fix(paypal): handle octet-stream NVP responses

[StaberindeZA]

Because

• PayPal's NVP endpoint intermittently returns application/octet-stream instead of text/plain. Superagent does not buffer unknown content types by default, leaving result.text and result.body empty and silently breaking every downstream NVP parse.

This pull request

• Adds .buffer(true) so superagent always reads the response body.
• Reads from result.text when populated, falling back to decoding result.body as utf-8 when it arrives as a Buffer.

Issue that this pull request solves

Closes: PAY-3709

Checklist

Put an x in the boxes that apply

• My commit is GPG signed.
• If applicable, I have modified or added tests which pass locally.
• I have added necessary documentation (if appropriate).
• I have verified that my changes render correctly in RTL (if appropriate).
• I have manually reviewed all AI generated code.

How to review (Optional)

• Key files/areas to focus on:
• Suggested review order:
• Risky or complex parts:

Screenshots (Optional)

Please attach the screenshots of the changes made in case of change in user interface.

Other information (Optional)

Any other information that is important to this pull request.

[Copilot]

Pull request overview

This PR hardens the PayPal NVP client against intermittent application/octet-stream responses by forcing Superagent to buffer responses and by decoding the raw response from either result.text or a UTF-8 stringified Buffer.

Changes:

• Force Superagent to always buffer the response body via .buffer(true).
• Parse NVP responses from result.text, falling back to decoding result.body as UTF-8 when it’s a Buffer.
• Reuse the computed raw payload string for downstream error construction.

Comments suppressed due to low confidence (1)

libs/payments/paypal/src/lib/paypal.client.ts:130

• raw uses nullish coalescing (result.text ?? ...). If result.text is an empty string (which is plausible in the reported failure mode), this will select '' and skip decoding result.body even when it contains the response bytes, leading to parsing an empty payload. Consider treating empty/whitespace result.text as absent (e.g., check result.text?.length > 0) before falling back to result.body.

    const raw =
      result.text ??
      (Buffer.isBuffer(result.body) ? result.body.toString('utf8') : '');

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[david1alvarez]

r+wc, copilot's suggestion to add a unit test would be great to implement