Patchwork lib/oe/lsb: Add basic LSB functions

login
register
mail settings
Submitter Richard Purdie
Date July 26, 2012, 1:05 p.m.
Message ID <1343307951.29991.27.camel@ted>
Download mbox | patch
Permalink /patch/33123/
State New
Headers show

Comments

Richard Purdie - July 26, 2012, 1:05 p.m.
This code was written by Christopher Larson <chris_larson@mentor.com> and
allows generation of the LSB release data based upon the lsb_release
command. It also includes a helper function to generate a string
representing a given distribution.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
---
ml@communistcode.co.uk - July 28, 2012, 9:03 a.m.
On 26/07/2012 14:05, Richard Purdie wrote:
> This code was written by Christopher Larson <chris_larson@mentor.com> and
> allows generation of the LSB release data based upon the lsb_release
> command. It also includes a helper function to generate a string
> representing a given distribution.
>
> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
> ---
> diff --git a/meta/lib/oe/lsb.py b/meta/lib/oe/lsb.py
> new file mode 100644
> index 0000000..a774169
> --- a/dev/null
> +++ b/meta/lib/oe/lsb.py
> @@ -0,0 +1,28 @@
> +def release_dict():
> +    """Return the output of lsb_release -a as a dictionary"""
> +    from subprocess import PIPE
> +
> +    try:
> +        output, err = bb.process.run(['lsb_release', '-a'], stderr=PIPE)
> +    except bb.process.CmdError as exc:
> +        return
> +
> +    data = {}
> +    for line in output.splitlines():
> +        try:
> +            key, value = line.split(":\t", 1)
> +        except ValueError:
> +            continue
> +        else:
> +            data[key] = value
> +    return data
> +
> +def distro_identifier(adjust_hook=None):
> +    """Return a distro identifier string based upon lsb_release -ri,
> +       with optional adjustment via a hook"""
> +
> +    lsb_data = release_dict()
> +    distro_id, release = lsb_data['Distributor ID'], lsb_data['Release']
> +    if adjust_hook:
> +        distro_id, release = adjust_hook(distro_id, release)
> +    return '{0}-{1}'.format(distro_id, release)
>
>
>
> _______________________________________________
> Openembedded-core mailing list
> Openembedded-core@lists.openembedded.org
> http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core

There is an unchecked error case here. I just had to fix this issue:

ERROR: Execution of event handler 'oe_import_eh' failed
Traceback (most recent call last):
   File "oe_import_eh(e)", line 4, in 
oe_import_eh(e=<bb.event.ConfigParsed object at 0x3834f90>)
   File "/mnt/yocto/poky.git/meta/lib/oe/lsb.py", line 25, in 
distro_identifier(adjust_hook=None):
          lsb_data = release_dict()
     >    distro_id, release = lsb_data['Distributor ID'], 
lsb_data['Release']
          if adjust_hook:
TypeError: 'NoneType' object has no attribute '__getitem__'


by installing lsb_release. Archlinux 64.

Should it check to see if it installed first and error otherwise? It 
looks like it tries to do it here:

+    try:
+        output, err = bb.process.run(['lsb_release', '-a'], stderr=PIPE)
+    except bb.process.CmdError as exc:
+        return


but it must be missing something.

Regards,
Jack
Martin Jansa - July 28, 2012, 9:16 a.m.
On Sat, Jul 28, 2012 at 10:03:11AM +0100, Jack Mitchell wrote:
> On 26/07/2012 14:05, Richard Purdie wrote:
> > This code was written by Christopher Larson <chris_larson@mentor.com> and
> > allows generation of the LSB release data based upon the lsb_release
> > command. It also includes a helper function to generate a string
> > representing a given distribution.
> >
> > Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
> > ---
> > diff --git a/meta/lib/oe/lsb.py b/meta/lib/oe/lsb.py
> > new file mode 100644
> > index 0000000..a774169
> > --- a/dev/null
> > +++ b/meta/lib/oe/lsb.py
> > @@ -0,0 +1,28 @@
> > +def release_dict():
> > +    """Return the output of lsb_release -a as a dictionary"""
> > +    from subprocess import PIPE
> > +
> > +    try:
> > +        output, err = bb.process.run(['lsb_release', '-a'], stderr=PIPE)
> > +    except bb.process.CmdError as exc:
> > +        return
> > +
> > +    data = {}
> > +    for line in output.splitlines():
> > +        try:
> > +            key, value = line.split(":\t", 1)
> > +        except ValueError:
> > +            continue
> > +        else:
> > +            data[key] = value
> > +    return data
> > +
> > +def distro_identifier(adjust_hook=None):
> > +    """Return a distro identifier string based upon lsb_release -ri,
> > +       with optional adjustment via a hook"""
> > +
> > +    lsb_data = release_dict()
> > +    distro_id, release = lsb_data['Distributor ID'], lsb_data['Release']
> > +    if adjust_hook:
> > +        distro_id, release = adjust_hook(distro_id, release)
> > +    return '{0}-{1}'.format(distro_id, release)
> >
> >
> >
> > _______________________________________________
> > Openembedded-core mailing list
> > Openembedded-core@lists.openembedded.org
> > http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core
> 
> There is an unchecked error case here. I just had to fix this issue:
> 
> ERROR: Execution of event handler 'oe_import_eh' failed
> Traceback (most recent call last):
>    File "oe_import_eh(e)", line 4, in 
> oe_import_eh(e=<bb.event.ConfigParsed object at 0x3834f90>)
>    File "/mnt/yocto/poky.git/meta/lib/oe/lsb.py", line 25, in 
> distro_identifier(adjust_hook=None):
>           lsb_data = release_dict()
>      >    distro_id, release = lsb_data['Distributor ID'], 
> lsb_data['Release']
>           if adjust_hook:
> TypeError: 'NoneType' object has no attribute '__getitem__'
> 
> 
> by installing lsb_release. Archlinux 64.
> 
> Should it check to see if it installed first and error otherwise? It 
> looks like it tries to do it here:
> 
> +    try:
> +        output, err = bb.process.run(['lsb_release', '-a'], stderr=PIPE)
> +    except bb.process.CmdError as exc:
> +        return
> 
> 
> but it must be missing something.

Same here, interesting is that it failed only during 2nd build, 1st
with this commit applied finished fine.

Cheers,
Martin Jansa - July 28, 2012, 9:37 a.m.
On Sat, Jul 28, 2012 at 11:16:14AM +0200, Martin Jansa wrote:
> > by installing lsb_release. Archlinux 64.
> > 
> > Should it check to see if it installed first and error otherwise? It 
> > looks like it tries to do it here:
> > 
> > +    try:
> > +        output, err = bb.process.run(['lsb_release', '-a'], stderr=PIPE)
> > +    except bb.process.CmdError as exc:
> > +        return
> > 
> > 
> > but it must be missing something.
> 
> Same here, interesting is that it failed only during 2nd build, 1st
> with this commit applied finished fine.

Ah no, it was there also first time I've overlooked it as it's not
fatal.
ml@communistcode.co.uk - July 28, 2012, 9:58 a.m.
On 28/07/2012 10:37, Martin Jansa wrote:
> On Sat, Jul 28, 2012 at 11:16:14AM +0200, Martin Jansa wrote:
>>> by installing lsb_release. Archlinux 64.
>>>
>>> Should it check to see if it installed first and error otherwise? It
>>> looks like it tries to do it here:
>>>
>>> +    try:
>>> +        output, err = bb.process.run(['lsb_release', '-a'], stderr=PIPE)
>>> +    except bb.process.CmdError as exc:
>>> +        return
>>>
>>>
>>> but it must be missing something.
>>
>> Same here, interesting is that it failed only during 2nd build, 1st
>> with this commit applied finished fine.
>
> Ah no, it was there also first time I've overlooked it as it's not
> fatal.
>

It was fatal for me as it trashed my sstate folder naming and then 
refused to build some packages. The directory in my sstate was created as:

${NATIVELSBSTRING}

Then when Bitbake tried to use shared state it expanded NATIVELSBSTRING 
to nothing and failed.
Denys Dmytriyenko - July 30, 2012, 7:50 p.m.
On Sat, Jul 28, 2012 at 10:58:16AM +0100, Jack Mitchell wrote:
> On 28/07/2012 10:37, Martin Jansa wrote:
> >On Sat, Jul 28, 2012 at 11:16:14AM +0200, Martin Jansa wrote:
> >>>by installing lsb_release. Archlinux 64.
> >>>
> >>>Should it check to see if it installed first and error otherwise? It
> >>>looks like it tries to do it here:
> >>>
> >>>+    try:
> >>>+        output, err = bb.process.run(['lsb_release', '-a'], stderr=PIPE)
> >>>+    except bb.process.CmdError as exc:
> >>>+        return
> >>>
> >>>
> >>>but it must be missing something.
> >>
> >>Same here, interesting is that it failed only during 2nd build, 1st
> >>with this commit applied finished fine.
> >
> >Ah no, it was there also first time I've overlooked it as it's not
> >fatal.
> >
> 
> It was fatal for me as it trashed my sstate folder naming and then
> refused to build some packages. The directory in my sstate was
> created as:
> 
> ${NATIVELSBSTRING}
> 
> Then when Bitbake tried to use shared state it expanded
> NATIVELSBSTRING to nothing and failed.

I have the same issue here - Gentoo/64. Can we get it fixed in the master, 
please? Thanks.
Richard Purdie - July 31, 2012, 10:47 a.m.
On Mon, 2012-07-30 at 15:50 -0400, Denys Dmytriyenko wrote:
> On Sat, Jul 28, 2012 at 10:58:16AM +0100, Jack Mitchell wrote:
> > On 28/07/2012 10:37, Martin Jansa wrote:
> > >On Sat, Jul 28, 2012 at 11:16:14AM +0200, Martin Jansa wrote:
> > >>>by installing lsb_release. Archlinux 64.
> > >>>
> > >>>Should it check to see if it installed first and error otherwise? It
> > >>>looks like it tries to do it here:
> > >>>
> > >>>+    try:
> > >>>+        output, err = bb.process.run(['lsb_release', '-a'], stderr=PIPE)
> > >>>+    except bb.process.CmdError as exc:
> > >>>+        return
> > >>>
> > >>>
> > >>>but it must be missing something.
> > >>
> > >>Same here, interesting is that it failed only during 2nd build, 1st
> > >>with this commit applied finished fine.
> > >
> > >Ah no, it was there also first time I've overlooked it as it's not
> > >fatal.
> > >
> > 
> > It was fatal for me as it trashed my sstate folder naming and then
> > refused to build some packages. The directory in my sstate was
> > created as:
> > 
> > ${NATIVELSBSTRING}
> > 
> > Then when Bitbake tried to use shared state it expanded
> > NATIVELSBSTRING to nothing and failed.
> 
> I have the same issue here - Gentoo/64. Can we get it fixed in the master, 
> please? Thanks.

I've pushed something which should resolve the problem and at least set
this to a known string ('Unknown').

Cheers,

Richard

Patch

diff --git a/meta/lib/oe/lsb.py b/meta/lib/oe/lsb.py
new file mode 100644
index 0000000..a774169
--- a/dev/null
+++ b/meta/lib/oe/lsb.py
@@ -0,0 +1,28 @@ 
+def release_dict():
+    """Return the output of lsb_release -a as a dictionary"""
+    from subprocess import PIPE
+
+    try:
+        output, err = bb.process.run(['lsb_release', '-a'], stderr=PIPE)
+    except bb.process.CmdError as exc:
+        return
+
+    data = {}
+    for line in output.splitlines():
+        try:
+            key, value = line.split(":\t", 1)
+        except ValueError:
+            continue
+        else:
+            data[key] = value
+    return data
+
+def distro_identifier(adjust_hook=None):
+    """Return a distro identifier string based upon lsb_release -ri,
+       with optional adjustment via a hook"""
+
+    lsb_data = release_dict()
+    distro_id, release = lsb_data['Distributor ID'], lsb_data['Release']
+    if adjust_hook:
+        distro_id, release = adjust_hook(distro_id, release)
+    return '{0}-{1}'.format(distro_id, release)