I&#39;d add my vote to picking up [`Test::More`][1] before going any further in Perl testing. The [Perl testing community][2] is fairly well united around the [Test Anything Protocol][3], and you&#39;ll want to play around with [`Test::More`][1] to understand how it works and how tools like [`prove`][4] and `Test::Harness::Archive` can help automate and distribute testing.

If you want to just &quot;jump right in&quot;, I think [`Test::Class`][5] provides xTest facilities with a [TAP][3] backend. I haven&#39;t used it at all (I&#39;m a [`Test::More`][1] person myself), but it&#39;s [very highly rated][6].


  [1]:  https://metacpan.org/pod/Test::More
  [2]: http://web.archive.org/web/20180628054858/http://perl-qa.hexten.net/wiki/index.php/Main_Page
  [3]: http://testanything.org/
  [4]: http://search.cpan.org/dist/Test-Harness/bin/prove
  [5]: https://metacpan.org/pod/Test::Class
  [6]:  http://cpanratings.perl.org/dist/Test-Class

Test::More should offer you more bang for your bucks once you get the hang of Test::Simple.

Judging by your comments on melaos answer, I&#39;d say [Test::Class][1] or [Test::Unit][2] is what you&#39;re looking for.


  [1]: https://metacpan.org/pod/Test::Class
  [2]: https://metacpan.org/pod/Test::Unit

Simple test example:


    #!/usr/bin/perl -w
    
    use strict;
    use warnings &#39;all&#39;;
    use Test::More plan =&gt; 4;  # or use Test::More &#39;no_plan&#39;;
    
    use_ok(&#39;My::Module&#39;, &#39;Loaded My::Module&#39;);
    ok( my $obj = My::Module-&gt;new(), &#39;Can create instance of My::Module&#39;);
    
    ok( $obj-&gt;value(&#39;hello&#39;), &#39;Set value to hello&#39; );
    is( $obj-&gt;value =&gt; &#39;hello&#39;, &#39;value is still hello&#39;);




[Test::Class](http://search.cpan.org/perldoc?Test::Class) usage you can see in this [example](https://stackoverflow.com/questions/499967/programmatically-determining-the-longest-similar-portion-of-several-strings/500518#500518).

My program need to parse css files into an in-memory object format. Any advice on how this should be done ?


Is there a CSS parser for C#?

As I have been testing sites, I have found [reCAPTCHA][1]s getting more and more difficult to read. Is it just me or are others having this problem too?

Along with this, I had a user this morning complain about receiving a Bristish Pound character in their reCAPTCHA. Of course the user didn&#39;t know what to do, even though I have message stating they can click the reload/refresh icon to get a new CAPTCHA.

Unfortunately, this implementation is on a site often used by people over 60 years of age, so more complicated or confusing CAPTCHAs are a problem, but the site still receives a lot of people attempting to produce spam.


  [1]: http://www.recaptcha.net

Are reCAPTCHA CAPTCHAs getting harder or is just me

I have been doing some OO Perl programming and I was wondering: which is the best way to perform unit tests?

So far I have been using the [Test::Simple][1] module to perform tests, but it feels 
insufficient for what I want. 

Can you point me to some nice modules for that?


  [1]: https://metacpan.org/pod/Test::Simple

How can I do unit testing in Perl?

I have been doing some OO Perl programming and I was wondering: which is the best way to perform unit tests?
So far I have been using the <a href="https://metacpan.org/pod/Test::Simple" target="_blank" rel="noopener noreferrer">Test::Simple</a> module to perform tests, but it feels
insufficient for what I want.
Can you point me to some nice modules for that?

Suppose I have a filehandle `$fh`.  I can check its existence with `-e $fh` or its file size with `-s $fh` or [a slew of additional information about the file][1].  How can I get its last modified time stamp?

[1]:http://perldoc.perl.org/functions/-X.html (Perl -X operands)

How do I get a file&#39;s last modified time in Perl?

I&#39;d like a daemonizer that can turn an arbitrary, generic script or command into a &lt;a href=&quot;http://en.wikipedia.org/wiki/Daemon_(computer_software)&quot;&gt;daemon&lt;/a&gt;.

There are two common cases I&#39;d like to deal with:

 1. I have a script that should run forever.  If it ever dies (or on reboot), restart it.  Don&#39;t let there ever be two copies running at once (detect if a copy is already running and don&#39;t launch it in that case).

 2. I have a simple script or command line command that I&#39;d like to keep executing repeatedly forever (with a short pause between runs).  Again, don&#39;t allow two copies of the script to ever be running at once.

Of course it&#39;s trivial to write a &quot;while(true)&quot; loop around the script in case 2 and then apply a solution for case 1, but a more general solution will just solve case 2 directly since that applies to the script in case 1 as well (you may just want a shorter or no pause if the script is not intended to ever die (of course if the script really *does* never die then the pause doesn&#39;t actually matter)).

Note that the solution should not involve, say, adding file-locking code or PID recording to the existing scripts.

More specifically, I&#39;d like a program &quot;daemonize&quot; that I can run like

    % daemonize myscript arg1 arg2

or, for example, 

    % daemonize &#39;echo `date` &gt;&gt; /tmp/times.txt&#39;

which would keep a growing list of dates appended to times.txt. (Note that if the argument(s) to daemonize is a script that runs forever as in case 1 above, then daemonize will still do the right thing, restarting it when necessary.)  I could then put a command like above in my .login and/or cron it hourly or minutely (depending on how worried I was about it dying unexpectedly).

NB: The daemonize script will need to remember the command string it is daemonizing so that if the same command string is daemonized again it does not launch a second copy.

Also, the solution should ideally work on both OS X and linux but solutions for one or the other are welcome.

EDIT: It&#39;s fine if you have to invoke it with `sudo daemonize myscript myargs`.

(If I&#39;m thinking of this all wrong or there are quick-and-dirty partial solutions, I&#39;d love to hear that too.)

&lt;hr&gt;

PS: In case it&#39;s useful, [here&#39;s](https://stackoverflow.com/questions/473620/how-do-you-create-a-daemon-in-python) a similar question specific to python.



And [this](https://stackoverflow.com/questions/38596/whats-the-best-way-to-keep-a-php-script-running-as-a-daemon/39609#39609) answer to a similar question has what appears to be a useful idiom for a quick-and-dirty demonizing of an arbitrary script:



How do I daemonize an arbitrary script in unix?

Most Stackoverflow answers that I have found in regards to the Perl build process and unit testing and code coverage simply point me to CPAN for the documentation there.  There&#39;s absolutely nothing wrong with pointing to CPAN modules because that&#39;s where the full documentation is supposed to reside.  I&#39;ve had trouble finding complete working code examples in many cases, though.  

I&#39;ve been searching all over the Internet for actual working code samples that I can download or paste into my IDE, like your typical tutorial &quot;Hello World&quot; example source code, but of an example that demonstrates the build process with unit testing and code coverage analysis.  Does anyone have a small example of a complete working project that demonstrates these technologies and processes?  

*(I do have a small working example and I will answer my own question with it, but there are probably other SO users who have better examples than the ones I came up with.)*


Perl build, unit testing, code coverage: A complete working example

What is a good way to **always** do integer division in Perl?

For example, I want:

    real / int = int
    
    int / real = int
    
    int / int = int

How should I do integer division in Perl?

I&#39;m using a hosted Linux machine so I don&#39;t have permissions to write
into the `/usr/lib` directory.

When I try to install a CPAN module by doing the usual:

    perl Makefile.PL
    make test
    make install

That module is extracted to a `blib/lib/` folder. I have kept `use
blib/lib/ModuleName` but it still the compiler says module can not be
found. I have tried copying the .pm file into local directory and kept
`require ModuleName` but still it gives me some error.

How can I install a module into some other directory and use it?

How can I install a CPAN module into a local directory?

I have a Django app that requires a `settings` attribute in the form of:

    RELATED_MODELS = (&#39;appname1.modelname1.attribute1&#39;,
                      &#39;appname1.modelname2.attribute2&#39;, 
                      &#39;appname2.modelname3.attribute3&#39;, ...)

Then hooks their post_save signal to update some other fixed model depending on the `attributeN` defined.

I would like to test this behaviour and tests should work even if this app is the only one in the project (except for its own dependencies, no other wrapper app need to be installed). How can I create and attach/register/activate mock models just for the test database? (or is it possible at all?)

Solutions that allow me to use test fixtures would be great.


Django: How to create a model dynamically just for testing

I&#39;m using Mockito to write a unit test in Java, and I&#39;d like to verify that a certain method is the *last one* called on an object.

I&#39;m doing something like this in the code under test:

    row.setSomething(value);
    row.setSomethingElse(anotherValue);
    row.editABunchMoreStuff();
    row.saveToDatabase();

In my mock, I don&#39;t care about the order in which I edit everything on the row, but it&#39;s very important that I *not* try to do anything more to it after I&#39;ve saved it. Is there a good way to do this? 

Note that I&#39;m not looking for verifyNoMoreInteractions: it doesn&#39;t confirm that saveToDatabase is the last thing called, and it also fails if I call anything on the row that I don&#39;t explicitly verify. I&#39;d like to be able to say something like:

    verify(row).setSomething(value);
    verify(row).setSomethingElse(anotherValue);
    verifyTheLastThingCalledOn(row).saveToDatabase();

If it helps, I&#39;m switching to Mockito from a JMock test that did this:

    row.expects(once()).method(&quot;saveToDatabase&quot;).id(&quot;save&quot;);
    row.expects(never()).method(ANYTHING).after(&quot;save&quot;);

Use Mockito to verify that nothing is called after a method

I am starting out with unit testing, I have a method that uses the web.config for a connection string.

I was hoping to be able to use

    [DeploymentItem(&quot;web.config&quot;)]

to get the web config file, this still leaves me with null reference exceptions (that&#39;d be what I write my next test for).

How do I use the config file included with the project I am trying to test?

I am using the testing framework included in VS 2008, if that makes any difference.

Thanks




How to use web.config when unit testing an asp .net application

What are unit tests, integration tests, smoke tests, and regression tests?  What are the differences between them and which tools can I use for each of them?

For example, I use [JUnit][1] and [NUnit][2] for *unit testing* and *integration testing*. Are there any tools for the last two, *smoke testing* or *regression testing*?

  [1]: https://en.wikipedia.org/wiki/JUnit
  [2]: https://en.wikipedia.org/wiki/NUnit



What are unit tests, integration tests, smoke tests, and regression tests?

It is not documented on the web site and people seem to be having problems setting up the framework. Can someone please show a step-by-step introduction for a sample project setup?

How to set up Google C++ Testing Framework (gtest) with Visual Studio 2005

I&#39;m trying to test a controller and I got this error. I understand the error, but don&#39;t know how to fix it.

    test: on CREATE to :user with completely invalid email should respond with 
      redirect
    (UsersControllerTest):ActionController::RedirectBackError: 
      No HTTP_REFERER was set in the request to this action, 
      so redirect_to :back could not be called successfully. 
    If this is a test, make sure to specify request.env[&quot;HTTP_REFERER&quot;].

Specify it where? I tried this:

    setup { post :create, { :user =&gt; { :email =&gt; &#39;invalid@abc&#39; } }, 
      { &#39;referer&#39; =&gt; &#39;/sessions/new&#39; } }

But got the same error.

Specify it with what, exactly? I guess the URI of the view I want it to go back to:

    &#39;/sessions/new&#39;

Is that what they mean?

-----

OK, so it turns out they mean do this:

    setup do
      @request.env[&#39;HTTP_REFERER&#39;] = &#39;http://localhost:3000/sessions/new&#39;
      post :create, { :user =&gt; { :email =&gt; &#39;invalid@abc&#39; } }, {}
    end

Can someone tell me where that&#39;s documented? I&#39;d like to read up on the context of that information.

What if the domain is not &quot;localhost:3000&quot;? What if it&#39;s &quot;localhost:3001&quot; or something? Any way to anticipate that?

Why doesn&#39;t this work:

    setup { post :create, { :user =&gt; { :email =&gt; &#39;invalid@abc&#39; } }, 
      { &#39;referer&#39; =&gt; &#39;/sessions/new&#39; } }

The Rails docs [specifically say](http://apidock.com/rails/ActionController/Integration/Session/post) that&#39;s how you set the headers.


How do I set HTTP_REFERER when testing in Rails?

In C#, what&#39;s the difference between 

    Assert.AreNotEqual

 

and 

    Assert.AreNotSame

 



What&#39;s the difference between Assert.AreNotEqual and Assert.AreNotSame?

I need to configure a SMTP server for testing my website which sends emails (for registration confirmation etc). 

I dont actually want the email to be sent, I just want to make sure that my code is correct. So I want to be able to check that the email is placed in a queue folder for example.

Can anybody recommend a SMTP server which is easy to configure?

Testing SMTP with .net

What are some of the nastiest, most difficult bugs you have had to track and fix and why?

I am both genuinely curious and knee deep in the process as we speak.  So as they say - misery likes company.

Content Type	Original Author	Original Content on Stackoverflow
Question	mandel	View Question on Stackoverflow
Solution 1 - Perl	Gaurav	View Answer on Stackoverflow
Solution 2 - Perl	melaos	View Answer on Stackoverflow
Solution 3 - Perl	Leon Timmermans	View Answer on Stackoverflow
Solution 4 - Perl	JDrago	View Answer on Stackoverflow
Solution 5 - Perl	Hynek -Pichi- Vychodil	View Answer on Stackoverflow

How can I do unit testing in Perl?

Perl Problem Overview

Perl Solutions

Solution 1 - Perl

Solution 2 - Perl

Solution 3 - Perl

Solution 4 - Perl

Solution 5 - Perl

Are reCAPTCHA CAPTCHAs getting harder or is just me

Is there a CSS parser for C#?

Attributions