Merge pull request #3493 from lonvia/clean-up-bdd-tests

Various cleanups of BDD tests

.github/workflows/ci-tests.yml

@@ -118,7 +118,8 @@ jobs:
 
             - name: BDD tests
               run: |
-                  python3 -m behave -DREMOVE_TEMPLATE=1 -DBUILDDIR=$GITHUB_WORKSPACE/build --format=progress3
+                  export PATH=$GITHUB_WORKSPACE/build/osm2pgsql:$PATH
+                  python3 -m behave -DREMOVE_TEMPLATE=1 --format=progress3
               working-directory: Nominatim/test/bdd
 
             - name: Install mypy and typechecking info
@@ -170,7 +171,8 @@ jobs:
 
             - name: BDD tests (legacy tokenizer)
               run: |
-                  python3 -m behave -DREMOVE_TEMPLATE=1 -DBUILDDIR=$GITHUB_WORKSPACE/build -DAPI_ENGINE=php -DTOKENIZER=legacy --format=progress3
+                  export PATH=$GITHUB_WORKSPACE/build/osm2pgsql:$PATH
+                  python3 -m behave -DREMOVE_TEMPLATE=1 -DSERVER_MODULE_PATH=$GITHUB_WORKSPACE/build/module -DAPI_ENGINE=php -DTOKENIZER=legacy --format=progress3
               working-directory: Nominatim/test/bdd
 
 
@@ -217,7 +219,8 @@ jobs:
 
             - name: BDD tests (php)
               run: |
-                  python3 -m behave -DREMOVE_TEMPLATE=1 -DBUILDDIR=$GITHUB_WORKSPACE/build -DAPI_ENGINE=php --format=progress3
+                  export PATH=$GITHUB_WORKSPACE/build/osm2pgsql:$PATH
+                  python3 -m behave -DREMOVE_TEMPLATE=1 -DAPI_ENGINE=php --format=progress3
               working-directory: Nominatim/test/bdd
 
 

docs/develop/Development-Environment.md

@@ -50,6 +50,10 @@ The documentation is built with mkdocs:
 * [mkdocstrings](https://mkdocstrings.github.io/) >= 0.25
 * [mkdocs-material](https://squidfunk.github.io/mkdocs-material/)
 
+Please be aware that tests always run against the globally installed
+osm2pgsql, so you need to have this set up. If you want to test against
+the vendored version of osm2pgsql, you need to set the PATH accordingly.
+
 ### Installing prerequisites on Ubuntu/Debian
 
 The Python tools should always be run with the most recent version.

docs/develop/Testing.md

@@ -78,7 +78,6 @@ To run the functional tests, do
 
 The tests can be configured with a set of environment variables (`behave -D key=val`):
 
- * `BUILDDIR` - build directory of Nominatim installation to test
  * `TEMPLATE_DB` - name of template database used as a skeleton for
                    the test databases (db tests)
  * `TEST_DB` - name of test database (db tests)
@@ -91,7 +90,7 @@ The tests can be configured with a set of environment variables (`behave -D key=
  * `DB_USER` - (optional) username of database login
  * `DB_PASS` - (optional) password for database login
  * `SERVER_MODULE_PATH` - (optional) path on the Postgres server to Nominatim
-                          module shared library file
+                          module shared library file (only needed for legacy tokenizer)
  * `REMOVE_TEMPLATE` - if true, the template and API database will not be reused
                        during the next run. Reusing the base templates speeds
                        up tests considerably but might lead to outdated errors
@@ -122,23 +121,6 @@ and compromises the following data:
 API tests should only be testing the functionality of the website PHP code.
 Most tests should be formulated as BDD DB creation tests (see below) instead.
 
-#### Code Coverage (PHP engine only)
-
-The API tests also support code coverage tests. You need to install
-[PHP_CodeCoverage](https://github.com/sebastianbergmann/php-code-coverage).
-On Debian/Ubuntu run:
-
-    apt-get install php-codecoverage php-xdebug
-
-Then run the API tests as follows:
-
-    behave api -DPHPCOV=<coverage output dir>
-
-The output directory must be an absolute path. To generate reports, you can use
-the [phpcov](https://github.com/sebastianbergmann/phpcov) tool:
-
-    phpcov merge --html=<report output dir> <coverage output dir>
-
 ### DB Creation Tests (`test/bdd/db`)
 
 These tests check the import and update of the Nominatim database. They do not

src/nominatim_api/search/db_search_builder.py

@@ -167,7 +167,12 @@ def build_housenumber_search(self, sdata: dbf.SearchData, hnrs: List[Token],
         expected_count = sum(t.count for t in hnrs)
 
         partials = {t.token: t.addr_count for trange in address
-                       for t in self.query.get_partials_list(trange)}
+                       for t in self.query.get_partials_list(trange)
+                       if t.is_indexed}
+
+        if not partials:
+            # can happen when none of the partials is indexed
+            return
 
         if expected_count < 8000:
             sdata.lookups.append(dbf.FieldLookup('nameaddress_vector',

src/nominatim_api/search/legacy_tokenizer.py

@@ -193,7 +193,7 @@ def make_token(self, row: SaRow) -> Tuple[LegacyToken, qmod.TokenType]:
                 lookup_word = row.word_token[1:]
             elif rowclass == 'place' and  row.type == 'postcode':
                 ttype = qmod.TokenType.POSTCODE
-                lookup_word = row.word_token[1:]
+                lookup_word = row.word
             else:
                 ttype = qmod.TokenType.NEAR_ITEM if row.operator in ('in', 'near')\
                         else qmod.TokenType.QUALIFIER
@@ -253,13 +253,14 @@ def rerank_tokens(self, query: qmod.QueryStruct) -> None:
 
 
 def _dump_word_tokens(query: qmod.QueryStruct) -> Iterator[List[Any]]:
-    yield ['type', 'token', 'word_token', 'lookup_word', 'penalty', 'count', 'info']
+    yield ['type', 'token', 'word_token', 'lookup_word', 'penalty', 'count', 'info', 'indexed']
     for node in query.nodes:
         for tlist in node.starting:
             for token in tlist.tokens:
                 t = cast(LegacyToken, token)
                 yield [tlist.ttype.name, t.token, t.word_token or '',
-                       t.lookup_word or '', t.penalty, t.count, t.info]
+                       t.lookup_word or '', t.penalty, t.count, t.info,
+                       'Y' if t.is_indexed else 'N']
 
 
 async def create_query_analyzer(conn: SearchConnection) -> AbstractQueryAnalyzer:

test/bdd/api/search/postcode.feature

@@ -3,11 +3,15 @@
 Feature: Searches with postcodes
     Various searches involving postcodes
 
+    @v1-api-php-only
     Scenario: US 5+4 ZIP codes are shortened to 5 ZIP codes if not found
         When sending json search query "36067 1111, us" with address
         Then result addresses contain
             | postcode |
             | 36067    |
+        And results contain
+            | type     |
+            | postcode |
 
     Scenario: Postcode search with address
         When sending json search query "9486, mauren"

test/bdd/db/query/postcodes.feature

@@ -76,6 +76,7 @@ Feature: Querying fo postcode variants
             | AD675    |
 
 
+    @fail-legacy
     Scenario: Different postcodes with the same normalization can both be found
         Given the places
            | osm | class | type  | addr+postcode | addr+housenumber | geometry |

test/bdd/db/query/reverse.feature

@@ -0,0 +1,23 @@
+@DB
+Feature: Reverse searches
+    Test results of reverse queries
+
+    @v1-api-python-only
+    Scenario: POI in POI area
+        Given the 0.0001 grid with origin 1,1
+          | 1 |   |  |  |  |  |  |  | 2 |
+          |   | 9 |  |  |  |  |  |  |   |
+          | 4 |   |  |  |  |  |  |  | 3 |
+        And the places
+          | osm | class   | type       | geometry    |
+          | W1  | aeroway | terminal   | (1,2,3,4,1) |
+          | N1  | amenity | restaurant | 9           |
+        When importing
+        And sending v1/reverse at 1.0001,1.0001
+        Then results contain
+         | osm |
+         | N1  |
+        When sending v1/reverse at 1.0003,1.0001
+        Then results contain
+         | osm |
+         | W1  |

test/bdd/db/query/search_simple.feature

@@ -77,6 +77,7 @@ Feature: Searching of simple objects
          | W1  |
 
 
+     @fail-legacy
      Scenario Outline: Special cased american states will be found
         Given the grid
          | 1 |    | 2 |

test/bdd/db/update/linked_places.feature

@@ -157,17 +157,17 @@ Feature: Updates of linked places
          | R1  | boundary | administrative | rel  | 8     | (10,11,12,13,10) |
         And the places
          | osm | class    | type        | name+name:de |
-         | N3  | place    | city        | pnt          |
+         | N3  | place    | city        | greeny       |
         And the relations
          | id | members  |
          | 1  | N3:label |
         When importing
         Then placex contains
          | object | linked_place_id | name+_place_name:de |
-         | R1     | -               | pnt  |
+         | R1     | -               | greeny  |
         And placex contains
          | object | linked_place_id | name+name:de |
-         | N3     | R1              | pnt  |
+         | N3     | R1              | greeny  |
         When updating places
          | osm | class    | type        | name+name:de |
          | N3  | place    | city        | newname      |
@@ -188,18 +188,18 @@ Feature: Updates of linked places
          | R1  | boundary | administrative | rel  | 8     | (10,11,12,13,10) |
         And the places
          | osm | class    | type           | name |
-         | N3  | place    | city           | pnt  |
+         | N3  | place    | city           | greeny  |
         And the relations
          | id | members  |
          | 1  | N3:label |
         When importing
         Then placex contains
          | object | linked_place_id | name+_place_name | name+name |
-         | R1     | -               | pnt              | rel       |
+         | R1     | -               | greeny              | rel       |
         And placex contains
          | object | linked_place_id | name+name |
-         | N3     | R1              | pnt  |
-        When sending search query "pnt"
+         | N3     | R1              | greeny  |
+        When sending search query "greeny"
         Then results contain
           | osm |
           | R1  |
@@ -212,7 +212,7 @@ Feature: Updates of linked places
         And placex contains
          | object | linked_place_id | name+_place_name:de | name+name |
          | R1     | -               | depnt               | rel       |
-        When sending search query "pnt"
+        When sending search query "greeny"
         Then exactly 0 results are returned
 
     Scenario: Updating linkee extratags keeps linker's extratags

test/bdd/environment.py

@@ -5,16 +5,18 @@
 # Copyright (C) 2024 by the Nominatim developer community.
 # For a full list of authors see the git log.
 from pathlib import Path
+import sys
 
 from behave import *
 
+sys.path.insert(1, str(Path(__file__, '..', '..', '..', 'src').resolve()))
+
 from steps.geometry_factory import GeometryFactory
 from steps.nominatim_environment import NominatimEnvironment
 
-TEST_BASE_DIR = Path(__file__) / '..' / '..'
+TEST_BASE_DIR = Path(__file__, '..', '..').resolve()
 
 userconfig = {
-    'BUILDDIR' : (TEST_BASE_DIR / '..' / 'build').resolve(),
     'REMOVE_TEMPLATE' : False,
     'KEEP_TEST_DB' : False,
     'DB_HOST' : None,
@@ -24,12 +26,11 @@
     'TEMPLATE_DB' : 'test_template_nominatim',
     'TEST_DB' : 'test_nominatim',
     'API_TEST_DB' : 'test_api_nominatim',
-    'API_TEST_FILE'  : (TEST_BASE_DIR / 'testdb' / 'apidb-test-data.pbf').resolve(),
+    'API_TEST_FILE'  : TEST_BASE_DIR / 'testdb' / 'apidb-test-data.pbf',
     'SERVER_MODULE_PATH' : None,
     'TOKENIZER' : None, # Test with a custom tokenizer
     'STYLE' : 'extratags',
-    'API_ENGINE': 'falcon',
-    'PHPCOV' : False, # set to output directory to enable code coverage
+    'API_ENGINE': 'falcon'
 }
 
 use_step_matcher("re")