similar to: [PATCH] Non-working support for GObject Introspection.

These are the changes I made to the spec file related to gobject bindings and gobject introspection. TBH I wasn't sure how to properly package this. Rich. diff --git a/libguestfs.spec b/libguestfs.spec index 095a848..1b6e43d 100644 --- a/libguestfs.spec +++ b/libguestfs.spec @@ -29,7 +29,7 @@ Summary: Access and modify virtual machine disk images Name: libguestfs Epoch:

This reverts commit 3f6ca541c7b24d4c86688a509582cb41a7e0078c. The original commit was reverted prematurely. --- generator/generator_actions.ml | 10 +++++----- generator/generator_checks.ml | 5 +++++ generator/generator_types.ml | 3 +++ 3 files changed, 13 insertions(+), 5 deletions(-) diff --git a/generator/generator_actions.ml b/generator/generator_actions.ml index

This is the first iteration of the GObject bindings. I have 'kicked the tyres' on these, meaning I have ensured that a bunch of basic manual tests work as expected. I'm in the process of adding more comprehensive tests. Here's an example simple javascript program which uses these bindings: === const Guestfs = imports.gi.Guestfs; print('Starting'); var g = new

Patch 1 cleans up the gobject tests. Patches 2 and 3 remote gtk-doc, which is slow and useless, and replace it with a simple manual page. https://bugzilla.redhat.com/show_bug.cgi?id=1465665 Rich.

Another test, the same as before.

Best of .. 4?

After enabling procmail debugging ...

Originally posted by Corentin Noël here: https://bugzilla.redhat.com/show_bug.cgi?id=1668307 -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-p2v converts physical machines to virtual machines. Boot with a live CD or over the network (PXE) and turn machines into KVM guests.

This is a repost so I can test my procmail script / automatic testing framework. The patch series is identical to v1. Rich.

Thanks for the quick review, I addressed the comments but couldn't run `make dist` because of some issue with `libguestfs.pot` while building. So please review it carefully (even if I followed the recommended steps to include Vala support as written here https://wiki.gnome.org/Projects/Vala/UpstreamGuide#Autotools_Integration ) Regards, Corentin >From

RConstOptString cannot return an error in the C api. This makes it a special case for the GObject api, as all other return types have a corresponding GError **err argument to return an error. This change removes this special case, and includes the possibility of an error return in the API. An error is indicated by setting *err to a non-NULL value. This change is in preparation for adding a close

I tried using Javascript (gjs) with GObject-Introspection and did a port of the create_disk example which is attached below. Quite a bit of time was spent looking up stuff in the GLib reference and guessing how certain idioms translate to gjs. I used the traditional add_drive_opts method since I haven't been able to find how to specify the optional arguments to add_drive_opts. How does that

This patch replaces patches 14 and 15 from my previous series. The gtk-doc output is now reasonable, and we can rely on an automatically generated guestfs-sections.txt. Matt

--- generator/generator_gobject.ml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/generator/generator_gobject.ml b/generator/generator_gobject.ml index e4c175b..48ddbf0 100644 --- a/generator/generator_gobject.ml +++ b/generator/generator_gobject.ml @@ -391,7 +391,7 @@ let generate_gobject_optargs_source name optargs flags () = pr "G_DEFINE_TYPE(%s, guestfs_%s,

I've summarised how the GObject bindings work below, which should hopefully help reviewing the generator/generated code. There are a couple of points in here I'm still not 100% happy with. Specifically the handling of FBuffer and the Cancellable flag. Both are explained below. I'm interested in suggestions. Return values: ************** All functions which can return an error

Here is the patch with the EXTRA_DIST moved. Note that the .deps file is useless without the .vapi file (that's why it was inside the if ENABLE_VAPIGEN ) Regards, Corentin --- .gitignore | 1 + configure.ac | 4 ++ gobject/Makefile.am | 17 +++++ gobject/libguestfs-gobject-1.0.deps | 2 + m4/vapigen.m4

Since we removed gtk doc, we might as well replace it with a manual page explaining the basics of how to run gjs. --- .gitignore | 2 ++ docs/guestfs-recipes.pod | 1 + erlang/examples/guestfs-erlang.pod | 1 + gobject/Makefile.am | 19 +++++++++++ gobject/guestfs-gobject.pod | 64 ++++++++++++++++++++++++++++++++++++++

The gobject bindings generate a large number of header files, which pollute /usr/include when installed. This patch moves them all into a guestfs-gobject/ subdirectory. guestfs-gobject.h remains in the same place. This change also moves generated source files into src/, because it makes the gobject directory a bit tidier. --- generator/Makefile.am | 3 +

--- generator/generator_actions.ml | 9 +++++---- 1 files changed, 5 insertions(+), 4 deletions(-) diff --git a/generator/generator_actions.ml b/generator/generator_actions.ml index 68a7bf6..fcf363f 100644 --- a/generator/generator_actions.ml +++ b/generator/generator_actions.ml @@ -4664,7 +4664,7 @@ This creates an ext2/3/4 filesystem on C<device> with an external journal on

This is my proposed alternative to the complicated test framework (https://www.redhat.com/archives/libguestfs/2015-August/msg00022.html). In this patch, when we want to run tests on installed packages we just copy in the installed binaries over the source directory, and run the test suite as normal. This is basically the same as the current (not used) contrib/make-check-on-installed.pl script,