Indexes and Subscriptions

Hash index

The hash index is a dictionary of the field’s hash values. Its filling (re-indexing) requires a certain time, but searching through the index is almost instantaneous. It is intended to be used only for unique values. At least it will always return only one document.

To register an index or change conditions, call register_hash_index function once. For example:

register_hash_index(<index name>,<field name>, <additional argument>)

For example:

db['query_test'].register_hash_index("my_hash_index","name",dynamic = True)

where the dynamic=True flag indicates that the index will not be stored on disk, it will only exist in memory. Those. In order to use the index, you need to re-index

After registering an index, all changes to the collection included in the index (insert, update, delete) will lead to a change in the index. This should be taken into account to understand the performance of operations. To disable such movements, you can specify the no_index=True flag in the arguments

You can also force re-index the entire index with the command

collection.reindex_hash(<index name>)

For example,

db['query_test'].reindex_hash('my_hash_index')

Searching by index is actually getting the value by the key in the dictionary and is performed by the get_by_index(<index>,<value>) function

For example,

r = db['query_test'].get_by_index(db["my_hash_index "],"apple")

Text index

Required to significantly speed up the search for a substring in a string. It is a variation of a balanced binary tree.

To register an index or change conditions, call the register_text_index function once. For example:

register_text_index(<index name>,<field name>, <additional argument>)

db[‘query_test’]. register_text_index(“my_text_index”,”name”,dynamic = True)

Note

As with hash indexes, registration leads to the fact that all operations begin to affect the index, but the generation time is somewhat longer, and the volume of files is quite significant. So it is recommended to keep it in memory ( dynamic = True)

You can also force re-index the entire index with the command

collection.reindex_text(<index name>)

For example,

db['query_test'].reindex_text('my_text_index')

To search by index, use the function

index. search_text_index(<search substring>)

For example:

db['goods'].search_text_index("appl")

Subscriptions

A subscription is a collection that automatically registers all changes to the selected collections. That is, during any operations, the IDs of the documents for which the data was changed will be added to this collection Subscriptions are registered by command:

Database.create_subscription(<subscription name>,[<list of collection names>])

For example

db.create_subscription("my_subscription",['income','operations'])

A subscription is a regular collection, so you can get data by any function for getting data - all, find, etc.