File: README

package info (click to toggle)
evolution-data-server 1.6.3-5
  • links: PTS
  • area: main
  • in suites: etch-m68k
  • size: 59,384 kB
  • ctags: 43,218
  • sloc: ansic: 319,315; tcl: 30,499; xml: 19,166; sh: 18,776; perl: 11,529; cpp: 8,259; java: 7,653; makefile: 6,448; awk: 1,338; yacc: 1,103; sed: 772; cs: 505; lex: 134; asm: 14
file content (37 lines) | stat: -rw-r--r-- 1,784 bytes parent folder | download | duplicates (3)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
# $Id: README,v 1.1.1.1 2003/11/20 22:14:05 toshok Exp $

Use the scripts testall or testone to run all, or just one of the Java
tests.  You must be in this directory to run them.  For example,

        $ export LD_LIBRARY_PATH=/usr/local/Berkeley3.3/lib
	$ ./testone TestAppendRecno
	$ ./testall

The scripts will use javac and java in your path.  Set environment
variables $JAVAC and $JAVA to override this.  It will also and honor
any $CLASSPATH that is already set, prepending ../../../../classes to
it, which is where the test .class files are put, and where the DB
.class files can normally be found after a build on Unix and Windows.
If none of these variables are set, everything will probably work
with whatever java/javac is in your path.

To run successfully, you will probably need to set $LD_LIBRARY_PATH
to be the directory containing libdb_java-X.Y.so

As an alternative, use the --prefix=<DIR> option, a la configure
to set the top of the BerkeleyDB install directory.  This forces
the proper options to be added to $LD_LIBRARY_PATH.
For example,

	$ ./testone --prefix=/usr/include/BerkeleyDB TestAppendRecno
	$ ./testall --prefix=/usr/include/BerkeleyDB

The test framework is pretty simple.  Any <name>.java file in this
directory that is not mentioned in the 'ignore' file represents a
test.  If the test is not compiled successfully, the compiler output
is left in <name>.compileout .  Otherwise, the java program is run in
a clean subdirectory using as input <name>.testin, or if that doesn't
exist, /dev/null.  Output and error from the test run are put into
<name>.out, <name>.err .  If <name>.testout, <name>.testerr exist,
they are used as reference files and any differences are reported.
If either of the reference files does not exist, /dev/null is used.