Importing test cases from /beatgammit/base64-js/

[prahaladbelavadi]

I'm trying to import test cases from beatgammit/base64-js/

In beatgammit/base64-js:

• big-data:
  • They generate a huge(8641028) uint8array
  • convert that array to base64
  • Convert base64 string back to bytearray and confirm if it matches
• url-safe:
  • converts a base64 string with '+','/','_','-' to a byte array
  • converts the byte array back to base64 string and check if it matches
• corrupt:
  • converts string to bytearray
  • checks for "=" in the bytearray that's not trailing
• Convert:
  • converts base64 string to byte array
  • converts byte array to base 64 string and checks if they are equal
  • uses an array to test different strings

To import the tests to LinusU/fast-base64-decode

• big data:
  • generate a big (8641028) uint8array
  • import a file containing the base64 string from the uint8array (what format, where should it be placed ?)
  • read the file
  • decode the contents using LinusU/fast-base64-decode (takes file content and byte length as params)
  • check if the generated uint8array and the decoded array matches
• url-safe:
  • check if string containing '+','/','_','-' decodes into [0xff, 0xff, 0xbe, 0xff, 0xef, 0xbf, 0xfb, 0xef, 0xff] array
• corrupt:
  • convert a base64 string containing padding bytes to a bytearray
  • iterate through the byte array checking for the "=" character in between and isn't trailing
• convert:
  • already has independent tests to confirm base64 to byte array decoding;
  • does not support encoding; so no tests for encoding

Do correct me if I'm wrong somewhere or missing something out.

I did have a couple of questions with regard to the big data tests;

• The javascript heap stack runs out while trying to operate on a big binary file
• While writing and reading to an external file, what format do I use to store/read the file ?
• where do I place the file ? Would it be okay if I made a folder for the tests and split them into different files ?
• Is importing a binary file to do big data tests a normal thing ? or are there better methods to test it ?
• [corrupt]:
  • how to handle splitting ? there apparently must be some kind of splitting when a padding byte is included in the middle of a string (Bad decoding of concatenated Base64 files beatgammit/base64-js#42)
  • how do we test this ?

Do let me know if I'm heading in the right direction or need to correct something.

Thank you @LinusU

[LinusU]

Sorry for the extremely slow response here. I've merged your PR, and I've added the "big data" test as well: 5cf93b3

Let me know if there is anything more you think we should do! Thank you so much for your contribution 🙏