Request timeout still triggered after an abandon, causing an uncaughtException

[glickel]

Hello,
I have an uncaughtException if I'm using the abandon with a client configured with a timeout.

Could you confirm if it's a bug or if it's just me wrongly using the client please?
And if it's a bug, could you fix it please?

Here is a an example:

const { createClient, EqualityFilter } = require('ldapjs');

const client = new createClient({
    url: '...',
    timeout: 2000,
});

process.on('uncaughtException', (error) => {
    console.log('Uncaught exception happened', error);

    client.unbind();
});

client.once('connect', () => {
    client.bind('...', '...', (bindError) => {
        if (bindError) {
            console.error('bind error', bindError);
            return;
        }

        client.search(
            '...',
            {
                attributes: ['member'],
                filter: new EqualityFilter({ attribute: 'objectClass', value: 'groupOfNames' }),
                sizeLimit: 100000,
            },
            (searchError, result) => {
                if (searchError) {
                    console.error('search error', searchError);
                    return;
                }

                let messageId;

                result.on('error', (listenerError) => {
                    console.error('listener error', listenerError);
                });

                result.on('searchRequest', (message) => {
                    messageId = message.messageId;
                });

                setTimeout(() => {
                    if (messageId !== undefined) {
                        client.abandon(messageId, () => {
                            console.log('Abandon done', messageId);
                        });
                    }
                }, 100);
            },
        );
    });
});

client.on('error', (error) => {
    console.error('client error', error);
});

It's causing this output:

Abandon done 2
Uncaught exception happened TypeError: Cannot destructure property 'callback' of 'tracker.fetch(...)' as it is null.
    at Timeout.onRequestTimeout [as _onTimeout] (/some/anonymized/path/node_modules/ldapjs/lib/client/client.js:1280:23)
    at listOnTimeout (node:internal/timers:569:17)
    at process.processTimers (node:internal/timers:512:7)

[jsumners]

It looks to me like you are attempting to send an abandon message for a request that has already been completed. Would you like to send a pull request to address this issue? Please remember to include unit tests.