fix docstring duplication error issue #2054

[bubbomb]

This commit fixes the issue #2054

[drewbanin]

hey @bubbomb - can you rebase this onto dev/barbara-gittings? We're cutting an RC of 0.15.2 tomorrow and are no longer accepting PRs to the dev/0.15.2 branch.

This looks really great so far! I saw your issue comment about adding tests. What do you think about adding a new test case for a sample project with two identically named docs block?

You can add a new test class like this one:
https://github.com/fishtown-analytics/dbt/blob/e080bfc79ab11a4d145de4c8da9d955c5e136d92/test/integration/035_docs_blocks/test_docs_blocks.py#L130-L147

And a new directory like this one (eg. duplicated_names/) that looks something like this:
https://github.com/fishtown-analytics/dbt/tree/dev/0.15.2/test/integration/035_docs_blocks/invalid_name_models

I can kick off the test suite once some tests are added here :)

[bubbomb]

thanks @drewbanin this is exactly what I was looking for. I messed around with a few other tests for a while but couldn't get things to work quite the way I wanted to. I'll give this a shot and let you know.

[bubbomb]

Would you like me to close this PR and submit a new PR to the dev/barbara-gittings branch?

[drewbanin]

It's up to you! You can totally do that, or if you're comfortable enough with git, you can also rebase your commits onto the latest dev/barbara-gittings and we can update the base branch in this PR!

I think that might look something like

git remote add fishtown https://github.com/fishtown-analytics/dbt.git
git fetch fishtown
git rebase -i fishtown/dev/barbara-gittings

If that becomes a big mess, you can certainly feel free to open a new PR for a branch off of dev/barbara-gittings :)

[drewbanin]

It looks like there are still some stray commits in here. If you'd prefer, please feel free to close this one and open up another PR for your change :)

[bubbomb]

Yea, that was a little odd. I just saw that too. No worries though I updated it. Will this work or do you want me to open a new PR?

[beckjake]

Hey @bubbomb, this PR looks great!

The new test is failing, I believe you need the commented out with self.assertRaises(...) - this code fails during compile time, not run time, so you have to catch an exception instead of expect_pass=False: Just one of those fun quirks of our integration test system.

[bubbomb]

@beckjake crud... I must have left that commented out after messing around with a few things. I'll get a fix up for it.

[drewbanin]

@bubbomb can you mark this PR as "ready for review" if it's ready to merge? Thanks!

[beckjake]

Thanks so much for your contribution @bubbomb! This will get in to 0.16.0 (but missed the cutoff for beta1).

[bubbomb]

@beckjake @drewbanin no problem you guys were great to work with!