Fix compilation issue under OTP 19

[paulcsmith]

There is an issue with ExMachina when used with OTP 19

== Compilation error on file test/support/test_factory.ex ==
** (ArgumentError) argument error
    (stdlib) erl_anno.erl:318: :erl_anno.set(:file, 'nofile', -1)
    (stdlib) erl_parse.yrl:1516: anonymous fn/3 in :erl_parse.map_anno/2
    (stdlib) erl_parse.yrl:1631: :erl_parse.modify_anno1/3
    (stdlib) erl_parse.yrl:1517: :erl_parse.map_anno/2
    (stdlib) erl_lint.erl:703: :erl_lint."-set_file/2-lc$^0/1-0-"/2
    (stdlib) erl_lint.erl:448: :erl_lint.exprs_opt/3
    (stdlib) erl_eval.erl:173: :erl_eval.check_command/

By removing the lineL -1 the issue goes away. There is also no warning
when doing this so it can be safely removed.

[jsteiner]

LGTM!

[nathany]

Thanks.

We are using line -1 because we don't want warnings coming from

save_record/1 when someone defines there own save_record/1 function.

I'm just curious what the difference in behaviour will be now (I don't quite understand this aspect of Elixir/Erlang).

Also note:
elixir-lang/elixir#4617 (comment)

[paulcsmith]

@nathany It appears there is no difference in behavior now because we no longer define factories with factory instead it's factory_#{name_of_factory}. Also because save_record is now deprecated and we are using ExMachina.Strategy. So now it's just there as a catch all for those upgrading from old versions that may be calling factory(:foo). So once they upgrade to the new factory definition syntax they shouldn't see any warnings. If people do see warnings from this then an issue should be opened, but I don't think they will.

Great question though!

[nathany]

Thanks @paulcsmith. Any plans to backport this fix to 0.6.x? (otherwise I may fork it and do that)

Or is master close to a point where 1.0.0-beta.1 could be uploaded to hex as a pre-release?

[paulcsmith]

I probably won’t back port it since OTP 19 is super new and 1.0 is about to be released to hex :D #139 #139

On Jun 24, 2016, at 2:07 PM, Nathan Youngman [email protected] wrote:

Thanks @paulcsmith https://github.com/paulcsmith. Any plans to backport this fix to 0.6.x? (otherwise I may fork it and do that)

Or is master close to a point where 1.0.0-beta.1 could be uploaded to hex as a pre-release?

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub #138 (comment), or mute the thread https://github.com/notifications/unsubscribe/AABXeslg2fx-4t77F8Qw-tthUUjpEzGeks5qPBz7gaJpZM4I99bQ.