FYI @dariowrk @shakaran @BlackthornYugen @armellarcier

I'm following your commits. Looks great, thanks!

You removed a "collection not found" error handling, is it handled somewhere else?

The exsitence check of a collection is here:

/lib/router.js

    if (!mongo.collections[req.dbName].includes(id)) {
      req.session.error = 'Collection not found!';
      return res.redirect(res.locals.baseHref + 'db/' + req.dbName);
    }

I think that if a collection exists, it also have the stats and, for a collection not found, there's a redirect in router (/lib/routes/collection.js isn't called).

Hi rtritto!

Thanks for all the fixes these last days. Amazing work, really appreciate it! Everything seems to work now. (Opening documents with Base64 BinaryIDs, Opening collections in Mongo 4.X and Opening rather large collections)

Regarding:
As default, use find instead of aggregate (fix #1086).

This works, i can open the large collections now. It's just slow as heck, i remember with a version of mongo-express a few months back (1.0.0-alpha4 IIRC) it was rather quick.

To open or change page in a collection of 5.666.031 documents it takes about 50 seconds in the most recent build.

@dariowrk created #1090

This works, i can open the large collections now. It's just slow as heck, i remember with a version of mongo-express a few months back (1.0.0-alpha4 IIRC) it was rather quick.

@dariowrk the 1.0.0-alpha4 use the count that is deprecated.
It's replaced with countDocuments and estimatedDocumentCount (source).

Under the hood, count uses find while countDocuments uses aggregate that is slower.