Buffer Base64 Encoding Truncates or Adds Extra Characters #6107

Activity

1. mscdex

   added

   bufferIssues and PRs related to the buffer subsystem.Issues and PRs related to the buffer subsystem.

   on Apr 7, 2016
2. kzc commented on Apr 7, 2016

   

   kzc

   on Apr 7, 2016

   More actions

   You're converting ill-formed base64 input to base64. Try this instead:

   Buffer(Buffer('hello', 'utf8').toString('base64'), 'base64').toString();
   
3. mscdex

   added

   questionIssues asking questions about Node.js.Issues asking questions about Node.js.

   on Apr 7, 2016
4. mscdex commented on Apr 7, 2016

   

   mscdex

   on Apr 7, 2016

   Contributor

   More actions

   @kzc is right, the second argument passed to Buffer() is the encoding for the string passed in as the first argument. 'helloIsItMeYourLookingFor' is not base64-encoded.
5. mscdex

   added

   invalidIssues and PRs that are invalid.Issues and PRs that are invalid.

   and removed

   questionIssues asking questions about Node.js.Issues asking questions about Node.js.

   on Apr 7, 2016
6. mscdex

   closed this as completedon Apr 7, 2016
7. davepacheco commented on Apr 8, 2016

   

   davepacheco

   on Apr 8, 2016

   Contributor

   More actions

   Shouldn't invalid input produce an error rather than garbage output?
8. addaleax commented on Apr 8, 2016

   

   addaleax

   on Apr 8, 2016

   Member

   More actions

   Base64 comes in various forms, the most common using one or two = signs as a form of final padding:

   Buffer.from([0x01, 0x02, 0x03, 0x04]).toString('base64') === 'AQIDBA=='

   This padding makes a buffer encoded with this variant of Base64 always have a length that is a multiple of 4, which is a nice property to have when writing a decoder for it.
   However, it’s not strictly necessary – there is no additional information in those = characters that wouldn’t also be conveyed by the length of the encoded string.

   So, whether you should consider this input invalid is a question you can probably argue about a lot. I’d say there’s no real downside to accepting it, especially since there are some uncommon Base64 encoders which actually leave out the padding; Wikipedia has a quite comprehensive listing of the different variants out there.