Merge remote-tracking branch 'upstream/master'

[nominatim.git] / test / bdd / steps / queries.py

diff --git a/test/bdd/steps/queries.py b/test/bdd/steps/queries.py
index d0cda77469bab1559260e3af3b6f3529ffdfa775..55285d6a6bac83c52fc6559c4e227dae5323b2bb 100644 (file)
--- a/test/bdd/steps/queries.py
+++ b/test/bdd/steps/queries.py
@@ -8,6 +8,7 @@ import json
 import os
 import io
 import re
 import os
 import io
 import re

+import logging

 from tidylib import tidy_document
 import xml.etree.ElementTree as ET
 import subprocess
 from tidylib import tidy_document
 import xml.etree.ElementTree as ET
 import subprocess


@@ -15,11 +16,12 @@ from urllib.parse import urlencode
 from collections import OrderedDict
 from nose.tools import * # for assert functions
 
 from collections import OrderedDict
 from nose.tools import * # for assert functions
 

+logger = logging.getLogger(__name__)
+

 BASE_SERVER_ENV = {
     'HTTP_HOST' : 'localhost',
     'HTTP_USER_AGENT' : 'Mozilla/5.0 (X11; Linux x86_64; rv:51.0) Gecko/20100101 Firefox/51.0',
     'HTTP_ACCEPT' : 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
 BASE_SERVER_ENV = {
     'HTTP_HOST' : 'localhost',
     'HTTP_USER_AGENT' : 'Mozilla/5.0 (X11; Linux x86_64; rv:51.0) Gecko/20100101 Firefox/51.0',
     'HTTP_ACCEPT' : 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',

-    'HTTP_ACCEPT_LANGUAGE' : 'en,de;q=0.5',

     'HTTP_ACCEPT_ENCODING' : 'gzip, deflate',
     'HTTP_CONNECTION' : 'keep-alive',
     'SERVER_SIGNATURE' : '<address>Nominatim BDD Tests</address>',
     'HTTP_ACCEPT_ENCODING' : 'gzip, deflate',
     'HTTP_CONNECTION' : 'keep-alive',
     'SERVER_SIGNATURE' : '<address>Nominatim BDD Tests</address>',


@@ -54,8 +56,40 @@ def compare(operator, op1, op2):
     else:
         raise Exception("unknown operator '%s'" % operator)
 
     else:
         raise Exception("unknown operator '%s'" % operator)
 

+class GenericResponse(object):
+
+    def match_row(self, row):
+        if 'ID' in row.headings:
+            todo = [int(row['ID'])]
+        else:
+            todo = range(len(self.result))
+
+        for i in todo:
+            res = self.result[i]
+            for h in row.headings:
+                if h == 'ID':
+                    pass
+                elif h == 'osm':
+                    assert_equal(res['osm_type'], row[h][0])
+                    assert_equal(res['osm_id'], row[h][1:])
+                elif h == 'centroid':
+                    x, y = row[h].split(' ')
+                    assert_almost_equal(float(y), float(res['lat']))
+                    assert_almost_equal(float(x), float(res['lon']))
+                elif row[h].startswith("^"):
+                    assert_in(h, res)
+                    assert_is_not_none(re.fullmatch(row[h], res[h]),
+                                       "attribute '%s': expected: '%s', got '%s'"
+                                          % (h, row[h], res[h]))
+                else:
+                    assert_in(h, res)
+                    assert_equal(str(res[h]), str(row[h]))
+
+    def property_list(self, prop):
+        return [ x[prop] for x in self.result ]
+

 
 

-class SearchResponse(object):
+class SearchResponse(GenericResponse):

 
     def __init__(self, page, fmt='json', errorcode=200):
         self.page = page
 
     def __init__(self, page, fmt='json', errorcode=200):
         self.page = page


@@ -117,36 +151,86 @@ class SearchResponse(object):
                 self.result[-1]['address'] = address
 
 
                 self.result[-1]['address'] = address
 
 

-    def match_row(self, row):
-        if 'ID' in row.headings:
-            todo = [int(row['ID'])]
+class ReverseResponse(GenericResponse):
+
+    def __init__(self, page, fmt='json', errorcode=200):
+        self.page = page
+        self.format = fmt
+        self.errorcode = errorcode
+        self.result = []
+        self.header = dict()
+
+        if errorcode == 200:
+            getattr(self, 'parse_' + fmt)()
+
+    def parse_html(self):
+        content, errors = tidy_document(self.page,
+                                        options={'char-encoding' : 'utf8'})
+        #eq_(len(errors), 0 , "Errors found in HTML document:\n%s" % errors)
+
+        b = content.find('nominatim_results =')
+        e = content.find('</script>')
+        content = content[b:e]
+        b = content.find('[')
+        e = content.rfind(']')
+
+        self.result = json.JSONDecoder(object_pairs_hook=OrderedDict).decode(content[b:e+1])
+
+    def parse_json(self):
+        m = re.fullmatch(r'([\w$][^(]*)\((.*)\)', self.page)
+        if m is None:
+            code = self.page

         else:
         else:

-            todo = range(len(self.result))
+            code = m.group(2)
+            self.header['json_func'] = m.group(1)
+        self.result = [json.JSONDecoder(object_pairs_hook=OrderedDict).decode(code)]

 
 

-        for i in todo:
-            res = self.result[i]
-            for h in row.headings:
-                if h == 'ID':
-                    pass
-                elif h == 'osm':
-                    assert_equal(res['osm_type'], row[h][0])
-                    assert_equal(res['osm_id'], row[h][1:])
-                elif h == 'centroid':
-                    x, y = row[h].split(' ')
-                    assert_almost_equal(float(y), float(res['lat']))
-                    assert_almost_equal(float(x), float(res['lon']))
-                elif row[h].startswith("^"):
-                    assert_in(h, res)
-                    assert_is_not_none(re.fullmatch(row[h], res[h]),
-                                       "attribute '%s': expected: '%s', got '%s'"
-                                          % (h, row[h], res[h]))
-                else:
-                    assert_in(h, res)
-                    assert_equal(str(res[h]), str(row[h]))
+    def parse_xml(self):
+        et = ET.fromstring(self.page)

 
 

-    def property_list(self, prop):
-        return [ x[prop] for x in self.result ]
+        self.header = dict(et.attrib)
+        self.result = []
+
+        for child in et:
+            if child.tag == 'result':
+                eq_(0, len(self.result), "More than one result in reverse result")
+                self.result.append(dict(child.attrib))
+            elif child.tag == 'addressparts':
+                address = {}
+                for sub in child:
+                    address[sub.tag] = sub.text
+                self.result[0]['address'] = address
+            elif child.tag == 'extratags':
+                self.result[0]['extratags'] = {}
+                for tag in child:
+                    self.result[0]['extratags'][tag.attrib['key']] = tag.attrib['value']
+            elif child.tag == 'namedetails':
+                self.result[0]['namedetails'] = {}
+                for tag in child:
+                    self.result[0]['namedetails'][tag.attrib['desc']] = tag.text
+            elif child.tag in ('geokml'):
+                self.result[0][child.tag] = True
+            else:
+                assert child.tag == 'error', \
+                        "Unknown XML tag %s on page: %s" % (child.tag, self.page)
+
+
+class DetailsResponse(GenericResponse):

 
 

+    def __init__(self, page, fmt='json', errorcode=200):
+        self.page = page
+        self.format = fmt
+        self.errorcode = errorcode
+        self.result = []
+        self.header = dict()
+
+        if errorcode == 200:
+            getattr(self, 'parse_' + fmt)()
+
+    def parse_html(self):
+        content, errors = tidy_document(self.page,
+                                        options={'char-encoding' : 'utf8'})
+        self.result = {}

 
 @when(u'searching for "(?P<query>.*)"(?P<dups> with dups)?')
 def query_cmd(context, query, dups):
 
 @when(u'searching for "(?P<query>.*)"(?P<dups> with dups)?')
 def query_cmd(context, query, dups):


@@ -172,16 +256,9 @@ def query_cmd(context, query, dups):
 
     context.response = SearchResponse(outp.decode('utf-8'), 'json')
 
 
     context.response = SearchResponse(outp.decode('utf-8'), 'json')
 

-
-@when(u'sending (?P<fmt>\S+ )?search query "(?P<query>.*)"(?P<addr> with address)?')
-def website_search_request(context, fmt, query, addr):
-    env = BASE_SERVER_ENV
-
-    params = { 'q' : query }
+def send_api_query(endpoint, params, fmt, context):

     if fmt is not None:
         params['format'] = fmt.strip()
     if fmt is not None:
         params['format'] = fmt.strip()

-    if addr is not None:
-        params['addressdetails'] = '1'

     if context.table:
         if context.table.headings[0] == 'param':
             for line in context.table:
     if context.table:
         if context.table.headings[0] == 'param':
             for line in context.table:


@@ -189,15 +266,23 @@ def website_search_request(context, fmt, query, addr):
         else:
             for h in context.table.headings:
                 params[h] = context.table[0][h]
         else:
             for h in context.table.headings:
                 params[h] = context.table[0][h]

+
+    env = dict(BASE_SERVER_ENV)

     env['QUERY_STRING'] = urlencode(params)
 
     env['QUERY_STRING'] = urlencode(params)
 

-    env['REQUEST_URI'] = '/search.php?' + env['QUERY_STRING']
-    env['SCRIPT_NAME'] = '/search.php'
+    env['SCRIPT_NAME'] = '/%s.php' % endpoint
+    env['REQUEST_URI'] = '%s?%s' % (env['SCRIPT_NAME'], env['QUERY_STRING'])

     env['CONTEXT_DOCUMENT_ROOT'] = os.path.join(context.nominatim.build_dir, 'website')
     env['CONTEXT_DOCUMENT_ROOT'] = os.path.join(context.nominatim.build_dir, 'website')

-    env['SCRIPT_FILENAME'] = os.path.join(context.nominatim.build_dir, 'website', 'search.php')
+    env['SCRIPT_FILENAME'] = os.path.join(env['CONTEXT_DOCUMENT_ROOT'],
+                                          '%s.php' % endpoint)

     env['NOMINATIM_SETTINGS'] = context.nominatim.local_settings_file
 
     env['NOMINATIM_SETTINGS'] = context.nominatim.local_settings_file
 

-    cmd = [ '/usr/bin/php-cgi', env['SCRIPT_FILENAME']]
+    logger.debug("Environment:" + json.dumps(env, sort_keys=True, indent=2))
+
+    if hasattr(context, 'http_headers'):
+        env.update(context.http_headers)
+
+    cmd = ['/usr/bin/php-cgi', env['SCRIPT_FILENAME']]

     for k,v in params.items():
         cmd.append("%s=%s" % (k, v))
 
     for k,v in params.items():
         cmd.append("%s=%s" % (k, v))
 


@@ -205,21 +290,44 @@ def website_search_request(context, fmt, query, addr):
                             stdout=subprocess.PIPE, stderr=subprocess.PIPE)
 
     (outp, err) = proc.communicate()
                             stdout=subprocess.PIPE, stderr=subprocess.PIPE)
 
     (outp, err) = proc.communicate()

+    outp = outp.decode('utf-8')
+
+    logger.debug("Result: \n===============================\n"
+                 + outp + "\n===============================\n")

 
     assert_equals(0, proc.returncode,
                   "query.php failed with message: %s\noutput: %s" % (err, outp))
 
     assert_equals(0, len(err), "Unexpected PHP error: %s" % (err))
 
 
     assert_equals(0, proc.returncode,
                   "query.php failed with message: %s\noutput: %s" % (err, outp))
 
     assert_equals(0, len(err), "Unexpected PHP error: %s" % (err))
 

-    outp = outp.decode('utf-8')
-

     if outp.startswith('Status: '):
         status = int(outp[8:11])
     else:
         status = 200
 
     content_start = outp.find('\r\n\r\n')
     if outp.startswith('Status: '):
         status = int(outp[8:11])
     else:
         status = 200
 
     content_start = outp.find('\r\n\r\n')

-    assert_less(11, content_start)
+
+    return outp[content_start + 4:], status
+
+@given(u'the HTTP header')
+def add_http_header(context):
+    if not hasattr(context, 'http_headers'):
+        context.http_headers = {}
+
+    for h in context.table.headings:
+        envvar = 'HTTP_' + h.upper().replace('-', '_')
+        context.http_headers[envvar] = context.table[0][h]
+
+
+@when(u'sending (?P<fmt>\S+ )?search query "(?P<query>.*)"(?P<addr> with address)?')
+def website_search_request(context, fmt, query, addr):
+    params = {}
+    if query:
+        params['q'] = query
+    if addr is not None:
+        params['addressdetails'] = '1'
+
+    outp, status = send_api_query('search', params, fmt, context)

 
     if fmt is None:
         outfmt = 'html'
 
     if fmt is None:
         outfmt = 'html'


@@ -228,7 +336,50 @@ def website_search_request(context, fmt, query, addr):
     else:
         outfmt = fmt.strip()
 
     else:
         outfmt = fmt.strip()
 

-    context.response = SearchResponse(outp[content_start + 4:], outfmt, status)
+    context.response = SearchResponse(outp, outfmt, status)
+
+@when(u'sending (?P<fmt>\S+ )?reverse coordinates (?P<lat>[0-9.-]+)?,(?P<lon>[0-9.-]+)?')
+def website_reverse_request(context, fmt, lat, lon):
+    params = {}
+    if lat is not None:
+        params['lat'] = lat
+    if lon is not None:
+        params['lon'] = lon
+
+    outp, status = send_api_query('reverse', params, fmt, context)
+
+    if fmt is None:
+        outfmt = 'xml'
+    elif fmt == 'jsonv2 ':
+        outfmt = 'json'
+    else:
+        outfmt = fmt.strip()
+
+    context.response = ReverseResponse(outp, outfmt, status)
+
+@when(u'sending (?P<fmt>\S+ )?details query for (?P<query>.*)')
+def website_details_request(context, fmt, query):
+    params = {}
+    if query[0] in 'NWR':
+        params['osmtype'] = query[0]
+        params['osmid'] = query[1:]
+    else:
+        params['place_id'] = query
+    outp, status = send_api_query('details', params, fmt, context)
+
+    context.response = DetailsResponse(outp, 'html', status)
+
+@when(u'sending (?P<fmt>\S+ )?lookup query for (?P<query>.*)')
+def website_lookup_request(context, fmt, query):
+    params = { 'osm_ids' : query }
+    outp, status = send_api_query('lookup', params, fmt, context)
+
+    if fmt == 'json ':
+        outfmt = 'json'
+    else:
+        outfmt = 'xml'
+
+    context.response = SearchResponse(outp, outfmt, status)

 
 
 @step(u'(?P<operator>less than|more than|exactly|at least|at most) (?P<number>\d+) results? (?:is|are) returned')
 
 
 @step(u'(?P<operator>less than|more than|exactly|at least|at most) (?P<number>\d+) results? (?:is|are) returned')


@@ -244,6 +395,7 @@ def check_http_return_status(context, status):
 
 @then(u'the result is valid (?P<fmt>\w+)')
 def step_impl(context, fmt):
 
 @then(u'the result is valid (?P<fmt>\w+)')
 def step_impl(context, fmt):

+    context.execute_steps("Then a HTTP 200 is returned")

     eq_(context.response.format, fmt)
 
 @then(u'result header contains')
     eq_(context.response.format, fmt)
 
 @then(u'result header contains')