Openembedded Bitbake Development
 help / color / mirror / Atom feed
From: Richard Purdie <richard.purdie@linuxfoundation.org>
To: Paul Eggleton <paul.eggleton@linux.intel.com>
Cc: Chris Larson <clarson@kergoth.com>, bitbake-devel@lists.openembedded.org
Subject: Re: [PATCH 2/3] lib/bb/command.py: ensure setVariable only sets values as strings
Date: Wed, 12 Sep 2012 15:55:16 +0100	[thread overview]
Message-ID: <1347461716.11710.45.camel@ted> (raw)
In-Reply-To: <1987076.DYWImTfRFL@helios>

On Wed, 2012-09-12 at 13:29 +0100, Paul Eggleton wrote:
> On Tuesday 11 September 2012 15:57:24 Chris Larson wrote:
> > On Tue, Sep 11, 2012 at 9:52 AM, Paul Eggleton
> > <paul.eggleton@linux.intel.com> wrote:
> > > This is the interface Hob uses to set variable values in many instances,
> > > and at the moment it is possible that some of the values it passes are
> > > not strings. If a non-string value gets into the datastore it can
> > > trigger exceptions during parsing when we attempt to expand the variable
> > > and substitute in the non-string value.
> > > 
> > > This fixes using the meta-ti layer within Hob - it currently has a
> > > reference to BB_NUMBER_THREADS within a shell function and since this
> > > is a variable that Hob was setting from its configuration as an integer,
> > > due to the above this was triggering an ExpansionError.
> > 
> > This is probably a good change, but I think that failing expansion due
> > to a non-string value is a bad behavior in bitbake. I think it should
> > call str() on it as appropriate, as that's the most idiomatic
> > solution: Don't check if something is a string, don't require only
> > strings, but call str() to get your string
> 
> Hmm, I suppose so; I do still worry about existing python code assuming the 
> result of getVar() is either None or a string though. Richard tells me that we 
> have some non-string values mostly internal to BitBake that are considered 
> legal, so that obviously isn't always 100% true, but it's a fair assumption to 
> have made for most situations.

If its being used in variable expansion, it should be a string or
convertible to a string so I have to admit I was thinking we probably
should have a str() conversion in there as well. I'm hoping that doesn't
have much overhead for something that is already a string. I'd also be
concerned about turning None into "None" which is probably the biggest
worry.

Cheers,

Richard






  reply	other threads:[~2012-09-12 15:07 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-09-11 16:51 [PATCH 0/3] More Hob fixes Paul Eggleton
2012-09-11 16:51 ` [PATCH 1/3] hob: don't reorder layers list Paul Eggleton
2012-09-11 16:52 ` [PATCH 2/3] lib/bb/command.py: ensure setVariable only sets values as strings Paul Eggleton
2012-09-11 22:57   ` Chris Larson
2012-09-12 12:29     ` Paul Eggleton
2012-09-12 14:55       ` Richard Purdie [this message]
2012-09-11 16:52 ` [PATCH 3/3] hob: sort base image drop-down list Paul Eggleton
2012-09-14  8:27 ` [PATCH 0/3] More Hob fixes Richard Purdie

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1347461716.11710.45.camel@ted \
    --to=richard.purdie@linuxfoundation.org \
    --cc=bitbake-devel@lists.openembedded.org \
    --cc=clarson@kergoth.com \
    --cc=paul.eggleton@linux.intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox