GNU bug report logs - #35880
[PATCH 0/7] Lzip support for 'guix publish' and 'guix substitute'

Previous Next

Full log

View this message in rfc822 format

From: Ludovic Courtès <ludo <at> gnu.org>
To: Pierre Neidhardt <mail <at> ambrevar.xyz>
Cc: 35880-done <at> debbugs.gnu.org
Subject: [bug#35880] [PATCH 1/7] lzlib: Add 'make-lzip-input-port/compressed'.
Date: Tue, 28 May 2019 15:47:13 +0200

Pierre Neidhardt <mail <at> ambrevar.xyz> skribis:

> Ludovic Courtès <ludo <at> gnu.org> writes:

[...]

>>> The thing is that we are not using lzlib as it is meant to be used
>>> (i.e. with the finish* functions) because of the functional approach of
>>> the binary ports which don't really play well with the procedural
>>> approach of the C library.
>>
>> I think we’re using it the way it’s meant to be used, roughly along the
>> lines of the examples of its manual (info "(lzlib) Examples").
>>
>> (I/O ports are not very “functional”.)
>
> In the sense that we define  "read!" and "write!" functions which don't
> allow us to call the "finish" functions properly.
>
> So maybe we are following the doc too "roughly" :p
>
> This has multiple implications, e.g. it impedes support for multiple
> members, (which is OK as long as we don't accept archives produced by a
> third-party) and more importantly it lifts the guarantee that the
> library is going to work as per the documentation.

I’m not sure I follow.  I think ‘make-lzip-input-port/compressed’
corresponds to Example 2 in the manual (info "(lzlib) Examples"),
‘make-lzip-output-port’ corresponds to Example 1, and
‘make-lzip-input-port’ corresponds to Example 4 (with the exception that
‘lzread!’ doesn’t call ‘lz-decompress-finished?’, but it has other means
to tell whether we’re done processing input.)

What do you think is not done as documented?

Thanks,
Ludo’.

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.