Merge remote-tracking branch 'upstream/master'

[nominatim.git] / test / python / api / search / test_token_assignment.py

diff --git a/test/python/api/search/test_token_assignment.py b/test/python/api/search/test_token_assignment.py
index 2ed55a0f80afb06372ef27f6a49fcf840e3854cc..fff8d471e246c8fdec6c72346a15c47c43d01fce 100644 (file)
--- a/test/python/api/search/test_token_assignment.py
+++ b/test/python/api/search/test_token_assignment.py
@@ -2,15 +2,16 @@
 #
 # This file is part of Nominatim. (https://nominatim.org)
 #
 #
 # This file is part of Nominatim. (https://nominatim.org)
 #

-# Copyright (C) 2023 by the Nominatim developer community.
+# Copyright (C) 2024 by the Nominatim developer community.

 # For a full list of authors see the git log.
 """
 Test for creation of token assignments from tokenized queries.
 """
 import pytest
 
 # For a full list of authors see the git log.
 """
 Test for creation of token assignments from tokenized queries.
 """
 import pytest
 

-from nominatim.api.search.query import QueryStruct, Phrase, PhraseType, BreakType, TokenType, TokenRange, Token
-from nominatim.api.search.token_assignment import yield_token_assignments, TokenAssignment, PENALTY_TOKENCHANGE
+from nominatim_api.search.query import QueryStruct, Phrase, TokenRange, Token
+import nominatim_api.search.query as qmod
+from nominatim_api.search.token_assignment import yield_token_assignments, TokenAssignment, PENALTY_TOKENCHANGE

 
 class MyToken(Token):
     def get_category(self):
 
 class MyToken(Token):
     def get_category(self):


@@ -19,11 +20,12 @@ class MyToken(Token):
 
 def make_query(*args):
     q = QueryStruct([Phrase(args[0][1], '')])
 
 def make_query(*args):
     q = QueryStruct([Phrase(args[0][1], '')])

-    dummy = MyToken(3.0, 45, 1, 'foo', True)
+    dummy = MyToken(penalty=3.0, token=45, count=1, addr_count=1,
+                    lookup_word='foo')

 
     for btype, ptype, _ in args[1:]:
         q.add_node(btype, ptype)
 
     for btype, ptype, _ in args[1:]:
         q.add_node(btype, ptype)

-    q.add_node(BreakType.END, PhraseType.NONE)
+    q.add_node(qmod.BREAK_END, qmod.PHRASE_ANY)

 
     for start, t in enumerate(args):
         for end, ttype in t[2]:
 
     for start, t in enumerate(args):
         for end, ttype in t[2]:


@@ -42,52 +44,52 @@ def check_assignments(actual, *expected):
 
 
 def test_query_with_missing_tokens():
 
 
 def test_query_with_missing_tokens():

-    q = QueryStruct([Phrase(PhraseType.NONE, '')])
-    q.add_node(BreakType.END, PhraseType.NONE)
+    q = QueryStruct([Phrase(qmod.PHRASE_ANY, '')])
+    q.add_node(qmod.BREAK_END, qmod.PHRASE_ANY)

 
     assert list(yield_token_assignments(q)) == []
 
 
 def test_one_word_query():
 
     assert list(yield_token_assignments(q)) == []
 
 
 def test_one_word_query():

-    q = make_query((BreakType.START, PhraseType.NONE,
-                    [(1, TokenType.PARTIAL),
-                     (1, TokenType.WORD),
-                     (1, TokenType.HOUSENUMBER)]))
+    q = make_query((qmod.BREAK_START, qmod.PHRASE_ANY,
+                    [(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))]
 
 
 def test_single_postcode():
 
     res = list(yield_token_assignments(q))
     assert res == [TokenAssignment(name=TokenRange(0, 1))]
 
 
 def test_single_postcode():

-    q = make_query((BreakType.START, PhraseType.NONE,
-                    [(1, TokenType.POSTCODE)]))
+    q = make_query((qmod.BREAK_START, qmod.PHRASE_ANY,
+                    [(1, qmod.TOKEN_POSTCODE)]))

 
     res = list(yield_token_assignments(q))
     assert res == [TokenAssignment(postcode=TokenRange(0, 1))]
 
 
 def test_single_country_name():
 
     res = list(yield_token_assignments(q))
     assert res == [TokenAssignment(postcode=TokenRange(0, 1))]
 
 
 def test_single_country_name():

-    q = make_query((BreakType.START, PhraseType.NONE,
-                    [(1, TokenType.COUNTRY)]))
+    q = make_query((qmod.BREAK_START, qmod.PHRASE_ANY,
+                    [(1, qmod.TOKEN_COUNTRY)]))

 
     res = list(yield_token_assignments(q))
     assert res == [TokenAssignment(country=TokenRange(0, 1))]
 
 
 def test_single_word_poi_search():
 
     res = list(yield_token_assignments(q))
     assert res == [TokenAssignment(country=TokenRange(0, 1))]
 
 
 def test_single_word_poi_search():

-    q = make_query((BreakType.START, PhraseType.NONE,
-                    [(1, TokenType.NEAR_ITEM),
-                     (1, TokenType.QUALIFIER)]))
+    q = make_query((qmod.BREAK_START, qmod.PHRASE_ANY,
+                    [(1, qmod.TOKEN_NEAR_ITEM),
+                     (1, qmod.TOKEN_QUALIFIER)]))

 
     res = list(yield_token_assignments(q))
     assert res == [TokenAssignment(near_item=TokenRange(0, 1))]
 
 
 
     res = list(yield_token_assignments(q))
     assert res == [TokenAssignment(near_item=TokenRange(0, 1))]
 
 

-@pytest.mark.parametrize('btype', [BreakType.WORD, BreakType.PART, BreakType.TOKEN])
+@pytest.mark.parametrize('btype', [qmod.BREAK_WORD, qmod.BREAK_PART, qmod.BREAK_TOKEN])

 def test_multiple_simple_words(btype):
 def test_multiple_simple_words(btype):

-    q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
-                   (btype, PhraseType.NONE, [(2, TokenType.PARTIAL)]),
-                   (btype, PhraseType.NONE, [(3, TokenType.PARTIAL)]))
+    q = make_query((qmod.BREAK_START, qmod.PHRASE_ANY, [(1, qmod.TOKEN_PARTIAL)]),
+                   (btype, qmod.PHRASE_ANY, [(2, qmod.TOKEN_PARTIAL)]),
+                   (btype, qmod.PHRASE_ANY, [(3, qmod.TOKEN_PARTIAL)]))

 
     penalty = PENALTY_TOKENCHANGE[btype]
 
 
     penalty = PENALTY_TOKENCHANGE[btype]
 


@@ -105,8 +107,8 @@ def test_multiple_simple_words(btype):
 
 
 def test_multiple_words_respect_phrase_break():
 
 
 def test_multiple_words_respect_phrase_break():

-    q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
-                   (BreakType.PHRASE, PhraseType.NONE, [(2, TokenType.PARTIAL)]))
+    q = make_query((qmod.BREAK_START, qmod.PHRASE_ANY, [(1, qmod.TOKEN_PARTIAL)]),
+                   (qmod.BREAK_PHRASE, qmod.PHRASE_ANY, [(2, qmod.TOKEN_PARTIAL)]))

 
     check_assignments(yield_token_assignments(q),
                       TokenAssignment(name=TokenRange(0, 1),
 
     check_assignments(yield_token_assignments(q),
                       TokenAssignment(name=TokenRange(0, 1),


@@ -116,8 +118,8 @@ def test_multiple_words_respect_phrase_break():
 
 
 def test_housenumber_and_street():
 
 
 def test_housenumber_and_street():

-    q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.HOUSENUMBER)]),
-                   (BreakType.PHRASE, PhraseType.NONE, [(2, TokenType.PARTIAL)]))
+    q = make_query((qmod.BREAK_START, qmod.PHRASE_ANY, [(1, qmod.TOKEN_HOUSENUMBER)]),
+                   (qmod.BREAK_PHRASE, qmod.PHRASE_ANY, [(2, qmod.TOKEN_PARTIAL)]))

 
     check_assignments(yield_token_assignments(q),
                       TokenAssignment(name=TokenRange(1, 2),
 
     check_assignments(yield_token_assignments(q),
                       TokenAssignment(name=TokenRange(1, 2),


@@ -127,8 +129,8 @@ def test_housenumber_and_street():
 
 
 def test_housenumber_and_street_backwards():
 
 
 def test_housenumber_and_street_backwards():

-    q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
-                   (BreakType.PHRASE, PhraseType.NONE, [(2, TokenType.HOUSENUMBER)]))
+    q = make_query((qmod.BREAK_START, qmod.PHRASE_ANY, [(1, qmod.TOKEN_PARTIAL)]),
+                   (qmod.BREAK_PHRASE, qmod.PHRASE_ANY, [(2, qmod.TOKEN_HOUSENUMBER)]))

 
     check_assignments(yield_token_assignments(q),
                       TokenAssignment(name=TokenRange(0, 1),
 
     check_assignments(yield_token_assignments(q),
                       TokenAssignment(name=TokenRange(0, 1),


@@ -138,10 +140,10 @@ def test_housenumber_and_street_backwards():
 
 
 def test_housenumber_and_postcode():
 
 
 def test_housenumber_and_postcode():

-    q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
-                   (BreakType.WORD, PhraseType.NONE, [(2, TokenType.HOUSENUMBER)]),
-                   (BreakType.WORD, PhraseType.NONE, [(3, TokenType.PARTIAL)]),
-                   (BreakType.WORD, PhraseType.NONE, [(4, TokenType.POSTCODE)]))
+    q = make_query((qmod.BREAK_START, qmod.PHRASE_ANY, [(1, qmod.TOKEN_PARTIAL)]),
+                   (qmod.BREAK_WORD, qmod.PHRASE_ANY, [(2, qmod.TOKEN_HOUSENUMBER)]),
+                   (qmod.BREAK_WORD, qmod.PHRASE_ANY, [(3, qmod.TOKEN_PARTIAL)]),
+                   (qmod.BREAK_WORD, qmod.PHRASE_ANY, [(4, qmod.TOKEN_POSTCODE)]))

 
     check_assignments(yield_token_assignments(q),
                       TokenAssignment(penalty=pytest.approx(0.3),
 
     check_assignments(yield_token_assignments(q),
                       TokenAssignment(penalty=pytest.approx(0.3),


@@ -155,10 +157,10 @@ def test_housenumber_and_postcode():
                                       postcode=TokenRange(3, 4)))
 
 def test_postcode_and_housenumber():
                                       postcode=TokenRange(3, 4)))
 
 def test_postcode_and_housenumber():

-    q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
-                   (BreakType.WORD, PhraseType.NONE, [(2, TokenType.POSTCODE)]),
-                   (BreakType.WORD, PhraseType.NONE, [(3, TokenType.PARTIAL)]),
-                   (BreakType.WORD, PhraseType.NONE, [(4, TokenType.HOUSENUMBER)]))
+    q = make_query((qmod.BREAK_START, qmod.PHRASE_ANY, [(1, qmod.TOKEN_PARTIAL)]),
+                   (qmod.BREAK_WORD, qmod.PHRASE_ANY, [(2, qmod.TOKEN_POSTCODE)]),
+                   (qmod.BREAK_WORD, qmod.PHRASE_ANY, [(3, qmod.TOKEN_PARTIAL)]),
+                   (qmod.BREAK_WORD, qmod.PHRASE_ANY, [(4, qmod.TOKEN_HOUSENUMBER)]))

 
     check_assignments(yield_token_assignments(q),
                       TokenAssignment(penalty=pytest.approx(0.3),
 
     check_assignments(yield_token_assignments(q),
                       TokenAssignment(penalty=pytest.approx(0.3),


@@ -173,38 +175,38 @@ def test_postcode_and_housenumber():
 
 
 def test_country_housenumber_postcode():
 
 
 def test_country_housenumber_postcode():

-    q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.COUNTRY)]),
-                   (BreakType.WORD, PhraseType.NONE, [(2, TokenType.PARTIAL)]),
-                   (BreakType.WORD, PhraseType.NONE, [(3, TokenType.HOUSENUMBER)]),
-                   (BreakType.WORD, PhraseType.NONE, [(4, TokenType.POSTCODE)]))
+    q = make_query((qmod.BREAK_START, qmod.PHRASE_ANY, [(1, qmod.TOKEN_COUNTRY)]),
+                   (qmod.BREAK_WORD, qmod.PHRASE_ANY, [(2, qmod.TOKEN_PARTIAL)]),
+                   (qmod.BREAK_WORD, qmod.PHRASE_ANY, [(3, qmod.TOKEN_HOUSENUMBER)]),
+                   (qmod.BREAK_WORD, qmod.PHRASE_ANY, [(4, qmod.TOKEN_POSTCODE)]))

 
     check_assignments(yield_token_assignments(q))
 
 
 
     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):
 def test_housenumber_with_only_special_terms(ttype):

-    q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.HOUSENUMBER)]),
-                   (BreakType.WORD, PhraseType.NONE, [(2, ttype)]))
+    q = make_query((qmod.BREAK_START, qmod.PHRASE_ANY, [(1, qmod.TOKEN_HOUSENUMBER)]),
+                   (qmod.BREAK_WORD, qmod.PHRASE_ANY, [(2, ttype)]))

 
     check_assignments(yield_token_assignments(q))
 
 
 
     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):
 def test_multiple_special_tokens(ttype):

-    q = make_query((BreakType.START, PhraseType.NONE, [(1, ttype)]),
-                   (BreakType.PHRASE, PhraseType.NONE, [(2, TokenType.PARTIAL)]),
-                   (BreakType.PHRASE, PhraseType.NONE, [(3, ttype)]))
+    q = make_query((qmod.BREAK_START, qmod.PHRASE_ANY, [(1, ttype)]),
+                   (qmod.BREAK_PHRASE, qmod.PHRASE_ANY, [(2, qmod.TOKEN_PARTIAL)]),
+                   (qmod.BREAK_PHRASE, qmod.PHRASE_ANY, [(3, ttype)]))

 
     check_assignments(yield_token_assignments(q))
 
 
 def test_housenumber_many_phrases():
 
     check_assignments(yield_token_assignments(q))
 
 
 def test_housenumber_many_phrases():

-    q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
-                   (BreakType.PHRASE, PhraseType.NONE, [(2, TokenType.PARTIAL)]),
-                   (BreakType.PHRASE, PhraseType.NONE, [(3, TokenType.PARTIAL)]),
-                   (BreakType.PHRASE, PhraseType.NONE, [(4, TokenType.HOUSENUMBER)]),
-                   (BreakType.WORD, PhraseType.NONE, [(5, TokenType.PARTIAL)]))
+    q = make_query((qmod.BREAK_START, qmod.PHRASE_ANY, [(1, qmod.TOKEN_PARTIAL)]),
+                   (qmod.BREAK_PHRASE, qmod.PHRASE_ANY, [(2, qmod.TOKEN_PARTIAL)]),
+                   (qmod.BREAK_PHRASE, qmod.PHRASE_ANY, [(3, qmod.TOKEN_PARTIAL)]),
+                   (qmod.BREAK_PHRASE, qmod.PHRASE_ANY, [(4, qmod.TOKEN_HOUSENUMBER)]),
+                   (qmod.BREAK_WORD, qmod.PHRASE_ANY, [(5, qmod.TOKEN_PARTIAL)]))

 
     check_assignments(yield_token_assignments(q),
                       TokenAssignment(penalty=0.1,
 
     check_assignments(yield_token_assignments(q),
                       TokenAssignment(penalty=0.1,


@@ -219,8 +221,8 @@ def test_housenumber_many_phrases():
 
 
 def test_country_at_beginning():
 
 
 def test_country_at_beginning():

-    q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.COUNTRY)]),
-                   (BreakType.WORD, PhraseType.NONE, [(2, TokenType.PARTIAL)]))
+    q = make_query((qmod.BREAK_START, qmod.PHRASE_ANY, [(1, qmod.TOKEN_COUNTRY)]),
+                   (qmod.BREAK_WORD, qmod.PHRASE_ANY, [(2, qmod.TOKEN_PARTIAL)]))

 
     check_assignments(yield_token_assignments(q),
                       TokenAssignment(penalty=0.1, name=TokenRange(1, 2),
 
     check_assignments(yield_token_assignments(q),
                       TokenAssignment(penalty=0.1, name=TokenRange(1, 2),


@@ -228,8 +230,8 @@ def test_country_at_beginning():
 
 
 def test_country_at_end():
 
 
 def test_country_at_end():

-    q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
-                   (BreakType.WORD, PhraseType.NONE, [(2, TokenType.COUNTRY)]))
+    q = make_query((qmod.BREAK_START, qmod.PHRASE_ANY, [(1, qmod.TOKEN_PARTIAL)]),
+                   (qmod.BREAK_WORD, qmod.PHRASE_ANY, [(2, qmod.TOKEN_COUNTRY)]))

 
     check_assignments(yield_token_assignments(q),
                       TokenAssignment(penalty=0.1, name=TokenRange(0, 1),
 
     check_assignments(yield_token_assignments(q),
                       TokenAssignment(penalty=0.1, name=TokenRange(0, 1),


@@ -237,16 +239,16 @@ def test_country_at_end():
 
 
 def test_country_in_middle():
 
 
 def test_country_in_middle():

-    q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
-                   (BreakType.WORD, PhraseType.NONE, [(2, TokenType.COUNTRY)]),
-                   (BreakType.WORD, PhraseType.NONE, [(3, TokenType.PARTIAL)]))
+    q = make_query((qmod.BREAK_START, qmod.PHRASE_ANY, [(1, qmod.TOKEN_PARTIAL)]),
+                   (qmod.BREAK_WORD, qmod.PHRASE_ANY, [(2, qmod.TOKEN_COUNTRY)]),
+                   (qmod.BREAK_WORD, qmod.PHRASE_ANY, [(3, qmod.TOKEN_PARTIAL)]))

 
     check_assignments(yield_token_assignments(q))
 
 
 def test_postcode_with_designation():
 
     check_assignments(yield_token_assignments(q))
 
 
 def test_postcode_with_designation():

-    q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.POSTCODE)]),
-                   (BreakType.PHRASE, PhraseType.NONE, [(2, TokenType.PARTIAL)]))
+    q = make_query((qmod.BREAK_START, qmod.PHRASE_ANY, [(1, qmod.TOKEN_POSTCODE)]),
+                   (qmod.BREAK_PHRASE, qmod.PHRASE_ANY, [(2, qmod.TOKEN_PARTIAL)]))

 
     check_assignments(yield_token_assignments(q),
                       TokenAssignment(penalty=0.1, name=TokenRange(1, 2),
 
     check_assignments(yield_token_assignments(q),
                       TokenAssignment(penalty=0.1, name=TokenRange(1, 2),


@@ -256,8 +258,8 @@ def test_postcode_with_designation():
 
 
 def test_postcode_with_designation_backwards():
 
 
 def test_postcode_with_designation_backwards():

-    q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
-                   (BreakType.PHRASE, PhraseType.NONE, [(2, TokenType.POSTCODE)]))
+    q = make_query((qmod.BREAK_START, qmod.PHRASE_ANY, [(1, qmod.TOKEN_PARTIAL)]),
+                   (qmod.BREAK_PHRASE, qmod.PHRASE_ANY, [(2, qmod.TOKEN_POSTCODE)]))

 
     check_assignments(yield_token_assignments(q),
                       TokenAssignment(name=TokenRange(0, 1),
 
     check_assignments(yield_token_assignments(q),
                       TokenAssignment(name=TokenRange(0, 1),


@@ -267,8 +269,8 @@ def test_postcode_with_designation_backwards():
 
 
 def test_near_item_at_beginning():
 
 
 def test_near_item_at_beginning():

-    q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.NEAR_ITEM)]),
-                   (BreakType.WORD, PhraseType.NONE, [(2, TokenType.PARTIAL)]))
+    q = make_query((qmod.BREAK_START, qmod.PHRASE_ANY, [(1, qmod.TOKEN_NEAR_ITEM)]),
+                   (qmod.BREAK_WORD, qmod.PHRASE_ANY, [(2, qmod.TOKEN_PARTIAL)]))

 
     check_assignments(yield_token_assignments(q),
                       TokenAssignment(penalty=0.1, name=TokenRange(1, 2),
 
     check_assignments(yield_token_assignments(q),
                       TokenAssignment(penalty=0.1, name=TokenRange(1, 2),


@@ -276,8 +278,8 @@ def test_near_item_at_beginning():
 
 
 def test_near_item_at_end():
 
 
 def test_near_item_at_end():

-    q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
-                   (BreakType.WORD, PhraseType.NONE, [(2, TokenType.NEAR_ITEM)]))
+    q = make_query((qmod.BREAK_START, qmod.PHRASE_ANY, [(1, qmod.TOKEN_PARTIAL)]),
+                   (qmod.BREAK_WORD, qmod.PHRASE_ANY, [(2, qmod.TOKEN_NEAR_ITEM)]))

 
     check_assignments(yield_token_assignments(q),
                       TokenAssignment(penalty=0.1, name=TokenRange(0, 1),
 
     check_assignments(yield_token_assignments(q),
                       TokenAssignment(penalty=0.1, name=TokenRange(0, 1),


@@ -285,17 +287,17 @@ def test_near_item_at_end():
 
 
 def test_near_item_in_middle():
 
 
 def test_near_item_in_middle():

-    q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
-                   (BreakType.WORD, PhraseType.NONE, [(2, TokenType.NEAR_ITEM)]),
-                   (BreakType.WORD, PhraseType.NONE, [(3, TokenType.PARTIAL)]))
+    q = make_query((qmod.BREAK_START, qmod.PHRASE_ANY, [(1, qmod.TOKEN_PARTIAL)]),
+                   (qmod.BREAK_WORD, qmod.PHRASE_ANY, [(2, qmod.TOKEN_NEAR_ITEM)]),
+                   (qmod.BREAK_WORD, qmod.PHRASE_ANY, [(3, qmod.TOKEN_PARTIAL)]))

 
     check_assignments(yield_token_assignments(q))
 
 
 def test_qualifier_at_beginning():
 
     check_assignments(yield_token_assignments(q))
 
 
 def test_qualifier_at_beginning():

-    q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.QUALIFIER)]),
-                   (BreakType.WORD, PhraseType.NONE, [(2, TokenType.PARTIAL)]),
-                   (BreakType.WORD, PhraseType.NONE, [(3, TokenType.PARTIAL)]))
+    q = make_query((qmod.BREAK_START, qmod.PHRASE_ANY, [(1, qmod.TOKEN_QUALIFIER)]),
+                   (qmod.BREAK_WORD, qmod.PHRASE_ANY, [(2, qmod.TOKEN_PARTIAL)]),
+                   (qmod.BREAK_WORD, qmod.PHRASE_ANY, [(3, qmod.TOKEN_PARTIAL)]))

 
 
     check_assignments(yield_token_assignments(q),
 
 
     check_assignments(yield_token_assignments(q),


@@ -307,11 +309,11 @@ def test_qualifier_at_beginning():
 
 
 def test_qualifier_after_name():
 
 
 def test_qualifier_after_name():

-    q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
-                   (BreakType.WORD, PhraseType.NONE, [(2, TokenType.PARTIAL)]),
-                   (BreakType.WORD, PhraseType.NONE, [(3, TokenType.QUALIFIER)]),
-                   (BreakType.WORD, PhraseType.NONE, [(4, TokenType.PARTIAL)]),
-                   (BreakType.WORD, PhraseType.NONE, [(5, TokenType.PARTIAL)]))
+    q = make_query((qmod.BREAK_START, qmod.PHRASE_ANY, [(1, qmod.TOKEN_PARTIAL)]),
+                   (qmod.BREAK_WORD, qmod.PHRASE_ANY, [(2, qmod.TOKEN_PARTIAL)]),
+                   (qmod.BREAK_WORD, qmod.PHRASE_ANY, [(3, qmod.TOKEN_QUALIFIER)]),
+                   (qmod.BREAK_WORD, qmod.PHRASE_ANY, [(4, qmod.TOKEN_PARTIAL)]),
+                   (qmod.BREAK_WORD, qmod.PHRASE_ANY, [(5, qmod.TOKEN_PARTIAL)]))

 
 
     check_assignments(yield_token_assignments(q),
 
 
     check_assignments(yield_token_assignments(q),


@@ -324,16 +326,27 @@ def test_qualifier_after_name():
 
 
 def test_qualifier_before_housenumber():
 
 
 def test_qualifier_before_housenumber():

-    q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.QUALIFIER)]),
-                   (BreakType.WORD, PhraseType.NONE, [(2, TokenType.HOUSENUMBER)]),
-                   (BreakType.WORD, PhraseType.NONE, [(3, TokenType.PARTIAL)]))
+    q = make_query((qmod.BREAK_START, qmod.PHRASE_ANY, [(1, qmod.TOKEN_QUALIFIER)]),
+                   (qmod.BREAK_WORD, qmod.PHRASE_ANY, [(2, qmod.TOKEN_HOUSENUMBER)]),
+                   (qmod.BREAK_WORD, qmod.PHRASE_ANY, [(3, qmod.TOKEN_PARTIAL)]))

 
     check_assignments(yield_token_assignments(q))
 
 
 def test_qualifier_after_housenumber():
 
     check_assignments(yield_token_assignments(q))
 
 
 def test_qualifier_after_housenumber():

-    q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.HOUSENUMBER)]),
-                   (BreakType.WORD, PhraseType.NONE, [(2, TokenType.QUALIFIER)]),
-                   (BreakType.WORD, PhraseType.NONE, [(3, TokenType.PARTIAL)]))
+    q = make_query((qmod.BREAK_START, qmod.PHRASE_ANY, [(1, qmod.TOKEN_HOUSENUMBER)]),
+                   (qmod.BREAK_WORD, qmod.PHRASE_ANY, [(2, qmod.TOKEN_QUALIFIER)]),
+                   (qmod.BREAK_WORD, qmod.PHRASE_ANY, [(3, qmod.TOKEN_PARTIAL)]))

 
     check_assignments(yield_token_assignments(q))
 
     check_assignments(yield_token_assignments(q))

+
+
+def test_qualifier_in_middle_of_phrase():
+    q = make_query((qmod.BREAK_START, qmod.PHRASE_ANY, [(1, qmod.TOKEN_PARTIAL)]),
+                   (qmod.BREAK_PHRASE, qmod.PHRASE_ANY, [(2, qmod.TOKEN_PARTIAL)]),
+                   (qmod.BREAK_WORD, qmod.PHRASE_ANY, [(3, qmod.TOKEN_QUALIFIER)]),
+                   (qmod.BREAK_WORD, qmod.PHRASE_ANY, [(4, qmod.TOKEN_PARTIAL)]),
+                   (qmod.BREAK_PHRASE, qmod.PHRASE_ANY, [(5, qmod.TOKEN_PARTIAL)]))
+
+    check_assignments(yield_token_assignments(q))
+