[enh] unit tests for wolframalpha

2 files changed, 61 insertions, 47 deletions

diff --git a/searx/engines/wolframalpha_api.py b/searx/engines/wolframalpha_api.py
index ad80fb0f4..9a13d7422 100644
--- a/searx/engines/wolframalpha_api.py
+++ b/searx/engines/wolframalpha_api.py
@@ -19,9 +19,10 @@ api_key = ''  # defined in settings.yml
 # xpath variables
 failure_xpath = '/queryresult[attribute::success="false"]'
 answer_xpath = '//pod[attribute::primary="true"]/subpod/plaintext'
-input_xpath = '//pod[starts-with(attribute::title, "Input")]/subpod/plaintext'
+input_xpath = '//pod[starts-with(attribute::id, "Input")]/subpod/plaintext'
 pods_xpath = '//pod'
 subpods_xpath = './subpod'
+pod_id_xpath = './@id'
 pod_title_xpath = './@title'
 plaintext_xpath = './plaintext'
 image_xpath = './img'
@@ -30,8 +31,8 @@ img_alt_xpath = './@alt'
 
 # pods to display as image in infobox
 # this pods do return a plaintext, but they look better and are more useful as images
-image_pods = {'Visual representation',
-              'Manipulatives illustration'}
+image_pods = {'VisualRepresentation',
+              'Illustration'}
 
 
 # do search-request
@@ -45,15 +46,15 @@ def request(query, params):
 
 # replace private user area characters to make text legible
 def replace_pua_chars(text):
-    pua_chars = {u'\uf522': u'\u2192',
-                 u'\uf7b1': u'\u2115',
-                 u'\uf7b4': u'\u211a',
-                 u'\uf7b5': u'\u211d',
-                 u'\uf7bd': u'\u2124',
-                 u'\uf74c': 'd',
-                 u'\uf74d': u'\u212f',
-                 u'\uf74e': 'i',
-                 u'\uf7d9': '='}
+    pua_chars = {u'\uf522': u'\u2192',  # rigth arrow
+                 u'\uf7b1': u'\u2115',  # set of natural numbers
+                 u'\uf7b4': u'\u211a',  # set of rational numbers
+                 u'\uf7b5': u'\u211d',  # set of real numbers
+                 u'\uf7bd': u'\u2124',  # set of integer numbers
+                 u'\uf74c': 'd',        # differential
+                 u'\uf74d': u'\u212f',  # euler's number
+                 u'\uf74e': 'i',        # imaginary number
+                 u'\uf7d9': '='}        # equals sign
 
     for k, v in pua_chars.iteritems():
         text = text.replace(k, v)
@@ -71,30 +72,35 @@ def response(resp):
     if search_results.xpath(failure_xpath):
         return []
 
-    infobox_title = search_results.xpath(input_xpath)
-    if infobox_title:
-        infobox_title = replace_pua_chars(infobox_title[0].text)
+    try:
+        infobox_title = search_results.xpath(input_xpath)[0].text
+    except:
+        infobox_title = None
 
     pods = search_results.xpath(pods_xpath)
     result_chunks = []
     for pod in pods:
-        pod_title = replace_pua_chars(pod.xpath(pod_title_xpath)[0])
+        pod_id = pod.xpath(pod_id_xpath)[0]
+        pod_title = pod.xpath(pod_title_xpath)[0]
 
         subpods = pod.xpath(subpods_xpath)
         if not subpods:
             continue
 
+        # Appends either a text or an image, depending on which one is more suitable
         for subpod in subpods:
             content = subpod.xpath(plaintext_xpath)[0].text
             image = subpod.xpath(image_xpath)
-            if content and pod_title not in image_pods:
-                content = replace_pua_chars(content)
-                result_chunks.append({'label': pod_title, 'value': content})
 
-                # if there's no input pod, infobox_title is content of first pod
+            if content and pod_id not in image_pods:
+
+                # if no input pod was found, title is first plaintext pod
                 if not infobox_title:
                     infobox_title = content
 
+                content = replace_pua_chars(content)
+                result_chunks.append({'label': pod_title, 'value': content})
+
             elif image:
                 result_chunks.append({'label': pod_title,
                                       'image': {'src': image[0].xpath(img_src_xpath)[0],
@@ -103,6 +109,7 @@ def response(resp):
     if not result_chunks:
         return []
 
+    # append infobox
     results.append({'infobox': infobox_title,
                     'attributes': result_chunks,
                     'urls': [{'title': 'Wolfram|Alpha', 'url': resp.request.headers['Referer']}]})
diff --git a/searx/engines/wolframalpha_noapi.py b/searx/engines/wolframalpha_noapi.py
index a5fbdeae7..d10716a2a 100644
--- a/searx/engines/wolframalpha_noapi.py
+++ b/searx/engines/wolframalpha_noapi.py
@@ -48,8 +48,8 @@ img_alt_xpath = './img/@alt'
 
 # pods to display as image in infobox
 # this pods do return a plaintext, but they look better and are more useful as images
-image_pods = {'Visual representation',
-              'Manipulatives illustration',
+image_pods = {'VisualRepresentation',
+              'Illustration',
               'Symbol'}
 
 
@@ -82,26 +82,35 @@ def request(query, params):
 # get additional pod
 # NOTE: this makes an additional requests to server, so the response will take longer and might reach timeout
 def get_async_pod(url):
-    pod = {'subpods': []}
-
     try:
         resp = http_get(url, timeout=2.0)
-
-        resp_pod = XML(resp.content)
-        if resp_pod.xpath(success_xpath):
-
-            for subpod in resp_pod:
-                plaintext = subpod.xpath(plaintext_xpath)[0].text
-                if plaintext:
-                    pod['subpods'].append({'title': subpod.xpath(title_xpath)[0],
-                                           'plaintext': plaintext})
-                elif subpod.xpath(image_xpath):
-                    pod['subpods'].append({'title': subpod.xpath(title_xpath)[0],
-                                           'plaintext': '',
-                                           'img': {'src': subpod.xpath(img_src_xpath)[0],
-                                                   'alt': subpod.xpath(img_alt_xpath)[0]}})
     except:
-        pass
+        return None
+
+    if resp:
+        return parse_async_pod(resp)
+
+
+def parse_async_pod(resp):
+    pod = {'subpods': []}
+
+    resp_pod = XML(resp.content)
+
+    if resp_pod.xpath(success_xpath):
+        for subpod in resp_pod:
+            new_subpod = {'title': subpod.xpath(title_xpath)[0]}
+
+            plaintext = subpod.xpath(plaintext_xpath)[0].text
+            if plaintext:
+                new_subpod['plaintext'] = plaintext
+            else:
+                new_subpod['plaintext'] = ''
+
+            if subpod.xpath(image_xpath):
+                new_subpod['img'] = {'src': subpod.xpath(img_src_xpath)[0],
+                                     'alt': subpod.xpath(img_alt_xpath)[0]}
+
+            pod['subpods'].append(new_subpod)
 
     return pod
 
@@ -119,6 +128,7 @@ def response(resp):
     result_chunks = []
     infobox_title = None
     for pod in resp_json['queryresult']['pods']:
+        pod_id = pod.get('id', '')
         pod_title = pod.get('title', '')
 
         if 'subpods' not in pod:
@@ -127,19 +137,16 @@ def response(resp):
                 result = get_async_pod(pod['async'])
                 if result:
                     pod = result
+                else:
+                    continue
             else:
                 continue
 
-        # infobox title is input or text content on first pod
-        if pod_title.startswith('Input') or not infobox_title:
-            try:
-                infobox_title = pod['subpods'][0]['plaintext']
-            except:
-                infobox_title = ''
-                pass
+        if pod_id == 'Input' or not infobox_title:
+            infobox_title = pod['subpods'][0]['plaintext']
 
         for subpod in pod['subpods']:
-            if subpod['plaintext'] != '' and pod_title not in image_pods:
+            if subpod['plaintext'] != '' and pod_id not in image_pods:
                 # append unless it's not an actual answer
                 if subpod['plaintext'] != '(requires interactivity)':
                     result_chunks.append({'label': pod_title, 'value': subpod['plaintext']})