Mixed front-end lib for auth0 authorization
The hosted Auth0 Lock signup page loads web-assets/auth0/js/util.js, published
as both util.js and the legacy country.js. Its capture-phase submit guard
allows handles from 3 through 15 characters inclusive and blocks out-of-range
handles before account creation. Login and password reset are unaffected. The
separate social-signup page applies the same bounds in web-assets/js/signup.js.
Configure the dev hosted template to load the topcoder-dev.com static host so
dev deployments are available to QA; the production host follows master.
web-assets/js/check_email.js completes account activation through
PUT /v6/users/activate. It displays v6 message strings/validation arrays as
well as legacy result.content errors, using text rather than HTML. Failed
activation keeps the entered code and enables retry; editing the code while a
request is pending cannot send a duplicate activation. Resend failures likewise
show the API error and retain the retry link. A successful response submits the
existing continuation form with the original Auth0 state.
Run npm test with Node 22 for dependency-free regression tests covering handle
boundaries, login/reset isolation, activation errors, retries and continuation.
The repository has no npm lint or build scripts. Validate JavaScript with
node --check; the CI builds are ./build.sh and ./buildstatic.sh. Run builds
in a disposable copy because they replace environment placeholders in source
files and create dist/diststatic directories.
QA should check an overlong handle followed by a corrected 15-character handle, then complete activation with a fresh emailed code. A remaining activation API failure now shows its actual reason (for example an expired or wrong code) instead of "Unknown Error"; the UI does not bypass activation checks.