redisearch-py is a python search engine library that utilizes the RediSearch Redis Module API.
It is the "official" client of redisearch, and should be regarded as its canonical client implementation.
The source code can be found at http://github.com/RedisLabs/redisearch-py
from redisearch import Client, TextField, NumericField, Query
# Creating a client with a given index name
client = Client('myIndex')
# Creating the index definition and schema
client.create_index([TextField('title', weight=5.0), TextField('body')])
# Indexing a document
client.add_document('doc1', title = 'RediSearch', body = 'Redisearch impements a search engine on top of redis')
# Simple search
res = client.search("search engine")
# the result has the total number of results, and a list of documents
print res.total # "1"
print res.docs[0].title
# Searching with complext parameters:
q = Query("search engine").verbatim().no_content().paging(0,5)
res = client.search(q)# Using the auto-completer
ac = AutoCompleter('ac')
# Adding some terms
ac.add_suggestions(Suggestion('foo', 5.0), Suggestion('bar', 1.0))
# Getting suggestions
suggs = ac.get_suggestions('goo') # returns nothing
suggs = ac.get_suggestions('goo', fuzzy = True) # returns ['foo']-
Install redis 4.0 RC2 or above
-
Install the python client
$ pip install redisearchA client to RediSearch's AutoCompleter API
It provides prefix searches with optionally fuzzy matching of prefixes
def __init__(self, key, host='localhost', port=6379, conn=None)Create a new AutoCompleter client for the given key, and optional host and port
If conn is not None, we employ an already existing redis connection
def add_suggestions(self, *suggestions, **kwargs)Add suggestion terms to the AutoCompleter engine. Each suggestion has a score and string.
If kwargs['increment'] is true and the terms are already in the server's dictionary, we increment their scores
def delete(self, string)Delete a string from the AutoCompleter index. Returns 1 if the string was found and deleted, 0 otherwise
def get_suggestions(self, prefix, fuzzy=False, num=10, with_scores=False, with_payloads=False)Get a list of suggestions from the AutoCompleter, for a given prefix
- prefix: the prefix we are searching. Must be valid ascii or utf-8
- fuzzy: If set to true, the prefix search is done in fuzzy mode. NOTE: Running fuzzy searches on short (<3 letters) prefixes can be very slow, and even scan the entire index.
- with_scores: if set to true, we also return the (refactored) score of each suggestion. This is normally not needed, and is NOT the original score inserted into the index
- with_payloads: Return suggestion payloads
- num: The maximum number of results we return. Note that we might return less. The algorithm trims irrelevant suggestions.
Returns a list of Suggestion objects. If with_scores was False, the score of all suggestions is 1.
def len(self)Return the number of entries in the AutoCompleter index
A client for the RediSearch module. It abstracts the API of the module and lets you just use the engine
def __init__(self, index_name, host='localhost', port=6379, conn=None)Create a new Client for the given index_name, and optional host and port
If conn is not None, we employ an already existing redis connection
def add_document(self, doc_id, nosave=False, score=1.0, payload=None, replace=False, partial=False, language=None, **fields)Add a single document to the index.
- doc_id: the id of the saved document.
- nosave: if set to true, we just index the document, and don't save a copy of it. This means that searches will just return ids.
- score: the document ranking, between 0.0 and 1.0
- payload: optional inner-index payload we can save for fast access in scoring functions
- replace: if True, and the document already is in the index, we perform an update and reindex the document
- partial: if True, the fields specified will be added to the existing document.
This has the added benefit that any fields specified with
no_indexwill not be reindexed again. Impliesreplace - language: Specify the language used for document tokenization.
- fields kwargs dictionary of the document fields to be saved and/or indexed. NOTE: Geo points shoule be encoded as strings of "lon,lat"
def add_document_hash(self, doc_id, score=1.0, language=None, replace=False)Add a hash document to the index.
- doc_id: the document's id. This has to be an existing HASH key in Redis that will hold the fields the index needs.
- score: the document ranking, between 0.0 and 1.0
- replace: if True, and the document already is in the index, we perform an update and reindex the document
- language: Specify the language used for document tokenization.
def aggregate(self, query)Issue an aggregation query
query: This can be either an AggeregateRequest, or a Cursor
An AggregateResult object is returned. You can access the rows from its
rows property, which will always yield the rows of the result
def alter_schema_add(self, fields)Alter the existing search index by adding new fields. The index must already exist.
- fields: a list of Field objects to add for the index
def batch_indexer(self, chunk_size=100)Create a new batch indexer from the client with a given chunk size
def create_index(self, fields, no_term_offsets=False, no_field_flags=False, stopwords=None)Create the search index. The index must not already exist.
- fields: a list of TextField or NumericField objects
- no_term_offsets: If true, we will not save term offsets in the index
- no_field_flags: If true, we will not save field flags that allow searching in specific fields
- stopwords: If not None, we create the index with this custom stopword list. The list can be empty
def delete_document(self, doc_id, conn=None, delete_actual_document=False)Delete a document from index Returns 1 if the document was deleted, 0 if not
- delete_actual_document: if set to True, RediSearch also delete the actual document if it is in the index
def drop_index(self)Drop the index if it exists
def explain(self, query)def info(self)Get info an stats about the the current index, including the number of documents, memory consumption, etc
def load_document(self, id)Load a single document by id
def search(self, query)Search the index for a given query, and return a result of documents
- query: the search query. Either a text for simple queries with default parameters, or a Query object for complex queries. See RediSearch's documentation on query format
A batch indexer allows you to automatically batch document indexeing in pipelines, flushing it every N documents.
def __init__(self, client, chunk_size=1000)def add_document(self, doc_id, nosave=False, score=1.0, payload=None, replace=False, partial=False, **fields)Add a document to the batch query
def add_document_hash(self, doc_id, score=1.0, language=None, replace=False)Add a hash document to the batch query
def commit(self)Manually commit and flush the batch indexing query
Represents a single document in a result set
def __init__(self, id, payload=None, **fields)GeoField is used to define a geo-indexing field in a schema defintion
def __init__(self, name)def redis_args(self)None
def __init__(self, field, lon, lat, radius, unit='km')NumericField is used to define a numeric field in a schema defintion
def __init__(self, name, sortable=False, no_index=False)def redis_args(self)None
def __init__(self, field, minval, maxval, minExclusive=False, maxExclusive=False)Query is used to build complex queries that have more parameters than just the query string. The query string is set in the constructor, and other options have setter functions.
The setter functions return the query object, so they can be chained,
i.e. Query("foo").verbatim().filter(...) etc.
def __init__(self, query_string)Create a new query object. The query string is set in the constructor, and other options have setter functions.
def add_filter(self, flt)Add a numeric or geo filter to the query. Currently only one of each filter is supported by the engine
- flt: A NumericFilter or GeoFilter object, used on a corresponding field
def get_args(self)Format the redis arguments for this query and return them
def highlight(self, fields=None, tags=None)Apply specified markup to matched term(s) within the returned field(s)
- fields If specified then only those mentioned fields are highlighted, otherwise all fields are highlighted
- tags A list of two strings to surround the match.
def in_order(self)Match only documents where the query terms appear in the same order in the document. i.e. for the query 'hello world', we do not match 'world hello'
def language(self, language)Analyze the query as being in the specified language
:param language: The language (e.g. chinese or english)
def limit_fields(self, *fields)Limit the search to specific TEXT fields only
- fields: A list of strings, case sensitive field names from the defined schema
def limit_ids(self, *ids)Limit the results to a specific set of pre-known document ids of any length
def no_content(self)Set the query to only return ids and not the document content
def no_stopwords(self)Prevent the query from being filtered for stopwords. Only useful in very big queries that you are certain contain no stopwords.
def paging(self, offset, num)Set the paging for the query (defaults to 0..10).
- offset: Paging offset for the results. Defaults to 0
- num: How many results do we want
def query_string(self)Return the query string of this query only
def return_fields(self, *fields)Only return values from these fields
def slop(self, slop)Allow a masimum of N intervening non matched terms between phrase terms (0 means exact phrase)
def sort_by(self, field, asc=True)Add a sortby field to the query
- field - the name of the field to sort by
- asc - when
True, sorting will be done in asceding order
def summarize(self, fields=None, context_len=None, num_frags=None, sep=None)Return an abridged format of the field, containing only the segments of the field which contain the matching term(s).
If fields is specified, then only the mentioned fields are
summarized; otherwise all results are summarized.
Server side defaults are used for each option (except fields) if not specified
- fields List of fields to summarize. All fields are summarized if not specified
- context_len Amount of context to include with each fragment
- num_frags Number of fragments per document
- sep Separator string to separate fragments
def verbatim(self)Set the query to be verbatim, i.e. use no query expansion or stemming
def with_payloads(self)Ask the engine to return document payloads
Represents the result of a search query, and has an array of Document objects
def __init__(self, res, hascontent, duration=0, has_payload=False)- snippets: An optional dictionary of the form {field: snippet_size} for snippet formatting
None
def __init__(self, field, asc=True)Represents a single suggestion being sent or returned from the auto complete server
def __init__(self, string, score=1.0, payload=None)TagField is a tag-indexing field with simpler compression and tokenization. See http://redisearch.io/Tags/
def __init__(self, name, separator=',', no_index=False)def redis_args(self)TextField is used to define a text field in a schema definition
def __init__(self, name, weight=1.0, sortable=False, no_stem=False, no_index=False)def redis_args(self)