Adjust the Indexing of Object Lots

robc

When I search for the exact Identifier of an object, I don't want this to match a related Lot. Is there a way to adjust the indexing configuration so that only the name of a Lot can match when searching?

darrigan

Hi,
I think you need to edit the search_indexing.conf file… but I haven’t tried making any changes to it yet. 🫤
Incidentally, could anyone confirm whether it is possible to set different indexing rules for Providence (curators) and Pawtucket (visitors)?
There is a documentation here: https://docs.collectiveaccess.org/providence/user/configuration/configuringProvidence/mainConfiguration/search_indexing (but examples should help to understand)

darrigan

Well... if I understand the documentation, you should modify the search_indexing.conf, block ca_object_lots = {...}.
By default, this block is:

ca_object_lots = {
	# ------------------------------------
	ca_object_lots = {
		fields = {
			idno_stub = { STORE, DONT_TOKENIZE, INDEX_AS_IDNO, BOOST = 100 },
			_metadata = { },
			lot_status_id = { STORE, DONT_TOKENIZE },
			type_id = { STORE, DONT_TOKENIZE },
			extent = { STORE, DONT_TOKENIZE },
			extent_units = { STORE, DONT_TOKENIZE },
			status = { STORE, DONT_TOKENIZE },
			access = { STORE, DONT_TOKENIZE },
			deleted = { STORE, DONT_TOKENIZE }
		}
	},
	# ------------------------------------
	ca_object_lot_labels = {
		key = lot_id,
		fields = {
			name = { BOOST = 100 },
			name_sort = { DONT_INCLUDE_IN_SEARCH_FORM }
		}
	},
	# ------------------------------------
	ca_entity_labels = {
		tables = {
			entities = [ca_object_lots_x_entities, ca_entities]
		},
		fields = {
			displayname = {}
		}
	},
	# ------------------------------------
	ca_list_item_labels = {
		tables = {
			vocabulary = [ca_object_lots_x_vocabulary_terms, ca_list_items]
		},
		fields = {
			item_id = { DONT_INCLUDE_IN_SEARCH_FORM },
			name_singular = {},
			name_plural = {}
		}
	},
	# ------------------------------------
	ca_objects = {
		key = lot_id,
		fields = {
			idno = { STORE, DONT_TOKENIZE, INDEX_AS_IDNO, BOOST = 100 }
		}
	},
	# ------------------------------------
	ca_storage_location_labels = {
		tables = {
			places = [ca_object_lots_x_storage_locations, ca_storage_locations]
		},
		fields = {
			location_id = { DONT_INCLUDE_IN_SEARCH_FORM },
			name = {}
		}
	},
	# ------------------------------------
	ca_sets = {
		tables = {
			sets = [ca_set_items],
		},
		keys = {		# manually override joins with explicit setting
			sets = {
				ca_set_items = {
					ca_object_lots = {
						left_key = row_id,
						left_table_num = table_num,
						right_key = lot_id
					}
				}
			}
		},
		fields = {
			set_code = { },
			set_id = { DONT_TOKENIZE, DONT_INCLUDE_IN_SEARCH_FORM }
		}
	},
	# ------------------------------------
	ca_media_upload_sessions = {
		key = session_id,
		fields = {
			session_key = { STORE, DONT_TOKENIZE },
			session_id = { STORE, DONT_TOKENIZE },
		}
	},
	# ------------------------------------
	_access_points = {
		basic = {
			fields = [ca_object_lot_labels.name, ca_object_lots.idno_stub],
			boolean = OR,
			name = _(Label + idno),
			options = {  }
		},
		set = {
			fields = [ca_sets.set_id, ca_sets.set_code],
			boolean = OR
		},
		mediaUploadSession = {
			fields = [ca_media_upload_sessions.session_key],
			options = { }
		}
	}
	# ------------------------------------
}

What you want (if I understand) is not find an object lot when you search for a string that corresponds to an object's metadata. Just find an object lot with its own metadata.
To do so, you may comment (or delete) the part related to ca_objects in this block:

ca_object_lots = {
	# ------------------------------------
	ca_object_lots = {
		fields = {
			idno_stub = { STORE, DONT_TOKENIZE, INDEX_AS_IDNO, BOOST = 100 },
			_metadata = { },
			lot_status_id = { STORE, DONT_TOKENIZE },
			type_id = { STORE, DONT_TOKENIZE },
			extent = { STORE, DONT_TOKENIZE },
			extent_units = { STORE, DONT_TOKENIZE },
			status = { STORE, DONT_TOKENIZE },
			access = { STORE, DONT_TOKENIZE },
			deleted = { STORE, DONT_TOKENIZE }
		}
	},
	# ------------------------------------
	ca_object_lot_labels = {
		key = lot_id,
		fields = {
			name = { BOOST = 100 },
			name_sort = { DONT_INCLUDE_IN_SEARCH_FORM }
		}
	},
	# ------------------------------------
	ca_entity_labels = {
		tables = {
			entities = [ca_object_lots_x_entities, ca_entities]
		},
		fields = {
			displayname = {}
		}
	},
	# ------------------------------------
	ca_list_item_labels = {
		tables = {
			vocabulary = [ca_object_lots_x_vocabulary_terms, ca_list_items]
		},
		fields = {
			item_id = { DONT_INCLUDE_IN_SEARCH_FORM },
			name_singular = {},
			name_plural = {}
		}
	},
	# ------------------------------------
#	ca_objects = {
#		key = lot_id,
#		fields = {
#			idno = { STORE, DONT_TOKENIZE, INDEX_AS_IDNO, BOOST = 100 }
#		}
#	},
	# ------------------------------------
	ca_storage_location_labels = {
		tables = {
			places = [ca_object_lots_x_storage_locations, ca_storage_locations]
		},
		fields = {
			location_id = { DONT_INCLUDE_IN_SEARCH_FORM },
			name = {}
		}
	},
	# ------------------------------------
	ca_sets = {
		tables = {
			sets = [ca_set_items],
		},
		keys = {		# manually override joins with explicit setting
			sets = {
				ca_set_items = {
					ca_object_lots = {
						left_key = row_id,
						left_table_num = table_num,
						right_key = lot_id
					}
				}
			}
		},
		fields = {
			set_code = { },
			set_id = { DONT_TOKENIZE, DONT_INCLUDE_IN_SEARCH_FORM }
		}
	},
	# ------------------------------------
	ca_media_upload_sessions = {
		key = session_id,
		fields = {
			session_key = { STORE, DONT_TOKENIZE },
			session_id = { STORE, DONT_TOKENIZE },
		}
	},
	# ------------------------------------
	_access_points = {
		basic = {
			fields = [ca_object_lot_labels.name, ca_object_lots.idno_stub],
			boolean = OR,
			name = _(Label + idno),
			options = {  }
		},
		set = {
			fields = [ca_sets.set_id, ca_sets.set_code],
			boolean = OR
		},
		mediaUploadSession = {
			fields = [ca_media_upload_sessions.session_key],
			options = { }
		}
	}
	# ------------------------------------
}

Then "Rebuild search indices" in web interface under Manage > Administrate > Maintenance (or using the command-line caUtils rebuild-search-index command)
Hope this will help...

seth

Just remove this block in search_indexing.conf (it's within the enclosing ca_object_lots block:

# ------------------------------------
	ca_objects = {
		key = lot_id,
		fields = {
			idno = { STORE, DONT_TOKENIZE, INDEX_AS_IDNO, BOOST = 100 }
		}
	},

and then reindex.

robc

Thanks for that. I'm trying it now. On a related topic, I seem to have specific items in the objects table where the rebuild hangs for several seconds, sometimes multiple minutes per object. Is there anything I should do to try to troubleshoot, or is that considered normal? I have about 9000 objects and there's something near the 7000 range that goes slower than all the others.

By the way, navigation is much easier now that the quick search is working for identifiers when the object is related to a lot. Before this conf change, it would take me to the list of results instead of the object editor.