ATTACH 'md:_share/duckdb_in_action_ch3_4/d0c08584-1d33-491c-8db7-cf9c6910eceb' AS duckdb_in_action_ch3_4;
USE duckdb_in_action_ch3_4;
Copy code

SELECT COLUMNS('valid.*') FROM prices LIMIT 3;
Copy code

SELECT max(COLUMNS('valid.*')) FROM prices;
Copy code

FROM prices
WHERE COLUMNS('valid.*') BETWEEN '2020-01-01' AND '2021-01-01';
Copy code

-- Install httpfs extension
INSTALL httpfs;
LOAD httpfs;
-- Install Scrooge extension https://github.com/pdet/Scrooge-McDuck
-- NOTE: You need to start DuckDB with `-unsigned` flag to authorize to install & load of 3rd party extension
SET custom_extension_repository='scrooge-duckdb.s3.us-west-2.amazonaws.com/scrooge/s3_deploy';
INSTALL scrooge;
LOAD scrooge;
-- Example of query 
FROM yahoo_finance("^GSPC", "2023-02-01", "2023-02-04", "1d");
Copy code

-- Read from a remote CSV file, and write partitioned Parquet files to local target
-- Queries like this are commonly used in Data Lakes
COPY (SELECT cloud_provider, cidr_block, ip_address, ip_address_mask, ip_address_cnt, region from read_csv_auto('https://raw.githubusercontent.com/tobilg/public-cloud-provider-ip-ranges/main/data/providers/all.csv')) TO '/tmp/ip-ranges' (FORMAT PARQUET, PARTITION_BY cloud_provider);
Copy code

duckdb -c "COPY (SELECT * FROM '~/Binance_Spot_Data/*.parquet') TO 'binance.parquet' (FORMAT 'PARQUET', CODEC 'zstd')"
Copy code

PRAGMA default_null_order='NULLS LAST';
Copy code

/* 
Background: If you have S3 Access Logging enabled on one of your S3 buckets, you'll have some useful information about requests to your bucket. Unfortunately, it's in a semistructured format that can be difficult to parse. This SQL query will can help in this manner, both pulling out individual fields and coersing them to native data types.

Usage: you'll want to search for the strings <bucket> and <prefix>, and insert the S3 bucket where your access logs are being delivered. Use (or delete) <prefix> to filter to a subset of your logs.

Also, these commented out configuration settings you can either run  yourself in the REPL and source this file using `.read parse_s3_access_logs.sql`, or you can uncomment them and supply values for yourself.
*/

-- install https;
-- load https;
-- SET s3_region='us-west-2';
-- SET s3_access_key_id='';
-- SET s3_secret_access_key='';

WITH parsed_logs AS (
    SELECT
        regexp_extract(col1, '^([0-9a-zA-Z]+)\s+([a-z0-9.\-]+)\s+\[([0-9/A-Za-z: +]+)\] ([^ ]+) ([^ ]+) ([^ ]+) ([^ ]+) ([^ ]+) ("[^"]*"|-) ([^ ]+) ([^ ]+) (\d+|-) (\d+|-) (\d+|-) (\d+|-) ("[^"]*"|-) ("[^"]*"|-) (\S+) (\S+) (\S+) (\S+) (\S+) (\S+) (\S+) (\S+) (\S+)(.*)$',
        ['bucket_owner', 'bucket', 'timestamp', 'remote_ip', 'request', 'request_id', 'operation', 's3_key', 'request_uri', 'http_status', 's3_errorcode', 'bytes_sent','object_size', 'total_time', 'turn_around_time', 'referer', 'user_agent', 'version_id', 'host_id', 'sigver', 'cyphersuite', 'auth_type', 'host_header', 'tls_version', 'access_point_arn', 'acl_required', 'extra']
    ) AS log_struct
    FROM
        -- Trick the CSV reader into reading as a single column
        read_csv(
            's3://<bucket>/<prefix>/*',
            columns={'col1': 'VARCHAR'},
            -- Use a *hopefully* nonsensical deliminator, so no ',' chars screw us up
            delim='\0'
        )
)
SELECT
        -- Grab everything from the struct that we want as strings, exclude stuff we'll coersce to diff types
        log_struct.* exclude (timestamp, bytes_sent, object_size, total_time, turn_around_time),
        strptime(log_struct.timestamp, '%d/%b/%Y:%H:%M:%S %z') AS timestamp,
        CASE
                WHEN log_struct.bytes_sent = '-' THEN NULL
                ELSE CAST(log_struct.bytes_sent AS INTEGER)
        END AS bytes_sent,
        CASE
                WHEN log_struct.object_size = '-' THEN NULL
                ELSE CAST(log_struct.object_size AS INTEGER)
        END AS object_size,
        CASE
                WHEN log_struct.total_time = '-' THEN NULL
                ELSE CAST(log_struct.total_time AS INTEGER)
        END AS total_time,
        CASE
                WHEN log_struct.turn_around_time = '-' THEN NULL
                ELSE CAST(log_struct.turn_around_time AS INTEGER)
        END AS turn_around_time
FROM parsed_logs;
Copy code

-- Install spatial extension
INSTALL spatial; 
LOAD spatial; 

-- Represent a latitude and longitude as a point
-- The Eiffel Tower in Paris, France has a 
-- latitude of 48.858935 and longitude of 2.293412
-- We can represent this location as a point
SELECT st_point(48.858935, 2.293412) AS Eiffel_Tower;
Copy code

-- Distance between the Eiffel Tower and the Arc de Triomphe in Paris
-- Using the EPSG spatial reference systems: 
-- EPSG:4326 geographic coordinates as latitude and longitude pairs
-- EPSG:27563 projection that covers northern France and uses meters 

SELECT
    st_point(48.858935, 2.293412) AS Eiffel_Tower, 
    st_point(48.873407, 2.295471) AS Arc_de_Triomphe,
    st_distance(
        st_transform(Eiffel_Tower, 'EPSG:4326', 'EPSG:27563'), 
        st_transform(Arc_de_Triomphe, 'EPSG:4326', 'EPSG:27563')
    ) AS Aerial_Distance_M;
Copy code

-- Load the geometry outline for each country
-- Save the country name and "geom" border in table world_boundaries
CREATE OR REPLACE TABLE world_boundaries
AS
SELECT *
FROM st_read('https://public.opendatasoft.com/api/explore/v2.1/catalog/datasets/world-administrative-boundaries/exports/geojson');

-- Find the enclosing country for a given point
-- We can which country the Eiffel Tower is in 
SELECT name, region
FROM world_boundaries
WHERE ST_Within(st_point(2.293412, 48.858935) , geom);
Copy code