Better support for masssive files

[Aditya94A]

There should be an option to simply "upload" (or point to) the input file and download the output file.

Using textareas means that the browser just gets stuck on large-ish files trying to show it all, making everything laggy.

Solution: Don't show text for large files.

[peterdresslar]

The support for massive files is already pretty amazing--I just ran a 75M file and it worked perfectly. The only thing I would change might be mentioning in the docs that the browser may complain a few times that it is frozen--not to worry, the file will process if you press "Wait" a few times.

[gabefair]

I was having problems getting it to work with large files as well, but I was able to determine that the following helped me:

1. Removing "_id":ObjectId("<number>") from the json
   In my case I had a massive mongodb dump from a journalist, and I removed them all using the following regex "_id" : \w+\("\w+"\), \n
2. Make sure the JSON is valid. Online validators can help. I found this one was able to take my 18Mb sized JSON dump I pasted via my clipboard.
3. Convert JSONL to JSON. By adding [ to the front and ] to the back I was able to convert the JSON Streaming (aka LDJSON) to valid JSON

My JSON had 13,122 urls and I was worried that this would break Eric's anonymous API request limit, but surprisingly it did not hit a limit and break.

I used Firefox Developer Edition with the script timeout setting changed to be its maximum value: 2147483647 and was able to download the CSV after leaving it to run for about 15 mins.