tests failing, also not all tests are running

[projectfluent]

tests failing, also not all tests are running

[zbraniecki]

The .only has been introduced in #81. With it, I see:

> make test
  59 passing (99ms)
  49 pending

when i remove it I see:

> make test
  196 passing (170ms)
  54 pending

@hkasemir - can you investigate which tests are failing and why? I suspect something is off with your node environment and its different than the one we have. It may be that you need Intl API, or it may be that we need to handle more variants of Node+Intl combinations, but I'd like to start with identifying tests and their failures.

[stasm]

Sorry about the regression introduced in #81. I use .only quite often to isolate a single test during development. I found https://www.npmjs.com/package/eslint-plugin-no-only-tests which would prevent such errors in the future.

[stasm]

There's also https://www.npmjs.com/package/eslint-plugin-mocha which has more configuration options.

[stasm]

I opened #94 to add a linting rule forbidding .only tests.

I have the same output as @zbraniecki before and after fixing that rogue .only in functions_builtin_test.js. I tried on a fresh clone and I was able to set it up by following the instructions in the README. I'm using node 8.9.1.

[hkasemir]

using node v6.11.2 and npm 5.4.1 fresh clone of the repo:

using node v8.9.1, seems all tests are passing 🤔 could be a good idea to define which version of node is required in the readme under instructions for local dev. Glad that was an easy fix. I did take the liberty of checking #93 to make sure it was not fixed by node version, but it is still an issue.

[stasm]

I found out why the tests fail. Node 6.x doesn't support Object.entries which is used here:

fluent.js/fluent/src/builtins.js

Line 29 in da0a580

for (const [name, opt] of Object.entries(opts)) {

Given that node 8.x has already become the active LTS version I think it makes sense to explicitly drop support for 6.x (and 7.x too, even though the tests actually pass on 7.x).

[stasm]

With #94 and #95 merged, this should be good to close. Thanks, @hkasemir, for the report!