Indexes: Map Indexes

Let's create an index that will help us search for Employees by their FirstName, LastName, or both.

• First, let's create an index called Employees/ByFirstAndLastName
  

  Note: The naming separator character "_" in your code will become "/" in the index name.
  In the following sample, "Employees_ByFirstAndLastName" will become "Employees/ByFirstAndLastName" in your indexes list.

class Employees_ByFirstAndLastName(AbstractIndexCreationTask):
    def __init__(self):
        super().__init__()
        # ...

class Employees_ByFirstAndLastName(AbstractJavaScriptIndexCreationTask):
    def __init__(self):
        super().__init__()
        # ...

You might notice that we're passing Employee as a generic parameter to AbstractIndexCreationTask. This gives our indexing function a strongly-typed syntax. If you are not familiar with AbstractIndexCreationTask, you can read this article before proceeding.

• The next step is to create the indexing function itself. This is done by setting the map property with our function in a parameterless constructor.

class Employees_ByFirstAndLastName(AbstractIndexCreationTask):
    def __init__(self):
        super().__init__()
        self.map = "from employee in docs.Employees select new { FirstName = employee.FirstName, LastName = employee.LastName }"

class Employees_ByFirstAndLastName(AbstractIndexCreationTask):
    def __init__(self):
        super().__init__()
        self.map = "from employee in docs.Employees select new { FirstName = employee.FirstName, LastName = employee.LastName }"

class Employees_ByFirstAndLastName(AbstractJavaScriptIndexCreationTask):
    def __init__(self):
        super().__init__()
        self.maps = {
            """
            map('Employees', function (employee){ 
                return { 
                    FirstName : employee.FirstName, 
                    LastName : employee.LastName
                };
            })
            """
        }

• The final step is to deploy it to the server and issue a query using the session Query method.
  To query an index, the name of the index must be called by the query.
  If the index isn't called, RavenDB will either use or create an auto index.

employees_1 = list(
    session.query_index_type(Employees_ByFirstAndLastName, Employee).where_equals(
        "first_name", "Robert"
    )
)
employees_2 = list(
    session.query_index("Employees/ByFirstAndLastName", Employee).where_equals("first_name", "Robert")
)

from index 'Employees/ByFirstAndLastName'
where FirstName = 'Robert'

This is how our final index looks like:

class Employees_ByFirstAndLastName(AbstractIndexCreationTask):
    def __init__(self):
        super().__init__()
        self.map = "from employee in docs.Employees select new { FirstName = employee.FirstName, LastName = employee.LastName }"

class Employees_ByFirstAndLastName(AbstractJavaScriptIndexCreationTask):
    def __init__(self):
        super().__init__()
        self.maps = {
            """
            map('Employees', function (employee){ 
                return { 
                    FirstName : employee.FirstName, 
                    LastName : employee.LastName
                };
            })
            
            """
        }

from employee in docs.Employees
select new
{
	Age = employee.Age
}

map('Employees', function(employee)
{
    return {
        Age : employee.Age
    };
})

from employee in docs.Employees
select new
{
	Age = employee.Age.ToString()
}

map('Employees', function(employee)
{
    return {
        Age : employee.Age.toString()
    };
})

from employee in docs.Employees
select new
{
	FirstName = employee.FirstName,
	LastName = employee.LastName
}

Since each index contains a function, you can combine multiple fields into one.

Example I

Index definition:

class Employees_ByFullName(AbstractIndexCreationTask):
    class Result:
        def __init__(self, full_name: str = None):
            self.full_name = full_name

    def __init__(self):
        super().__init__()
        self.map = (
            'from employee in docs.Employees select new { full_name = employee.FirstName + " " + employee.LastName }'
        )

class Employees_ByFullName(AbstractJavaScriptIndexCreationTask):
    class Result:
        def __init__(self, full_name: str = None):
            self.full_name = full_name

    def __init__(self):
        super().__init__()
        self.maps = {
            """
            map('Employees', function (employee){ 
                return { 
                    FullName  : employee.FirstName + ' ' + employee.LastName
                };
            })
            """
        }

Query the index:

employees = list(
    session.query_index_type(Employees_ByFullName, Employee).where_equals("full_name", "Robert King")
)

from index 'Employees/ByFullName'
where FullName = 'Robert King'

Example II

Index definition:

class Companies_ByAddress_Country(AbstractIndexCreationTask):
    class Result:
        def __init__(self, city: str = None, company: str = None, phone: str = None):
            self.city = city
            self.company = company
            self.phone = phone

    def __init__(self):
        super().__init__()
        self.map = (
            'from company in docs.Companies where company.Address.Country == "USA"'
            "select new { company = company.Name, city = company.Address.City, phone = company.Phone }"
        )

class Employees_Query(AbstractJavaScriptIndexCreationTask):
    class Result:
        def __init__(self, query: List[str] = None):
            self.query = query

    def __init__(self):
        super().__init__()
        self.maps = {
            """
            map('Employees', function (employee) { 
            return { 
                query : [employee.FirstName, 
                         employee.LastName,
                         employee.Title,
                         employee.Address.City] 
                    } 
            })
            """
        }

        self.fields = {"query": IndexFieldOptions(indexing=FieldIndexing.SEARCH)}

Query the index:

employees = list(
    session.query_index_type(Employees_Query, Employees_Query.Result)
    .search("query", "John Doe")
    .of_type(Employee)
)

from index 'Employees/Query'
where search(Query, 'John Doe')

Imagine that you would like to return all employees that were born in a specific year. You can do it by indexing Birthday from Employee, then specify the year in Birthday as you query the index:

Index definition:

class Employees_ByBirthday(AbstractIndexCreationTask):
    class Result:
        def __init__(self, birthday: datetime.datetime = None):
            self.birthday = birthday

        @classmethod
        def from_json(cls, json_dict: Dict[str, Any]) -> "Employees_ByBirthday.Result":
            # import 'Utils' from 'ravendb.tools.utils' to convert C# datetime strings to Python datetime objects
            return cls(Utils.string_to_datetime(json_dict["Birthday"]))

class Employees_ByBirthday(AbstractJavaScriptIndexCreationTask):
    class Result:
        def __init__(self, birthday: datetime.datetime = None):
            self.birthday = birthday

        @classmethod
        def from_json(cls, json_dict: Dict[str, Any]) -> "Employees_ByBirthday.Result":
            return cls(json_dict["Birthday"])

    def __init__(self):
        super().__init__()
        self.maps = {
            """
            map('Employees', function (employee){ 
                return {
                    Birthday : employee.Birthday 
                } 
            })
            """
        }

Query the index:

start_date = datetime.datetime(1963, 1, 1)
end_date = start_date + datetime.timedelta(days=365) - datetime.timedelta(milliseconds=1)
employees = list(
    session.query_index_type(Employees_ByBirthday, Employees_ByBirthday.Result)
    .where_between("birthday", start_date, end_date)
    .of_type(Employee)
)

from index 'Employees/ByBirthday '
where Birthday between '1963-01-01' and '1963-12-31T23:59:59.9990000'

RavenDB gives you the ability to extract field data and to index by it. A different way to achieve our goal will look as follows:

Index definition:

class Employees_ByYearOfBirth(AbstractIndexCreationTask):
    class Result:
        def __init__(self, year_of_birth: int = None):
            self.year_of_birth = year_of_birth

    def __init__(self):
        super().__init__()
        self.map = "from employee in docs.Employees select new { year_of_birth = employee.Birthday.Year }"

class Employees_ByYearOfBirth(AbstractJavaScriptIndexCreationTask):
    class Result:
        def __init__(self, year_of_birth: int = None):
            self.year_of_birth = year_of_birth

        @classmethod
        def from_json(cls, json_dict: Dict[str, Any]) -> "Employees_ByYearOfBirth.Result":
            return cls(json_dict["Birthday"])

    def __init__(self):
        super().__init__()
        self.maps = {
            """
            map('Employees', function (employee){ 
                return {
                    Birthday : employee.Birthday.Year 
                } 
            })
            """
        }

Query the index:

employees = list(
    session.query_index_type(Employees_ByYearOfBirth, Employees_ByYearOfBirth.Result)
    .where_equals("year_of_birth", 1963)
    .of_type(Employee)
)

from index 'Employees/ByYearOfBirth'
where YearOfBirth = 1963

In the examples above, where_equals is used in the query to filter the results.
If you consistently want to filter with the same filtering conditions, you can use where_equals in the index definition to narrow the index terms that the query must scan.

This can save query-time but narrows the terms available to query.

Example I

For logic that has to do with special import rules that only apply to the USA
where can be used to filter the Companies collection Address.Country field.
Thus, we only index documents where company.Address.Country == "USA" .

Index definition:

class Companies_ByAddress_Country(AbstractIndexCreationTask):
    class Result:
        def __init__(self, city: str = None, company: str = None, phone: str = None):
            self.city = city
            self.company = company
            self.phone = phone

    def __init__(self):
        super().__init__()
        self.map = (
            'from company in docs.Companies where company.Address.Country == "USA"'
            "select new { company = company.Name, city = company.Address.City, phone = company.Phone }"
        )

Query the index:

orders = list(
    session.query_index_type(Companies_ByAddress_Country, Companies_ByAddress_Country.Result).of_type(
        Company
    )
)

from index 'Companies_ByAddress_Country'

Example II

Imagine a seed company that needs to categorize its customers by latitude-based growing zones.

They can create a different index for each zone and filter their customers in the index with
where (company.Address.Location.Latitude > 20 && company.Address.Location.Latitude < 50) .

Index definition:

class Companies_ByAddress_Latitude(AbstractIndexCreationTask):
    class Result:
        def __init__(
            self,
            latitude: float = None,
            longitude: float = None,
            company_name: str = None,
            company_address: str = None,
            company_phone: str = None,
        ):
            self.latitude = latitude
            self.longitude = longitude
            self.company_name = company_name
            self.company_address = company_address
            self.company_phone = company_phone

    def __init__(self):
        super().__init__()
        self.map = (
            "from company in companies"
            "where (company.Address.Location.Latitude > 20 && company.Address.Location.Latitude < 50"
            "select new"
            "{"
            "    latitude = company.Address.Location.Latitude,"
            "    longitude = company.Address.Location.Longitude,"
            "    company_name = company.Name,"
            "    company_address = company.Address,"
            "    company_phone = company.Phone"
            "}"
        )

Query the index:

orders = list(
    session.query_index_type(Companies_ByAddress_Latitude, Companies_ByAddress_Latitude.Result).of_type(
        Company
    )
)

from index 'Companies_ByAddress_Latitude'

If your document contains nested data, e.g. Employee contains Address, you can index on its fields by accessing them directly in the index. Let's say that we would like to create an index that returns all employees that were born in a specific Country:

Index definition:

class Employees_ByCountry(AbstractIndexCreationTask):
    class Result:
        def __init__(self, country: str = None):
            self.country = country

    def __init__(self):
        super().__init__()
        self.map = "from employee in docs.Employees select new { country = employee.Address.Country }"

class Employees_ByCountry(AbstractJavaScriptIndexCreationTask):
    class Result:
        def __init__(self, country: str = None):
            self.country = country

    def __init__(self):
        super().__init__()
        self.maps = {
            """
            map('Employees', function (employee){ 
                return {
                    country : employee.Address.Country 
                 } 
            })
            """
        }

Query the index:

employees = list(
    session.query_index_type(Employees_ByCountry, Employees_ByCountry.Result)
    .where_equals("country", "USA")
    .of_type(Employee)
)

from index 'Employees/ByCountry'
where Country = 'USA'

If a document relationship is represented by the document's ID, you can use the LoadDocument method to retrieve such a document.
Learn more here.

By default, indexes will not index a document that contains none of the specified fields. This behavior can be changed using the Indexing.IndexEmptyEntries configuration option.

The option Indexing.IndexMissingFieldsAsNull determines whether missing fields in documents are indexed with the value null, or not indexed at all.