def _update_obj_index_iter(self, obj):
"""Either updates the given object index, or yields an unsaved search entry."""
model = obj.__class__
adapter = self.get_adapter(model)
content_type = ContentType.objects.get_for_model(model)
object_id = force_text(obj.pk)
# Create the search entry data.
search_entry_data = {
"engine_slug": self._engine_slug,
"title": adapter.get_title(obj),
"description": adapter.get_description(obj),
"content": adapter.get_content(obj),
"url": adapter.get_url(obj),
"meta_encoded": json.dumps(adapter.get_meta(obj)),
}
# Try to get the existing search entry.
object_id_int, search_entries = self._get_entries_for_obj(obj)
# Attempt to update the search entries.
update_count = search_entries.update(**search_entry_data)
if update_count == 0:
# This is the first time the entry was created.
search_entry_data.update((
("content_type", content_type),
("object_id", object_id),
("object_id_int", object_id_int),
))
yield SearchEntry(**search_entry_data)
elif update_count > 1:
# Oh no! Somehow we've got duplicated search entries!
search_entries.exclude(id=search_entries[0].id).delete()
def get_title(self, obj):
"""
Returns the title of this search result. This is given high priority in search result ranking.
You can access the title of the search entry as `entry.title` in your search results.
The default implementation returns `force_text(obj)`.
"""
return force_text(obj)
def testSiteSearchJSON(self):
# Test a search that should find everything.
response = self.client.get("/simple/json/?q=title")
self.assertEqual(response["Content-Type"], "application/json; charset=utf-8")
results = set(result["title"] for result in json.loads(force_text(response.content))["results"])
self.assertEqual(len(results), 4)
self.assertTrue("title model1 instance11" in results)
self.assertTrue("title model1 instance12" in results)
self.assertTrue("title model2 instance21" in results)
self.assertTrue("title model2 instance22" in results)
def _resolve_field(self, obj, name):
"""Resolves the content of the given model field."""
name_parts = name.split("__", 1)
prefix = name_parts[0]
# If we're at the end of the resolve chain, return.
if obj is None:
return ""
# Try to get the attribute from the object.
try:
value = getattr(obj, prefix)
except ObjectDoesNotExist:
return ""
except AttributeError:
# Try to get the attribute from the search adapter.
try:
value = getattr(self, prefix)
except AttributeError:
raise SearchAdapterError("Could not find a property called {name!r} on either {obj!r} or {search_adapter!r}".format(
name = prefix,
obj = obj,
search_adapter = self,
))
else:
# Run the attribute on the search adapter, if it's callable.
if not isinstance(value, (QuerySet, models.Manager)):
if callable(value):
value = value(obj)
else:
# Run the attribute on the object, if it's callable.
if not isinstance(value, (QuerySet, models.Manager)):
if callable(value):
value = value()
# Look up recursive fields.
if len(name_parts) == 2:
if isinstance(value, (QuerySet, models.Manager)):
return " ".join(force_text(self._resolve_field(obj, name_parts[1])) for obj in value.all())
return self._resolve_field(value, name_parts[1])
# Resolve querysets.
if isinstance(value, (QuerySet, models.Manager)):
value = " ".join(force_text(related) for related in value.all())
# Resolution complete!
return value
def testSiteSearchCustomJSON(self):
# Test a search that should find everything.
response = self.client.get("/custom/json/?fooo=title&page=last")
self.assertEqual(response["Content-Type"], "application/json; charset=utf-8")
results = set(result["title"] for result in json.loads(force_text(response.content))["results"])
self.assertEqual(len(results), 4)
self.assertTrue("title model1 instance11" in results)
self.assertTrue("title model1 instance12" in results)
self.assertTrue("title model2 instance21" in results)
self.assertTrue("title model2 instance22" in results)
# Test a search with an invalid page.
response = self.client.get("/custom/json/?fooo=title&page=200")
self.assertEqual(response.status_code, 404)
def get_content(self, obj):
"""
Returns the content of this search result. This is given low priority in search result ranking.
You can access the content of the search entry as `entry.content` in your search results, although
this field generally contains a big mess of search data so is less suitable for frontend display.
The default implementation returns all the registered fields in your model joined together.
"""
# Get the field names to look up.
field_names = self.fields or (field.name for field in self.model._meta.fields if isinstance(field, (models.CharField, models.TextField)))
# Exclude named fields.
field_names = (field_name for field_name in field_names if field_name not in self.exclude)
# Create the text.
return self.prepare_content(" ".join(
force_text(self._resolve_field(obj, field_name))
for field_name in field_names
))
def resolve_url(to, *args, **kwargs):
"""
Return a URL appropriate for the arguments passed.
The arguments could be:
* A model: the model's `get_absolute_url()` function will be called.
* A view name, possibly with arguments: `urlresolvers.reverse()` will
be used to reverse-resolve the name.
* A URL, which will be returned as-is.
"""
from compat import six, force_text
# If it's a model, use get_absolute_url()
if hasattr(to, 'get_absolute_url'):
return to.get_absolute_url()
if isinstance(to, Promise):
# Expand the lazy instance, as it can cause issues when it is passed
# further to some Python functions like urlparse.
to = force_text(to)
if isinstance(to, six.string_types):
# Handle relative URLs
if any(to.startswith(path) for path in ('./', '../')):
return to
# Next try a reverse URL resolution.
try:
return urlresolvers.reverse(to, args=args, kwargs=kwargs)
except urlresolvers.NoReverseMatch:
# If this is a callable, re-raise.
if callable(to):
raise
# If this doesn't "feel" like a URL, re-raise.
if '/' not in to and '.' not in to:
raise
# Finally, fall back and assume it's a URL
return to
def _get_entries_for_obj(self, obj):
"""Returns a queryset of entries associate with the given obj."""
model = obj.__class__
content_type = ContentType.objects.get_for_model(model)
object_id = force_text(obj.pk)
# Get the basic list of search entries.
search_entries = SearchEntry.objects.filter(
content_type = content_type,
engine_slug = self._engine_slug,
)
if has_int_pk(model):
# Do a fast indexed lookup.
object_id_int = int(obj.pk)
search_entries = search_entries.filter(
object_id_int = object_id_int,
)
else:
# Alas, have to do a slow unindexed lookup.
object_id_int = None
search_entries = search_entries.filter(
object_id = object_id,
)
return object_id_int, search_entries
def escape_query(text):
text = force_text(text)
text = RE_SPACE.sub(" ", text) # Standardize spacing.
text = RE_NON_WORD.sub("", text) # Remove non-word characters.
return text
def __str__(self):
return force_text(self.title)