sqlite-cloudflare-d1
v0.0.2
Published
Ergonomic interface for writing sqlite queries against cloudflare d1
Downloads
4
Readme
sqlite-cloudflare-d1
This library provides a minimal and flexible interface for cloudflare d1.
The CURD
operations are supported.
insert
Inserts a single row into and returns the inserted row.
Example:
insert(db, {
into: "users",
data: { name: "bob", age: 30 }
});
// { id: 1, name: "bob", age: 30 }
query
Query database for one or more rows. See the last example on how to join tables.
Example 1: Returns every row in albums table.
await query(db, { from: "albums" });
// [{id: 1, title: "Carry On"}, ...]
Example 2: Returns every album that starts with "Carry".
await query(db, {
from: "albums",
where: {
"Title LIKE ?": "Carry%",
},
});
// [{id: 1, title: "Carry On"}]
Example 3: A more complex query. Note that in "where" clauses objects mean "AND" and arrays mean "OR".
await query(db, {
from: "albums",
where: [
{ Title: "Transmission" },
{
"Title LIKE ?": "%On",
"ArtistId > ?": 57,
},
],
});
// This would effectively run the following query:
// "SELECT * FROM albums WHERE (Title = ?) OR (Title LIKE ? AND ArtisID > ?)"
Example 4: A join example with select and group by.
await query(db, {
select: {
"a.Name": "Name",
"count(*)": "count",
},
from: "artists a join albums b on a.ArtistId = b.ArtistId",
group_by: "a.ArtistId",
having: { "count > ?": 12 },
});
update
Update the matching rows to the given values. Similar to other functions, update returns the affected rows.
Example:
await update(db, {
table: "albums",
where: { title: "Carry On" },
set: { title: "Carry On!" },
});
// [{id: 1, title: "Carry On!"}]
remove
Remove one or more rows from a table, and returns the deleted rows.
Example 1: Deletes every row that matches the condition.
await remove(db, {
from: "albums",
where: { title: "Carry On" },
});
// [{id: 1, title: "Carry On"}]
More examples of the "where:" clause
Example 1: The keys are expressions and they reference the values by "?". Object keys are "AND"ed together.
query({ ...,
where: {
name: "bob",
"age > ?": 31,
}
})
// { sql: "name = ? AND age > ?", values: ["bob", 31] }
Example 2: Arrays represent an "OR".
query({ ...
[
{
"name LIKE ?": "Frank%",
"age > ?": 31,
},
{ height: 181 },
]
})
// { sql: "(name = ? AND age > ?) OR (height = ?)", values: ["Frank%", 31, 181] }
Note that height: 31
is a shorthand for "height = ?": 31
.