\section{\module{doctest} --- Test interactive Python examples} \declaremodule{standard}{doctest} \moduleauthor{Tim Peters}{tim@python.org} \sectionauthor{Tim Peters}{tim@python.org} \sectionauthor{Moshe Zadka}{moshez@debian.org} \sectionauthor{Edward Loper}{edloper@users.sourceforge.net} \modulesynopsis{A framework for verifying interactive Python examples.} The \module{doctest} module searches for pieces of text that look like interactive Python sessions, and then executes those sessions to verify that they work exactly as shown. There are several common ways to use doctest: \begin{itemize} \item To check that a module's docstrings are up-to-date by verifying that all interactive examples still work as documented. \item To perform regression testing by verifying that interactive examples from a test file or a test object work as expected. \item To write tutorial documentation for a package, liberally illustrated with input-output examples. Depending on whether the examples or the expository text are emphasized, this has the flavor of "literate testing" or "executable documentation". \end{itemize} Here's a complete but small example module: \begin{verbatim} """ This is the "example" module. The example module supplies one function, factorial(). For example, >>> factorial(5) 120 """ def factorial(n): """Return the factorial of n, an exact integer >= 0. If the result is small enough to fit in an int, return an int. Else return a long. >>> [factorial(n) for n in range(6)] [1, 1, 2, 6, 24, 120] >>> [factorial(long(n)) for n in range(6)] [1, 1, 2, 6, 24, 120] >>> factorial(30) 265252859812191058636308480000000L >>> factorial(30L) 265252859812191058636308480000000L >>> factorial(-1) Traceback (most recent call last): ... ValueError: n must be >= 0 Factorials of floats are OK, but the float must be an exact integer: >>> factorial(30.1) Traceback (most recent call last): ... ValueError: n must be exact integer >>> factorial(30.0) 265252859812191058636308480000000L It must also not be ridiculously large: >>> factorial(1e100) Traceback (most recent call last): ... OverflowError: n too large """ \end{verbatim} % allow LaTeX to break here. \begin{verbatim} import math if not n >= 0: raise ValueError("n must be >= 0") if math.floor(n) != n: raise ValueError("n must be exact integer") if n+1 == n: # catch a value like 1e300 raise OverflowError("n too large") result = 1 factor = 2 while factor <= n: result *= factor factor += 1 return result def _test(): import doctest doctest.testmod() if __name__ == "__main__": _test() \end{verbatim} If you run \file{example.py} directly from the command line, \module{doctest} works its magic: \begin{verbatim} $ python example.py $ \end{verbatim} There's no output! That's normal, and it means all the examples worked. Pass \programopt{-v} to the script, and \module{doctest} prints a detailed log of what it's trying, and prints a summary at the end: \begin{verbatim} $ python example.py -v Trying: factorial(5) Expecting: 120 ok Trying: [factorial(n) for n in range(6)] Expecting: [1, 1, 2, 6, 24, 120] ok Trying: [factorial(long(n)) for n in range(6)] Expecting: [1, 1, 2, 6, 24, 120] ok \end{verbatim} And so on, eventually ending with: \begin{verbatim} Trying: factorial(1e100) Expecting: Traceback (most recent call last): ... OverflowError: n too large ok 1 items had no tests: __main__._test 2 items passed all tests: 1 tests in __main__ 8 tests in __main__.factorial 9 tests in 3 items. 9 passed and 0 failed. Test passed. $ \end{verbatim} That's all you need to know to start making productive use of \module{doctest}! Jump in. The following sections provide full details. Note that there are many examples of doctests in the standard Python test suite and libraries. Especially useful examples can be found in the standard test file \file{Lib/test/test_doctest.py}. \subsection{Simple Usage: Checking Examples in Docstrings\label{doctest-simple-testmod}} The simplest way to start using doctest (but not necessarily the way you'll continue to do it) is to end each module \module{M} with: \begin{verbatim} def _test(): import doctest doctest.testmod() if __name__ == "__main__": _test() \end{verbatim} \module{doctest} then examines docstrings in module \module{M}. Running the module as a script causes the examples in the docstrings to get executed and verified: \begin{verbatim} python M.py \end{verbatim} This won't display anything unless an example fails, in which case the failing example(s) and the cause(s) of the failure(s) are printed to stdout, and the final line of output is \samp{***Test Failed*** \var{N} failures.}, where \var{N} is the number of examples that failed. Run it with the \programopt{-v} switch instead: \begin{verbatim} python M.py -v \end{verbatim} and a detailed report of all examples tried is printed to standard output, along with assorted summaries at the end. You can force verbose mode by passing \code{verbose=True} to \function{testmod()}, or prohibit it by passing \code{verbose=False}. In either of those cases, \code{sys.argv} is not examined by \function{testmod()} (so passing \programopt{-v} or not has no effect). For more information on \function{testmod()}, see section~\ref{doctest-basic-api}. \subsection{Simple Usage: Checking Examples in a Text File\label{doctest-simple-testfile}} Another simple application of doctest is testing interactive examples in a text file. This can be done with the \function{testfile()} function: \begin{verbatim} import doctest doctest.testfile("example.txt") \end{verbatim} That short script executes and verifies any interactive Python examples contained in the file \file{example.txt}. The file content is treated as if it were a single giant docstring; the file doesn't need to contain a Python program! For example, perhaps \file{example.txt} contains this: \begin{verbatim} The ``example`` module ====================== Using ``factorial`` ------------------- This is an example text file in reStructuredText format. First import ``factorial`` from the ``example`` module: >>> from example import factorial Now use it: >>> factorial(6) 120 \end{verbatim} Running \code{doctest.testfile("example.txt")} then finds the error in this documentation: \begin{verbatim} File "./example.txt", line 14, in example.txt Failed example: factorial(6) Expected: 120 Got: 720 \end{verbatim} As with \function{testmod()}, \function{testfile()} won't display anything unless an example fails. If an example does fail, then the failing example(s) and the cause(s) of the failure(s) are printed to stdout, using the same format as \function{testmod()}. By default, \function{testfile()} looks for files in the calling module's directory. See section~\ref{doctest-basic-api} for a description of the optional arguments that can be used to tell it to look for files in other locations. Like \function{testmod()}, \function{testfile()}'s verbosity can be set with the \programopt{-v} command-line switch or with the optional keyword argument \var{verbose}. For more information on \function{testfile()}, see section~\ref{doctest-basic-api}. \subsection{How It Works\label{doctest-how-it-works}} This section examines in detail how doctest works: which docstrings it looks at, how it finds interactive examples, what execution context it uses, how it handles exceptions, and how option flags can be used to control its behavior. This is the information that you need to know to write doctest examples; for information about actually running doctest on these examples, see the following sections. \subsubsection{Which Docstrings Are Examined?\label{doctest-which-docstrings}} The module docstring, and all function, class and method docstrings are searched. Objects imported into the module are not searched. In addition, if \code{M.__test__} exists and "is true", it must be a dict, and each entry maps a (string) name to a function object, class object, or string. Function and class object docstrings found from \code{M.__test__} are searched, and strings are treated as if they were docstrings. In output, a key \code{K} in \code{M.__test__} appears with name \begin{verbatim} .__test__.K \end{verbatim} Any classes found are recursively searched similarly, to test docstrings in their contained methods and nested classes. \versionchanged[A "private name" concept is deprecated and no longer documented]{2.4} \subsubsection{How are Docstring Examples Recognized?\label{doctest-finding-examples}} In most cases a copy-and-paste of an interactive console session works fine, but doctest isn't trying to do an exact emulation of any specific Python shell. All hard tab characters are expanded to spaces, using 8-column tab stops. If you don't believe tabs should mean that, too bad: don't use hard tabs, or write your own \class{DocTestParser} class. \versionchanged[Expanding tabs to spaces is new; previous versions tried to preserve hard tabs, with confusing results]{2.4} \begin{verbatim} >>> # comments are ignored >>> x = 12 >>> x 12 >>> if x == 13: ... print "yes" ... else: ... print "no" ... print "NO" ... print "NO!!!" ... no NO NO!!! >>> \end{verbatim} Any expected output must immediately follow the final \code{'>\code{>}>~'} or \code{'...~'} line containing the code, and the expected output (if any) extends to the next \code{'>\code{>}>~'} or all-whitespace line. The fine print: \begin{itemize} \item Expected output cannot contain an all-whitespace line, since such a line is taken to signal the end of expected output. If expected output does contain a blank line, put \code{} in your doctest example each place a blank line is expected. \versionchanged[\code{} was added; there was no way to use expected output containing empty lines in previous versions]{2.4} \item Output to stdout is captured, but not output to stderr (exception tracebacks are captured via a different means). \item If you continue a line via backslashing in an interactive session, or for any other reason use a backslash, you should use a raw docstring, which will preserve your backslashes exactly as you type them: \begin{verbatim} >>> def f(x): ... r'''Backslashes in a raw docstring: m\n''' >>> print f.__doc__ Backslashes in a raw docstring: m\n \end{verbatim} Otherwise, the backslash will be interpreted as part of the string. For example, the "{\textbackslash}" above would be interpreted as a newline character. Alternatively, you can double each backslash in the doctest version (and not use a raw string): \begin{verbatim} >>> def f(x): ... '''Backslashes in a raw docstring: m\\n''' >>> print f.__doc__ Backslashes in a raw docstring: m\n \end{verbatim} \item The starting column doesn't matter: \begin{verbatim} >>> assert "Easy!" >>> import math >>> math.floor(1.9) 1.0 \end{verbatim} and as many leading whitespace characters are stripped from the expected output as appeared in the initial \code{'>\code{>}>~'} line that started the example. \end{itemize} \subsubsection{What's the Execution Context?\label{doctest-execution-context}} By default, each time \module{doctest} finds a docstring to test, it uses a \emph{shallow copy} of \module{M}'s globals, so that running tests doesn't change the module's real globals, and so that one test in \module{M} can't leave behind crumbs that accidentally allow another test to work. This means examples can freely use any names defined at top-level in \module{M}, and names defined earlier in the docstring being run. Examples cannot see names defined in other docstrings. You can force use of your own dict as the execution context by passing \code{globs=your_dict} to \function{testmod()} or \function{testfile()} instead. \subsubsection{What About Exceptions?\label{doctest-exceptions}} No problem, provided that the traceback is the only output produced by the example: just paste in the traceback. Since tracebacks contain details that are likely to change rapidly (for example, exact file paths and line numbers), this is one case where doctest works hard to be flexible in what it accepts. Simple example: \begin{verbatim} >>> [1, 2, 3].remove(42) Traceback (most recent call last): File "", line 1, in ? ValueError: list.remove(x): x not in list \end{verbatim} That doctest succeeds if \exception{ValueError} is raised, with the \samp{list.remove(x): x not in list} detail as shown. The expected output for an exception must start with a traceback header, which may be either of the following two lines, indented the same as the first line of the example: \begin{verbatim} Traceback (most recent call last): Traceback (innermost last): \end{verbatim} The traceback header is followed by an optional traceback stack, whose contents are ignored by doctest. The traceback stack is typically omitted, or copied verbatim from an interactive session. The traceback stack is followed by the most interesting part: the line(s) containing the exception type and detail. This is usually the last line of a traceback, but can extend across multiple lines if the exception has a multi-line detail: \begin{verbatim} >>> raise ValueError('multi\n line\ndetail') Traceback (most recent call last): File "", line 1, in ? ValueError: multi line detail \end{verbatim} The last three lines (starting with \exception{ValueError}) are compared against the exception's type and detail, and the rest are ignored. Best practice is to omit the traceback stack, unless it adds significant documentation value to the example. So the last example is probably better as: \begin{verbatim} >>> raise ValueError('multi\n line\ndetail') Traceback (most recent call last): ... ValueError: multi line detail \end{verbatim} Note that tracebacks are treated very specially. In particular, in the rewritten example, the use of \samp{...} is independent of doctest's \constant{ELLIPSIS} option. The ellipsis in that example could be left out, or could just as well be three (or three hundred) commas or digits, or an indented transcript of a Monty Python skit. Some details you should read once, but won't need to remember: \begin{itemize} \item Doctest can't guess whether your expected output came from an exception traceback or from ordinary printing. So, e.g., an example that expects \samp{ValueError: 42 is prime} will pass whether \exception{ValueError} is actually raised or if the example merely prints that traceback text. In practice, ordinary output rarely begins with a traceback header line, so this doesn't create real problems. \item Each line of the traceback stack (if present) must be indented further than the first line of the example, \emph{or} start with a non-alphanumeric character. The first line following the traceback header indented the same and starting with an alphanumeric is taken to be the start of the exception detail. Of course this does the right thing for genuine tracebacks. \item When the \constant{IGNORE_EXCEPTION_DETAIL} doctest option is is specified, everything following the leftmost colon is ignored. \end{itemize} \versionchanged[The ability to handle a multi-line exception detail, and the \constant{IGNORE_EXCEPTION_DETAIL} doctest option, were added]{2.4} \subsubsection{Option Flags and Directives\label{doctest-options}} A number of option flags control various aspects of doctest's behavior. Symbolic names for the flags are supplied as module constants, which can be or'ed together and passed to various functions. The names can also be used in doctest directives (see below). The first group of options define test semantics, controlling aspects of how doctest decides whether actual output matches an example's expected output: \begin{datadesc}{DONT_ACCEPT_TRUE_FOR_1} By default, if an expected output block contains just \code{1}, an actual output block containing just \code{1} or just \code{True} is considered to be a match, and similarly for \code{0} versus \code{False}. When \constant{DONT_ACCEPT_TRUE_FOR_1} is specified, neither substitution is allowed. The default behavior caters to that Python changed the return type of many functions from integer to boolean; doctests expecting "little integer" output still work in these cases. This option will probably go away, but not for several years. \end{datadesc} \begin{datadesc}{DONT_ACCEPT_BLANKLINE} By default, if an expected output block contains a line containing only the string \code{}, then that line will match a blank line in the actual output. Because a genuinely blank line delimits the expected output, this is the only way to communicate that a blank line is expected. When \constant{DONT_ACCEPT_BLANKLINE} is specified, this substitution is not allowed. \end{datadesc} \begin{datadesc}{NORMALIZE_WHITESPACE} When specified, all sequences of whitespace (blanks and newlines) are treated as equal. Any sequence of whitespace within the expected output will match any sequence of whitespace within the actual output. By default, whitespace must match exactly. \constant{NORMALIZE_WHITESPACE} is especially useful when a line of expected output is very long, and you want to wrap it across multiple lines in your source. \end{datadesc} \begin{datadesc}{ELLIPSIS} When specified, an ellipsis marker (\code{...}) in the expected output can match any substring in the actual output. This includes substrings that span line boundaries, and empty substrings, so it's best to keep usage of this simple. Complicated uses can lead to the same kinds of "oops, it matched too much!" surprises that \regexp{.*} is prone to in regular expressions. \end{datadesc} \begin{datadesc}{IGNORE_EXCEPTION_DETAIL} When specified, an example that expects an exception passes if an exception of the expected type is raised, even if the exception detail does not match. For example, an example expecting \samp{ValueError: 42} will pass if the actual exception raised is \samp{ValueError: 3*14}, but will fail, e.g., if \exception{TypeError} is raised. Note that a similar effect can be obtained using \constant{ELLIPSIS}, and \constant{IGNORE_EXCEPTION_DETAIL} may go away when Python releases prior to 2.4 become uninteresting. Until then, \constant{IGNORE_EXCEPTION_DETAIL} is the only clear way to write a doctest that doesn't care about the exception detail yet continues to pass under Python releases prior to 2.4 (doctest directives appear to be comments to them). For example, \begin{verbatim} >>> (1, 2)[3] = 'moo' #doctest: +IGNORE_EXCEPTION_DETAIL Traceback (most recent call last): File "", line 1, in ? TypeError: object doesn't support item assignment \end{verbatim} passes under Python 2.4 and Python 2.3. The detail changed in 2.4, to say "does not" instead of "doesn't". \end{datadesc} \begin{datadesc}{COMPARISON_FLAGS} A bitmask or'ing together all the comparison flags above. \end{datadesc} The second group of options controls how test failures are reported: \begin{datadesc}{REPORT_UDIFF} When specified, failures that involve multi-line expected and actual outputs are displayed using a unified diff. \end{datadesc} \begin{datadesc}{REPORT_CDIFF} When specified, failures that involve multi-line expected and actual outputs will be displayed using a context diff. \end{datadesc} \begin{datadesc}{REPORT_NDIFF} When specified, differences are computed by \code{difflib.Differ}, using the same algorithm as the popular \file{ndiff.py} utility. This is the only method that marks differences within lines as well as across lines. For example, if a line of expected output contains digit \code{1} where actual output contains letter \code{l}, a line is inserted with a caret marking the mismatching column positions. \end{datadesc} \begin{datadesc}{REPORT_ONLY_FIRST_FAILURE} When specified, display the first failing example in each doctest, but suppress output for all remaining examples. This will prevent doctest from reporting correct examples that break because of earlier failures; but it might also hide incorrect examples that fail independently of the first failure. When \constant{REPORT_ONLY_FIRST_FAILURE} is specified, the remaining examples are still run, and still count towards the total number of failures reported; only the output is suppressed. \end{datadesc} \begin{datadesc}{REPORTING_FLAGS} A bitmask or'ing together all the reporting flags above. \end{datadesc} "Doctest directives" may be used to modify the option flags for individual examples. Doctest directives are expressed as a special Python comment following an example's source code: \begin{productionlist}[doctest] \production{directive} {"\#" "doctest:" \token{directive_options}} \production{directive_options} {\token{directive_option} ("," \token{directive_option})*} \production{directive_option} {\token{on_or_off} \token{directive_option_name}} \production{on_or_off} {"+" | "-"} \production{directive_option_name} {"DONT_ACCEPT_BLANKLINE" | "NORMALIZE_WHITESPACE" | ...} \end{productionlist} Whitespace is not allowed between the \code{+} or \code{-} and the directive option name. The directive option name can be any of the option flag names explained above. An example's doctest directives modify doctest's behavior for that single example. Use \code{+} to enable the named behavior, or \code{-} to disable it. For example, this test passes: \begin{verbatim} >>> print range(20) #doctest: +NORMALIZE_WHITESPACE [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19] \end{verbatim} Without the directive it would fail, both because the actual output doesn't have two blanks before the single-digit list elements, and because the actual output is on a single line. This test also passes, and also requires a directive to do so: \begin{verbatim} >>> print range(20) # doctest:+ELLIPSIS [0, 1, ..., 18, 19] \end{verbatim} Multiple directives can be used on a single physical line, separated by commas: \begin{verbatim} >>> print range(20) # doctest: +ELLIPSIS, +NORMALIZE_WHITESPACE [0, 1, ..., 18, 19] \end{verbatim} If multiple directive comments are used for a single example, then they are combined: \begin{verbatim} >>> print range(20) # doctest: +ELLIPSIS ... # doctest: +NORMALIZE_WHITESPACE [0, 1, ..., 18, 19] \end{verbatim} As the previous example shows, you can add \samp{...} lines to your example containing only directives. This can be useful when an example is too long for a directive to comfortably fit on the same line: \begin{verbatim} >>> print range(5) + range(10,20) + range(30,40) + range(50,60) ... # doctest: +ELLIPSIS [0, ..., 4, 10, ..., 19, 30, ..., 39, 50, ..., 59] \end{verbatim} Note that since all options are disabled by default, and directives apply only to the example they appear in, enabling options (via \code{+} in a directive) is usually the only meaningful choice. However, option flags can also be passed to functions that run doctests, establishing different defaults. In such cases, disabling an option via \code{-} in a directive can be useful. \versionchanged[Constants \constant{DONT_ACCEPT_BLANKLINE}, \constant{NORMALIZE_WHITESPACE}, \constant{ELLIPSIS}, \constant{IGNORE_EXCEPTION_DETAIL}, \constant{REPORT_UDIFF}, \constant{REPORT_CDIFF}, \constant{REPORT_NDIFF}, \constant{REPORT_ONLY_FIRST_FAILURE}, \constant{COMPARISON_FLAGS} and \constant{REPORTING_FLAGS} were added; by default \code{} in expected output matches an empty line in actual output; and doctest directives were added]{2.4} \subsubsection{Warnings\label{doctest-warnings}} \module{doctest} is serious about requiring exact matches in expected output. If even a single character doesn't match, the test fails. This will probably surprise you a few times, as you learn exactly what Python does and doesn't guarantee about output. For example, when printing a dict, Python doesn't guarantee that the key-value pairs will be printed in any particular order, so a test like % Hey! What happened to Monty Python examples? % Tim: ask Guido -- it's his example! \begin{verbatim} >>> foo() {"Hermione": "hippogryph", "Harry": "broomstick"} \end{verbatim} is vulnerable! One workaround is to do \begin{verbatim} >>> foo() == {"Hermione": "hippogryph", "Harry": "broomstick"} True \end{verbatim} instead. Another is to do \begin{verbatim} >>> d = foo().items() >>> d.sort() >>> d [('Harry', 'broomstick'), ('Hermione', 'hippogryph')] \end{verbatim} There are others, but you get the idea. Another bad idea is to print things that embed an object address, like \begin{verbatim} >>> id(1.0) # certain to fail some of the time 7948648 >>> class C: pass >>> C() # the default repr() for instances embeds an address <__main__.C instance at 0x00AC18F0> \end{verbatim} The \constant{ELLIPSIS} directive gives a nice approach for the last example: \begin{verbatim} >>> C() #doctest: +ELLIPSIS <__main__.C instance at 0x...> \end{verbatim} Floating-point numbers are also subject to small output variations across platforms, because Python defers to the platform C library for float formatting, and C libraries vary widely in quality here. \begin{verbatim} >>> 1./7 # risky 0.14285714285714285 >>> print 1./7 # safer 0.142857142857 >>> print round(1./7, 6) # much safer 0.142857 \end{verbatim} Numbers of the form \code{I/2.**J} are safe across all platforms, and I often contrive doctest examples to produce numbers of that form: \begin{verbatim} >>> 3./4 # utterly safe 0.75 \end{verbatim} Simple fractions are also easier for people to understand, and that makes for better documentation. \subsection{Basic API\label{doctest-basic-api}} The functions \function{testmod()} and \function{testfile()} provide a simple interface to doctest that should be sufficient for most basic uses. For a less formal introduction to these two functions, see sections \ref{doctest-simple-testmod} and \ref{doctest-simple-testfile}. \begin{funcdesc}{testfile}{filename\optional{, module_relative}\optional{, name}\optional{, package}\optional{, globs}\optional{, verbose}\optional{, report}\optional{, optionflags}\optional{, extraglobs}\optional{, raise_on_error}} All arguments except \var{filename} are optional, and should be specified in keyword form. Test examples in the file named \var{filename}. Return \samp{(\var{failure_count}, \var{test_count})}. Optional argument \var{module_relative} specifies how the filename should be interpreted: \begin{itemize} \item If \var{module_relative} is \code{True} (the default), then \var{filename} specifies an OS-independent module-relative path. By default, this path is relative to the calling module's directory; but if the \var{package} argument is specified, then it is relative to that package. To ensure OS-independence, \var{filename} should use \code{/} characters to separate path segments, and may not be an absolute path (i.e., it may not begin with \code{/}). \item If \var{module_relative} is \code{False}, then \var{filename} specifies an OS-specific path. The path may be absolute or relative; relative paths are resolved with respect to the current working directory. \end{itemize} Optional argument \var{name} gives the name of the test; by default, or if \code{None}, \code{os.path.basename(\var{filename})} is used. Optional argument \var{package} is a Python package or the name of a Python package whose directory should be used as the base directory for a module-relative filename. If no package is specified, then the calling module's directory is used as the base directory for module-relative filenames. It is an error to specify \var{package} if \var{module_relative} is \code{False}. Optional argument \var{globs} gives a dict to be used as the globals when executing examples. A new shallow copy of this dict is created for the doctest, so its examples start with a clean slate. By default, or if \code{None}, a new empty dict is used. Optional argument \var{extraglobs} gives a dict merged into the globals used to execute examples. This works like \method{dict.update()}: if \var{globs} and \var{extraglobs} have a common key, the associated value in \var{extraglobs} appears in the combined dict. By default, or if \code{None}, no extra globals are used. This is an advanced feature that allows parameterization of doctests. For example, a doctest can be written for a base class, using a generic name for the class, then reused to test any number of subclasses by passing an \var{extraglobs} dict mapping the generic name to the subclass to be tested. Optional argument \var{verbose} prints lots of stuff if true, and prints only failures if false; by default, or if \code{None}, it's true if and only if \code{'-v'} is in \code{sys.argv}. Optional argument \var{report} prints a summary at the end when true, else prints nothing at the end. In verbose mode, the summary is detailed, else the summary is very brief (in fact, empty if all tests passed). Optional argument \var{optionflags} or's together option flags. See see section~\ref{doctest-options}. Optional argument \var{raise_on_error} defaults to false. If true, an exception is raised upon the first failure or unexpected exception in an example. This allows failures to be post-mortem debugged. Default behavior is to continue running examples. \versionadded{2.4} \end{funcdesc} \begin{funcdesc}{testmod}{\optional{m}\optional{, name}\optional{, globs}\optional{, verbose}\optional{, isprivate}\optional{, report}\optional{, optionflags}\optional{, extraglobs}\optional{, raise_on_error}\optional{, exclude_empty}} All arguments are optional, and all except for \var{m} should be specified in keyword form. Test examples in docstrings in functions and classes reachable from module \var{m} (or module \module{__main__} if \var{m} is not supplied or is \code{None}), starting with \code{\var{m}.__doc__}. Also test examples reachable from dict \code{\var{m}.__test__}, if it exists and is not \code{None}. \code{\var{m}.__test__} maps names (strings) to functions, classes and strings; function and class docstrings are searched for examples; strings are searched directly, as if they were docstrings. Only docstrings attached to objects belonging to module \var{m} are searched. Return \samp{(\var{failure_count}, \var{test_count})}. Optional argument \var{name} gives the name of the module; by default, or if \code{None}, \code{\var{m}.__name__} is used. Optional argument \var{exclude_empty} defaults to false. If true, objects for which no doctests are found are excluded from consideration. The default is a backward compatibility hack, so that code still using \method{doctest.master.summarize()} in conjunction with \function{testmod()} continues to get output for objects with no tests. The \var{exclude_empty} argument to the newer \class{DocTestFinder} constructor defaults to true. Optional arguments \var{extraglobs}, \var{verbose}, \var{report}, \var{optionflags}, \var{raise_on_error}, and \var{globs} are the same as for function \function{testfile()} above, except that \var{globs} defaults to \code{\var{m}.__dict__}. Optional argument \var{isprivate} specifies a function used to determine whether a name is private. The default function treats all names as public. \var{isprivate} can be set to \code{doctest.is_private} to skip over names that are private according to Python's underscore naming convention. \deprecated{2.4}{\var{isprivate} was a stupid idea -- don't use it. If you need to skip tests based on name, filter the list returned by \code{DocTestFinder.find()} instead.} \versionchanged[The parameter \var{optionflags} was added]{2.3} \versionchanged[The parameters \var{extraglobs}, \var{raise_on_error} and \var{exclude_empty} were added]{2.4} \end{funcdesc} \subsection{Unittest API\label{doctest-unittest-api}} Doctest provides several functions that can be used to create \module{unittest} test suites from doctest examples. These test suites can then be run using \module{unittest} test runners: \begin{verbatim} import unittest import doctest import my_module_with_doctests suite = doctest.DocTestSuite(my_module_with_doctests) runner = unittest.TextTestRunner() runner.run(suite) \end{verbatim} \begin{funcdesc}{DocFileSuite}{*paths, **kw} Convert doctest tests from one or more text files to a \class{\refmodule{unittest}.TestSuite}. The returned \class{TestSuite} is to be run by the unittest framework and runs the interactive examples in each file. If an example in any file fails, then the synthesized unit test fails, and a \exception{failureException} exception is raised showing the name of the file containing the test and a (sometimes approximate) line number. A number of options may be provided as keyword arguments: The optional argument \var{module_relative} specifies how the the filenames in \var{paths} should be interpreted: \begin{itemize} \item If \var{module_relative} is \code{True} (the default), then each filename specifies an OS-independent module-relative path. By default, this path is relative to the calling module's directory; but if the \var{package} argument is specified, then it is relative to that package. To ensure OS-independence, each filename should use \code{/} characters to separate path segments, and may not be an absolute path (i.e., it may not begin with \code{/}). \item If \var{module_relative} is \code{False}, then each filename specifies an OS-specific path. The path may be absolute or relative; relative paths are resolved with respect to the current working directory. \end{itemize} The optional argument \var{package} is a Python package or the name of a Python package whose directory should be used as the base directory for module-relative filenames. If no package is specified, then the calling module's directory is used as the base directory for module-relative filenames. It is an error to specify \var{package} if \var{module_relative} is \code{False}. The optional argument \var{setUp} specifies a set-up function for the test suite. This is called before running the tests in each file. The \var{setUp} function will be passed a \class{DocTest} object. The setUp function can access the test globals as the \var{globs} attribute of the test passed. The optional argument \var{tearDown} specifies a tear-down function for the test suite. This is called after running the tests in each file. The \var{tearDown} function will be passed a \class{DocTest} object. The setUp function can access the test globals as the \var{globs} attribute of the test passed. The optional argument \var{globs} is a dictionary containing the initial global variables for the tests. A new copy of this dictionary is created for each test. By default, \var{globs} is empty. The optional argument \var{optionflags} specifies the default doctest options for the tests. It is created by or-ing together individual option flags. \versionadded{2.4} \end{funcdesc} \begin{funcdesc}{DocTestSuite}{\optional{module}\optional{, globs}\optional{, extraglobs}\optional{, test_finder}\optional{, setUp}\optional{, tearDown}\optional{, checker}} Convert doctest tests for a module to a \class{\refmodule{unittest}.TestSuite}. The returned \class{TestSuite} is to be run by the unittest framework and runs each doctest in the module. If any of the doctests fail, then the synthesized unit test fails, and a \exception{failureException} exception is raised showing the name of the file containing the test and a (sometimes approximate) line number. The optional argument \var{module} provides the module to be tested. It can be a module object or a (possibly dotted) module name. If not specified, the module calling this function is used. The optional argument \var{globs} is a dictionary containing the initial global variables for the tests. A new copy of this dictionary is created for each test. By default, \var{globs} is empty. The optional argument \var{extraglobs} specifies an extra set of global variables, which is merged into \var{globs}. By default, no extra globals are used. The optional argument \var{test_finder} is the \class{DocTestFinder} object (or a drop-in replacement) that is used to extract doctests from the module. The optional argument \var{setUp} specifies a set-up function for the test suite. This is called before running the tests in each file. The \var{setUp} function will be passed a \class{DocTest} object. The setUp function can access the test globals as the \var{globs} attribute of the test passed. The optional argument \var{tearDown} specifies a tear-down function for the test suite. This is called after running the tests in each file. The \var{tearDown} function will be passed a \class{DocTest} object. The setUp function can access the test globals as the \var{globs} attribute of the test passed. The optional argument \var{optionflags} specifies the default doctest options for the tests. It is created by or-ing together individual option flags. \versionadded{2.3} \versionchanged[The parameters \var{globs}, \var{extraglobs}, \var{test_finder}, \var{setUp}, \var{tearDown}, and \var{optionflags} were added]{2.4} \versionchanged[This function now uses the same search technique as \function{testmod()}.]{2.4} \end{funcdesc} \subsection{Advanced API\label{doctest-advanced-api}} The basic API is a simple wrapper that's intended to make doctest easy to use. It is fairly flexible, and should meet most user's needs; however, if you require more fine grained control over testing, or wish to extend doctest's capabilities, then you should use the advanced API. The advanced API revolves around two container classes, which are used to store the interactive examples extracted from doctest cases: \begin{itemize} \item \class{Example}: A single python statement, paired with its expected output. \item \class{DocTest}: A collection of \class{Example}s, typically extracted from a single docstring or text file. \end{itemize} Additional processing classes are defined to find, parse, and run, and check doctest examples: \begin{itemize} \item \class{DocTestFinder}: Finds all docstrings in a given module, and uses a \class{DocTestParser} to create a \class{DocTest} from every docstring that contains interactive examples. \item \class{DocTestParser}: Creates a \class{DocTest} object from a string (such as an object's docstring). \item \class{DocTestRunner}: Executes the examples in a \class{DocTest}, and uses an \class{OutputChecker} to verify their output. \item \class{OutputChecker}: Compares the actual output from a doctest example with the expected output, and decides whether they match. \end{itemize} The relationship between these processing classes is summarized in the following diagram: \begin{verbatim} list of: +------+ +---------+ |module| --DocTestFinder-> | DocTest | --DocTestRunner-> results +------+ | ^ +---------+ | ^ (printed) | | | Example | | | V | | ... | V | DocTestParser | Example | OutputChecker +---------+ \end{verbatim} \subsubsection{DocTest Objects\label{doctest-DocTest}} \begin{classdesc}{DocTest}{examples, globs, name, filename, lineno, docstring} A collection of doctest examples that should be run in a single namespace. The constructor arguments are used to initialize the member variables of the same names. \versionadded{2.4} \end{classdesc} \class{DocTest} defines the following member variables. They are initialized by the constructor, and should not be modified directly. \begin{memberdesc}{examples} A list of \class{Example} objects encoding the individual interactive Python examples that should be run by this test. \end{memberdesc} \begin{memberdesc}{globs} The namespace (aka globals) that the examples should be run in. This is a dictionary mapping names to values. Any changes to the namespace made by the examples (such as binding new variables) will be reflected in \member{globs} after the test is run. \end{memberdesc} \begin{memberdesc}{name} A string name identifying the \class{DocTest}. Typically, this is the name of the object or file that the test was extracted from. \end{memberdesc} \begin{memberdesc}{filename} The name of the file that this \class{DocTest} was extracted from; or \code{None} if the filename is unknown, or if the \class{DocTest} was not extracted from a file. \end{memberdesc} \begin{memberdesc}{lineno} The line number within \member{filename} where this \class{DocTest} begins, or \code{None} if the line number is unavailable. This line number is zero-based with respect to the beginning of the file. \end{memberdesc} \begin{memberdesc}{docstring} The string that the test was extracted from, or `None` if the string is unavailable, or if the test was not extracted from a string. \end{memberdesc} \subsubsection{Example Objects\label{doctest-Example}} \begin{classdesc}{Example}{source, want\optional{, exc_msg}\optional{, lineno}\optional{, indent}\optional{, options}} A single interactive example, consisting of a Python statement and its expected output. The constructor arguments are used to initialize the member variables of the same names. \versionadded{2.4} \end{classdesc} \class{Example} defines the following member variables. They are initialized by the constructor, and should not be modified directly. \begin{memberdesc}{source} A string containing the example's source code. This source code consists of a single Python statement, and always ends with a newline; the constructor adds a newline when necessary. \end{memberdesc} \begin{memberdesc}{want} The expected output from running the example's source code (either from stdout, or a traceback in case of exception). \member{want} ends with a newline unless no output is expected, in which case it's an empty string. The constructor adds a newline when necessary. \end{memberdesc} \begin{memberdesc}{exc_msg} The exception message generated by the example, if the example is expected to generate an exception; or \code{None} if it is not expected to generate an exception. This exception message is compared against the return value of \function{traceback.format_exception_only()}. \member{exc_msg} ends with a newline unless it's \code{None}. The constructor adds a newline if needed. \end{memberdesc} \begin{memberdesc}{lineno} The line number within the string containing this example where the example begins. This line number is zero-based with respect to the beginning of the containing string. \end{memberdesc} \begin{memberdesc}{indent} The example's indentation in the containing string. I.e., the number of space characters that preceed the example's first prompt. \end{memberdesc} \begin{memberdesc}{options} A dictionary mapping from option flags to \code{True} or \code{False}, which is used to override default options for this example. Any option flags not contained in this dictionary are left at their default value (as specified by the \class{DocTestRunner}'s \member{optionflags}). By default, no options are set. \end{memberdesc} \subsubsection{DocTestFinder objects\label{doctest-DocTestFinder}} \begin{classdesc}{DocTestFinder}{\optional{verbose}\optional{, parser}\optional{, recurse}\optional{, exclude_empty}} A processing class used to extract the \class{DocTest}s that are relevant to a given object, from its docstring and the docstrings of its contained objects. \class{DocTest}s can currently be extracted from the following object types: modules, functions, classes, methods, staticmethods, classmethods, and properties. The optional argument \var{verbose} can be used to display the objects searched by the finder. It defaults to \code{False} (no output). The optional argument \var{parser} specifies the \class{DocTestParser} object (or a drop-in replacement) that is used to extract doctests from docstrings. If the optional argument \var{recurse} is false, then \method{DocTestFinder.find()} will only examine the given object, and not any contained objects. If the optional argument \var{exclude_empty} is false, then \method{DocTestFinder.find()} will include tests for objects with empty docstrings. \versionadded{2.4} \end{classdesc} \class{DocTestFinder} defines the following method: \begin{methoddesc}{find}{obj\optional{, name}\optional{, module}\optional{, globs}\optional{, extraglobs}} Return a list of the \class{DocTest}s that are defined by \var{obj}'s docstring, or by any of its contained objects' docstrings. The optional argument \var{name} specifies the object's name; this name will be used to construct names for the returned \class{DocTest}s. If \var{name} is not specified, then \code{var.__name__} is used. The optional parameter \var{module} is the module that contains the given object. If the module is not specified or is None, then the test finder will attempt to automatically determine the correct module. The object's module is used: \begin{itemize} \item As a default namespace, if `globs` is not specified. \item To prevent the DocTestFinder from extracting DocTests from objects that are imported from other modules. (Contained objects with modules other than \var{module} are ignored.) \item To find the name of the file containing the object. \item To help find the line number of the object within its file. \end{itemize} If \var{module} is \code{False}, no attempt to find the module will be made. This is obscure, of use mostly in testing doctest itself: if \var{module} is \code{False}, or is \code{None} but cannot be found automatically, then all objects are considered to belong to the (non-existent) module, so all contained objects will (recursively) be searched for doctests. The globals for each \class{DocTest} is formed by combining \var{globs} and \var{extraglobs} (bindings in \var{extraglobs} override bindings in \var{globs}). A new copy of the globals dictionary is created for each \class{DocTest}. If \var{globs} is not specified, then it defaults to the module's \var{__dict__}, if specified, or \code{\{\}} otherwise. If \var{extraglobs} is not specified, then it defaults to \code{\{\}}. \end{methoddesc} \subsubsection{DocTestParser objects\label{doctest-DocTestParser}} \begin{classdesc}{DocTestParser}{} A processing class used to extract interactive examples from a string, and use them to create a \class{DocTest} object. \versionadded{2.4} \end{classdesc} \class{DocTestParser} defines the following methods: \begin{methoddesc}{get_doctest}{string, globs, name, filename, lineno} Extract all doctest examples from the given string, and collect them into a \class{DocTest} object. \var{globs}, \var{name}, \var{filename}, and \var{lineno} are attributes for the new \class{DocTest} object. See the documentation for \class{DocTest} for more information. \end{methoddesc} \begin{methoddesc}{get_examples}{string\optional{, name}} Extract all doctest examples from the given string, and return them as a list of \class{Example} objects. Line numbers are 0-based. The optional argument \var{name} is a name identifying this string, and is only used for error messages. \end{methoddesc} \begin{methoddesc}{parse}{string\optional{, name}} Divide the given string into examples and intervening text, and return them as a list of alternating \class{Example}s and strings. Line numbers for the \class{Example}s are 0-based. The optional argument \var{name} is a name identifying this string, and is only used for error messages. \end{methoddesc} \subsubsection{DocTestRunner objects\label{doctest-DocTestRunner}} \begin{classdesc}{DocTestRunner}{\optional{checker}\optional{, verbose}\optional{, optionflags}} A processing class used to execute and verify the interactive examples in a \class{DocTest}. The comparison between expected outputs and actual outputs is done by an \class{OutputChecker}. This comparison may be customized with a number of option flags; see section~\ref{doctest-options} for more information. If the option flags are insufficient, then the comparison may also be customized by passing a subclass of \class{OutputChecker} to the constructor. The test runner's display output can be controlled in two ways. First, an output function can be passed to \method{TestRunner.run()}; this function will be called with strings that should be displayed. It defaults to \code{sys.stdout.write}. If capturing the output is not sufficient, then the display output can be also customized by subclassing DocTestRunner, and overriding the methods \method{report_start}, \method{report_success}, \method{report_unexpected_exception}, and \method{report_failure}. The optional keyword argument \var{checker} specifies the \class{OutputChecker} object (or drop-in replacement) that should be used to compare the expected outputs to the actual outputs of doctest examples. The optional keyword argument \var{verbose} controls the \class{DocTestRunner}'s verbosity. If \var{verbose} is \code{True}, then information is printed about each example, as it is run. If \var{verbose} is \code{False}, then only failures are printed. If \var{verbose} is unspecified, or \code{None}, then verbose output is used iff the command-line switch \programopt{-v} is used. The optional keyword argument \var{optionflags} can be used to control how the test runner compares expected output to actual output, and how it displays failures. For more information, see section~\ref{doctest-options}. \versionadded{2.4} \end{classdesc} \class{DocTestParser} defines the following methods: \begin{methoddesc}{report_start}{out, test, example} Report that the test runner is about to process the given example. This method is provided to allow subclasses of \class{DocTestRunner} to customize their output; it should not be called directly. \var{example} is the example about to be processed. \var{test} is the test containing \var{example}. \var{out} is the output function that was passed to \method{DocTestRunner.run()}. \end{methoddesc} \begin{methoddesc}{report_success}{out, test, example, got} Report that the given example ran successfully. This method is provided to allow subclasses of \class{DocTestRunner} to customize their output; it should not be called directly. \var{example} is the example about to be processed. \var{got} is the actual output from the example. \var{test} is the test containing \var{example}. \var{out} is the output function that was passed to \method{DocTestRunner.run()}. \end{methoddesc} \begin{methoddesc}{report_failure}{out, test, example, got} Report that the given example failed. This method is provided to allow subclasses of \class{DocTestRunner} to customize their output; it should not be called directly. \var{example} is the example about to be processed. \var{got} is the actual output from the example. \var{test} is the test containing \var{example}. \var{out} is the output function that was passed to \method{DocTestRunner.run()}. \end{methoddesc} \begin{methoddesc}{report_unexpected_exception}{out, test, example, exc_info} Report that the given example raised an unexpected exception. This method is provided to allow subclasses of \class{DocTestRunner} to customize their output; it should not be called directly. \var{example} is the example about to be processed. \var{exc_info} is a tuple containing information about the unexpected exception (as returned by \function{sys.exc_info()}). \var{test} is the test containing \var{example}. \var{out} is the output function that was passed to \method{DocTestRunner.run()}. \end{methoddesc} \begin{methoddesc}{run}{test\optional{, compileflags}\optional{, out}\optional{, clear_globs}} Run the examples in \var{test} (a \class{DocTest} object), and display the results using the writer function \var{out}. The examples are run in the namespace \code{test.globs}. If \var{clear_globs} is true (the default), then this namespace will be cleared after the test runs, to help with garbage collection. If you would like to examine the namespace after the test completes, then use \var{clear_globs=False}. \var{compileflags} gives the set of flags that should be used by the Python compiler when running the examples. If not specified, then it will default to the set of future-import flags that apply to \var{globs}. The output of each example is checked using the \class{DocTestRunner}'s output checker, and the results are formatted by the \method{DocTestRunner.report_*} methods. \end{methoddesc} \begin{methoddesc}{summarize}{\optional{verbose}} Print a summary of all the test cases that have been run by this DocTestRunner, and return a tuple \samp{(\var{failure_count}, \var{test_count})}. The optional \var{verbose} argument controls how detailed the summary is. If the verbosity is not specified, then the \class{DocTestRunner}'s verbosity is used. \end{methoddesc} \subsubsection{OutputChecker objects\label{doctest-OutputChecker}} \begin{classdesc}{OutputChecker}{} A class used to check the whether the actual output from a doctest example matches the expected output. \class{OutputChecker} defines two methods: \method{check_output}, which compares a given pair of outputs, and returns true if they match; and \method{output_difference}, which returns a string describing the differences between two outputs. \versionadded{2.4} \end{classdesc} \class{OutputChecker} defines the following methods: \begin{methoddesc}{check_output}{want, got, optionflags} Return \code{True} iff the actual output from an example (\var{got}) matches the expected output (\var{want}). These strings are always considered to match if they are identical; but depending on what option flags the test runner is using, several non-exact match types are also possible. See section~\ref{doctest-options} for more information about option flags. \end{methoddesc} \begin{methoddesc}{output_difference}{example, got, optionflags} Return a string describing the differences between the expected output for a given example (\var{example}) and the actual output (\var{got}). \var{optionflags} is the set of option flags used to compare \var{want} and \var{got}. \end{methoddesc} \subsection{Debugging\label{doctest-debugging}} Doctest provides three mechanisms for debugging doctest examples: \begin{enumerate} \item The \function{debug()} function converts a specified doctest to a Python script, and executes that script using \module{pdb}. \item The \class{DebugRunner} class is a subclass of \class{DocTestRunner} that raises an exception for the first failing example, containing information about that example. This information can be used to perform post-mortem debugging on the example. \item The unittest cases generated by \function{DocTestSuite()} support the \method{debug} method defined by \class{unittest.TestCase}. \end{enumerate} \begin{funcdesc}{debug}{module, name} Debug a single doctest docstring. Provide the \var{module} (or dotted name of the module) containing the docstring to be debugged and the fully qualified dotted \var{name} of the object with the docstring to be debugged. The doctest examples are extracted (see function \function{testsource()}), and written to a temporary file. The Python debugger, \refmodule{pdb}, is then invoked on that file. \versionadded{2.3} \end{funcdesc} \begin{classdesc}{DebugRunner}{\optional{checker}\optional{, verbose}\optional{, optionflags}} A subclass of \class{DocTestRunner} that raises an exception as soon as a failure is encountered. If an unexpected exception occurs, an \exception{UnexpectedException} exception is raised, containing the test, the example, and the original exception. If the output doesn't match, then a \exception{DocTestFailure} exception is raised, containing the test, the example, and the actual output. For information about the constructor parameters and methods, see the documentation for \class{DocTestRunner} in section~\ref{doctest-advanced-api}. \end{classdesc} \begin{excclassdesc}{DocTestFailure}{test, example, got} An exception thrown by \class{DocTestRunner} to signal that a doctest example's actual output did not match its expected output. The constructor arguments are used to initialize the member variables of the same names. \end{excclassdesc} \exception{DocTestFailure} defines the following member variables: \begin{memberdesc}{test} The \class{DocTest} object that was being run when the example failed. \end{memberdesc} \begin{memberdesc}{example} The \class{Example} that failed. \end{memberdesc} \begin{memberdesc}{got} The example's actual output. \end{memberdesc} \begin{funcdesc}{testsource}{module, name} Extract the doctest examples from a docstring. Provide the \var{module} (or dotted name of the module) containing the tests to be extracted and the \var{name} (within the module) of the object with the docstring containing the tests to be extracted. The doctest examples are returned as a string containing Python code. The expected output blocks in the examples are converted to Python comments. \versionadded{2.3} \end{funcdesc} \begin{excclassdesc}{UnexpectedException}{test, example, got} An exception thrown by \class{DocTestRunner} to signal that a doctest example raised an unexpected exception. The constructor arguments are used to initialize the member variables of the same names. \end{excclassdesc} \exception{UnexpectedException} defines the following member variables: \begin{memberdesc}{test} The \class{DocTest} object that was being run when the example failed. \end{memberdesc} \begin{memberdesc}{example} The \class{Example} that failed. \end{memberdesc} \begin{memberdesc}{exc_info} A tuple containing information about the unexpected exception, as returned by \function{sys.exc_info()}. \end{memberdesc} \begin{funcdesc}{register_optionflag}{name} Create a new option flag with a given name, and return the new flag's integer value. \function{register_optionflag()} can be used when subclassing \class{OutputChecker} or \class{DocTestRunner} to create new options that are supported by your subclasses. \function{register_optionflag} should always be called using the following idiom: \begin{verbatim} MY_FLAG = register_optionflag('MY_FLAG') \end{verbatim} \end{funcdesc} \subsection{Soapbox\label{doctest-soapbox}} As mentioned in the introduction, \module{doctest} has two primary uses: \begin{enumerate} \item Checking examples in docstrings. \item Regression testing. \end{enumerate} These two uses have different requirements, and it is important to distinguish them. In particular, filling your docstrings with obscure test cases makes for bad documentation. When writing a docstring, choose docstring examples with care. There's an art to this that needs to be learned---it may not be natural at first. Examples should add genuine value to the documentation. A good example can often be worth many words. % [edloper] I think this may be excessive for many cases; let's % just leave it to the user's judgement: %% If possible, show just a few normal cases, show endcases, show %% interesting subtle cases, and show an example of each kind of %% exception that can be raised. You're probably testing for endcases %% and subtle cases anyway in an interactive shell: %% \refmodule{doctest} wants to make it as easy as possible to capture %% those sessions, and will verify they continue to work as designed %% forever after. If done with care, the examples will be invaluable for your users, and will pay back the time it takes to collect them many times over as the years go by and things change. I'm still amazed at how often one of my \refmodule{doctest} examples stops working after a ``harmless'' change. Doctest also makes an excellent tool for writing regression testing. By interleaving prose and examples, it becomes much easier to keep track of what's actually being tested, and why. When a test fails, the prose descriptions makes it much easier to figure out what the problem is, and how it should be fixed. Regression testing is best confined to dedicated objects or files. There are several options for organizing regressions: \begin{itemize} \item Define functions named \code{_regrtest_\textit{topic}} that consist of single docstrings, containing test cases for the named topics. These functions can be included in the same file as the module, or separated out into a separate test file. \item Define a \code{__test__} dictionary mapping from regression test topics to docstrings containing test cases. \item Write a text file containing test cases as interactive examples, and test that file using \function{testfunc()}. \end{itemize}