Seems that starting in perl 5.17, keys() to obtain the keys in a hash is now truly randomized, which means that even within the same process, calling keys() twice on the same hash will result in a different key order.

In previous versions, this used to be just per-process, so you couldn't rely on the sort order from one process invocation to the next, but calling keys() twice on the same hash within the same process always returned consistent results.

This seems like a major challenge for testing applications. If my hash serializer produces different results every time, should my CPAN module have to deal with the daunting task of creating every single permutation of the keys() result that's used underneath, leading to an explosion of application-layer results to test against?

So far, in the test suite for https://github.com/mschilli/php-httpbuildquery-perl I've worked around this issue by running keys() once to determine the order, to test against the result generated by my second call by the application. With perl 5.17, this is no longer possible.

What do people do to test against unpredictable core functions? Shouldn't the core provide some kind of API to figure out what the order is/was for testing?

---

---

• Are you posting in the right place? Check out Where do I post X? to know for sure.
• Posts may use any of the Perl Monks Approved HTML tags. Currently these include the following:

  <code> <a> <b> <big> <blockquote> <br /> <dd> <dl> <dt> <em> <font> <h1> <h2> <h3> <h4> <h5> <h6> <hr /> <i> <li> <nbsp> <ol> <p> <small> <strike> <strong> <sub> <sup> <table> <td> <th> <tr> <tt> <u> <ul>

• Snippets of code should be wrapped in <code> tags not <pre> tags. In fact, <pre> tags should generally be avoided. If they must be used, extreme care should be taken to ensure that their contents do not have long lines (<70 chars), in order to prevent horizontal scrolling (and possible janitor intervention).
• Want more info? How to link or How to display code and escape characters are good places to start.