|
| 1 | +# Licensed to Elasticsearch B.V. under one or more contributor |
| 2 | +# license agreements. See the NOTICE file distributed with |
| 3 | +# this work for additional information regarding copyright |
| 4 | +# ownership. Elasticsearch B.V. licenses this file to you under |
| 5 | +# the Apache License, Version 2.0 (the "License"); you may |
| 6 | +# not use this file except in compliance with the License. |
| 7 | +# You may obtain a copy of the License at |
| 8 | +# |
| 9 | +# http://www.apache.org/licenses/LICENSE-2.0 |
| 10 | +# |
| 11 | +# Unless required by applicable law or agreed to in writing, |
| 12 | +# software distributed under the License is distributed on an |
| 13 | +# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY |
| 14 | +# KIND, either express or implied. See the License for the |
| 15 | +# specific language governing permissions and limitations |
| 16 | +# under the License. |
| 17 | + |
| 18 | +""" |
| 19 | +# ES|QL query builder example |
| 20 | +
|
| 21 | +Requirements: |
| 22 | +
|
| 23 | +$ pip install "elasticsearch[async]" faker |
| 24 | +
|
| 25 | +To run the example: |
| 26 | +
|
| 27 | +$ python esql_employees.py "name to search" |
| 28 | +
|
| 29 | +The index will be created automatically with a list of 1000 randomly generated |
| 30 | +employees if it does not exist. Add `--recreate-index` or `-r` to the command |
| 31 | +to regenerate it. |
| 32 | +
|
| 33 | +Examples: |
| 34 | +
|
| 35 | +$ python esql_employees "Mark" # employees named Mark (first or last names) |
| 36 | +$ python esql_employees "Sarah" --limit 10 # up to 10 employees named Sarah |
| 37 | +$ python esql_employees "Sam" --sort height # sort results by height |
| 38 | +$ python esql_employees "Sam" --sort name # sort results by last name |
| 39 | +""" |
| 40 | + |
| 41 | +import argparse |
| 42 | +import asyncio |
| 43 | +import os |
| 44 | +import random |
| 45 | + |
| 46 | +from faker import Faker |
| 47 | + |
| 48 | +from elasticsearch.dsl import AsyncDocument, InnerDoc, M, async_connections |
| 49 | +from elasticsearch.esql import ESQLBase |
| 50 | +from elasticsearch.esql.functions import concat, multi_match |
| 51 | + |
| 52 | +fake = Faker() |
| 53 | + |
| 54 | + |
| 55 | +class Address(InnerDoc): |
| 56 | + address: M[str] |
| 57 | + city: M[str] |
| 58 | + zip_code: M[str] |
| 59 | + |
| 60 | + |
| 61 | +class Employee(AsyncDocument): |
| 62 | + emp_no: M[int] |
| 63 | + first_name: M[str] |
| 64 | + last_name: M[str] |
| 65 | + height: M[float] |
| 66 | + still_hired: M[bool] |
| 67 | + address: M[Address] |
| 68 | + |
| 69 | + class Index: |
| 70 | + name = "employees" |
| 71 | + |
| 72 | + @property |
| 73 | + def name(self) -> str: |
| 74 | + return f"{self.first_name} {self.last_name}" |
| 75 | + |
| 76 | + def __repr__(self) -> str: |
| 77 | + return f"<Employee[{self.meta.id}]: {self.first_name} {self.last_name}>" |
| 78 | + |
| 79 | + |
| 80 | +async def create(num_employees: int = 1000) -> None: |
| 81 | + print("Creating a new employee index...") |
| 82 | + if await Employee._index.exists(): |
| 83 | + await Employee._index.delete() |
| 84 | + await Employee.init() |
| 85 | + |
| 86 | + for i in range(num_employees): |
| 87 | + address = Address( |
| 88 | + address=fake.address(), city=fake.city(), zip_code=fake.zipcode() |
| 89 | + ) |
| 90 | + emp = Employee( |
| 91 | + emp_no=10000 + i, |
| 92 | + first_name=fake.first_name(), |
| 93 | + last_name=fake.last_name(), |
| 94 | + height=int((random.random() * 0.8 + 1.5) * 1000) / 1000, |
| 95 | + still_hired=random.random() >= 0.5, |
| 96 | + address=address, |
| 97 | + ) |
| 98 | + await emp.save() |
| 99 | + await Employee._index.refresh() |
| 100 | + |
| 101 | + |
| 102 | +async def search(query: str, limit: int, sort: str) -> None: |
| 103 | + q: ESQLBase = ( |
| 104 | + Employee.esql_from() |
| 105 | + .where(multi_match(query, Employee.first_name, Employee.last_name)) |
| 106 | + .eval(full_name=concat(Employee.first_name, " ", Employee.last_name)) |
| 107 | + ) |
| 108 | + if sort == "height": |
| 109 | + q = q.sort(Employee.height.desc()) |
| 110 | + else: |
| 111 | + q = q.sort(Employee.last_name.asc()) |
| 112 | + q = q.limit(limit) |
| 113 | + async for result in Employee.esql_execute(q, return_additional=True): |
| 114 | + assert type(result) == tuple |
| 115 | + employee = result[0] |
| 116 | + full_name = result[1]["full_name"] |
| 117 | + print( |
| 118 | + f"{full_name:<20}", |
| 119 | + f"{'Hired' if employee.still_hired else 'Not hired':<10}", |
| 120 | + f"{employee.height:5.2f}m", |
| 121 | + f"{employee.address.city:<20}", |
| 122 | + ) |
| 123 | + |
| 124 | + |
| 125 | +def parse_args() -> argparse.Namespace: |
| 126 | + parser = argparse.ArgumentParser(description="Employee ES|QL example") |
| 127 | + parser.add_argument( |
| 128 | + "--recreate-index", |
| 129 | + "-r", |
| 130 | + action="store_true", |
| 131 | + help="Recreate and populate the index", |
| 132 | + ) |
| 133 | + parser.add_argument( |
| 134 | + "--limit", |
| 135 | + action="store", |
| 136 | + type=int, |
| 137 | + default=100, |
| 138 | + help="Maximum number or employees to return (default: 100)", |
| 139 | + ) |
| 140 | + parser.add_argument( |
| 141 | + "--sort", |
| 142 | + action="store", |
| 143 | + type=str, |
| 144 | + default="name", |
| 145 | + help='Sort by "name" (ascending) or by "height" (descending) (default: name)', |
| 146 | + ) |
| 147 | + parser.add_argument( |
| 148 | + "query", action="store", help="The name or partial name to search for" |
| 149 | + ) |
| 150 | + return parser.parse_args() |
| 151 | + |
| 152 | + |
| 153 | +async def main() -> None: |
| 154 | + args = parse_args() |
| 155 | + |
| 156 | + # initiate the default connection to elasticsearch |
| 157 | + async_connections.create_connection(hosts=[os.environ["ELASTICSEARCH_URL"]]) |
| 158 | + |
| 159 | + if args.recreate_index or not await Employee._index.exists(): |
| 160 | + await create() |
| 161 | + await Employee.init() |
| 162 | + |
| 163 | + await search(args.query, args.limit, args.sort) |
| 164 | + |
| 165 | + # close the connection |
| 166 | + await async_connections.get_connection().close() |
| 167 | + |
| 168 | + |
| 169 | +if __name__ == "__main__": |
| 170 | + asyncio.run(main()) |
0 commit comments