As an alternative to compiling C extensions like extension-functions.c and sqlean into mattn/go-sqlite3, this package implements many of these functions (and more from PostgreSQL) in Go.
These are in addition to all builtin functions provided by SQLite.
Continue reading for all functions, notes and examples.
This library is used in DataStation and dsq to simplify and power data analysis in SQL.
Read the DataStation blog post to better understand the background.
package main
import (
"fmt"
"database/sql"
_ "github.com/mattn/go-sqlite3"
stdlib "github.com/multiprocessio/go-sqlite3-stdlib"
)
func main() {
stdlib.Register("sqlite3_ext")
db, err := sql.Open("sqlite3_ext", ":memory:")
if err != nil {
panic(err)
}
var s string
err = db.QueryRow("SELECT repeat('x', 2)").Scan(&s)
if err != nil {
panic(err)
}
fmt.Println(s)
}
Alternatively if you want to be able to add your own additional
extensions you can just use the ConnectHook
:
package main
import (
"database/sql"
"fmt"
sqlite3 "github.com/mattn/go-sqlite3"
stdlib "github.com/multiprocessio/go-sqlite3-stdlib"
)
func main() {
sql.Register("sqlite3_ext",
&sqlite3.SQLiteDriver{
ConnectHook: stdlib.ConnectHook,
})
db, err := sql.Open("sqlite3_ext", ":memory:")
if err != nil {
panic(err)
}
var s string
err = db.QueryRow("SELECT repeat('x', 2)").Scan(&s)
if err != nil {
panic(err)
}
fmt.Println(s)
}
Name(s) | Notes | Example |
---|---|---|
repeat, replicate | repeat('f', 5) = 'fffff' |
|
strpos, charindex | 0-indexed position of substring in string | strpos('abc', 'b') = 1 |
reverse | reverse('abc') = 'cba' |
|
lpad | Omit the third argument to default to padding with spaces | lpad('22', 3, '0') = '022' |
rpad | Omit the third argument to default to padding with spaces | rpad('22', 3, '0') = '220' |
len | Shorthand for length |
len('my string') = '9' |
split_part | Split string an take nth split piece | split('1,2,3', ',', 0) = '1' , split('1,2,3', ',' -1) = '3' |
regexp | Go's regexp package, not PCRE | x REGEXP '[a-z]+$' , REGEXP('[a-z]+$', x) |
regexp_count | Number of times the regexp matches in string | regexp_count('abc1', '[a-z]1') = '1' |
regexp_split_part | Regexp equivalent of split_part |
regexp_split_part('ab12', '[a-z]1', 0) = 'a' |
Most of these are implemented as bindings to gonum.
Name(s) | Notes | Example |
---|---|---|
stddev, stdev, stddev_pop | stddev(n) |
|
mode | mode(n) |
|
median | median(n) |
|
percentile, perc | Discrete | perc(response_time, 95) |
percentile_25, perc_25, percentile_50, perc_50, percentile_75, perc_75, percentile_90, perc_90, percentile_95, perc_95, percentile_99, perc_99 | Discrete | perc_99(response_time) |
percentile_cont, perc_cont | Continuous | perc_cont(response_time, 95) |
percentile_cont_25, perc_cont_25, percentile_cont_50, perc_cont_50, percentile_cont_75, perc_cont_75, percentile_cont_90, perc_cont_90, percentile_cont_95, perc_cont_95, percentile_cont_99, perc_cont_99 | Continuous | perc_cont_99(response_time) |
Name(s) | Notes | Example |
---|---|---|
url_scheme | url_scheme('https://x.com:90/home.html') = 'https' |
|
url_host | url_host('https://x.com:90/home.html') = 'x.com:90' |
|
url_port | url_port('https://x.com:90/home.html') = '90' |
|
url_path | url_path('https://x.com/some/path.html?p=123') = '/some/path.html' |
|
url_param | url_param('https://x.com/home.html?p=123&z=%5B1%2C2%5D#section-1', 'z') = '[1,2]' |
|
url_fragment | url_fragment('https://x.com/home.html?p=123&z=%5B1%2C2%5D#section-1') = 'section-1' |
Best effort family of date parsing (uses dateparse) and date part retrieval. Results will differ depending on your computer's timezone.
Name(s) | Notes | Example |
---|---|---|
date_year | date_year('2021-04-05') = 2021 |
|
date_month | January is 1, not 0 | date_month('May 6, 2021') = 5 |
date_day | date_day('May 6, 2021') = 6 |
|
date_yearday | Day offset in year | date_yearday('May 6, 2021') = 127 |
date_hour | 24-hour | date_hour('May 6, 2021 4:50 PM') = 16 |
date_minute | date_minute('May 6, 2021 4:50') = 50 |
|
date_second | date_second('May 6, 2021 4:50:20') = 20 |
|
date_unix | date_unix('May 6, 2021 4:50:20') = 1588740620 |
|
date_rfc3339 | date_rfc3339('May 6, 2021 4:50:20') = 2020-05-06T04:50:20Z |
Name(s) | Notes | Example |
---|---|---|
acos | acos(n) |
|
acosh | acosh(n) |
|
asin | asin(n) |
|
asinh | asinh(n) |
|
atan | atan(n) |
|
atanh | atanh(n) |
|
ceil, ceiling | ceil(n) |
|
cos | ceil(n) |
|
cosh | cosh(n) |
|
degrees | degrees(radians) |
|
exp | e^n | exp(n) |
floor | floor(n) |
|
ln, log | log(x) |
|
log10 | log10(x) |
|
log2 | log2(x) |
|
mod | mod(num, denom) |
|
pi | pi() |
|
pow, power | pow(base, exp) |
|
radians | radians(degrees) |
|
sin | sin(n) |
|
sinh | sinh(n) |
|
sqrt | sqrt(n) |
|
tan | tan(n) |
|
tanh | tanh(n) |
|
trunc, truncate | Rounds up to zero if negative, down to zero if positive. | trunc(-10.9) = -10 , trunc(10.4) = 10.0 |
Name(s) | Notes | Example |
---|---|---|
base64 | Convert string to base64 | base64(s) |
from_base64 | Convert string from base64 | from_base64(s) |
base32 | Convert string to base32 | base32(s) |
from_base32 | Convert string from base32 | from_base32(s) |
md5 | Hex md5 sum of string | md5(s) |
sha1 | Hex sha1 sum of string | sha1(s) |
sha256 | Hex sha256 sum of string | sha256(s) |
sha512 | Hex sha512 sum of string | sha512(s) |
sha3_256 | Hex sha3_256 sum of string | sha3_256(s) |
sha3_512 | Hex sha3_512 sum of string | sha3_512(s) |
blake2b_256 | Hex blake2b_256 sum of string | blake2b_256(s) |
blake2b_512 | Hex blake2b_512 sum of string | blake2b_512(s) |
There is 95% test coverage and automated tests on Windows, macOS and Linux.
See dsq (a command-line tool for executing SQL on data files) and DataStation, a GUI application for querying and building reports with data from databases, servers, and files.
Join the #dev channel on the Multiprocess Labs Discord.
If you have an idea for a new function, say so on the Discord channel or open an issue here.
Make sure the function doesn't already exist in dsq (or the sqlite3 CLI).
This software is licensed under an Apache 2.0 license.