Enhancement reassemble chunks

[paulklemm]

In the node.js example, files are now assembled in the uploads folder after the upload is finished. Took me quite a while to figure this out how to do this today, so I would like to improve the example. This is discussed in Issue #17.

[AidasK]

Looks ok, but will it work with simultaneousUploads > 1 ? Is status done returned only one time for files bigger than chunkSize?

[paulklemm]

Is status done returned only one time for files bigger than chunkSize?

Yes, it is only called once. See here the output for a large file: https://gist.github.com/Powernap/bcca6e5a5ffc37cca630.

[...] will it work with simultaneousUploads > 1?

It should work, since the unique identifier is used for assembling the files.

[AidasK]

Great, thanks!
How about updating file download logic too? https://github.com/Powernap/flow.js/blob/3e23096679f8e041024c3056088b5925f7afe115/samples/Node.js/app.js#L61
It should download file from uploads dir instead of merging chunks again.

And maybe we could clean all the chunks after the upload with https://github.com/Powernap/flow.js/blob/3e23096679f8e041024c3056088b5925f7afe115/samples/Node.js/flow-node.js#L180?

[paulklemm]

And maybe we could clean all the chunks after the upload [...]
Done.

I'm with you, the download request should provide the assembled file! I could not find a way to get the filename from the identifier without writing a new method for it in the flow-node.js. What would be the proper way to do get the filename?

[AidasK]

https://github.com/Powernap/flow.js/blob/3e23096679f8e041024c3056088b5925f7afe115/samples/Node.js/app.js#L23
Maybe we could save file by its ID and not by the file name?

[AidasK]

For file download we could use modified route with given file name:

app.get('/download/:identifier/:name', function(req, res) {
   res.download('uploads/' +req.params.identifier, req.params.name);
});

Downloaded file will have its original name.
http://expressjs.com/api.html#res.download

[paulklemm]

You are right, saving the file by ID makes much more sense. I will update the pull request.

[AidasK]

How it is going with a pull request?

[paulklemm]

Sorry, my Job and other responsibilities keep me busy, I did not forget it! I will push the request asap!

[AidasK]

Glad to hear that you are on to it, but no need to hurry, pull request can wait, take your time. Thanks!

[weiway]

How it going?

[paulklemm]

So sorry for the delay. @AidasK Is this pull request still relevant? If so, I'll be happy to finish it asap.

[AidasK]

Please do, thanks

[aidenyang]

Any intention on merging this pull request? Just checking in since it looks like it's good to go; just hasn't been merged yet.

[paulklemm]

Apparently the only change missing is saving the file by ID, other than that it should be ready. Sorry for taking so much time. Should be easy to do, but it is quite a while ago since I touched that code. I will look into it this week.

[paulklemm]

Sorry for the delay. The files are now saved using their ID.

@AidasK The download routine does not work. It provides the proper path and file ID but the download fails. Do you have time to have a look at it?