Question

I have been trying to get xapian working django haystack for a project im working on that requires some search functionality but have run into a bit of a wall!! Basically i got everything installed per the instructions, so:

ran make install for xapian-core and the xapian bindings
ran pip install haystack and pip install xapian-haystack and everything installed correctly

As im using the django cms app i simply copied thier example over to give the search functionality a test and ran into this error

InvalidIndexError at /search/
Unable to open index at /home/mike/sites/xapian_search

I have tried several different paths for the HAYSTACK_XAPIAN_PATH setting and have also encountered another error

OSError at /
(13, 'Permission denied')

the folder xapain_search has been given full perms (chmod 777) and theres an xapian_index.php file with full perms too. Im not sure what im missing here but im desperate to try and get this working!!

my haystack settings look like

HAYSTACK_SITECONF = 'lactoseintolerant.lactose_search'
HAYSTACK_SEARCH_ENGINE = 'xapian'
HAYSTACK_XAPIAN_PATH = '/home/mike/sites/xapian_search'
HAYSTACK_SEARCH_RESULTS_PER_PAGE = 50

Any advice would be greatly appreciated!!

edit

Hey again
i think this error is relating to the fact that there are no indexes(is that right?) i have run the commands update_index rebuild_index clear_index all of which dont seem to do anything, there are no errors outputted but still the indexs never seem to be built when the commands are run

i have an app called lactose_search which my HAYSTACK_SITECONF points to like so projectname.lactose_search in this app folder i have a file called search_indexs.py. For now i have simply c+p the example from the django cms site as it is the cms_app content i want to search
this file looks like

from django.conf import settings
from django.utils.translation import string_concat, ugettext_lazy

from haystack import indexes, site

from cms.models.managers import PageManager
from cms.models.pagemodel import Page

def page_index_factory(lang, lang_name):
if isinstance(lang_name, basestring):
    lang_name = ugettext_lazy(lang_name)

def get_absolute_url(self):
    return '/%s%s' % (lang, Page.get_absolute_url(self))

class Meta:
    proxy = True
    app_label = 'cms'
    verbose_name = string_concat(Page._meta.verbose_name, ' (', lang_name, ')')
    verbose_name_plural = string_concat(Page._meta.verbose_name_plural, ' (', lang_name, ')')
    
attrs = {'__module__': Page.__module__, 
         'Meta': Meta,
         'objects': PageManager(),
         'get_absolute_url': get_absolute_url}

_PageProxy = type("Page%s" % lang.title() , (Page,), attrs)

_PageProxy._meta.parent_attr = 'parent'
_PageProxy._meta.left_attr = 'lft'
_PageProxy._meta.right_attr = 'rght'
_PageProxy._meta.tree_id_attr = 'tree_id'

class _PageIndex(indexes.SearchIndex):
    language = lang
    
    text = indexes.CharField(document=True, use_template=False)
    pub_date = indexes.DateTimeField(model_attr='publication_date')
    login_required = indexes.BooleanField(model_attr='login_required')
    url = indexes.CharField(stored=True, indexed=False, model_attr='get_absolute_url')
    title = indexes.CharField(stored=True, indexed=False, model_attr='get_title')
    
    def prepare(self, obj):
        self.prepared_data = super(_PageIndex, self).prepare(obj)
        plugins = obj.cmsplugin_set.filter(language=lang)
        text = ''
        for plugin in plugins:
            instance, _ = plugin.get_plugin_instance()
            if hasattr(instance, 'search_fields'):
                text += ''.join(getattr(instance, field) for field in instance.search_fields)
        self.prepared_data['text'] = text
        return self.prepared_data
    
    def get_queryset(self):
        return _PageProxy.objects.published().filter(title_set__language=lang, publisher_is_draft=False).distinct()

return _PageProxy, _PageIndex

for lang_tuple in settings.LANGUAGES:
   lang, lang_name = lang_tuple
    site.register(*page_index_factory(lang, lang_name))

and can be found here http://docs.django-cms.org/en/2.1.3/extending_cms/searchdocs.html

Hope this extra info may make answering this question abit easier!

Answer 1

I figured out what my issue was here, when i installed the packages to my env i ran sudo pip install instead of simply using pip. I cant explain why this affected the haystack install but once i removed all the packages and re-installed them i managed to get haystack running