]> git.openstreetmap.org Git - nominatim.git/commitdiff
Merge pull request #2145 from lonvia/cli-query-functions
authorSarah Hoffmann <lonvia@denofr.de>
Wed, 20 Jan 2021 08:00:45 +0000 (09:00 +0100)
committerGitHub <noreply@github.com>
Wed, 20 Jan 2021 08:00:45 +0000 (09:00 +0100)
Add interface to search via command line tool

CMakeLists.txt
cmake/tool.tmpl
nominatim/admin/exec_utils.py [deleted file]
nominatim/cli.py
nominatim/tools/__init__.py [moved from nominatim/admin/__init__.py with 100% similarity]
nominatim/tools/exec_utils.py [new file with mode: 0644]

index 3a9ceed91a66e7e56600d01eef3e23a95395ab29..c6f63a8b4366e146570ad2b34d33e486087886b4 100644 (file)
@@ -86,8 +86,19 @@ if (BUILD_API OR BUILD_IMPORTER)
     # sanity check if PHP binary exists
     if (NOT EXISTS ${PHP_BIN})
         message(FATAL_ERROR "PHP binary not found. Install php or provide location with -DPHP_BIN=/path/php ")
+    else()
+        message (STATUS "Using PHP binary " ${PHP_BIN})
+    endif()
+    if (NOT PHPCGI_BIN)
+        find_program (PHPCGI_BIN php-cgi)
+    endif()
+    # sanity check if PHP binary exists
+    if (NOT EXISTS ${PHPCGI_BIN})
+        message(WARNING "php-cgi binary not found. nominatim tool will not provide query functions.")
+        set (PHPCGI_BIN "")
+    else()
+        message (STATUS "Using php-cgi binary " ${PHPCGI_BIN})
     endif()
-    message (STATUS "Using PHP binary " ${PHP_BIN})
 endif()
 
 #-----------------------------------------------------------------------------
index 43646792d19d8824753a331f39b3111fa1d6f357..3bba72c3a585e81d2b44539e64f4a6ff47bf3eab 100755 (executable)
@@ -11,4 +11,5 @@ from nominatim import cli
 exit(cli.nominatim(module_dir='@CMAKE_BINARY_DIR@/module',
                    osm2pgsql_path='@CMAKE_BINARY_DIR@/osm2pgsql/osm2pgsql',
                    phplib_dir='@CMAKE_SOURCE_DIR@/lib',
-                   data_dir='@CMAKE_SOURCE_DIR@'))
+                   data_dir='@CMAKE_SOURCE_DIR@',
+                   phpcgi_path='@PHPCGI_BIN@'))
diff --git a/nominatim/admin/exec_utils.py b/nominatim/admin/exec_utils.py
deleted file mode 100644 (file)
index 0158227..0000000
+++ /dev/null
@@ -1,27 +0,0 @@
-"""
-Helper functions for executing external programs.
-"""
-import subprocess
-
-def run_legacy_script(script, *args, nominatim_env=None, throw_on_fail=False):
-    """ Run a Nominatim PHP script with the given arguments.
-
-        Returns the exit code of the script. If `throw_on_fail` is True
-        then throw a `CalledProcessError` on a non-zero exit.
-    """
-    cmd = ['/usr/bin/env', 'php', '-Cq',
-           nominatim_env.phplib_dir / 'admin' / script]
-    cmd.extend([str(a) for a in args])
-
-    env = nominatim_env.config.get_os_env()
-    env['NOMINATIM_DATADIR'] = str(nominatim_env.data_dir)
-    env['NOMINATIM_BINDIR'] = str(nominatim_env.data_dir / 'utils')
-    if not env['NOMINATIM_DATABASE_MODULE_PATH']:
-        env['NOMINATIM_DATABASE_MODULE_PATH'] = nominatim_env.module_dir
-    if not env['NOMINATIM_OSM2PGSQL_BINARY']:
-        env['NOMINATIM_OSM2PGSQL_BINARY'] = nominatim_env.osm2pgsql_path
-
-    proc = subprocess.run(cmd, cwd=str(nominatim_env.project_dir), env=env,
-                          check=throw_on_fail)
-
-    return proc.returncode
index 65ea90bbc3534b4421558e883ecaf31796e5b0b5..34b19d7da41205142a90483b2a3bb86d5046f7be 100644 (file)
@@ -9,7 +9,7 @@ import logging
 from pathlib import Path
 
 from .config import Configuration
-from .admin.exec_utils import run_legacy_script
+from .tools.exec_utils import run_legacy_script, run_api_script
 
 from .indexer.indexer import Indexer
 
@@ -73,7 +73,7 @@ class CommandlineParser:
         if args.subcommand is None:
             return self.parser.print_help()
 
-        for arg in ('module_dir', 'osm2pgsql_path', 'phplib_dir', 'data_dir'):
+        for arg in ('module_dir', 'osm2pgsql_path', 'phplib_dir', 'data_dir', 'phpcgi_path'):
             setattr(args, arg, Path(kwargs[arg]))
         args.project_dir = Path(args.project_dir)
 
@@ -435,7 +435,7 @@ class AdminWarm:
 
 class QueryExport:
     """\
-    Export addresses as CSV file from a Nominatim database.
+    Export addresses as CSV file from the database.
     """
 
     @staticmethod
@@ -488,17 +488,269 @@ class QueryExport:
 
         return run_legacy_script(*params, nominatim_env=args)
 
-class QueryTodo:
+STRUCTURED_QUERY = (
+    ('street', 'housenumber and street'),
+    ('city', 'city, town or village'),
+    ('county', 'county'),
+    ('state', 'state'),
+    ('country', 'country'),
+    ('postcalcode', 'postcode')
+)
+
+EXTRADATA_PARAMS = (
+    ('addressdetails', 'Include a breakdown of the address into elements.'),
+    ('extratags', """Include additional information if available
+                     (e.g. wikipedia link, opening hours)."""),
+    ('namedetails', 'Include a list of alternative names.')
+)
+
+DETAILS_SWITCHES = (
+    ('addressdetails', 'Include a breakdown of the address into elements.'),
+    ('keywords', 'Include a list of name keywords and address keywords.'),
+    ('linkedplaces', 'Include a details of places that are linked with this one.'),
+    ('hierarchy', 'Include details of places lower in the address hierarchy.'),
+    ('group_hierarchy', 'Group the places by type.'),
+    ('polygon_geojson', 'Include geometry of result.')
+)
+
+class APISearch:
     """\
-    Todo
+    Execute API search query.
     """
+
     @staticmethod
     def add_args(parser):
-        pass
+        group = parser.add_argument_group('Query arguments')
+        group.add_argument('--query',
+                           help='Free-form query string')
+        for name, desc in STRUCTURED_QUERY:
+            group.add_argument('--' + name, help='Structured query: ' + desc)
+
+        group = parser.add_argument_group('Output arguments')
+        group.add_argument('--format', default='jsonv2',
+                           choices=['xml', 'json', 'jsonv2', 'geojson', 'geocodejson'],
+                           help='Format of result')
+        for name, desc in EXTRADATA_PARAMS:
+            group.add_argument('--' + name, action='store_true', help=desc)
+
+        group.add_argument('--lang', '--accept-language', metavar='LANGS',
+                           help='Preferred language order for presenting search results')
+        group.add_argument('--polygon-output',
+                           choices=['geojson', 'kml', 'svg', 'text'],
+                           help='Output geometry of results as a GeoJSON, KML, SVG or WKT.')
+        group.add_argument('--polygon-threshold', type=float, metavar='TOLERANCE',
+                           help="""Simplify output geometry.
+                                   Parameter is difference tolerance in degrees.""")
+
+        group = parser.add_argument_group('Result limitation')
+        group.add_argument('--countrycodes', metavar='CC,..',
+                           help='Limit search results to one or more countries.')
+        group.add_argument('--exclude_place_ids', metavar='ID,..',
+                           help='List of search object to be excluded')
+        group.add_argument('--limit', type=int,
+                           help='Limit the number of returned results')
+        group.add_argument('--viewbox', metavar='X1,Y1,X2,Y2',
+                           help='Preferred area to find search results')
+        group.add_argument('--bounded', action='store_true',
+                           help='Strictly restrict results to viewbox area')
+
+        group = parser.add_argument_group('Other arguments')
+        group.add_argument('--no-dedupe', action='store_false', dest='dedupe',
+                           help='Do not remove duplicates from the result list')
+
 
     @staticmethod
-    def run(args): # pylint: disable=W0613
-        print("TODO: searching")
+    def run(args):
+        if args.query:
+            params = dict(q=args.query)
+        else:
+            params = {k : getattr(args, k) for k, _ in STRUCTURED_QUERY if getattr(args, k)}
+
+        for param, _ in EXTRADATA_PARAMS:
+            if getattr(args, param):
+                params[param] = '1'
+        for param in ('format', 'countrycodes', 'exclude_place_ids', 'limit', 'viewbox'):
+            if getattr(args, param):
+                params[param] = getattr(args, param)
+        if args.lang:
+            params['accept-language'] = args.lang
+        if args.polygon_output:
+            params['polygon_' + args.polygon_output] = '1'
+        if args.polygon_threshold:
+            params['polygon_threshold'] = args.polygon_threshold
+        if args.bounded:
+            params['bounded'] = '1'
+        if not args.dedupe:
+            params['dedupe'] = '0'
+
+        return run_api_script('search', args.project_dir,
+                              phpcgi_bin=args.phpcgi_path, params=params)
+
+class APIReverse:
+    """\
+    Execute API reverse query.
+    """
+
+    @staticmethod
+    def add_args(parser):
+        group = parser.add_argument_group('Query arguments')
+        group.add_argument('--lat', type=float, required=True,
+                           help='Latitude of coordinate to look up (in WGS84)')
+        group.add_argument('--lon', type=float, required=True,
+                           help='Longitude of coordinate to look up (in WGS84)')
+        group.add_argument('--zoom', type=int,
+                           help='Level of detail required for the address')
+
+        group = parser.add_argument_group('Output arguments')
+        group.add_argument('--format', default='jsonv2',
+                           choices=['xml', 'json', 'jsonv2', 'geojson', 'geocodejson'],
+                           help='Format of result')
+        for name, desc in EXTRADATA_PARAMS:
+            group.add_argument('--' + name, action='store_true', help=desc)
+
+        group.add_argument('--lang', '--accept-language', metavar='LANGS',
+                           help='Preferred language order for presenting search results')
+        group.add_argument('--polygon-output',
+                           choices=['geojson', 'kml', 'svg', 'text'],
+                           help='Output geometry of results as a GeoJSON, KML, SVG or WKT.')
+        group.add_argument('--polygon-threshold', type=float, metavar='TOLERANCE',
+                           help="""Simplify output geometry.
+                                   Parameter is difference tolerance in degrees.""")
+
+
+    @staticmethod
+    def run(args):
+        params = dict(lat=args.lat, lon=args.lon)
+        if args.zoom is not None:
+            params['zoom'] = args.zoom
+
+        for param, _ in EXTRADATA_PARAMS:
+            if getattr(args, param):
+                params[param] = '1'
+        if args.format:
+            params['format'] = args.format
+        if args.lang:
+            params['accept-language'] = args.lang
+        if args.polygon_output:
+            params['polygon_' + args.polygon_output] = '1'
+        if args.polygon_threshold:
+            params['polygon_threshold'] = args.polygon_threshold
+
+        return run_api_script('reverse', args.project_dir,
+                              phpcgi_bin=args.phpcgi_path, params=params)
+
+
+class APILookup:
+    """\
+    Execute API reverse query.
+    """
+
+    @staticmethod
+    def add_args(parser):
+        group = parser.add_argument_group('Query arguments')
+        group.add_argument('--id', metavar='OSMID',
+                           action='append', required=True, dest='ids',
+                           help='OSM id to lookup in format <NRW><id> (may be repeated)')
+
+        group = parser.add_argument_group('Output arguments')
+        group.add_argument('--format', default='jsonv2',
+                           choices=['xml', 'json', 'jsonv2', 'geojson', 'geocodejson'],
+                           help='Format of result')
+        for name, desc in EXTRADATA_PARAMS:
+            group.add_argument('--' + name, action='store_true', help=desc)
+
+        group.add_argument('--lang', '--accept-language', metavar='LANGS',
+                           help='Preferred language order for presenting search results')
+        group.add_argument('--polygon-output',
+                           choices=['geojson', 'kml', 'svg', 'text'],
+                           help='Output geometry of results as a GeoJSON, KML, SVG or WKT.')
+        group.add_argument('--polygon-threshold', type=float, metavar='TOLERANCE',
+                           help="""Simplify output geometry.
+                                   Parameter is difference tolerance in degrees.""")
+
+
+    @staticmethod
+    def run(args):
+        params = dict(osm_ids=','.join(args.ids))
+
+        for param, _ in EXTRADATA_PARAMS:
+            if getattr(args, param):
+                params[param] = '1'
+        if args.format:
+            params['format'] = args.format
+        if args.lang:
+            params['accept-language'] = args.lang
+        if args.polygon_output:
+            params['polygon_' + args.polygon_output] = '1'
+        if args.polygon_threshold:
+            params['polygon_threshold'] = args.polygon_threshold
+
+        return run_api_script('lookup', args.project_dir,
+                              phpcgi_bin=args.phpcgi_path, params=params)
+
+
+class APIDetails:
+    """\
+    Execute API lookup query.
+    """
+
+    @staticmethod
+    def add_args(parser):
+        group = parser.add_argument_group('Query arguments')
+        objs = group.add_mutually_exclusive_group(required=True)
+        objs.add_argument('--node', '-n', type=int,
+                          help="Look up the OSM node with the given ID.")
+        objs.add_argument('--way', '-w', type=int,
+                          help="Look up the OSM way with the given ID.")
+        objs.add_argument('--relation', '-r', type=int,
+                          help="Look up the OSM relation with the given ID.")
+        objs.add_argument('--place_id', '-p', type=int,
+                          help='Database internal identifier of the OSM object to look up.')
+        group.add_argument('--class', dest='object_class',
+                           help="""Class type to disambiguated multiple entries
+                                   of the same object.""")
+
+        group = parser.add_argument_group('Output arguments')
+        for name, desc in DETAILS_SWITCHES:
+            group.add_argument('--' + name, action='store_true', help=desc)
+        group.add_argument('--lang', '--accept-language', metavar='LANGS',
+                           help='Preferred language order for presenting search results')
+
+    @staticmethod
+    def run(args):
+        if args.node:
+            params = dict(osmtype='N', osmid=args.node)
+        elif args.way:
+            params = dict(osmtype='W', osmid=args.node)
+        elif args.relation:
+            params = dict(osmtype='R', osmid=args.node)
+        else:
+            params = dict(place_id=args.place_id)
+        if args.object_class:
+            params['class'] = args.object_class
+        for name, _ in DETAILS_SWITCHES:
+            params[name] = '1' if getattr(args, name) else '0'
+
+        return run_api_script('details', args.project_dir,
+                              phpcgi_bin=args.phpcgi_path, params=params)
+
+
+class APIStatus:
+    """\
+    Execute API status query.
+    """
+
+    @staticmethod
+    def add_args(parser):
+        group = parser.add_argument_group('API parameters')
+        group.add_argument('--format', default='text', choices=['text', 'json'],
+                           help='Format of result')
+
+    @staticmethod
+    def run(args):
+        return run_api_script('status', args.project_dir,
+                              phpcgi_bin=args.phpcgi_path,
+                              params=dict(format=args.format))
 
 
 def nominatim(**kwargs):
@@ -522,10 +774,14 @@ def nominatim(**kwargs):
     parser.add_subcommand('refresh', UpdateRefresh)
 
     parser.add_subcommand('export', QueryExport)
-    parser.add_subcommand('search', QueryTodo)
-    parser.add_subcommand('reverse', QueryTodo)
-    parser.add_subcommand('lookup', QueryTodo)
-    parser.add_subcommand('details', QueryTodo)
-    parser.add_subcommand('status', QueryTodo)
+
+    if kwargs.get('phpcgi_path'):
+        parser.add_subcommand('search', APISearch)
+        parser.add_subcommand('reverse', APIReverse)
+        parser.add_subcommand('lookup', APILookup)
+        parser.add_subcommand('details', APIDetails)
+        parser.add_subcommand('status', APIStatus)
+    else:
+        parser.parser.epilog = 'php-cgi not found. Query commands not available.'
 
     return parser.run(**kwargs)
diff --git a/nominatim/tools/exec_utils.py b/nominatim/tools/exec_utils.py
new file mode 100644 (file)
index 0000000..ca30b2f
--- /dev/null
@@ -0,0 +1,79 @@
+"""
+Helper functions for executing external programs.
+"""
+import logging
+import subprocess
+from urllib.parse import urlencode
+
+def run_legacy_script(script, *args, nominatim_env=None, throw_on_fail=False):
+    """ Run a Nominatim PHP script with the given arguments.
+
+        Returns the exit code of the script. If `throw_on_fail` is True
+        then throw a `CalledProcessError` on a non-zero exit.
+    """
+    cmd = ['/usr/bin/env', 'php', '-Cq',
+           nominatim_env.phplib_dir / 'admin' / script]
+    cmd.extend([str(a) for a in args])
+
+    env = nominatim_env.config.get_os_env()
+    env['NOMINATIM_DATADIR'] = str(nominatim_env.data_dir)
+    env['NOMINATIM_BINDIR'] = str(nominatim_env.data_dir / 'utils')
+    if not env['NOMINATIM_DATABASE_MODULE_PATH']:
+        env['NOMINATIM_DATABASE_MODULE_PATH'] = nominatim_env.module_dir
+    if not env['NOMINATIM_OSM2PGSQL_BINARY']:
+        env['NOMINATIM_OSM2PGSQL_BINARY'] = nominatim_env.osm2pgsql_path
+
+    proc = subprocess.run(cmd, cwd=str(nominatim_env.project_dir), env=env,
+                          check=throw_on_fail)
+
+    return proc.returncode
+
+def run_api_script(endpoint, project_dir, extra_env=None, phpcgi_bin=None,
+                   params=None):
+    """ Execute a Nominiatim API function.
+
+        The function needs a project directory that contains the website
+        directory with the scripts to be executed. The scripts will be run
+        using php_cgi. Query parameters can be addd as named arguments.
+
+        Returns the exit code of the script.
+    """
+    log = logging.getLogger()
+    webdir = str(project_dir / 'website')
+    query_string = urlencode(params or {})
+
+    env = dict(QUERY_STRING=query_string,
+               SCRIPT_NAME='/{}.php'.format(endpoint),
+               REQUEST_URI='/{}.php?{}'.format(endpoint, query_string),
+               CONTEXT_DOCUMENT_ROOT=webdir,
+               SCRIPT_FILENAME='{}/{}.php'.format(webdir, endpoint),
+               HTTP_HOST='localhost',
+               HTTP_USER_AGENT='nominatim-tool',
+               REMOTE_ADDR='0.0.0.0',
+               DOCUMENT_ROOT=webdir,
+               REQUEST_METHOD='GET',
+               SERVER_PROTOCOL='HTTP/1.1',
+               GATEWAY_INTERFACE='CGI/1.1',
+               REDIRECT_STATUS='CGI')
+
+    if extra_env:
+        env.update(extra_env)
+
+    if phpcgi_bin is None:
+        cmd = ['/usr/bin/env', 'php-cgi']
+    else:
+        cmd = [str(phpcgi_bin)]
+
+    proc = subprocess.run(cmd, cwd=str(project_dir), env=env, capture_output=True,
+                          check=False)
+
+    if proc.returncode != 0 or proc.stderr:
+        log.error(proc.stderr.decode('utf-8').replace('\\n', '\n'))
+        return proc.returncode or 1
+
+    result = proc.stdout.decode('utf-8')
+    content_start = result.find('\r\n\r\n')
+
+    print(result[content_start + 4:].replace('\\n', '\n'))
+
+    return 0