[rfc][icedtea-web] allow internal server frm makefile

[Jiri Vanek]

On 04/19/2013 03:27 PM, Adam Domurad wrote:
> On 04/19/2013 07:38 AM, Jiri Vanek wrote:
>> hi!
>> I'm running internal server - used for reproducers runs - also outside of testengine to test some particular cases.
>> This command, can be maybe done more easy accessible via makefile?
>>
>> J.
>
> OK it does make sense.
>
>> diff -r fbb6b3605538 Makefile.am
>> --- a/Makefile.am Thu Apr 04 11:21:04 2013 +0200
>> +++ b/Makefile.am Fri Apr 19 13:24:17 2013 +0200
>> @@ -641,7 +641,7 @@
>> mkdir -p stamps && \
>> touch $@
>>
>> -stamps/netx-dist-tests-prepare-reproducers.stamp: stamps/junit-jnlp-dist-dirs stamps/liveconnect-dist.stamp
>> +stamps/netx-dist-tests-prepare-reproducers.stamp: stamps/junit-jnlp-dist-dirs stamps/liveconnect-dist.stamp stamps/netx-dist.stamp stamps/plugin.stamp
>
> Smuggled ? Looks OK though.
>
>> types=($(ALL_NONCUSTOM_REPRODUCERS)); \
>> for which in "$${types[@]}" ; do \
>> . $(abs_top_srcdir)/NEW_LINE_IFS ; \
>> @@ -754,7 +754,7 @@
>> test-extensions-source-files.txt:
>> find $(TEST_EXTENSIONS_SRCDIR) -name '*.java' | sort > $@
>>
>> -stamps/test-extensions-compile.stamp: stamps/junit-jnlp-dist-dirs test-extensions-source-files.txt
>> +stamps/test-extensions-compile.stamp: stamps/netx-dist.stamp stamps/plugin.stamp stamps/junit-jnlp-dist-dirs test-extensions-source-files.txt
>> mkdir -p $(TEST_EXTENSIONS_DIR);
>> mkdir -p $(NETX_TEST_DIR);
>> ln -s $(TEST_EXTENSIONS_DIR) $(TEST_EXTENSIONS_COMPATIBILITY_SYMLINK);
>> @@ -777,7 +777,7 @@
>> mkdir -p stamps && \
>> touch $@
>>
>> -stamps/compile-reproducers-testcases.stamp: stamps/netx.stamp stamps/junit-jnlp-dist-dirs \
>> +stamps/compile-reproducers-testcases.stamp: stamps/netx-dist.stamp stamps/plugin.stamp stamps/junit-jnlp-dist-dirs \
>> test-extensions-source-files.txt stamps/test-extensions-compile.stamp stamps/test-extensions-tests-compile.stamp
>
> Also smuggled ?
>
>> types=($(ALL_REPRODUCERS)); \
>> for which in "$${types[@]}" ; do \
>> @@ -1380,6 +1380,20 @@
>> exit 5
>> endif
>>
>> +run-test-server-on-44321: stamps/netx.stamp stamps/junit-jnlp-dist-dirs stamps/netx-dist-tests-import-cert-to-public \
>> + stamps/test-extensions-compile.stamp stamps/compile-reproducers-testcases.stamp $(JUNIT_RUNNER_JAR) stamps/copy-reproducers-resources.stamp
>> + cd $(TEST_EXTENSIONS_DIR) ; \
>> + CLASSPATH=$(NETX_DIR)/lib/classes.jar:$(JUNIT_JAR):$(JUNIT_RUNNER_JAR):.:$(TEST_EXTENSIONS_TESTS_DIR) \
>> + $(BOOT_DIR)/bin/java $(REPRODUCERS_DPARAMETERS) \
>> + -Xbootclasspath:$(RUNTIME) net.sourceforge.jnlp.ServerAccess
>> +
>> +run-test-server-on-random-port: stamps/netx.stamp stamps/junit-jnlp-dist-dirs stamps/netx-dist-tests-import-cert-to-public \
>> + stamps/test-extensions-compile.stamp stamps/compile-reproducers-testcases.stamp $(JUNIT_RUNNER_JAR) stamps/copy-reproducers-resources.stamp
>> + cd $(TEST_EXTENSIONS_DIR) ; \
>> + CLASSPATH=$(NETX_DIR)/lib/classes.jar:$(JUNIT_JAR):$(JUNIT_RUNNER_JAR):.:$(TEST_EXTENSIONS_TESTS_DIR) \
>> + $(BOOT_DIR)/bin/java $(REPRODUCERS_DPARAMETERS) \
>> + -Xbootclasspath:$(RUNTIME) net.sourceforge.jnlp.ServerAccess randomport
>> +
>> clean-netx-tests: clean-netx-unit-tests clean-junit-runner clean-netx-dist-tests clean-test-code-coverage-jacoco clean-test-code-coverage
>> if [ -e $(TESTS_DIR)/netx ]; then \
>> rmdir $(TESTS_DIR)/netx ; \
>> diff -r fbb6b3605538 tests/test-extensions/net/sourceforge/jnlp/ServerAccess.java
>> --- a/tests/test-extensions/net/sourceforge/jnlp/ServerAccess.java Thu Apr 04 11:21:04 2013 +0200
>> +++ b/tests/test-extensions/net/sourceforge/jnlp/ServerAccess.java Fri Apr 19 13:24:17 2013 +0200
>> @@ -138,8 +138,10 @@
>> System.exit(0);
>> } else {
>> int port = 44321;
>> - if (args.length > 0) {
>> - port=new Integer(args[0]);
>> + if (args.length > 0 && args[0].equalsIgnoreCase("randomport")) {
>> + port = findFreePort();
>> + } else if (args.length > 0) {
>> + port = new Integer(args[0]);
>> }
>> getIndependentInstance(port);
>> while (true) {
>
> Looks good to me, I could see myself using it as well. OK to push.
>
> Thanks,
> -Adam
Going to push, thanx for review.
The "smuggled" parts are not smuggled, but  real fixes. The affected targets wee working just because they were called from reprroducers machinery "after" new depndences.
so whnen  I called them from new, target, without "the parts" then they started to fail.

Still ok?