This library is a wrapper around ldapjs providing convenience with a few core principles:
- Focus on promises and async iterators, do away with callbacks and event-emitting streams
- Always use a connection pool
- Hide everything having to do with acquiring/releasing connections
- Provide an easy way to configure with environment variables
- Provide lots of convenience methods and restructure return data for easier lookups
You WILL have to alter your usage after upgrading to 2.0. The major breaking change is the return object from search queries. See the section below titled "Return Object".
An Ldap instance represents a connection pool. You will want to make a single pool and export it so that it can be imported in any other code file in your project.
import Ldap from 'ldap-async'
export const ldap = new Ldap({
// either
url: 'ldap://yourhost:10389',
// or
host: 'yourhost',
port: 10389,
secure: false,
// optional pool size (default is 5 simultaneous connections)
poolSize: 5,
// then your login and password
bindDN: 'cn=root',
bindCredentials: 'secret',
// and any other options supported by ldapjs
timeout: 30000
})
async function main() {
const person = await ldap.get('cn=you,ou=people,dc=yourdomain,dc=com')
}
main().catch(e => console.error(e))
When working in docker, it's common to keep configuration in environment variables. In order to make that easy, this library provides a convenient way to import a singleton pool created with the following environment variables:
LDAP_HOST
LDAP_PORT // default is 389 or 636 if you set LDAP_SECURE
LDAP_SECURE // set truthy to use ldaps protocol
LDAP_DN // the DN with which to bind
LDAP_PASS // the password for the bind DN
LDAP_POOLSIZE (default: 5)
This way, connecting is very simple, and you don't have to worry about creating a singleton pool for the rest of your codebase to import, because it's done for you:
import ldap from 'ldap-async/client'
async function main() {
const person = await ldap.get('cn=you,ou=people,dc=yourdomain,dc=com')
}
main().catch(e => console.error(e))
You must refer to .default
when importing with require
:
const Ldap = require('ldap-async').default
// or the instance created with environment variables (see above)
const ldap = require('ldap-async/client').default
Convenience methods are provided that allow you to specify the kind of operation you are about to do and the type of return data you expect.
const person = await ldap.get('cn=you,ou=people,dc=yourdomain,dc=com')
console.log(person.toJSON()) // { givenName: 'John', ... }
const people = await ldap.search('ou=people,dc=yourdomain,dc=com', { scope: 'sub', filter: 'objectclass=person' })
console.log(people.map(p => p.toJSON())) // [{ givenName: 'John', ... }, { givenName: 'Mary', ... }]
// return full LdapEntry objects for members of a group and members of all subgroups
const people = await ldap.getMembers('cn=yourgroup,ou=groups,dc=yourdomain,dc=com')
console.log(people.map(p => p.toJSON())) // [{ givenName: 'John', ... }, { givenName: 'Mary', ... }]
In ldap-async v2.0 the return object changed to give you greater control over the return type you want/expect. Now you get a special LdapEntry class with methods for getting the entry attributes:
const entry = await ldap.get(... whatever ...)
entry.get('givenName') // 'John' - you can also use entry.one('givenName') or entry.first('givenName')
entry.all('givenName') // ['John']
entry.buffer('givenName') // Buffer.from('John', 'utf8')
entry.buffers('givenName') // [Buffer.from('John', 'utf8')]
If you want something more like the ldap-async v1.0 return object, use the .toJSON()
method. You'll
get back an object with attribute names as the keys and the values will be a mixture of string and
string[]. Attributes with only one value will be string
, attributes with multiple values will
be string[]
. Attributes with at least one value that is not valid UTF-8 (usually binaries
like image data) will be base64 encoded strings.
// change the value of a single attribute on a record
await ldap.setAttribute('cn=you,ou=people,dc=yourdomain,dc=com', 'email', 'newemail@company.com')
// change the value of multiple attributes in one round trip
await ldap.setAttributes('cn=you,ou=people,dc=yourdomain,dc=com', { email: 'newemail@company.com', sn: 'Smith' })
// pushes value onto an array attribute unless it's already there
await ldap.pushAttribute('cn=you,ou=people,dc=yourdomain,dc=com', 'email', 'newemail@company.com')
// remove a value from an array attribute (returns true without doing anything if value wasn't there)
await ldap.pullAttribute('cn=you,ou=people,dc=yourdomain,dc=com', 'email', ['newemail@company.com'])
// remove an attribute entirely
await ldap.removeAttribute('cn=you,ou=people,dc=yourdomain,dc=com', 'customAttr')
// add a full record
await ldap.add('cn=you,ou=people,dc=yourdomain,dc=com', { /* a person record */ })
// remove a full record
await ldap.remove('cn=you,ou=people,dc=yourdomain,dc=com')
// rename a record (in this example only the cn changes, the ou,dc entries are preserved)
await ldap.modifyDN('cn=you,ou=people,dc=yourdomain,dc=com', 'cn=yourself')
// special group membership functions
await ldap.addMember('cn=you,ou=people,dc=yourdomain,dc=com', 'cn=yourgroup,ou=groups,dc=yourdomain,dc=com')
await ldap.removeMember('cn=you,ou=people,dc=yourdomain,dc=com', 'cn=yourgroup,ou=groups,dc=yourdomain,dc=com')
When you construct LDAP search query strings, it's important to escape any input strings to prevent injection attacks. LDAP has two kinds of strings with different escaping requirements, so we provide a template literal helper for each.
For DN strings, use ldap.dn
:
const person = await ldap.get(ldap.dn`cn=${myCN},ou=people,dc=yourdomain,dc=com`)
For filter strings, use ldap.filter
:
const people = await ldap.search('ou=people,dc=yourdomain,dc=com', {
scope: 'sub',
filter: ldap.filter`givenName=${n}`
})
More complex queries may also use ldap.filter
inside a map function, such as this one that finds many users by their names:
const people = await ldap.search('ou=people,dc=yourdomain,dc=com', {
scope: 'sub',
filter: `(|${myNames.map(n => ldap.filter`(givenName=${n})`).join('')})`
})
For convenience, a few helper functions are provided to help you construct LDAP filters: in
, any
, all
, and anyall
. These functions take care of escaping for you.
- Everyone named John or Mary:
ldap.in(['John', 'Mary'], 'givenName') // => '(|(givenName=John)(givenName=Mary))
- Everyone named John or with the surname Smith
ldap.any({ givenName: 'John', sn: 'Smith' }) // => '(|(givenName=John)(sn=Smith))
- Everyone named John Smith
ldap.all({ givenName: 'John', sn: 'Smith' }) // => '(&(givenName=John)(sn=Smith))
- Everyone named John Smith or Mary Scott
ldap.anyall([{ givenName: 'John', sn: 'Smith' }, { givenName: 'Mary', sn: 'Scott' }]) // => '(|(&(givenName=John)(sn=Smith))(&(givenName=Mary)(sn=Scott)))'
Note that any
, all
and anyall
can accept an optional wildcard
parameter if you want users to be able to provide wildcards. Other special characters like parentheses will be properly escaped.
- Everyone named John whose surname starts with S
ldap.all({ givenName: 'John', sn: 'S*' }, true) // => '(&(givenName=John)(sn=S*))
ldapjs added a "Filters API" in their 3.0 version that helps you create (and parse) filters. You're free to use that instead of the filter helpers provided by ldap-async. Just make a filter object with their Filters API and give it to any appropriate method in ldap-async:
const people = await ldap.search('ou=people,dc=yourdomain,dc=com', {
scope: 'sub',
filter: new EqualityFilter({ attribute: 'givenName', value: n })
})
For more information, see the Filters API documentation.
To avoid using too much memory on huge datasets, we provide a stream
method that performs the same as search
but returns a node Readable
. It is recommended to use the async iterator pattern:
const stream = ldap.stream('ou=people,dc=yourdomain,dc=com', {
scope: 'sub',
filter: ldap.in(myNames, 'givenName')
})
for await (const person of stream) { /* do some work on the person */ }
for await
is very safe, as break
ing the loop or throwing an error inside the loop will clean up the stream appropriately.
Since .stream()
returns a Readable
in object mode, you can easily do other things with
it like .pipe()
it to another stream processor. When using the stream without for await
, you must call stream.destroy()
if you do not want to finish processing it and carefully use try {} finally {}
to destroy it in case your code throws an error. Failure to do so will leak a connection from the pool.
There is also a getMemberStream
function which performs the same as getMembers
:
const people = await ldap.getMemberStream('cn=yourgroup,ou=groups,dc=yourdomain,dc=com')
for await (const p of people) { /* do some work on the person */ }
Some LDAP services store binary data as properties of records (e.g. user profile photos). In ldap-async v1.0,
we provided a _raw
property to work around this, but in v2.0 we support it with the new LdapEntry
return
object. So now you simply have to ask for the Buffer
for the attribute in question.
For example, to convert profile photos to data URLs, you could do something like this:
const user = await ldap.get(userDn)
const convertedUser = {
...user.toJSON(),
jpegPhoto: `data:image/jpeg;base64,${user.buffer('jpegPhoto').toString('base64')}`,
}
Generally you want to let the pool do its thing for the entire life of your process, but if you are sure you're done with it, you can call await client.close()
and it will wait for all existing requests to finish, then empty the pool so that everything can be garbage collected. The pool is still valid, so if you make another request, the pool will open back up and work normally.
This library is written in typescript and provides its own types. For added convenience, methods that return
objects will accept a generic so that you can specify the return type you expect from the .toJSON()
method:
interface LDAPPerson {
cn: string
givenName: string
}
const person = ldap.get<LDAPPerson>(ldap.dn`cn=${myCN},ou=people,dc=yourdomain,dc=com`)
// person.toJSON() will be an LDAPPerson