COMPOUND WTF TEST



  • Application I'm working on is basically a message bus.  Upstream server sends us a message, we insert it in the database, package it up and send it on to somebody else.  The following code snippet is used to read data from a file and put it on the input queue (testing purposes, fixing production issues).

    <FONT color=#7f0055>if</FONT>(((source.getMessageData()).toString()).contains(<FONT color=#2a00ff>"<WTFOrder xmlns:xsi="</FONT>) == <FONT color=#7f0055>true</FONT>)
    {

    <FONT color=#0000c0>session</FONT>.begin();
    endpoint.putMessage(source, <FONT color=#7f0055>false</FONT>, 0);
    <FONT color=#3f7f5f>// COMPOUND TRANSACTION TEST
    </FONT><FONT color=#3f7f5f>// endpoint2.putMessage(outmsg, false, 0);
    </FONT><FONT color=#0000c0>session</FONT>.commit();

    }

    <FONT color=#7f0055>if</FONT>(((source.getMessageData()).toString()).contains(<FONT color=#2a00ff>"<WTFWrittenData xmlns:xsi="</FONT>) == <FONT color=#7f0055>true</FONT>)
    {

    <FONT color=#0000c0>session</FONT>.begin();
    endpoint.putMessage(source, <FONT color=#7f0055>false</FONT>, 0);
    <FONT color=#3f7f5f>// COMPOUND TRANSACTION TEST
    </FONT><FONT color=#3f7f5f>// endpoint2.putMessage(outmsg, false, 0);
    </FONT><FONT color=#0000c0>session</FONT>.commit();

    }
    <FONT color=#7f0055>if</FONT>(((source.getMessageData()).toString()).contains(<FONT color=#2a00ff>"<WTFOrder>"</FONT>) == <FONT color=#7f0055>true</FONT>)
    {

    <FONT color=#0000c0>session</FONT>.begin();
    endpoint.putMessage(source, <FONT color=#7f0055>false</FONT>, 0);
    <FONT color=#3f7f5f>// COMPOUND TRANSACTION TEST
    </FONT><FONT color=#3f7f5f>// endpoint2.putMessage(outmsg, false, 0);
    </FONT><FONT color=#0000c0>session</FONT>.commit();

    }

    (WTF replaces acronym to protect the guilty, obviously)

    I really want to assume there used to be different code in each of those blocks, but the fact that the commented out stuff is identical too....?



  • Well, it is possible they used to be different and then changed so they were all the same, resulting in a copy past that included comments.  But still...



  • Hmmm...

                if (someObject.toString().contains("xxx")) { ...}

    as opposed to

                if (someObject.getXxx() != null) { ... }

    Seems like a lot of pointless toString'ing


  • Considered Harmful

    I like the fact that if the input string contains more than one matching value, it will be inserted multiple times.



  • Re: CPOUND WTF TEST

    Did anyone else read this as a CPound WTF TEST?



  • Maybe their compiler doesn't support the light-pole without a light operator.



  • if (expression == true) { /* shoot myself */ ; }



  • if if == then then then() else else()



  • if (x != y) {

        i = x;

        dosomething(c);

    } else {

        i = y;

        dosome...err hold on, why did I have an if here?!

     



  • The real WTF is that the session handling is not guarded by try-finally. Otherwise it looks good :D



  • The real WTF is that they are not using an XML parser.



  • @dasluq said:

    The real WTF is that they are not using an XML parser.

    Nah, the data's coming in as a string and going out as a string.  Parsing the whole thing to search for that one substring would lower performance.

     

    The real WTF is that the session handling is not guarded by try-finally.

    That code came from the inside of a try block.


Log in to reply