Failing test - It seems that PPI has problems when heredoc terminator is... #17
Conversation
… is last in file and there is 0 or 1 newline after it (not more).
|
I investigated the relevant PPI code and found the issue - the section of code that is designed to handle here-doc terminators without a trailing newline can never be reached due to a bug in the parsing loop. I sent a pull request to PPI's repository yesterday to address the issue: Perl-Critic/PPI#72. I'm going to keep your pull request open until the underlying issue is addressed in PPI, and once that is done I will re-run your pull request against Travis to make sure that it is completely fixed. |
That PR was merged and released in PPI 1.222 https://metacpan.org/changes/distribution/PPI#L162 |
|
With the latest |
- Added test for heredoc terminators
guillaumeaubert#17
(Victor Efimov)
- Reset `_sqlsafe` when scanning multiple documents; added tests
guillaumeaubert#22
(Charlie Garrison)
- New test cases related to anonymous functions
guillaumeaubert#18
(Victor Efimov)
- Add safe_context configuration option
guillaumeaubert#23
(Nelson Ferraz)
- Added a new "prefer_upper_case_keywords" configuration option (Olaf Alders)
- Added detection for ALTER, CREATE and DROP statements.
... last in file and there is 0 or 1 newline after it (not more).