You can execute Blitz from the command line by means of the
following command:
java -DBLITZ_HOME=%BLITZ_HOME% -cp %BLITZ_HOME%\lib\blitz.jar;%BLITZ_HOME%\lib\log4j-1.2.12.jar;%CLASSPATH% blitz.main.Blitz <arg1> <arg2> <arg3>
However, this is fairly long and difficult to remember, so several shell
scripts have been provided to simplify the process.
NOTE: If you run this command with no arguments (i.e. <arg1> ,
<arg2> , and <arg3> are omitted), Blitz will display a usage
screen to explain all of the options.
The basic process for instantiating and executing a test is to invoke
Blitz from the command line and provide, as its single argument,
the name of the root-level test file. This is accomplished through the use
of a batch script file.
Blitz.bat executes the test (scenario, suite, or case) in
the specified file, and produces a result summary file containing details of
the outcomes of all the individual test cases encountered during execution.
The usage for this script is:
Blitz <test file>
where <test file> [REQUIRED] is the name of a file
containing a testCase (".tst "), a testSuite (".ste "),
or a testScenario (".sno "). The script will execute the
test contained in the given file, and produce a detailed result with an
extension of ".result " in the directory containing the test
file.
Examples:
Blitz C:\MyTestDirectory\MyTest.tst – causes the execution
of the test case in file MyTest.tst from directory C:\MyTestDirectory .
On completion of the test, a result file with the name
MyTest.tst.result is written to the same directory.
Blitz MySuite.ste – causes the execution of the test suite
in file MySuite.ste to be executed from the "current"
directory. All nested test cases and test suites are assumed to be in the
same "current" directory, unless they explicitly override this default
setting. On completion of the test suite, the accumulated results are
detailed in a file with the name MySuite.ste.result (also in
the "current" directory).
Notes:
- If any validation, parsing, or execution error occurs, Blitz
will terminate and display an appropriate exception message on the screen.
- If the number of failures or errors exceeds the prescribed maximum (as
specified in a test scenario), Blitz will terminate early
with an appropriate message on the console.
- If the
keepBackups.outcome system property is set to
"Yes", and a result file for the specified test already exists, the old
result file will be backed up first. The backup file will have a
name like MyTest.tst.20030215.161503.330.result, where the embedded
numbers are of the form "YYYYMMDD.HHMMSS.MMM" representing the file’s
creation date and time to the nearest millisecond. This allows for
preservation of all result files during multiple runs of the same test.
If the system property is set to "No", then any previously existing result
file will be overwritten.
In planning mode, batch script files are used to perform administrative
tasks. These scripts do not actually execute the specified test files;
instead, they perform functions which provide a sanity check on the validity
of test files.
BlitzValidate.bat performs a quick evaluation of the
specified file to verify that it is well-formed and valid XML that conforms
to the Blitz schema. The usage for this script is:
BlitzValidate <test file>
where <test file> [REQUIRED] is the name of a file
containing a testCase (".tst "), a testSuite (".ste "),
or a testScenario (".sno "). The script will create
output on the console indicating whether the specified file is valid; if it
is not valid, an appropriate exception message is displayed on the console
to indicate the nature of the problem.
BlitzCertify.bat performs an evaluation of the specified file to
construct and display a fully-formed hierarchical test object, and to
display a description of any problems that might exist within the test. The
usage for this script is:
BlitzCertify <test file>
where <test file> [REQUIRED] is the name of a file containing a
testCase (".tst "), a testSuite (".ste "),
or a testScenario (".sno "). The script will produce
textual output describing the hierarchical test; if any error occurs, an
appropriate exception message is displayed on the console to indicate the
nature of the problem.
BlitzTrace.bat parses and analyzes the test in the specified
file to provide an overview of how the test will proceed when it is
executed. This does not actually execute the test so it cannot predict the
outcome of a test or determine the correctness of a test (although validity
of the XML file is checked); its only purpose is to provide a detailed view
of the environment and configuration under which a set of test cases will
run. The report generated by this utility, produced as a text file with a ".trace "
extension, can be used to verify that a scenario will do what was originally
expected.
Usage for this script is:
BlitzTrace [<detailLevel>] <test file>
where <detailLevel> [OPTIONAL] specifies the amount of
detail to be included in the file. Allowable values are as follows:
-terse | -t : minimum detail; only include test file
names and concurrency/repeat details
-normal | -n : normal detail; include test file names,
descriptions, and explicitly defined or inherited characteristics; exclude
anything that is undefined
-verbose | -v : maximum detail; include test file names,
descriptions, and all other characteristics, whether they are defined or
not
and <test file> [REQUIRED] is the name of a file that
contains a test scenario, a test suite, or a test case. Blitz
analyzes the given file and produces an identically named file with an
extension of ".trace ", located in the same directory as the
source file.
If the keepBackups.trace system property is set to "Yes",
and a trace file already exists for the named test, the old trace file will
be backed up first. The backup will have a name like
MyTest.tst.20030215.161503.330.trace , where the embedded numbers are
of the form "YYYYMMDD.HHMMSS.MMM" representing the file’s creation date and
time to the nearest millisecond. If the system property is set to "No", then
any previously existing trace file will be overwritten.
Example:
BlitzTrace –v C:\MyTestScenarios\MyTest.sno – causes
Blitz to analyze the specified file and create a verbose trace file
called C:\MyTestScenarios\MyTest.sno.trace .
BlitzPromote.bat converts the actual result from a previous
run of a test into an expected result for subsequent executions of the same
test. The usage for this script is:
BlitzPromote <source> [<target>]
where <source> [REQUIRED] is the name of a particular file
containing an actual result. If the last component of the pathname ends with
".act ", the object is assumed to be a file, and the utility
will attempt to convert only this file; otherwise, the object is treated as
a directory, and the utility attempts to convert all files in that directory
that end with the ".act " extension;
and <target> [OPTIONAL] is the name of a directory to which
the corresponding ".exp " file should be written; if omitted,
Blitz will store the result file(s) in same directory that
contained the source file(s).
Examples:
BlitzPromote C:\Actual\ThisFile.act C:\Expected – causes the
specific file "ThisFile.act " in the first named directory (C:\Actual )
to be copied as "ThisFile.exp " into the second named directory
(C:\Expected ); any pre-existing file with the name "ThisFile.exp "
is overwritten.
BlitzPromote C:\Actual C:\Expected – causes all ".act "
files in the first named directory (C:\Actual ) to be copied as
".exp " files in the second named directory (C:\Expected );
any pre-existing ".exp " files may be overwritten.
BlitzPromote C:\MyTestArea – causes all ".act "
files in the specified directory to be copied as ".exp " files
in the same directory. Any existing ".exp " files may be
overwritten.
TOP
|