Weird errors with ddgc_populate_ia_dev

[GuiltyDolphin]

@jbarrett I was trying to run the ddgc_populate_ia_dev.pl script to get up-to-date with duck.co (as suggested by @jdorweiler), but I kept running into this issue:

It doesn't appear to be dependency related - I've updated fairly recently, and the script is definitely getting part-way through (I'm seeing a few IAs pop up locally - as well as having put some printing in there to confirm) - but it seems to then just give up!

Any ideas?

[jdorweiler]

I'm seeing a similar error when I run it.

Use of uninitialized value in concatenation (.) or string at /home/ddgc/community-platform/script/../lib/DDGC/DB/Base/Result.pm line 280.
SQL::Abstract::puke(): [SQL::Abstract::__ANON__] Fatal: Operator calls in update must be in the form { -op => $arg } at /home/ddgc/community-platform/script/../lib/DDGC/DB/Result/InstantAnswer.pm line 549
DBIx::Class::Storage::TxnScopeGuard::DESTROY(): A DBIx::Class::Storage::TxnScopeGuard went out of scope without explicit commit or error. Rolling back. at /opt/perlbrew/perls/perl-5.16.3/lib/site_perl/5.16.3/DBIx/Class/Exception.pm line 77

[GuiltyDolphin]

@jbarrett @jdorweiler I think it is to do with the maintainer field.

I checked the $data->{$ia_id} hash, and for the ones that were getting through the maintainer field was undefined, but it failed whenever one came up in the form:

maintainer => {
    duckco => ...,
    github => ...
}

I set the maintainer field to undef for each of them, and it ran much more smoothly (albeit without any maintainers). I guess that explains why it was complaining about a hash!

[MariagraziaAlastra]

Fixed on #1284. Thanks @GuiltyDolphin for the bug report!

[GuiltyDolphin]

@MariagraziaAlastra @jbarrett Just tried it (#1284) out locally and I'm getting a different error now:

[kentfredric]

When you have errors like this, use some tool like Devel::Confess to get a more comprehensive stack trace, including context and arguments.

Also, don't screenshot errors, that's the least useful way you could ever report text-based error.

Copy paste the whole text, and put it in a block with

``` 
text here
```

Then you might be able to work out what's going on. ( And people who have similar errors will be able to search for them, because search can't index the text content of images... )

[GuiltyDolphin]

@jbarrett @jdorweiler I've just re-run it again and it went all the way without crashing (there were a bunch of concatenation errors though - but nothing fatal). It's now got ~ 800 live IAs showing 😀

I'll let you know if the errors start popping up again.

[jbarrett]

Thanks @GuiltyDolphin - there's a stringification override causing the concat warnings. You can safely ignore them, but I wonder if the override is worth keeping.

It is kind of sort of useful when working in a REPL and such.

[GuiltyDolphin]

@jbarrett Might it be worth returning another string when the override fails? So we don't get all the errors but can keep the functionality when it works.