[webkit-reviews] review canceled: [Bug 36983] new-run-webkit-tests ignores trailing EOL differences in text tests : [Attachment 76634] Fix and unit tests

[bugzilla-daemon at webkit.org]

Cosmin Truta <ctruta at chromium.org> has canceled Cosmin Truta
<ctruta at chromium.org>'s request for review:
Bug 36983: new-run-webkit-tests ignores trailing EOL differences in text tests
https://bugs.webkit.org/show_bug.cgi?id=36983

Attachment 76634: Fix and unit tests
https://bugs.webkit.org/attachment.cgi?id=76634&action=review

------- Additional Comments from Cosmin Truta <ctruta at chromium.org>
Resubmitting to comply with Dirk's comments.

(In reply to comment #11)
> 4) What happens if the actual output contains "\r\r\r\n"? It looks like that
gets would get changed to '\r\n'. Is that what we want? I'd like to see a test
with at least one more \r than the number of replace() calls. A test with 10
\r's instead of 3 would be fine as well.

I don't think we "want" \r\r\r\n to change to \r\n. I found no tests to require
such lines.
Instead, I preferred to say that we don't want too many \r...\r\n to be
absorbed into \n, because that might mean something's wrong. So I tested this
as an expected failure, not as a pass.