fix(conversion): prevent Base64Decode panic and Base64Encode input mutation

[SAY-5]

Description of Change

Base64Decode indexes input[i+1]/input[i+2]/input[i+3] for every step of 4, so any input whose length is not a multiple of 4 panics with index out of range instead of returning. The decoder now iterates over whole 4-character groups only, so a malformed-length payload is ignored at the tail rather than reading past the end of the string.

Base64Encode padded short final groups by appending zero bytes to the input slice. When the caller passes a slice with spare capacity (e.g. Base64Encode(buf[:n])), append writes those zeros into the caller's backing array and corrupts data observed through other slice headers. The padding bytes are now read locally, so the argument is left untouched.

Added regression tests for both: decoding inputs of length 1, 2, 3 and 5, and encoding a sub-slice whose backing array must remain unchanged.

Closes #807

Checklist

• Added description of change
• Added file name matches File name guidelines
• Added tests and example, test must pass
• Added documentation so that the program is self-explanatory and educational - GoDoc guidelines
• Relevant documentation/comments is changed or added
• PR title follows semantic commit guidelines
• Search previous suggestions before making a new one, as yours may be a duplicate.
• I acknowledge that all my contributions will be made under the project's license.

Notes: Guard base64 decode/encode against malformed-length input and caller-slice mutation.

[codacy-production]

Not up to standards ⛔

_{NEW Get contextual insights on your PRs based on Codacy's metrics, along with PR and Jira context, without leaving GitHub. Enable AI reviewer
TIP This summary will be updated as you push new changes.}