[PATCH] 35 Added option to build without Javadoc.
The build scripts have been modified to provide an option to build without Javadoc to speed up development builds. The option can be used as follows: compose_pki_core_packages --without-javadoc hybrid_rpms Ticket #111 -- Endi S. Dewata
On 03/23/12 12:53, Endi Sukma Dewata wrote:
The build scripts have been modified to provide an option to build without Javadoc to speed up development builds. The option can be used as follows:
compose_pki_core_packages --without-javadoc hybrid_rpms
Ticket #111
_______________________________________________ Pki-devel mailing list Pki-devel@redhat.com https://www.redhat.com/mailman/listinfo/pki-devel CONDITIONAL ACK with CAVEATS:
The following invocations succeeded building the RPMS without building the javadoc RPMS (it should have failed with a Usage message): * pki/scripts/compose_pki_core_packages *--w* hybrid_rpms * pki/scripts/compose_pki_core_packages *--without* hybrid_rpms (note that '*--without*' was used instead of '*--without-javadoc*') while the following invocations failed as expected and issued a Usage message: * pki/scripts/compose_pki_core_packages *--foo* hybrid_rpms * pki/scripts/compose_pki_core_packages *--without-foo* hybrid_rpms * pki/scripts/compose_pki_core_packages *--without-javadocs* hybrid_rpms (note that '*--without-javadocs*' (plural) was used instead of '*--without-javadoc*') *Please feel free to check-in this patch once this issue has been corrected.* Although I think that it is un-necessary to change code nor necessarily document this, the '*--without-javadoc*' option is /*ONLY*/ useful for building the '*pki-core*' component (no other '*compose*' scripts generate any 'javadoc' packages, and certain packages such as 'pki-tps' will NEVER generate javadocs as this is a non-java component). Additionally, this option would only be utilized when building 'RPMS' (e. g. - 'rpms', 'hybrid_rpms', and/or 'patched_rpms'; it is completely ignored when building 'srpm', 'hybrid_srpm', and/or 'patched_srpm') regardless of component. However, I don't think that invocation necessarily needs to fail which building SRPMS.
On 3/23/2012 8:48 PM, Matthew Harmsen wrote:
On 03/23/12 12:53, Endi Sukma Dewata wrote:
The build scripts have been modified to provide an option to build without Javadoc to speed up development builds. The option can be used as follows:
compose_pki_core_packages --without-javadoc hybrid_rpms
Ticket #111
CONDITIONAL ACK with CAVEATS:
The following invocations succeeded building the RPMS without building the javadoc RPMS (it should have failed with a Usage message):
* pki/scripts/compose_pki_core_packages *--w* hybrid_rpms * pki/scripts/compose_pki_core_packages *--without* hybrid_rpms (note that '*--without*' was used instead of '*--without-javadoc*')
while the following invocations failed as expected and issued a Usage message:
* pki/scripts/compose_pki_core_packages *--foo* hybrid_rpms * pki/scripts/compose_pki_core_packages *--without-foo* hybrid_rpms * pki/scripts/compose_pki_core_packages *--without-javadocs* hybrid_rpms (note that '*--without-javadocs*' (plural) was used instead of '*--without-javadoc*')
*Please feel free to check-in this patch once this issue has been corrected.*
It seems to be a feature of the getopt. According to getopt man page: Long options may be abbreviated, as long as the abbreviation is not ambiguous. There's already a request to allow disabling abbreviation, but it won't be fixed: http://sourceware.org/bugzilla/show_bug.cgi?id=6863 I've tested adding another option (e.g. --without-test). If you specify an abbreviated option (e.g. --without) it will find the first defined option that matches it (e.g. --without-javadoc) even thought it could be considered ambiguous. -- Endi S. Dewata
On 03/26/12 08:41, Endi Sukma Dewata wrote:
On 3/23/2012 8:48 PM, Matthew Harmsen wrote:
On 03/23/12 12:53, Endi Sukma Dewata wrote:
The build scripts have been modified to provide an option to build without Javadoc to speed up development builds. The option can be used as follows:
compose_pki_core_packages --without-javadoc hybrid_rpms
Ticket #111
CONDITIONAL ACK with CAVEATS:
The following invocations succeeded building the RPMS without building the javadoc RPMS (it should have failed with a Usage message):
* pki/scripts/compose_pki_core_packages *--w* hybrid_rpms * pki/scripts/compose_pki_core_packages *--without* hybrid_rpms (note that '*--without*' was used instead of '*--without-javadoc*')
while the following invocations failed as expected and issued a Usage message:
* pki/scripts/compose_pki_core_packages *--foo* hybrid_rpms * pki/scripts/compose_pki_core_packages *--without-foo* hybrid_rpms * pki/scripts/compose_pki_core_packages *--without-javadocs* hybrid_rpms (note that '*--without-javadocs*' (plural) was used instead of '*--without-javadoc*')
*Please feel free to check-in this patch once this issue has been corrected.*
It seems to be a feature of the getopt. According to getopt man page:
Long options may be abbreviated, as long as the abbreviation is not ambiguous.
There's already a request to allow disabling abbreviation, but it won't be fixed:
http://sourceware.org/bugzilla/show_bug.cgi?id=6863
I've tested adding another option (e.g. --without-test). If you specify an abbreviated option (e.g. --without) it will find the first defined option that matches it (e.g. --without-javadoc) even thought it could be considered ambiguous.
okay -- looks like there is no intention to fix this. Go ahead and check this in. -- Matt
On 3/26/2012 11:28 AM, Matthew Harmsen wrote:
It seems to be a feature of the getopt. According to getopt man page:
Long options may be abbreviated, as long as the abbreviation is not ambiguous.
There's already a request to allow disabling abbreviation, but it won't be fixed:
http://sourceware.org/bugzilla/show_bug.cgi?id=6863
I've tested adding another option (e.g. --without-test). If you specify an abbreviated option (e.g. --without) it will find the first defined option that matches it (e.g. --without-javadoc) even thought it could be considered ambiguous.
okay -- looks like there is no intention to fix this.
Go ahead and check this in.
Pushed to master. Thanks. -- Endi S. Dewata
participants (2)
-
Endi Sukma Dewata -
Matthew Harmsen