalties to tokens that depend on presence of other token.
"""
for i, node, tlist in query.iter_token_lists():
if tlist.ttype == qmod.TokenType.POSTCODE:
if tlist.ttype == qmod.TOKEN_POSTCODE:
for repl in node.starting:
if repl.end == tlist.end and repl.ttype != qmod.TokenType.POSTCODE \
and (repl.ttype != qmod.TokenType.HOUSENUMBER
if repl.end == tlist.end and repl.ttype != qmod.TOKEN_POSTCODE \
and (repl.ttype != qmod.TOKEN_HOUSENUMBER
or len(tlist.tokens[0].lookup_word) > 4):
repl.add_penalty(0.39)
elif (tlist.ttype == qmod.TokenType.HOUSENUMBER
elif (tlist.ttype == qmod.TOKEN_HOUSENUMBER
and len(tlist.tokens[0].lookup_word) <= 3):
if any(c.isdigit() for c in tlist.tokens[0].lookup_word):
for repl in node.starting:
if repl.end == tlist.end and repl.ttype != qmod.TokenType.HOUSENUMBER:
if repl.end == tlist.end and repl.ttype != qmod.TOKEN_HOUSENUMBER:
repl.add_penalty(0.5 - tlist.tokens[0].penalty)
elif tlist.ttype not in (qmod.TokenType.COUNTRY, qmod.TokenType.PARTIAL):
elif tlist.ttype not in (qmod.TOKEN_COUNTRY, qmod.TOKEN_PARTIAL):
norm = parts[i].normalized
for j in range(i + 1, tlist.end):
if node.btype != qmod.BREAK_TOKEN:
@@ -341,7 +341,7 @@ def _dump_word_tokens(query: qmod.QueryStruct) -> Iterator[List[Any]]:
for tlist in node.starting:
for token in tlist.tokens:
t = cast(ICUToken, token)
yield [tlist.ttype.name, t.token, t.word_token or '',
yield [tlist.ttype, t.token, t.word_token or '',
t.lookup_word or '', t.penalty, t.count, t.info]

54
src/nominatim_api/search/query.py
View File

@@ -38,23 +38,23 @@ BREAK_TOKEN = '`'
"""
class TokenType(enum.Enum):
""" Type of token.
"""
WORD = enum.auto()
""" Full name of a place. """
PARTIAL = enum.auto()
""" Word term without breaks, does not necessarily represent a full name. """
HOUSENUMBER = enum.auto()
""" Housenumber term. """
POSTCODE = enum.auto()
""" Postal code term. """
COUNTRY = enum.auto()
""" Country name or reference. """
QUALIFIER = enum.auto()
""" Special term used together with name (e.g. _Hotel_ Bellevue). """
NEAR_ITEM = enum.auto()
""" Special term used as searchable object(e.g. supermarket in ...). """
TokenType = str
""" Type of token.
"""
TOKEN_WORD = 'W'
""" Full name of a place. """
TOKEN_PARTIAL = 'w'
""" Word term without breaks, does not necessarily represent a full name. """
TOKEN_HOUSENUMBER = 'H'
""" Housenumber term. """
TOKEN_POSTCODE = 'P'
""" Postal code term. """
TOKEN_COUNTRY = 'C'
""" Country name or reference. """
TOKEN_QUALIFIER = 'Q'
""" Special term used together with name (e.g. _Hotel_ Bellevue). """
TOKEN_NEAR_ITEM = 'N'
""" Special term used as searchable object(e.g. supermarket in ...). """
class PhraseType(enum.Enum):
@@ -82,19 +82,19 @@ class PhraseType(enum.Enum):
""" Check if the given token type can be used with the phrase type.
"""
if self == PhraseType.NONE:
return not is_full_phrase or ttype != TokenType.QUALIFIER
return not is_full_phrase or ttype != TOKEN_QUALIFIER
if self == PhraseType.AMENITY:
return ttype in (TokenType.WORD, TokenType.PARTIAL)\
or (is_full_phrase and ttype == TokenType.NEAR_ITEM)\
or (not is_full_phrase and ttype == TokenType.QUALIFIER)
return ttype in (TOKEN_WORD, TOKEN_PARTIAL)\
or (is_full_phrase and ttype == TOKEN_NEAR_ITEM)\
or (not is_full_phrase and ttype == TOKEN_QUALIFIER)
if self == PhraseType.STREET:
return ttype in (TokenType.WORD, TokenType.PARTIAL, TokenType.HOUSENUMBER)
return ttype in (TOKEN_WORD, TOKEN_PARTIAL, TOKEN_HOUSENUMBER)
if self == PhraseType.POSTCODE:
return ttype == TokenType.POSTCODE
return ttype == TOKEN_POSTCODE
if self == PhraseType.COUNTRY:
return ttype == TokenType.COUNTRY
return ttype == TOKEN_COUNTRY
return ttype in (TokenType.WORD, TokenType.PARTIAL)
return ttype in (TOKEN_WORD, TOKEN_PARTIAL)
@dataclasses.dataclass
@@ -265,7 +265,7 @@ class QueryStruct:
going to the subsequent node. Such PARTIAL tokens are
assumed to exist.
"""
return [next(iter(self.get_tokens(TokenRange(i, i+1), TokenType.PARTIAL)))
return [next(iter(self.get_tokens(TokenRange(i, i+1), TOKEN_PARTIAL)))
for i in range(trange.start, trange.end)]
def iter_token_lists(self) -> Iterator[Tuple[int, QueryNode, TokenList]]:
@@ -285,5 +285,5 @@ class QueryStruct:
for tlist in node.starting:
for t in tlist.tokens:
if t.token == token:
return f"[{tlist.ttype.name[0]}]{t.lookup_word}"
return f"[{tlist.ttype}]{t.lookup_word}"
return 'None'

68
src/nominatim_api/search/token_assignment.py
View File

@@ -56,17 +56,17 @@ class TokenAssignment:
"""
out = TokenAssignment()
for token in ranges:
if token.ttype == qmod.TokenType.PARTIAL:
if token.ttype == qmod.TOKEN_PARTIAL:
out.address.append(token.trange)
elif token.ttype == qmod.TokenType.HOUSENUMBER:
elif token.ttype == qmod.TOKEN_HOUSENUMBER:
out.housenumber = token.trange
elif token.ttype == qmod.TokenType.POSTCODE:
elif token.ttype == qmod.TOKEN_POSTCODE:
out.postcode = token.trange
elif token.ttype == qmod.TokenType.COUNTRY:
elif token.ttype == qmod.TOKEN_COUNTRY:
out.country = token.trange
elif token.ttype == qmod.TokenType.NEAR_ITEM:
elif token.ttype == qmod.TOKEN_NEAR_ITEM:
out.near_item = token.trange
elif token.ttype == qmod.TokenType.QUALIFIER:
elif token.ttype == qmod.TOKEN_QUALIFIER:
out.qualifier = token.trange
return out
@@ -84,7 +84,7 @@ class _TokenSequence:
self.penalty = penalty
def __str__(self) -> str:
seq = ''.join(f'[{r.trange.start} - {r.trange.end}: {r.ttype.name}]' for r in self.seq)
seq = ''.join(f'[{r.trange.start} - {r.trange.end}: {r.ttype}]' for r in self.seq)
return f'{seq} (dir: {self.direction}, penalty: {self.penalty})'
@property
@@ -105,7 +105,7 @@ class _TokenSequence:
"""
# Country and category must be the final term for left-to-right
return len(self.seq) > 1 and \
self.seq[-1].ttype in (qmod.TokenType.COUNTRY, qmod.TokenType.NEAR_ITEM)
self.seq[-1].ttype in (qmod.TOKEN_COUNTRY, qmod.TOKEN_NEAR_ITEM)
def appendable(self, ttype: qmod.TokenType) -> Optional[int]:
""" Check if the give token type is appendable to the existing sequence.
@@ -114,23 +114,23 @@ class _TokenSequence:
new direction of the sequence after adding such a type. The
token is not added.
"""
if ttype == qmod.TokenType.WORD:
if ttype == qmod.TOKEN_WORD:
return None
if not self.seq:
# Append unconditionally to the empty list
if ttype == qmod.TokenType.COUNTRY:
if ttype == qmod.TOKEN_COUNTRY:
return -1
if ttype in (qmod.TokenType.HOUSENUMBER, qmod.TokenType.QUALIFIER):
if ttype in (qmod.TOKEN_HOUSENUMBER, qmod.TOKEN_QUALIFIER):
return 1
return self.direction
# Name tokens are always acceptable and don't change direction
if ttype == qmod.TokenType.PARTIAL:
if ttype == qmod.TOKEN_PARTIAL:
# qualifiers cannot appear in the middle of the query. They need
# to be near the next phrase.
if self.direction == -1 \
and any(t.ttype == qmod.TokenType.QUALIFIER for t in self.seq[:-1]):
and any(t.ttype == qmod.TOKEN_QUALIFIER for t in self.seq[:-1]):
return None
return self.direction
@@ -138,54 +138,54 @@ class _TokenSequence:
if self.has_types(ttype):
return None
if ttype == qmod.TokenType.HOUSENUMBER:
if ttype == qmod.TOKEN_HOUSENUMBER:
if self.direction == 1:
if len(self.seq) == 1 and self.seq[0].ttype == qmod.TokenType.QUALIFIER:
if len(self.seq) == 1 and self.seq[0].ttype == qmod.TOKEN_QUALIFIER:
return None
if len(self.seq) > 2 \
or self.has_types(qmod.TokenType.POSTCODE, qmod.TokenType.COUNTRY):
or self.has_types(qmod.TOKEN_POSTCODE, qmod.TOKEN_COUNTRY):
return None # direction left-to-right: housenumber must come before anything
elif (self.direction == -1
or self.has_types(qmod.TokenType.POSTCODE, qmod.TokenType.COUNTRY)):
or self.has_types(qmod.TOKEN_POSTCODE, qmod.TOKEN_COUNTRY)):
return -1 # force direction right-to-left if after other terms
return self.direction
if ttype == qmod.TokenType.POSTCODE:
if ttype == qmod.TOKEN_POSTCODE:
if self.direction == -1:
if self.has_types(qmod.TokenType.HOUSENUMBER, qmod.TokenType.QUALIFIER):
if self.has_types(qmod.TOKEN_HOUSENUMBER, qmod.TOKEN_QUALIFIER):
return None
return -1
if self.direction == 1:
return None if self.has_types(qmod.TokenType.COUNTRY) else 1
if self.has_types(qmod.TokenType.HOUSENUMBER, qmod.TokenType.QUALIFIER):
return None if self.has_types(qmod.TOKEN_COUNTRY) else 1
if self.has_types(qmod.TOKEN_HOUSENUMBER, qmod.TOKEN_QUALIFIER):
return 1
return self.direction
if ttype == qmod.TokenType.COUNTRY:
if ttype == qmod.TOKEN_COUNTRY:
return None if self.direction == -1 else 1
if ttype == qmod.TokenType.NEAR_ITEM:
if ttype == qmod.TOKEN_NEAR_ITEM:
return self.direction
if ttype == qmod.TokenType.QUALIFIER:
if ttype == qmod.TOKEN_QUALIFIER:
if self.direction == 1:
if (len(self.seq) == 1
and self.seq[0].ttype in (qmod.TokenType.PARTIAL, qmod.TokenType.NEAR_ITEM)) \
and self.seq[0].ttype in (qmod.TOKEN_PARTIAL, qmod.TOKEN_NEAR_ITEM)) \
or (len(self.seq) == 2
and self.seq[0].ttype == qmod.TokenType.NEAR_ITEM
and self.seq[1].ttype == qmod.TokenType.PARTIAL):
and self.seq[0].ttype == qmod.TOKEN_NEAR_ITEM
and self.seq[1].ttype == qmod.TOKEN_PARTIAL):
return 1
return None
if self.direction == -1:
return -1
tempseq = self.seq[1:] if self.seq[0].ttype == qmod.TokenType.NEAR_ITEM else self.seq
tempseq = self.seq[1:] if self.seq[0].ttype == qmod.TOKEN_NEAR_ITEM else self.seq
if len(tempseq) == 0:
return 1
if len(tempseq) == 1 and self.seq[0].ttype == qmod.TokenType.HOUSENUMBER:
if len(tempseq) == 1 and self.seq[0].ttype == qmod.TOKEN_HOUSENUMBER:
return None
if len(tempseq) > 1 or self.has_types(qmod.TokenType.POSTCODE, qmod.TokenType.COUNTRY):
if len(tempseq) > 1 or self.has_types(qmod.TOKEN_POSTCODE, qmod.TOKEN_COUNTRY):
return -1
return 0
@@ -240,18 +240,18 @@ class _TokenSequence:
# housenumbers may not be further than 2 words from the beginning.
# If there are two words in front, give it a penalty.
hnrpos = next((i for i, tr in enumerate(self.seq)
if tr.ttype == qmod.TokenType.HOUSENUMBER),
if tr.ttype == qmod.TOKEN_HOUSENUMBER),
None)
if hnrpos is not None:
if self.direction != -1:
priors = sum(1 for t in self.seq[:hnrpos] if t.ttype == qmod.TokenType.PARTIAL)
priors = sum(1 for t in self.seq[:hnrpos] if t.ttype == qmod.TOKEN_PARTIAL)
if not self._adapt_penalty_from_priors(priors, -1):
return False
if self.direction != 1:
priors = sum(1 for t in self.seq[hnrpos+1:] if t.ttype == qmod.TokenType.PARTIAL)
priors = sum(1 for t in self.seq[hnrpos+1:] if t.ttype == qmod.TOKEN_PARTIAL)
if not self._adapt_penalty_from_priors(priors, 1):
return False
if any(t.ttype == qmod.TokenType.NEAR_ITEM for t in self.seq):
if any(t.ttype == qmod.TOKEN_NEAR_ITEM for t in self.seq):
self.penalty += 1.0
return True

90
test/python/api/search/test_api_search_query.py
View File

@@ -22,42 +22,42 @@ def mktoken(tid: int):
lookup_word='foo')
@pytest.mark.parametrize('ptype,ttype', [('NONE', 'WORD'),
('AMENITY', 'QUALIFIER'),
('STREET', 'PARTIAL'),
('CITY', 'WORD'),
('COUNTRY', 'COUNTRY'),
('POSTCODE', 'POSTCODE')])
@pytest.mark.parametrize('ptype,ttype', [('NONE', 'W'),
('AMENITY', 'Q'),
('STREET', 'w'),
('CITY', 'W'),
('COUNTRY', 'C'),
('POSTCODE', 'P')])
def test_phrase_compatible(ptype, ttype):
assert query.PhraseType[ptype].compatible_with(query.TokenType[ttype], False)
assert query.PhraseType[ptype].compatible_with(ttype, False)
@pytest.mark.parametrize('ptype', ['COUNTRY', 'POSTCODE'])
def test_phrase_incompatible(ptype):
assert not query.PhraseType[ptype].compatible_with(query.TokenType.PARTIAL, True)
assert not query.PhraseType[ptype].compatible_with(query.TOKEN_PARTIAL, True)
def test_query_node_empty():
qn = query.QueryNode(query.BREAK_PHRASE, query.PhraseType.NONE)
assert not qn.has_tokens(3, query.TokenType.PARTIAL)
assert qn.get_tokens(3, query.TokenType.WORD) is None
assert not qn.has_tokens(3, query.TOKEN_PARTIAL)
assert qn.get_tokens(3, query.TOKEN_WORD) is None
def test_query_node_with_content():
qn = query.QueryNode(query.BREAK_PHRASE, query.PhraseType.NONE)
qn.starting.append(query.TokenList(2, query.TokenType.PARTIAL, [mktoken(100), mktoken(101)]))
qn.starting.append(query.TokenList(2, query.TokenType.WORD, [mktoken(1000)]))
qn.starting.append(query.TokenList(2, query.TOKEN_PARTIAL, [mktoken(100), mktoken(101)]))
qn.starting.append(query.TokenList(2, query.TOKEN_WORD, [mktoken(1000)]))
assert not qn.has_tokens(3, query.TokenType.PARTIAL)
assert not qn.has_tokens(2, query.TokenType.COUNTRY)
assert qn.has_tokens(2, query.TokenType.PARTIAL)
assert qn.has_tokens(2, query.TokenType.WORD)
assert not qn.has_tokens(3, query.TOKEN_PARTIAL)
assert not qn.has_tokens(2, query.TOKEN_COUNTRY)
assert qn.has_tokens(2, query.TOKEN_PARTIAL)
assert qn.has_tokens(2, query.TOKEN_WORD)
assert qn.get_tokens(3, query.TokenType.PARTIAL) is None
assert qn.get_tokens(2, query.TokenType.COUNTRY) is None
assert len(qn.get_tokens(2, query.TokenType.PARTIAL)) == 2
assert len(qn.get_tokens(2, query.TokenType.WORD)) == 1
assert qn.get_tokens(3, query.TOKEN_PARTIAL) is None
assert qn.get_tokens(2, query.TOKEN_COUNTRY) is None
assert len(qn.get_tokens(2, query.TOKEN_PARTIAL)) == 2
assert len(qn.get_tokens(2, query.TOKEN_WORD)) == 1
def test_query_struct_empty():
@@ -73,13 +73,13 @@ def test_query_struct_with_tokens():
assert q.num_token_slots() == 2
q.add_token(query.TokenRange(0, 1), query.TokenType.PARTIAL, mktoken(1))
q.add_token(query.TokenRange(1, 2), query.TokenType.PARTIAL, mktoken(2))
q.add_token(query.TokenRange(1, 2), query.TokenType.WORD, mktoken(99))
q.add_token(query.TokenRange(1, 2), query.TokenType.WORD, mktoken(98))
q.add_token(query.TokenRange(0, 1), query.TOKEN_PARTIAL, mktoken(1))
q.add_token(query.TokenRange(1, 2), query.TOKEN_PARTIAL, mktoken(2))
q.add_token(query.TokenRange(1, 2), query.TOKEN_WORD, mktoken(99))
q.add_token(query.TokenRange(1, 2), query.TOKEN_WORD, mktoken(98))
assert q.get_tokens(query.TokenRange(0, 2), query.TokenType.WORD) == []
assert len(q.get_tokens(query.TokenRange(1, 2), query.TokenType.WORD)) == 2
assert q.get_tokens(query.TokenRange(0, 2), query.TOKEN_WORD) == []
assert len(q.get_tokens(query.TokenRange(1, 2), query.TOKEN_WORD)) == 2
partials = q.get_partials_list(query.TokenRange(0, 2))
@@ -95,24 +95,24 @@ def test_query_struct_incompatible_token():
q.add_node(query.BREAK_WORD, query.PhraseType.COUNTRY)
q.add_node(query.BREAK_END, query.PhraseType.NONE)
q.add_token(query.TokenRange(0, 1), query.TokenType.PARTIAL, mktoken(1))
q.add_token(query.TokenRange(1, 2), query.TokenType.COUNTRY, mktoken(100))
q.add_token(query.TokenRange(0, 1), query.TOKEN_PARTIAL, mktoken(1))
q.add_token(query.TokenRange(1, 2), query.TOKEN_COUNTRY, mktoken(100))
assert q.get_tokens(query.TokenRange(0, 1), query.TokenType.PARTIAL) == []
assert len(q.get_tokens(query.TokenRange(1, 2), query.TokenType.COUNTRY)) == 1
assert q.get_tokens(query.TokenRange(0, 1), query.TOKEN_PARTIAL) == []
assert len(q.get_tokens(query.TokenRange(1, 2), query.TOKEN_COUNTRY)) == 1
def test_query_struct_amenity_single_word():
q = query.QueryStruct([query.Phrase(query.PhraseType.AMENITY, 'bar')])
q.add_node(query.BREAK_END, query.PhraseType.NONE)
q.add_token(query.TokenRange(0, 1), query.TokenType.PARTIAL, mktoken(1))
q.add_token(query.TokenRange(0, 1), query.TokenType.NEAR_ITEM, mktoken(2))
q.add_token(query.TokenRange(0, 1), query.TokenType.QUALIFIER, mktoken(3))
q.add_token(query.TokenRange(0, 1), query.TOKEN_PARTIAL, mktoken(1))
q.add_token(query.TokenRange(0, 1), query.TOKEN_NEAR_ITEM, mktoken(2))
q.add_token(query.TokenRange(0, 1), query.TOKEN_QUALIFIER, mktoken(3))
assert len(q.get_tokens(query.TokenRange(0, 1), query.TokenType.PARTIAL)) == 1
assert len(q.get_tokens(query.TokenRange(0, 1), query.TokenType.NEAR_ITEM)) == 1
assert len(q.get_tokens(query.TokenRange(0, 1), query.TokenType.QUALIFIER)) == 0
assert len(q.get_tokens(query.TokenRange(0, 1), query.TOKEN_PARTIAL)) == 1
assert len(q.get_tokens(query.TokenRange(0, 1), query.TOKEN_NEAR_ITEM)) == 1
assert len(q.get_tokens(query.TokenRange(0, 1), query.TOKEN_QUALIFIER)) == 0
def test_query_struct_amenity_two_words():
@@ -121,15 +121,15 @@ def test_query_struct_amenity_two_words():
q.add_node(query.BREAK_END, query.PhraseType.NONE)
for trange in [(0, 1), (1, 2)]:
q.add_token(query.TokenRange(*trange), query.TokenType.PARTIAL, mktoken(1))
q.add_token(query.TokenRange(*trange), query.TokenType.NEAR_ITEM, mktoken(2))
q.add_token(query.TokenRange(*trange), query.TokenType.QUALIFIER, mktoken(3))
q.add_token(query.TokenRange(*trange), query.TOKEN_PARTIAL, mktoken(1))
q.add_token(query.TokenRange(*trange), query.TOKEN_NEAR_ITEM, mktoken(2))
q.add_token(query.TokenRange(*trange), query.TOKEN_QUALIFIER, mktoken(3))
assert len(q.get_tokens(query.TokenRange(0, 1), query.TokenType.PARTIAL)) == 1
assert len(q.get_tokens(query.TokenRange(0, 1), query.TokenType.NEAR_ITEM)) == 0
assert len(q.get_tokens(query.TokenRange(0, 1), query.TokenType.QUALIFIER)) == 1
assert len(q.get_tokens(query.TokenRange(0, 1), query.TOKEN_PARTIAL)) == 1
assert len(q.get_tokens(query.TokenRange(0, 1), query.TOKEN_NEAR_ITEM)) == 0
assert len(q.get_tokens(query.TokenRange(0, 1), query.TOKEN_QUALIFIER)) == 1
assert len(q.get_tokens(query.TokenRange(1, 2), query.TokenType.PARTIAL)) == 1
assert len(q.get_tokens(query.TokenRange(1, 2), query.TokenType.NEAR_ITEM)) == 0
assert len(q.get_tokens(query.TokenRange(1, 2), query.TokenType.QUALIFIER)) == 1
assert len(q.get_tokens(query.TokenRange(1, 2), query.TOKEN_PARTIAL)) == 1
assert len(q.get_tokens(query.TokenRange(1, 2), query.TOKEN_NEAR_ITEM)) == 0
assert len(q.get_tokens(query.TokenRange(1, 2), query.TOKEN_QUALIFIER)) == 1

118
test/python/api/search/test_db_search_builder.py
View File

@@ -9,7 +9,7 @@ Tests for creating abstract searches from token assignments.
"""
import pytest
from nominatim_api.search.query import Token, TokenRange, PhraseType, TokenType, QueryStruct, Phrase
from nominatim_api.search.query import Token, TokenRange, PhraseType, QueryStruct, Phrase
import nominatim_api.search.query as qmod
from nominatim_api.search.db_search_builder import SearchBuilder
from nominatim_api.search.token_assignment import TokenAssignment
@@ -32,7 +32,7 @@ def make_query(*args):
for end, ttype, tinfo in tlist:
for tid, word in tinfo:
q.add_token(TokenRange(start, end), ttype,
MyToken(penalty=0.5 if ttype == TokenType.PARTIAL else 0.0,
MyToken(penalty=0.5 if ttype == qmod.TOKEN_PARTIAL else 0.0,
token=tid, count=1, addr_count=1,
lookup_word=word))
@@ -41,7 +41,7 @@ def make_query(*args):
def test_country_search():
q = make_query([(1, TokenType.COUNTRY, [(2, 'de'), (3, 'en')])])
q = make_query([(1, qmod.TOKEN_COUNTRY, [(2, 'de'), (3, 'en')])])
builder = SearchBuilder(q, SearchDetails())
searches = list(builder.build(TokenAssignment(country=TokenRange(0, 1))))
@@ -55,7 +55,7 @@ def test_country_search():
def test_country_search_with_country_restriction():
q = make_query([(1, TokenType.COUNTRY, [(2, 'de'), (3, 'en')])])
q = make_query([(1, qmod.TOKEN_COUNTRY, [(2, 'de'), (3, 'en')])])
builder = SearchBuilder(q, SearchDetails.from_kwargs({'countries': 'en,fr'}))
searches = list(builder.build(TokenAssignment(country=TokenRange(0, 1))))
@@ -69,7 +69,7 @@ def test_country_search_with_country_restriction():
def test_country_search_with_conflicting_country_restriction():
q = make_query([(1, TokenType.COUNTRY, [(2, 'de'), (3, 'en')])])
q = make_query([(1, qmod.TOKEN_COUNTRY, [(2, 'de'), (3, 'en')])])
builder = SearchBuilder(q, SearchDetails.from_kwargs({'countries': 'fr'}))
searches = list(builder.build(TokenAssignment(country=TokenRange(0, 1))))
@@ -78,7 +78,7 @@ def test_country_search_with_conflicting_country_restriction():
def test_postcode_search_simple():
q = make_query([(1, TokenType.POSTCODE, [(34, '2367')])])
q = make_query([(1, qmod.TOKEN_POSTCODE, [(34, '2367')])])
builder = SearchBuilder(q, SearchDetails())
searches = list(builder.build(TokenAssignment(postcode=TokenRange(0, 1))))
@@ -94,8 +94,8 @@ def test_postcode_search_simple():
def test_postcode_with_country():
q = make_query([(1, TokenType.POSTCODE, [(34, '2367')])],
[(2, TokenType.COUNTRY, [(1, 'xx')])])
q = make_query([(1, qmod.TOKEN_POSTCODE, [(34, '2367')])],
[(2, qmod.TOKEN_COUNTRY, [(1, 'xx')])])
builder = SearchBuilder(q, SearchDetails())
searches = list(builder.build(TokenAssignment(postcode=TokenRange(0, 1),
@@ -112,8 +112,8 @@ def test_postcode_with_country():
def test_postcode_with_address():
q = make_query([(1, TokenType.POSTCODE, [(34, '2367')])],
[(2, TokenType.PARTIAL, [(100, 'word')])])
q = make_query([(1, qmod.TOKEN_POSTCODE, [(34, '2367')])],
[(2, qmod.TOKEN_PARTIAL, [(100, 'word')])])
builder = SearchBuilder(q, SearchDetails())
searches = list(builder.build(TokenAssignment(postcode=TokenRange(0, 1),
@@ -130,9 +130,9 @@ def test_postcode_with_address():
def test_postcode_with_address_with_full_word():
q = make_query([(1, TokenType.POSTCODE, [(34, '2367')])],
[(2, TokenType.PARTIAL, [(100, 'word')]),
(2, TokenType.WORD, [(1, 'full')])])
q = make_query([(1, qmod.TOKEN_POSTCODE, [(34, '2367')])],
[(2, qmod.TOKEN_PARTIAL, [(100, 'word')]),
(2, qmod.TOKEN_WORD, [(1, 'full')])])
builder = SearchBuilder(q, SearchDetails())
searches = list(builder.build(TokenAssignment(postcode=TokenRange(0, 1),
@@ -151,7 +151,7 @@ def test_postcode_with_address_with_full_word():
@pytest.mark.parametrize('kwargs', [{'viewbox': '0,0,1,1', 'bounded_viewbox': True},
{'near': '10,10'}])
def test_near_item_only(kwargs):
q = make_query([(1, TokenType.NEAR_ITEM, [(2, 'foo')])])
q = make_query([(1, qmod.TOKEN_NEAR_ITEM, [(2, 'foo')])])
builder = SearchBuilder(q, SearchDetails.from_kwargs(kwargs))
searches = list(builder.build(TokenAssignment(near_item=TokenRange(0, 1))))
@@ -167,7 +167,7 @@ def test_near_item_only(kwargs):
@pytest.mark.parametrize('kwargs', [{'viewbox': '0,0,1,1'},
{}])
def test_near_item_skipped(kwargs):
q = make_query([(1, TokenType.NEAR_ITEM, [(2, 'foo')])])
q = make_query([(1, qmod.TOKEN_NEAR_ITEM, [(2, 'foo')])])
builder = SearchBuilder(q, SearchDetails.from_kwargs(kwargs))
searches = list(builder.build(TokenAssignment(near_item=TokenRange(0, 1))))
@@ -176,8 +176,8 @@ def test_near_item_skipped(kwargs):
def test_name_only_search():
q = make_query([(1, TokenType.PARTIAL, [(1, 'a')]),
(1, TokenType.WORD, [(100, 'a')])])
q = make_query([(1, qmod.TOKEN_PARTIAL, [(1, 'a')]),
(1, qmod.TOKEN_WORD, [(100, 'a')])])
builder = SearchBuilder(q, SearchDetails())
searches = list(builder.build(TokenAssignment(name=TokenRange(0, 1))))
@@ -195,9 +195,9 @@ def test_name_only_search():
def test_name_with_qualifier():
q = make_query([(1, TokenType.PARTIAL, [(1, 'a')]),
(1, TokenType.WORD, [(100, 'a')])],
[(2, TokenType.QUALIFIER, [(55, 'hotel')])])
q = make_query([(1, qmod.TOKEN_PARTIAL, [(1, 'a')]),
(1, qmod.TOKEN_WORD, [(100, 'a')])],
[(2, qmod.TOKEN_QUALIFIER, [(55, 'hotel')])])
builder = SearchBuilder(q, SearchDetails())
searches = list(builder.build(TokenAssignment(name=TokenRange(0, 1),
@@ -216,9 +216,9 @@ def test_name_with_qualifier():
def test_name_with_housenumber_search():
q = make_query([(1, TokenType.PARTIAL, [(1, 'a')]),
(1, TokenType.WORD, [(100, 'a')])],
[(2, TokenType.HOUSENUMBER, [(66, '66')])])
q = make_query([(1, qmod.TOKEN_PARTIAL, [(1, 'a')]),
(1, qmod.TOKEN_WORD, [(100, 'a')])],
[(2, qmod.TOKEN_HOUSENUMBER, [(66, '66')])])
builder = SearchBuilder(q, SearchDetails())
searches = list(builder.build(TokenAssignment(name=TokenRange(0, 1),
@@ -236,12 +236,12 @@ def test_name_with_housenumber_search():
def test_name_and_address():
q = make_query([(1, TokenType.PARTIAL, [(1, 'a')]),
(1, TokenType.WORD, [(100, 'a')])],
[(2, TokenType.PARTIAL, [(2, 'b')]),
(2, TokenType.WORD, [(101, 'b')])],
[(3, TokenType.PARTIAL, [(3, 'c')]),
(3, TokenType.WORD, [(102, 'c')])]
q = make_query([(1, qmod.TOKEN_PARTIAL, [(1, 'a')]),
(1, qmod.TOKEN_WORD, [(100, 'a')])],
[(2, qmod.TOKEN_PARTIAL, [(2, 'b')]),
(2, qmod.TOKEN_WORD, [(101, 'b')])],
[(3, qmod.TOKEN_PARTIAL, [(3, 'c')]),
(3, qmod.TOKEN_WORD, [(102, 'c')])]
)
builder = SearchBuilder(q, SearchDetails())
@@ -261,13 +261,13 @@ def test_name_and_address():
def test_name_and_complex_address():
q = make_query([(1, TokenType.PARTIAL, [(1, 'a')]),
(1, TokenType.WORD, [(100, 'a')])],
[(2, TokenType.PARTIAL, [(2, 'b')]),
(3, TokenType.WORD, [(101, 'bc')])],
[(3, TokenType.PARTIAL, [(3, 'c')])],
[(4, TokenType.PARTIAL, [(4, 'd')]),
(4, TokenType.WORD, [(103, 'd')])]
q = make_query([(1, qmod.TOKEN_PARTIAL, [(1, 'a')]),
(1, qmod.TOKEN_WORD, [(100, 'a')])],
[(2, qmod.TOKEN_PARTIAL, [(2, 'b')]),
(3, qmod.TOKEN_WORD, [(101, 'bc')])],
[(3, qmod.TOKEN_PARTIAL, [(3, 'c')])],
[(4, qmod.TOKEN_PARTIAL, [(4, 'd')]),
(4, qmod.TOKEN_WORD, [(103, 'd')])]
)
builder = SearchBuilder(q, SearchDetails())
@@ -287,9 +287,9 @@ def test_name_and_complex_address():
def test_name_only_near_search():
q = make_query([(1, TokenType.NEAR_ITEM, [(88, 'g')])],
[(2, TokenType.PARTIAL, [(1, 'a')]),
(2, TokenType.WORD, [(100, 'a')])])
q = make_query([(1, qmod.TOKEN_NEAR_ITEM, [(88, 'g')])],
[(2, qmod.TOKEN_PARTIAL, [(1, 'a')]),
(2, qmod.TOKEN_WORD, [(100, 'a')])])
builder = SearchBuilder(q, SearchDetails())
searches = list(builder.build(TokenAssignment(name=TokenRange(1, 2),
@@ -303,8 +303,8 @@ def test_name_only_near_search():
def test_name_only_search_with_category():
q = make_query([(1, TokenType.PARTIAL, [(1, 'a')]),
(1, TokenType.WORD, [(100, 'a')])])
q = make_query([(1, qmod.TOKEN_PARTIAL, [(1, 'a')]),
(1, qmod.TOKEN_WORD, [(100, 'a')])])
builder = SearchBuilder(q, SearchDetails.from_kwargs({'categories': [('foo', 'bar')]}))
searches = list(builder.build(TokenAssignment(name=TokenRange(0, 1))))
@@ -317,9 +317,9 @@ def test_name_only_search_with_category():
def test_name_with_near_item_search_with_category_mismatch():
q = make_query([(1, TokenType.NEAR_ITEM, [(88, 'g')])],
[(2, TokenType.PARTIAL, [(1, 'a')]),
(2, TokenType.WORD, [(100, 'a')])])
q = make_query([(1, qmod.TOKEN_NEAR_ITEM, [(88, 'g')])],
[(2, qmod.TOKEN_PARTIAL, [(1, 'a')]),
(2, qmod.TOKEN_WORD, [(100, 'a')])])
builder = SearchBuilder(q, SearchDetails.from_kwargs({'categories': [('foo', 'bar')]}))
searches = list(builder.build(TokenAssignment(name=TokenRange(1, 2),
@@ -329,9 +329,9 @@ def test_name_with_near_item_search_with_category_mismatch():
def test_name_with_near_item_search_with_category_match():
q = make_query([(1, TokenType.NEAR_ITEM, [(88, 'g')])],
[(2, TokenType.PARTIAL, [(1, 'a')]),
(2, TokenType.WORD, [(100, 'a')])])
q = make_query([(1, qmod.TOKEN_NEAR_ITEM, [(88, 'g')])],
[(2, qmod.TOKEN_PARTIAL, [(1, 'a')]),
(2, qmod.TOKEN_WORD, [(100, 'a')])])
builder = SearchBuilder(q, SearchDetails.from_kwargs({'categories': [('foo', 'bar'),
('this', 'that')]}))
@@ -346,9 +346,9 @@ def test_name_with_near_item_search_with_category_match():
def test_name_with_qualifier_search_with_category_mismatch():
q = make_query([(1, TokenType.QUALIFIER, [(88, 'g')])],
[(2, TokenType.PARTIAL, [(1, 'a')]),
(2, TokenType.WORD, [(100, 'a')])])
q = make_query([(1, qmod.TOKEN_QUALIFIER, [(88, 'g')])],
[(2, qmod.TOKEN_PARTIAL, [(1, 'a')]),
(2, qmod.TOKEN_WORD, [(100, 'a')])])
builder = SearchBuilder(q, SearchDetails.from_kwargs({'categories': [('foo', 'bar')]}))
searches = list(builder.build(TokenAssignment(name=TokenRange(1, 2),
@@ -358,9 +358,9 @@ def test_name_with_qualifier_search_with_category_mismatch():
def test_name_with_qualifier_search_with_category_match():
q = make_query([(1, TokenType.QUALIFIER, [(88, 'g')])],
[(2, TokenType.PARTIAL, [(1, 'a')]),
(2, TokenType.WORD, [(100, 'a')])])
q = make_query([(1, qmod.TOKEN_QUALIFIER, [(88, 'g')])],
[(2, qmod.TOKEN_PARTIAL, [(1, 'a')]),
(2, qmod.TOKEN_WORD, [(100, 'a')])])
builder = SearchBuilder(q, SearchDetails.from_kwargs({'categories': [('foo', 'bar'),
('this', 'that')]}))
@@ -375,8 +375,8 @@ def test_name_with_qualifier_search_with_category_match():
def test_name_only_search_with_countries():
q = make_query([(1, TokenType.PARTIAL, [(1, 'a')]),
(1, TokenType.WORD, [(100, 'a')])])
q = make_query([(1, qmod.TOKEN_PARTIAL, [(1, 'a')]),
(1, qmod.TOKEN_WORD, [(100, 'a')])])
builder = SearchBuilder(q, SearchDetails.from_kwargs({'countries': 'de,en'}))
searches = list(builder.build(TokenAssignment(name=TokenRange(0, 1))))
@@ -397,14 +397,14 @@ def make_counted_searches(name_part, name_full, address_part, address_full,
q.add_node(qmod.BREAK_WORD, PhraseType.NONE)
q.add_node(qmod.BREAK_END, PhraseType.NONE)
q.add_token(TokenRange(0, 1), TokenType.PARTIAL,
q.add_token(TokenRange(0, 1), qmod.TOKEN_PARTIAL,
MyToken(0.5, 1, name_part, 1, 'name_part'))
q.add_token(TokenRange(0, 1), TokenType.WORD,
q.add_token(TokenRange(0, 1), qmod.TOKEN_WORD,
MyToken(0, 101, name_full, 1, 'name_full'))
for i in range(num_address_parts):
q.add_token(TokenRange(i + 1, i + 2), TokenType.PARTIAL,
q.add_token(TokenRange(i + 1, i + 2), qmod.TOKEN_PARTIAL,
MyToken(0.5, 2, address_part, 1, 'address_part'))
q.add_token(TokenRange(i + 1, i + 2), TokenType.WORD,
q.add_token(TokenRange(i + 1, i + 2), qmod.TOKEN_WORD,
MyToken(0, 102, address_full, 1, 'address_full'))
builder = SearchBuilder(q, SearchDetails())

22
test/python/api/search/test_icu_query_analyzer.py
View File

@@ -11,7 +11,7 @@ import pytest
import pytest_asyncio
from nominatim_api import NominatimAPIAsync
from nominatim_api.search.query import Phrase, PhraseType, TokenType
from nominatim_api.search.query import Phrase, PhraseType
import nominatim_api.search.query as qmod
import nominatim_api.search.icu_tokenizer as tok
from nominatim_api.logging import set_log_output, get_and_disable
@@ -101,8 +101,8 @@ async def test_splitting_in_transliteration(conn):
@pytest.mark.asyncio
@pytest.mark.parametrize('term,order', [('23456', ['POSTCODE', 'HOUSENUMBER', 'WORD', 'PARTIAL']),
('3', ['HOUSENUMBER', 'POSTCODE', 'WORD', 'PARTIAL'])
@pytest.mark.parametrize('term,order', [('23456', ['P', 'H', 'W', 'w']),
('3', ['H', 'P', 'W', 'w'])
])
async def test_penalty_postcodes_and_housenumbers(conn, term, order):
ana = await tok.create_query_analyzer(conn)
@@ -116,7 +116,7 @@ async def test_penalty_postcodes_and_housenumbers(conn, term, order):
assert query.num_token_slots() == 1
torder = [(tl.tokens[0].penalty, tl.ttype.name) for tl in query.nodes[0].starting]
torder = [(tl.tokens[0].penalty, tl.ttype) for tl in query.nodes[0].starting]
torder.sort()
assert [t[1] for t in torder] == order
@@ -132,7 +132,7 @@ async def test_category_words_only_at_beginning(conn):
assert query.num_token_slots() == 3
assert len(query.nodes[0].starting) == 1
assert query.nodes[0].starting[0].ttype == TokenType.NEAR_ITEM
assert query.nodes[0].starting[0].ttype == qmod.TOKEN_NEAR_ITEM
assert not query.nodes[2].starting
@@ -146,7 +146,7 @@ async def test_freestanding_qualifier_words_become_category(conn):
assert query.num_token_slots() == 1
assert len(query.nodes[0].starting) == 1
assert query.nodes[0].starting[0].ttype == TokenType.NEAR_ITEM
assert query.nodes[0].starting[0].ttype == qmod.TOKEN_NEAR_ITEM
@pytest.mark.asyncio
@@ -159,9 +159,9 @@ async def test_qualifier_words(conn):
query = await ana.analyze_query(make_phrase('foo BAR foo BAR foo'))
assert query.num_token_slots() == 5
assert set(t.ttype for t in query.nodes[0].starting) == {TokenType.QUALIFIER}
assert set(t.ttype for t in query.nodes[2].starting) == {TokenType.QUALIFIER}
assert set(t.ttype for t in query.nodes[4].starting) == {TokenType.QUALIFIER}
assert set(t.ttype for t in query.nodes[0].starting) == {qmod.TOKEN_QUALIFIER}
assert set(t.ttype for t in query.nodes[2].starting) == {qmod.TOKEN_QUALIFIER}
assert set(t.ttype for t in query.nodes[4].starting) == {qmod.TOKEN_QUALIFIER}
@pytest.mark.asyncio
@@ -173,10 +173,10 @@ async def test_add_unknown_housenumbers(conn):
query = await ana.analyze_query(make_phrase('466 23 99834 34a'))
assert query.num_token_slots() == 4
assert query.nodes[0].starting[0].ttype == TokenType.HOUSENUMBER
assert query.nodes[0].starting[0].ttype == qmod.TOKEN_HOUSENUMBER
assert len(query.nodes[0].starting[0].tokens) == 1
assert query.nodes[0].starting[0].tokens[0].token == 0
assert query.nodes[1].starting[0].ttype == TokenType.HOUSENUMBER
assert query.nodes[1].starting[0].ttype == qmod.TOKEN_HOUSENUMBER
assert len(query.nodes[1].starting[0].tokens) == 1
assert query.nodes[1].starting[0].tokens[0].token == 1
assert not query.nodes[2].starting

152
test/python/api/search/test_token_assignment.py
View File

@@ -9,7 +9,7 @@ Test for creation of token assignments from tokenized queries.
"""
import pytest
from nominatim_api.search.query import QueryStruct, Phrase, PhraseType, TokenType, TokenRange, Token
from nominatim_api.search.query import QueryStruct, Phrase, PhraseType, TokenRange, Token
import nominatim_api.search.query as qmod
from nominatim_api.search.token_assignment import yield_token_assignments, TokenAssignment, PENALTY_TOKENCHANGE
@@ -52,9 +52,9 @@ def test_query_with_missing_tokens():
def test_one_word_query():
q = make_query((qmod.BREAK_START, PhraseType.NONE,
[(1, TokenType.PARTIAL),
(1, TokenType.WORD),
(1, TokenType.HOUSENUMBER)]))
[(1, qmod.TOKEN_PARTIAL),
(1, qmod.TOKEN_WORD),
(1, qmod.TOKEN_HOUSENUMBER)]))
res = list(yield_token_assignments(q))
assert res == [TokenAssignment(name=TokenRange(0, 1))]
@@ -62,7 +62,7 @@ def test_one_word_query():
def test_single_postcode():
q = make_query((qmod.BREAK_START, PhraseType.NONE,
[(1, TokenType.POSTCODE)]))
[(1, qmod.TOKEN_POSTCODE)]))
res = list(yield_token_assignments(q))
assert res == [TokenAssignment(postcode=TokenRange(0, 1))]
@@ -70,7 +70,7 @@ def test_single_postcode():
def test_single_country_name():
q = make_query((qmod.BREAK_START, PhraseType.NONE,
[(1, TokenType.COUNTRY)]))
[(1, qmod.TOKEN_COUNTRY)]))
res = list(yield_token_assignments(q))
assert res == [TokenAssignment(country=TokenRange(0, 1))]
@@ -78,8 +78,8 @@ def test_single_country_name():
def test_single_word_poi_search():
q = make_query((qmod.BREAK_START, PhraseType.NONE,
[(1, TokenType.NEAR_ITEM),
(1, TokenType.QUALIFIER)]))
[(1, qmod.TOKEN_NEAR_ITEM),
(1, qmod.TOKEN_QUALIFIER)]))
res = list(yield_token_assignments(q))
assert res == [TokenAssignment(near_item=TokenRange(0, 1))]
@@ -87,9 +87,9 @@ def test_single_word_poi_search():
@pytest.mark.parametrize('btype', [qmod.BREAK_WORD, qmod.BREAK_PART, qmod.BREAK_TOKEN])
def test_multiple_simple_words(btype):
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(btype, PhraseType.NONE, [(2, TokenType.PARTIAL)]),
(btype, PhraseType.NONE, [(3, TokenType.PARTIAL)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, qmod.TOKEN_PARTIAL)]),
(btype, PhraseType.NONE, [(2, qmod.TOKEN_PARTIAL)]),
(btype, PhraseType.NONE, [(3, qmod.TOKEN_PARTIAL)]))
penalty = PENALTY_TOKENCHANGE[btype]
@@ -107,8 +107,8 @@ def test_multiple_simple_words(btype):
def test_multiple_words_respect_phrase_break():
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(2, TokenType.PARTIAL)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, qmod.TOKEN_PARTIAL)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(2, qmod.TOKEN_PARTIAL)]))
check_assignments(yield_token_assignments(q),
TokenAssignment(name=TokenRange(0, 1),
@@ -118,8 +118,8 @@ def test_multiple_words_respect_phrase_break():
def test_housenumber_and_street():
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.HOUSENUMBER)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(2, TokenType.PARTIAL)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, qmod.TOKEN_HOUSENUMBER)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(2, qmod.TOKEN_PARTIAL)]))
check_assignments(yield_token_assignments(q),
TokenAssignment(name=TokenRange(1, 2),
@@ -129,8 +129,8 @@ def test_housenumber_and_street():
def test_housenumber_and_street_backwards():
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(2, TokenType.HOUSENUMBER)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, qmod.TOKEN_PARTIAL)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(2, qmod.TOKEN_HOUSENUMBER)]))
check_assignments(yield_token_assignments(q),
TokenAssignment(name=TokenRange(0, 1),
@@ -140,10 +140,10 @@ def test_housenumber_and_street_backwards():
def test_housenumber_and_postcode():
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, TokenType.HOUSENUMBER)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(3, TokenType.PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(4, TokenType.POSTCODE)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, qmod.TOKEN_PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, qmod.TOKEN_HOUSENUMBER)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(3, qmod.TOKEN_PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(4, qmod.TOKEN_POSTCODE)]))
check_assignments(yield_token_assignments(q),
TokenAssignment(penalty=pytest.approx(0.3),
@@ -157,10 +157,10 @@ def test_housenumber_and_postcode():
postcode=TokenRange(3, 4)))
def test_postcode_and_housenumber():
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, TokenType.POSTCODE)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(3, TokenType.PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(4, TokenType.HOUSENUMBER)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, qmod.TOKEN_PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, qmod.TOKEN_POSTCODE)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(3, qmod.TOKEN_PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(4, qmod.TOKEN_HOUSENUMBER)]))
check_assignments(yield_token_assignments(q),
TokenAssignment(penalty=pytest.approx(0.3),
@@ -175,38 +175,38 @@ def test_postcode_and_housenumber():
def test_country_housenumber_postcode():
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.COUNTRY)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, TokenType.PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(3, TokenType.HOUSENUMBER)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(4, TokenType.POSTCODE)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, qmod.TOKEN_COUNTRY)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, qmod.TOKEN_PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(3, qmod.TOKEN_HOUSENUMBER)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(4, qmod.TOKEN_POSTCODE)]))
check_assignments(yield_token_assignments(q))
@pytest.mark.parametrize('ttype', [TokenType.POSTCODE, TokenType.COUNTRY,
TokenType.NEAR_ITEM, TokenType.QUALIFIER])
@pytest.mark.parametrize('ttype', [qmod.TOKEN_POSTCODE, qmod.TOKEN_COUNTRY,
qmod.TOKEN_NEAR_ITEM, qmod.TOKEN_QUALIFIER])
def test_housenumber_with_only_special_terms(ttype):
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.HOUSENUMBER)]),
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, qmod.TOKEN_HOUSENUMBER)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, ttype)]))
check_assignments(yield_token_assignments(q))
@pytest.mark.parametrize('ttype', [TokenType.POSTCODE, TokenType.HOUSENUMBER, TokenType.COUNTRY])
@pytest.mark.parametrize('ttype', [qmod.TOKEN_POSTCODE, qmod.TOKEN_HOUSENUMBER, qmod.TOKEN_COUNTRY])
def test_multiple_special_tokens(ttype):
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, ttype)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(2, TokenType.PARTIAL)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(2, qmod.TOKEN_PARTIAL)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(3, ttype)]))
check_assignments(yield_token_assignments(q))
def test_housenumber_many_phrases():
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(2, TokenType.PARTIAL)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(3, TokenType.PARTIAL)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(4, TokenType.HOUSENUMBER)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(5, TokenType.PARTIAL)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, qmod.TOKEN_PARTIAL)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(2, qmod.TOKEN_PARTIAL)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(3, qmod.TOKEN_PARTIAL)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(4, qmod.TOKEN_HOUSENUMBER)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(5, qmod.TOKEN_PARTIAL)]))
check_assignments(yield_token_assignments(q),
TokenAssignment(penalty=0.1,
@@ -221,8 +221,8 @@ def test_housenumber_many_phrases():
def test_country_at_beginning():
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.COUNTRY)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, TokenType.PARTIAL)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, qmod.TOKEN_COUNTRY)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, qmod.TOKEN_PARTIAL)]))
check_assignments(yield_token_assignments(q),
TokenAssignment(penalty=0.1, name=TokenRange(1, 2),
@@ -230,8 +230,8 @@ def test_country_at_beginning():
def test_country_at_end():
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, TokenType.COUNTRY)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, qmod.TOKEN_PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, qmod.TOKEN_COUNTRY)]))
check_assignments(yield_token_assignments(q),
TokenAssignment(penalty=0.1, name=TokenRange(0, 1),
@@ -239,16 +239,16 @@ def test_country_at_end():
def test_country_in_middle():
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, TokenType.COUNTRY)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(3, TokenType.PARTIAL)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, qmod.TOKEN_PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, qmod.TOKEN_COUNTRY)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(3, qmod.TOKEN_PARTIAL)]))
check_assignments(yield_token_assignments(q))
def test_postcode_with_designation():
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.POSTCODE)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(2, TokenType.PARTIAL)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, qmod.TOKEN_POSTCODE)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(2, qmod.TOKEN_PARTIAL)]))
check_assignments(yield_token_assignments(q),
TokenAssignment(penalty=0.1, name=TokenRange(1, 2),
@@ -258,8 +258,8 @@ def test_postcode_with_designation():
def test_postcode_with_designation_backwards():
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(2, TokenType.POSTCODE)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, qmod.TOKEN_PARTIAL)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(2, qmod.TOKEN_POSTCODE)]))
check_assignments(yield_token_assignments(q),
TokenAssignment(name=TokenRange(0, 1),
@@ -269,8 +269,8 @@ def test_postcode_with_designation_backwards():
def test_near_item_at_beginning():
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.NEAR_ITEM)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, TokenType.PARTIAL)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, qmod.TOKEN_NEAR_ITEM)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, qmod.TOKEN_PARTIAL)]))
check_assignments(yield_token_assignments(q),
TokenAssignment(penalty=0.1, name=TokenRange(1, 2),
@@ -278,8 +278,8 @@ def test_near_item_at_beginning():
def test_near_item_at_end():
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, TokenType.NEAR_ITEM)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, qmod.TOKEN_PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, qmod.TOKEN_NEAR_ITEM)]))
check_assignments(yield_token_assignments(q),
TokenAssignment(penalty=0.1, name=TokenRange(0, 1),
@@ -287,17 +287,17 @@ def test_near_item_at_end():
def test_near_item_in_middle():
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, TokenType.NEAR_ITEM)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(3, TokenType.PARTIAL)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, qmod.TOKEN_PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, qmod.TOKEN_NEAR_ITEM)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(3, qmod.TOKEN_PARTIAL)]))
check_assignments(yield_token_assignments(q))
def test_qualifier_at_beginning():
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.QUALIFIER)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, TokenType.PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(3, TokenType.PARTIAL)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, qmod.TOKEN_QUALIFIER)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, qmod.TOKEN_PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(3, qmod.TOKEN_PARTIAL)]))
check_assignments(yield_token_assignments(q),
@@ -309,11 +309,11 @@ def test_qualifier_at_beginning():
def test_qualifier_after_name():
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, TokenType.PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(3, TokenType.QUALIFIER)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(4, TokenType.PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(5, TokenType.PARTIAL)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, qmod.TOKEN_PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, qmod.TOKEN_PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(3, qmod.TOKEN_QUALIFIER)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(4, qmod.TOKEN_PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(5, qmod.TOKEN_PARTIAL)]))
check_assignments(yield_token_assignments(q),
@@ -326,27 +326,27 @@ def test_qualifier_after_name():
def test_qualifier_before_housenumber():
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.QUALIFIER)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, TokenType.HOUSENUMBER)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(3, TokenType.PARTIAL)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, qmod.TOKEN_QUALIFIER)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, qmod.TOKEN_HOUSENUMBER)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(3, qmod.TOKEN_PARTIAL)]))
check_assignments(yield_token_assignments(q))
def test_qualifier_after_housenumber():
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.HOUSENUMBER)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, TokenType.QUALIFIER)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(3, TokenType.PARTIAL)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, qmod.TOKEN_HOUSENUMBER)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, qmod.TOKEN_QUALIFIER)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(3, qmod.TOKEN_PARTIAL)]))
check_assignments(yield_token_assignments(q))
def test_qualifier_in_middle_of_phrase():
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(2, TokenType.PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(3, TokenType.QUALIFIER)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(4, TokenType.PARTIAL)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(5, TokenType.PARTIAL)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, qmod.TOKEN_PARTIAL)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(2, qmod.TOKEN_PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(3, qmod.TOKEN_QUALIFIER)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(4, qmod.TOKEN_PARTIAL)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(5, qmod.TOKEN_PARTIAL)]))
check_assignments(yield_token_assignments(q))