[Date Prev][Date Next][Thread Prev][Thread Next]
- Subject: Re: 'struct' library formal releases or stable URLs
- From: "Thomas Harning Jr." <harningt@...>
- Date: Wed, 18 Mar 2009 17:24:35 -0400
On Wed, Mar 18, 2009 at 3:24 PM, Thomas Harning Jr. <email@example.com> wrote:
> On Mon, Feb 23, 2009 at 10:19 AM, Jerome Vuarand
> <firstname.lastname@example.org> wrote:
>> I'm using Roberto Ierusalimschy's struct library in some of my
>> projects. I have a build system that automatically downloads it from
>> the struct website  and patch it on the fly for my needs. I noticed
>> that the file changed recently, and my patches fail to apply. This is
>> not a big concern, I can update them. What I'd like to know is if you
>> plan to do a formal (versionned) release anytime soon. Otherwise is
>> there another stable url where I could retrieve older versions ?
> Having a versioned struct library w/ changelogs would be nice...
> I also noticed the docs no longer match the code... struct.size has
> been removed.
> Also... I have made some additional patches to support BER-style
> encoded number and tags.
> Attached is a patch to the latest struct.c out there for the following:
> * Add 't' for encoding ASN.1 tag values
> * Accepts an integer as the whole representation of the tag
> (includes class, constructed or not, and tag-value)
> * Add 'G' and 'g' for encoding ASN.1 BER lengths
> * 'G' expects a number after it to specify the minimum size
> * 'g' uses a number > 0 after it to specify the required size (too
> large results in error)
> * NOTE: t, G, and g ignore endian-ness as ASN.1 is supposed to be
> endian agnostic
> * Correct spelling in main comment
> * Re-add b_size (with the added elements, t, g0 and G are marked
> to have an undefined size)
> * Removes alignment usage of 's' value - didn't think that this was a
> correct implementation...
> If there are any alterations to these that would be
> important/useful... please let me know.
> If there is any credit assigned anywhere, please assign it to:
> Thomas Harning Jr <email@example.com> - TrustBearer Labs
> since this patchwork was performed during employment for them.
I forgot to add the change that used a signed version of 'optsize' to
handle the minimum length that 'G' uses. I don't expect option sizes
of things to wrap around the signed/unsigned range....
Thomas Harning Jr.
Description: Binary data