runtime/doc/testing.txt

*testing.txt*	For Vim version 8.2.  Last change: 2020 Dec 12


		  VIM REFERENCE MANUAL	  by Bram Moolenaar


Testing Vim and Vim script			*testing-support*

Expression evaluation is explained in |eval.txt|.  This file goes into details
about writing tests in Vim script.  This can be used for testing Vim itself
and for testing plugins.

1. Testing Vim				|testing|
2. Test functions			|test-functions-details|
3. Assert functions			|assert-functions-details|

==============================================================================
1. Testing Vim						*testing*

Vim can be tested after building it, usually with "make test".
The tests are located in the directory "src/testdir".

There are two types of tests added over time:
	test20.in		oldest, only for tiny and small builds
	test_something.vim	new style tests

						*new-style-testing*
New tests should be added as new style tests.  The test scripts are named
test_<feature>.vim (replace <feature> with the feature under test). These use
functions such as |assert_equal()| to keep the test commands and the expected
result in one place.
						*old-style-testing*
These tests are used only for testing Vim without the |+eval| feature.

Find more information in the file src/testdir/README.txt.

==============================================================================
2. Test functions				*test-functions-details*

test_alloc_fail({id}, {countdown}, {repeat})		*test_alloc_fail()*
		This is for testing: If the memory allocation with {id} is
		called, then decrement {countdown}, and when it reaches zero
		let memory allocation fail {repeat} times.  When {repeat} is
		smaller than one it fails one time.

		Can also be used as a |method|: >
			GetAllocId()->test_alloc_fail()

test_autochdir()					*test_autochdir()*
		Set a flag to enable the effect of 'autochdir' before Vim
		startup has finished.


test_feedinput({string})				*test_feedinput()*
		Characters in {string} are queued for processing as if they
		were typed by the user. This uses a low level input buffer.
		This function works only when with |+unix| or GUI is running.

		Can also be used as a |method|: >
			GetText()->test_feedinput()

test_garbagecollect_now()			 *test_garbagecollect_now()*
		Like garbagecollect(), but executed right away.  This must
		only be called directly to avoid any structure to exist
		internally, and |v:testing| must have been set before calling
		any function.


test_garbagecollect_soon()			 *test_garbagecollect_soon()*
		Set the flag to call the garbagecollector as if in the main
		loop.  Only to be used in tests.


test_getvalue({name})					*test_getvalue()*
		Get the value of an internal variable.  These values for
		{name} are supported:
			need_fileinfo

		Can also be used as a |method|: >
			GetName()->test_getvalue()

test_ignore_error({expr})			 *test_ignore_error()*
		Ignore any error containing {expr}.  A normal message is given
		instead.
		This is only meant to be used in tests, where catching the
		error with try/catch cannot be used (because it skips over
		following code).
		{expr} is used literally, not as a pattern.
		When the {expr} is the string "RESET" then the list of ignored
		errors is made empty.

		Can also be used as a |method|: >
			GetErrorText()->test_ignore_error()

test_null_blob()					*test_null_blob()*
		Return a |Blob| that is null. Only useful for testing.


test_null_channel()					*test_null_channel()*
		Return a |Channel| that is null. Only useful for testing.
		{only available when compiled with the +channel feature}


test_null_dict()					*test_null_dict()*
		Return a |Dict| that is null. Only useful for testing.


test_null_function()					*test_null_function()*
		Return a |Funcref| that is null. Only useful for testing.


test_null_job()						*test_null_job()*
		Return a |Job| that is null. Only useful for testing.
		{only available when compiled with the +job feature}


test_null_list()					*test_null_list()*
		Return a |List| that is null. Only useful for testing.


test_null_partial()					*test_null_partial()*
		Return a |Partial| that is null. Only useful for testing.


test_null_string()					*test_null_string()*
		Return a |String| that is null. Only useful for testing.


test_unknown()						*test_unknown()*
		Return a value with unknown type. Only useful for testing.

test_void()						*test_void()*
		Return a value with void type. Only useful for testing.


test_option_not_set({name})				*test_option_not_set()*
		Reset the flag that indicates option {name} was set.  Thus it
		looks like it still has the default value. Use like this: >
			set ambiwidth=double
			call test_option_not_set('ambiwidth')
<		Now the 'ambiwidth' option behaves like it was never changed,
		even though the value is "double".
		Only to be used for testing!

		Can also be used as a |method|: >
			GetOptionName()->test_option_not_set()


test_override({name}, {val})				*test_override()*
		Overrides certain parts of Vim's internal processing to be able
		to run tests. Only to be used for testing Vim!
		The override is enabled when {val} is non-zero and removed
		when {val} is zero.
		Current supported values for name are:

		name	     effect when {val} is non-zero ~
		redraw       disable the redrawing() function
		redraw_flag  ignore the RedrawingDisabled flag
		char_avail   disable the char_avail() function
		starting     reset the "starting" variable, see below
		nfa_fail     makes the NFA regexp engine fail to force a
			     fallback to the old engine
		no_query_mouse  do not query the mouse position for "dec"
				terminals
		no_wait_return	set the "no_wait_return" flag.  Not restored
				with "ALL".
		ui_delay     time in msec to use in ui_delay(); overrules a
			     wait time of up to 3 seconds for messages
		term_props   reset all terminal properties when the version
			     string is detected
		ALL	     clear all overrides ({val} is not used)

		"starting" is to be used when a test should behave like
		startup was done.  Since the tests are run by sourcing a
		script the "starting" variable is non-zero. This is usually a
		good thing (tests run faster), but sometimes changes behavior
		in a way that the test doesn't work properly.
		When using: >
			call test_override('starting', 1)
<		The value of "starting" is saved.  It is restored by: >
			call test_override('starting', 0)

<		Can also be used as a |method|: >
			GetOverrideVal()-> test_override('starting')

test_refcount({expr})					*test_refcount()*
		Return the reference count of {expr}.  When {expr} is of a
		type that does not have a reference count, returns -1.  Only
		to be used for testing.

		Can also be used as a |method|: >
			GetVarname()->test_refcount()


test_scrollbar({which}, {value}, {dragging})		*test_scrollbar()*
		Pretend using scrollbar {which} to move it to position
		{value}.  {which} can be:
			left	Left scrollbar of the current window
			right	Right scrollbar of the current window
			hor	Horizontal scrollbar

		For the vertical scrollbars {value} can be 1 to the
		line-count of the buffer.  For the horizontal scrollbar the
		{value} can be between 1 and the maximum line length, assuming
		'wrap' is not set.

		When {dragging} is non-zero it's like dragging the scrollbar,
		otherwise it's like clicking in the scrollbar.
		Only works when the {which} scrollbar actually exists,
		obviously only when using the GUI.

		Can also be used as a |method|: >
			GetValue()->test_scrollbar('right', 0)

test_setmouse({row}, {col})				*test_setmouse()*
		Set the mouse position to be used for the next mouse action.
		{row} and {col} are one based.
		For example: >
			call test_setmouse(4, 20)
			call feedkeys("\<LeftMouse>", "xt")

test_settime({expr})					*test_settime()*
		Set the time Vim uses internally.  Currently only used for
		timestamps in the history, as they are used in viminfo, and
		for undo.
		Using a value of 1 makes Vim not sleep after a warning or
		error message.
		{expr} must evaluate to a number.  When the value is zero the
		normal behavior is restored.

		Can also be used as a |method|: >
			GetTime()->test_settime()

test_srand_seed([seed])					*test_srand_seed()*
		When [seed] is given this sets the seed value used by
		`srand()`.  When omitted the test seed is removed.

==============================================================================
3. Assert functions				*assert-functions-details*


assert_beeps({cmd})					*assert_beeps()*
		Run {cmd} and add an error message to |v:errors| if it does
		NOT produce a beep or visual bell.
		Also see |assert_fails()| and |assert-return|.

		Can also be used as a |method|: >
			GetCmd()->assert_beeps()
<
							*assert_equal()*
assert_equal({expected}, {actual} [, {msg}])
		When {expected} and {actual} are not equal an error message is
		added to |v:errors| and 1 is returned.  Otherwise zero is
		returned |assert-return|.
		There is no automatic conversion, the String "4" is different
		from the Number 4.  And the number 4 is different from the
		Float 4.0.  The value of 'ignorecase' is not used here, case
		always matters.
		When {msg} is omitted an error in the form "Expected
		{expected} but got {actual}" is produced.
		Example: >
	assert_equal('foo', 'bar')
<		Will result in a string to be added to |v:errors|:
	test.vim line 12: Expected 'foo' but got 'bar' ~

		Can also be used as a |method|, the base is passed as the
		second argument: >
			mylist->assert_equal([1, 2, 3])

<							*assert_equalfile()*
assert_equalfile({fname-one}, {fname-two} [, {msg}])
		When the files {fname-one} and {fname-two} do not contain
		exactly the same text an error message is added to |v:errors|.
		Also see |assert-return|.
		When {fname-one} or {fname-two} does not exist the error will
		mention that.
		Mainly useful with |terminal-diff|.

		Can also be used as a |method|: >
			GetLog()->assert_equalfile('expected.log')

assert_exception({error} [, {msg}])			*assert_exception()*
		When v:exception does not contain the string {error} an error
		message is added to |v:errors|.  Also see |assert-return|.
		This can be used to assert that a command throws an exception.
		Using the error number, followed by a colon, avoids problems
		with translations: >
			try
			  commandthatfails
			  call assert_false(1, 'command should have failed')
			catch
			  call assert_exception('E492:')
			endtry
<
							*assert_fails()*
assert_fails({cmd} [, {error} [, {msg} [, {lnum} [, {context}]]]])
		Run {cmd} and add an error message to |v:errors| if it does
		NOT produce an error or when {error} is not found in the
		error message.  Also see |assert-return|.

		When {error} is a string it must be found literally in the
		first reported error. Most often this will be the error code,
		including the colon, e.g. "E123:". >
			assert_fails('bad cmd', 'E987:')
<
		When {error} is a |List| with one or two strings, these are
		used as patterns.  The first pattern is matched against the
		first reported error: >
			assert_fails('cmd', ['E987:.*expected bool'])
<		The second pattern, if present, is matched against the last
		reported error.
		If there is only one error then both patterns must match. This
		can be used to check that there is only one error.
		To only match the last error use an empty string for the first
		error: >
			assert_fails('cmd', ['', 'E987:'])
<
		If {msg} is empty then it is not used.  Do this to get the
		default message when passing the {lnum} argument.

		When {lnum} is present and not negative, and the {error}
		argument is present and matches, then this is compared with
		the line number at which the error was reported. That can be
		the line number in a function or in a script.

		When {context} is present it is used as a pattern and matched
		against the context (script name or function name) where
		{lnum} is located in.

		Note that beeping is not considered an error, and some failing
		commands only beep.  Use |assert_beeps()| for those.

		Can also be used as a |method|: >
			GetCmd()->assert_fails('E99:')

assert_false({actual} [, {msg}])			*assert_false()*
		When {actual} is not false an error message is added to
		|v:errors|, like with |assert_equal()|.
		Also see |assert-return|.
		A value is false when it is zero. When {actual} is not a
		number the assert fails.
		When {msg} is omitted an error in the form
		"Expected False but got {actual}" is produced.

		Can also be used as a |method|: >
			GetResult()->assert_false()

assert_inrange({lower}, {upper}, {actual} [, {msg}])	 *assert_inrange()*
		This asserts number and |Float| values.  When {actual}  is lower
		than {lower} or higher than {upper} an error message is added
		to |v:errors|.  Also see |assert-return|.
		When {msg} is omitted an error in the form
		"Expected range {lower} - {upper}, but got {actual}" is
		produced.

								*assert_match()*
assert_match({pattern}, {actual} [, {msg}])
		When {pattern} does not match {actual} an error message is
		added to |v:errors|.  Also see |assert-return|.

		{pattern} is used as with |=~|: The matching is always done
		like 'magic' was set and 'cpoptions' is empty, no matter what
		the actual value of 'magic' or 'cpoptions' is.

		{actual} is used as a string, automatic conversion applies.
		Use "^" and "$" to match with the start and end of the text.
		Use both to match the whole text.

		When {msg} is omitted an error in the form
		"Pattern {pattern} does not match {actual}" is produced.
		Example: >
	assert_match('^f.*o$', 'foobar')
<		Will result in a string to be added to |v:errors|:
	test.vim line 12: Pattern '^f.*o$' does not match 'foobar' ~

		Can also be used as a |method|: >
			getFile()->assert_match('foo.*')
<
							*assert_notequal()*
assert_notequal({expected}, {actual} [, {msg}])
		The opposite of `assert_equal()`: add an error message to
		|v:errors| when {expected} and {actual} are equal.
		Also see |assert-return|.

		Can also be used as a |method|: >
			mylist->assert_notequal([1, 2, 3])

<							*assert_notmatch()*
assert_notmatch({pattern}, {actual} [, {msg}])
		The opposite of `assert_match()`: add an error message to
		|v:errors| when {pattern} matches {actual}.
		Also see |assert-return|.

		Can also be used as a |method|: >
			getFile()->assert_notmatch('bar.*')


assert_report({msg})					*assert_report()*
		Report a test failure directly, using {msg}.
		Always returns one.

		Can also be used as a |method|: >
			GetMessage()->assert_report()


assert_true({actual} [, {msg}])				*assert_true()*
		When {actual} is not true an error message is added to
		|v:errors|, like with |assert_equal()|.
		Also see |assert-return|.
		A value is TRUE when it is a non-zero number.  When {actual}
		is not a number the assert fails.
		When {msg} is omitted an error in the form "Expected True but
		got {actual}" is produced.

		Can also be used as a |method|: >
			GetResult()->assert_true()
<

 vim:tw=78:ts=8:noet:ft=help:norl: