fix(OpenRosa): correct detection of root node tag name for XForm surv…

…eys TASK-1258 (#5272)

### 📣 Summary
Fixes the detection of the root node tag name of the survey in the XForm
XML, respecting the name field in XLSForm settings

### 📖 Description
This bugfix addresses two issues caused by wrong detection of the root
node tag name in the "survey" section of XForm XML:
- an empty `formhub/uuid` on submission creation
- a 500 error when a form is open in Enketo (preview/collection) when a
disclaimer is set

The update ensures that the name field, if provided in the XLSForm
settings, is respected as the root node's tag name.


### 👀 Preview steps

Bug template:
1. Log in as a regular user
2. Create a project (add `name` field with custom value in the
`settings` worksheet)
3. Deploy and submit data
  1. 🔴 [on main] notice that this `formhub/uuid` is empty
  2. 🟢 [on PR] notice that this is `formhub/uuid` matches XForm uuid
  
1. Go to admin and add a disclaimer
2. Try to open the project in Enketo (preview and collection)
  1. 🔴 [on main] notice that it raises a 500 error
  2. 🟢 [on PR] notice that the form opens

### 💭 Notes
This PR relies on the fact that `id_string` and `name` are always
provided in the XForm.json field (according to existing data on
production servers).
In case of preview, If the `name` field is not supplied, the fallback
mechanism uses the asset's UID to determine the tag name (as it is in
`main`).

kobo/apps/openrosa/apps/logger/models/xform.py

@@ -291,6 +291,18 @@ def url(self):
             'download_xform', kwargs={'username': self.user.username, 'pk': self.pk}
         )
 
+    @property
+    def xform_root_node_name(self):
+        """
+        Retrieves the name of the XML tag representing the root node of the "survey"
+        in the XForm XML structure.
+
+        It should always be present in `self.json`.
+        """
+
+        form_json = json.loads(self.json)
+        return form_json['name']
+
     @property
     def xml_with_disclaimer(self):
         return XMLFormWithDisclaimer(self).get_object().xml

kobo/apps/openrosa/apps/logger/tests/test_simple_submission.py

@@ -32,7 +32,7 @@ def _get_xml_for_form(self, xform):
         builder = SurveyElementBuilder()
         sss = builder.create_survey_element_from_json(xform.json)
         xform.xml = sss.to_xml()
-        xform._mark_start_time_boolean()
+        xform.mark_start_time_boolean()
         xform.save()
 
     def _submit_at_hour(self, hour):

kobo/apps/openrosa/apps/viewer/management/commands/mark_start_times.py

@@ -11,7 +11,7 @@ class Command(BaseCommand):
     def handle(self, *args, **kwargs):
         for dd in DataDictionary.objects.all():
             try:
-                dd._mark_start_time_boolean()
+                dd.mark_start_time_boolean()
                 dd.save()
             except Exception:
                 print ("Could not mark start time for DD: %(data)s" % {

kobo/apps/openrosa/apps/viewer/management/commands/set_uuid_in_xml.py

@@ -14,7 +14,7 @@ def handle(self, *args, **kwargs):
         for i, dd in enumerate(
                 queryset_iterator(DataDictionary.objects.all())):
             if dd.xls:
-                dd.set_uuid_in_xml(id_string=dd.id_string)
+                dd.set_uuid_in_xml()
                 super(DataDictionary, dd).save()
             if (i + 1) % 10 == 0:
                 print('Updated %(nb)d XForms...' % {'nb': i})

kobo/apps/openrosa/apps/viewer/models/data_dictionary.py

@@ -1,4 +1,3 @@
-# coding: utf-8
 import os
 import re
 from xml.dom import Node
@@ -65,17 +64,11 @@ def __init__(self, *args, **kwargs):
         self.instances_for_export = lambda d: d.instances.all()
         super().__init__(*args, **kwargs)
 
-    def set_uuid_in_xml(self, file_name=None, id_string=None):
+    def set_uuid_in_xml(self):
         """
         Add bind to automatically set UUID node in XML.
         """
-
-        if id_string:
-            root_node = id_string
-        else:
-            if not file_name:
-                file_name = self.file_name()
-            root_node, _ = os.path.splitext(file_name)
+        root_node = self.xform_root_node_name
 
         doc = clean_and_parse_xml(self.xml)
         model_nodes = doc.getElementsByTagName("model")
@@ -98,19 +91,21 @@ def set_uuid_in_xml(self, file_name=None, id_string=None):
         instance_node = instance_nodes[0]
 
         # get the first child whose id attribute matches our id_string
-        survey_nodes = [node for node in instance_node.childNodes
-                        if node.nodeType == Node.ELEMENT_NODE and
-                        (node.tagName == root_node or
-                         node.attributes.get('id'))]
-
+        survey_nodes = [
+            node
+            for node in instance_node.childNodes
+            if node.nodeType == Node.ELEMENT_NODE
+            and node.tagName == root_node
+        ]
         if len(survey_nodes) != 1:
-            raise Exception(
-                "Multiple survey nodes with the id '{}'".format(root_node))
+            raise Exception(f'Multiple survey nodes `{root_node}`')
 
         survey_node = survey_nodes[0]
-        formhub_nodes = [n for n in survey_node.childNodes
-                         if n.nodeType == Node.ELEMENT_NODE and
-                         n.tagName == "formhub"]
+        formhub_nodes = [
+            n
+            for n in survey_node.childNodes
+            if n.nodeType == Node.ELEMENT_NODE and n.tagName == 'formhub'
+        ]
 
         if len(formhub_nodes) > 1:
             raise Exception(
@@ -130,10 +125,9 @@ def set_uuid_in_xml(self, file_name=None, id_string=None):
         if len(formhub_nodes) == 0:
             # append the calculate bind node
             calculate_node = doc.createElement("bind")
-            calculate_node.setAttribute(
-                "nodeset", "/%s/formhub/uuid" % root_node)
-            calculate_node.setAttribute("type", "string")
-            calculate_node.setAttribute("calculate", "'%s'" % self.uuid)
+            calculate_node.setAttribute('nodeset', f'/{root_node}/formhub/uuid')
+            calculate_node.setAttribute('type', 'string')
+            calculate_node.setAttribute('calculate', f"'{self.uuid}'")
             model_node.appendChild(calculate_node)
 
         self.xml = smart_str(doc.toprettyxml(indent="  ", encoding='utf-8'))
@@ -166,9 +160,9 @@ def save(self, *args, **kwargs):
                 survey.name = survey.id_string
             self.json = survey.to_json()
             self.xml = survey.to_xml()
-            self._mark_start_time_boolean()
+            self.mark_start_time_boolean()
             set_uuid(self)
-            self.set_uuid_in_xml(id_string=survey.id_string)
+            self.set_uuid_in_xml()
         super().save(*args, **kwargs)
 
     def file_name(self):
@@ -418,7 +412,7 @@ def get_data_for_excel(self):
                 self._expand_geocodes(d, key, e)
             yield d
 
-    def _mark_start_time_boolean(self):
+    def mark_start_time_boolean(self):
         starttime_substring = 'jr:preloadParams="start"'
         if self.xml.find(starttime_substring) != -1:
             self.has_start_time = True

kobo/apps/openrosa/libs/utils/logger_tools.py

@@ -523,26 +523,16 @@ def publish_xls_form(xls_file, user, id_string=None):
         return dd
 
 
-def publish_xml_form(xml_file, user, id_string=None):
+def publish_xml_form(xml_file, user):
     xml = smart_str(xml_file.read())
     survey = create_survey_element_from_xml(xml)
     form_json = survey.to_json()
-    if id_string:
-        dd = DataDictionary.objects.get(user=user, id_string=id_string)
-        dd.xml = xml
-        dd.json = form_json
-        dd._mark_start_time_boolean()
-        set_uuid(dd)
-        dd.set_uuid_in_xml()
-        dd.save()
-        return dd
-    else:
-        dd = DataDictionary(user=user, xml=xml, json=form_json)
-        dd._mark_start_time_boolean()
-        set_uuid(dd)
-        dd.set_uuid_in_xml(file_name=xml_file.name)
-        dd.save()
-        return dd
+    dd = DataDictionary(user=user, xml=xml, json=form_json)
+    dd.mark_start_time_boolean()
+    set_uuid(dd)
+    dd.set_uuid_in_xml()
+    dd.save()
+    return dd
 
 
 def report_exception(subject, info, exc_info=None):

kpi/models/asset_snapshot.py

@@ -242,3 +242,18 @@ def generate_xml_from_source(self,
                 'warnings': warnings,
             })
         return xml, details
+
+    @property
+    def xform_root_node_name(self):
+        """
+        Retrieves the name of the XML tag representing the root node of the "survey"
+        in the XForm XML structure.
+
+        This method uses the `name` setting from the XLSForm to determine the tag name.
+        If no name is provided, it falls back to using the asset UID.
+        """
+
+        try:
+            return self.asset.content['settings']['name']
+        except KeyError:
+            return self.asset.uid

kpi/utils/xml.py

@@ -1,4 +1,3 @@
-# coding: utf-8
 from __future__ import annotations
 
 import re
@@ -373,6 +372,10 @@ class XMLFormWithDisclaimer:
     def __init__(self, obj: Union['kpi.AssetSnapshot', 'logger.XForm']):
         self._object = obj
         self._unique_id = obj.asset.uid
+
+        # Avoid accessing the `xform_root_node_name` property immediately to prevent
+        # extra database queries. It will be set only when it is actually needed.
+        self._root_tag_name = None
         self._add_disclaimer()
 
     def get_object(self):
@@ -391,6 +394,7 @@ def _add_disclaimer(self):
         translated, disclaimers_dict, default_language_code = value
 
         self._root_node = minidom.parseString(self._object.xml)
+        self._root_tag_name = self._object.xform_root_node_name
 
         if translated:
             self._add_translation_nodes(disclaimers_dict, default_language_code)
@@ -413,17 +417,17 @@ def _add_instance_and_bind_nodes(self):
         # Inject <bind nodeset /> inside <model odk:xforms-version="1.0.0">
         bind_node = self._root_node.createElement('bind')
         bind_node.setAttribute(
-            'nodeset', f'/{self._unique_id}/_{self._unique_id}__disclaimer'
+            'nodeset', f'/{self._root_tag_name}/_{self._unique_id}__disclaimer'
         )
         bind_node.setAttribute('readonly', 'true()')
         bind_node.setAttribute('required', 'false()')
         bind_node.setAttribute('type', 'string')
         bind_node.setAttribute('relevant', 'false()')
         model_node.appendChild(bind_node)
 
-        # Inject note node inside <{self._unique_id}>
+        # Inject note node inside <{self._root_tag_name}>
         instance_node = model_node.getElementsByTagName('instance')[0]
-        instance_node = instance_node.getElementsByTagName(self._unique_id)[0]
+        instance_node = instance_node.getElementsByTagName(self._root_tag_name)[0]
         instance_node.appendChild(
             self._root_node.createElement(f'_{self._unique_id}__disclaimer')
         )
@@ -442,11 +446,11 @@ def _add_disclaimer_input(
         disclaimer_input_label = self._root_node.createElement('label')
         disclaimer_input.setAttribute('appearance', 'kobo-disclaimer')
         disclaimer_input.setAttribute(
-            'ref', f'/{self._unique_id}/_{self._unique_id}__disclaimer'
+            'ref', f'/{self._root_tag_name}/_{self._unique_id}__disclaimer'
         )
 
         if translated:
-            itext = f'/{self._unique_id}/_{self._unique_id}__disclaimer:label'
+            itext = f'/{self._root_tag_name}/_{self._unique_id}__disclaimer:label'
             disclaimer_input_label.setAttribute(
                 'ref',
                 f"jr:itext('{itext}')",
@@ -474,7 +478,7 @@ def _add_translation_nodes(
                 disclaimer_translation = self._root_node.createElement('text')
                 disclaimer_translation.setAttribute(
                     'id',
-                    f'/{self._unique_id}/_{self._unique_id}__disclaimer:label',
+                    f'/{self._root_tag_name}/_{self._unique_id}__disclaimer:label',
                 )
                 value = self._root_node.createElement('value')
                 language = n.getAttribute('lang').lower().strip()