Intro

There is implemented simple test enviroment. There are test which requires 'valgrind'.

• Debian based distro: apt-get install valgrind

You can run it using ./runTest TEST_NAME, or use help: ./runTest -h to see all options.

fruh@htfuws:~/projects/BitPunch$ ./runTest -h
Usage: runTest [options]

Options:
  --version             show program's version number and exit
  -h, --help            show this help message and exit
  -a, --all             Run all registered tests
  -t RUN_TEST, --test=RUN_TEST
                                                Run named tests
  -r RUN_LEVEL, --runLevel=RUN_LEVEL
                                                Run all tests <= runLevel
  -v VERBOSE_LEVEL, --verbose=VERBOSE_LEVEL
                                                Verbose level, can be ERROR, WARNING, INFO, DEBUG
  -l, --list            List available tests

This test enviroment creates on every run snapshot of current workspace and copies it into runtest/results/ISO_TIME_FORMAT/ folder.

Test is considered as PASSED when return code is 0 (zero) and 'ERROR' string not occured at the beggining of the line. Otherwise, test is considered as FAILED.

Example

Here is the guide, how to create dummy test using C and bash sources. C source represents test body and bash source represents control section.

We can create new suite called "foosuite" or use existing. "Create" means to make directory "foosuite" in our test root:

• runtest/tests/foosuite/

Then we have to create our source files:

• runtest/tests/foosuite/testFoo.sh (with EXECUTABLE RIGHTS)
• runtest/tests/foosuite/testFoo.c

Now we have to register this test, it means we have to add following record into TESTS python structure defined in:

• runtest/tests/register.py

  TESTS = {
          "testFooName" : {
                  "suite" : "foosuite",
                  "file" : "testFoo.sh",
                  "args" : "testFoo fooarg2",
                  "runLevel" : "regular",
          },
          ... heere are other tests ...
  }
  

Now we can set up body of the control bash source. The first lines are required (until cd $1 (included)), first argument is still set to the path of the current workspace snapshot (this is done automatically by test enviroment):

• runtest/tests/foosuite/testFoo.sh

  #!/bin/bash
  if [ "$1" = "" ]; then
          echo "First argument is result folder."
          exit 1
  fi
  
  TESTFILE_DIR=$(pwd)
  echo "INFO: $TESTFILE_DIR"
  
  # now we are in current snapshot
  cd "$1"
  
  APP="./dist/test/fooBuild"
  
  # control body
  echo "INFO: fooarg1: $2"
  echo "INFO: fooarg2: $3"
  
  echo "WARNING: warning message"
  
  echo "INFO: cleaning build..."
  make clean
  if [ $? != 0 ]; then
          echo "ERROR: make clean"
          exit 1
  fi
  
  echo "INFO: building static-lib..."
  make static-lib
  if [ $? != 0 ]; then
          echo "ERROR: building target static-lib"
          exit 2
  fi
  gcc -Isrc/ $TESTFILE_DIR/testFoo.c ./dist/libbpumecs.a -o $APP || exit 1
  
  # run app
  $APP
  
  # return code
  exit $?
  

And now we set test body:

• runtest/tests/foosuite/testFoo.c

  #include <bitpunch/bitpunch.h>
  
  #include <stdio.h>
  
  int main(int argc, char **argv) {
          int rc = 0;
          // MUST BE NULL
          BPU_T_Mecs_Ctx *ctx = NULL;
          BPU_T_UN_Mecs_Params params;
  
          /***************************************/
          // mce initialisation t = 50, m = 11
          if (BPU_mecsInitParamsGoppa(&params, 11, 50, 0)) {
                  return 1;
          }
          fprintf(stderr, "Basic GOPPA Initialisation...\n");
          if (BPU_mecsInitCtx(&ctx, &params, BPU_EN_MECS_BASIC_GOPPA)) {
                  return 1;
          }
          BPU_mecsFreeCtx(&ctx);
          BPU_mecsFreeParamsGoppa(&params);
  
          return rc;
  }
  

We can see our registered test in the list of the tests:

• ./runTest -l

  fruh@htfuws:~/projects/BitPunch$ ./runTest -l
  Test: testMemRun                     runLevel: regular
  Test: testFooName                    runLevel: regular
  Test: testMemLeakWithH               runLevel: regular
  Test: testBuild                      runLevel: express
  Test: testMemLeakWoH                 runLevel: regular
  Test: testLibSize                    runLevel: express
  

We can run our dummy test using:

• ./runTest testFooName

   fruh@htfuws:~/projects/BitPunch$ ./runTest testFooName
   Tests to run (1): testFooName
   ======== BEGIN::testFooName ========
   INFO: /media/fruh/data/projects/BitPunch/runtest/tests/foosuite
   INFO: fooarg1: testFoo
   INFO: fooarg2: fooarg2
   WARNING: warning message
   INFO: cleaning build...
   INFO: building static-lib...
   PASSED
   --------   END::testFooName --------
   ======== RESULTS ========
   Result folder: /media/fruh/data/projects/BitPunch/runtest/results/2015-05-17T23:51:36.709632
   FAILED: 0
   SKIPPED: 0
   PASSED: 1
   -------------------------