test: fix path to module for repl test on Windows

[mcornac]

Use path join to construct the path instead of concatenating strings.
Replace backslash with double backslash so that they are escaped
correctly in the string passed to REPL.

[john-yan]

This is to fix the following message on windows:

not ok 8 test.js
#
#assert.js:89
# throw new assert.AssertionError({
# ^
#AssertionError: false == true
# at process.<anonymous> (D:\build\jenkins-new\workspace\ibmnode-v4-test-win\SUITE\node\arch\x64\label\bvt\os\win\node\test\addons\repl-domain-abort\test.js:11:3)
# at emitOne (events.js:82:20)
# at process.emit (events.js:169:7)
---
duration_ms: 0.499
...

[john-yan]

Can someone add a 'lts-v4.x-watch' tag please?

[jasnell]

@john-yan ... added... likely won't go in until after v4.2.2 tho

[jasnell]

PR LGTM

[john-yan]

@jasnell Thanks for letting me know.

[thefourtheye]

Please use common.isWindows

[mcornac]

Updated, thanks

[thefourtheye]

Windows accepts slash separated paths as well, right?

[mcornac]

@thefourtheye
I tried replacing with forward slash now that I have access to the Windows machine again using:
if (common.isWindows) buildPath = buildPath.replace(//g, '/');
The test passes like this as well.

The reason the original test did not pass is that it has a mixture of (not escaped) backslashes from __dirname with hard coded forward slashes. I.e. "D:\node-1\test\addons\repl-domain-abort/build/Release/binding".

Should I make the change from '\' to '/' ?

[mcornac]

Can someone give a suggestion or merge this in please?

[cjihrig]

Can you split this across two lines please.

[bnoordhuis]

LGTM although I would have replaced them with forward slashes. It should work the same and is arguably easier to read / less ambiguous.

[cjihrig]

Can't really merge this just yet, as the CI is under repairs, but LGTM.

[mcornac]

I split the if statement to two lines and replaced with forward slash.

[bnoordhuis]

Comment is wrong now. Don't worry, I'll update it when I land it. :-)

[mcornac]

Oops, thanks!

[bnoordhuis]

LGTM with a comment. CI: https://ci.nodejs.org/job/node-test-pull-request/694/

[cjihrig]

CI: https://ci.nodejs.org/job/node-test-pull-request/706/

[john-yan]

@thinkingdust Could you please take a look at the failures on the CI and see if any failure are related.

[mcornac]

I took a look. I'm not sure if I'm parsing the output right but I can't see anything related. Many of them look like this one on Windows with a Java exception.
https://ci.nodejs.org/job/node-test-binary-windows/RUN_SUBSET=1,VS_VERSION=vs2015,label=win10/249/console
Caused by: java.io.IOException: Connection timed out

[cjihrig]

Yea, those are issues with the CI itself. Will try running it again tomorrow, if no one else does by then.

[cjihrig]

Let's see how the CI is feeling today: https://ci.nodejs.org/job/node-test-pull-request/723/

[joaocgreis]

@cjihrig CI was bad, but should be better now: https://ci.nodejs.org/job/node-test-pull-request/729/

[cjihrig]

Thanks @joaocgreis.

This is failing for me locally on OS X:

$ ./node test/addons/repl-domain-abort/test.js assert.js:89
throw new assert.AssertionError({
^
AssertionError: false == true
at process.<anonymous> (/Users/cjihrig/iojs/node/test/addons/repl-domain-abort/test.js:15:3)
at emitOne (events.js:82:20)
at process.emit (events.js:169:7)

[jasnell]

New CI Run: https://ci.nodejs.org/job/node-test-pull-request/737/

[jasnell]

CI is mostly green except for the known CI failures. @cjihrig it appears to be running fine for me on OS X and is green on OS X in CI (https://ci.nodejs.org/job/node-test-commit-osx/1186/). Can you run the test again and verify if you're still seeing errors?

[cjihrig]

If the CI is happy, don't let me hold this back. LGTM

[jasnell]

Landed in c0bac95

[MylesBorins]

landed in v4.x-staging as c10f17f