hans/Nominatim
1
0
Fork 1
mirror of https://github.com/osm-search/Nominatim.git synced 2026-02-26 11:08:13 +00:00
Code Issues Packages Projects Releases Wiki Activity

add tests for new full name computation with ICU

Browse Source
This commit is contained in:
Sarah Hoffmann
2021-05-24 10:29:21 +02:00
parent 4f4d15c28a
commit 24c986c842
3 changed files with 7 additions and 16 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
nominatim/tokenizer/legacy_icu_tokenizer.py
View File

@@ -537,8 +537,6 @@ class _TokenInfo:
""" """
# Start with all partial names # Start with all partial names
terms = set((part for ns in names for part in ns.split())) terms = set((part for ns in names for part in ns.split()))
# Add partials for the full terms (TO BE REMOVED)
terms.update((n for n in names))
# Add the full names # Add the full names
terms.update((' ' + n for n in names)) terms.update((' ' + n for n in names))

4
test/Makefile
View File

@@ -4,8 +4,8 @@ no-test-db: bdd-no-test-db php
bdd: bdd:
cd bdd && behave -DREMOVE_TEMPLATE=1 cd bdd && behave -DREMOVE_TEMPLATE=1
bdd-no-test-db: icu:
cd bdd && behave -DREMOVE_TEMPLATE=1 db osm2pgsql cd bdd && behave -DREMOVE_TEMPLATE=1 -DTOKENIZER=legacy_icu
php: php:
cd php && phpunit ./ cd php && phpunit ./

17
test/python/test_tokenizer_legacy_icu.py
View File

@@ -232,23 +232,16 @@ def test_process_place_names(analyzer, getorcreate_term_id):
@pytest.mark.parametrize('sep', [',' , ';']) @pytest.mark.parametrize('sep', [',' , ';'])
def test_full_names_with_separator(analyzer, getorcreate_term_id, sep): def test_full_names_with_separator(analyzer, getorcreate_term_id, sep):
with analyzer() as anl: with analyzer() as anl:
full_names = names = anl._compute_full_names({'name' : sep.join(('New York', 'Big Apple'))})
anl._compute_full_names({'name' : sep.join(('New York', 'Big Apple'))})
expect = set((anl.make_standard_word(w) for w in ('New York', 'Big Apple'))) assert names == set(('NEW YORK', 'BIG APPLE'))
assert full_names == expect
def test_process_place_names_with_bracket(analyzer, getorcreate_term_id): def test_full_names_with_bracket(analyzer, getorcreate_term_id):
with analyzer() as anl: with analyzer() as anl:
info = anl.process_place({'name' : names = anl._compute_full_names({'name' : 'Houseboat (left)'})
{'name' : 'Houseboat (left)'}})
expect = set((anl.make_standard_word(w) for w in assert names == set(('HOUSEBOAT (LEFT)', 'HOUSEBOAT'))
(' houseboat', ' houseboat left', 'houseboat', 'left')))
assert eval(info['names']) == expect
@pytest.mark.parametrize('pcode', ['12345', 'AB 123', '34-345']) @pytest.mark.parametrize('pcode', ['12345', 'AB 123', '34-345'])