Debian Bug report logs - #460350
[frontend] please support configuration option to make --display-info default

Toggle useless messages

Message #5 received at submit@bugs.debian.org (full text, mbox, reply):

From: David Bremner <bremner@unb.ca>

To: submit@bugs.debian.org

Subject: please support configuration option to make --display-info default

Date: Sat, 12 Jan 2008 11:39:43 +0100

Package: lintian
Version: 1.23.42
Severity: wishlist

--- Please enter the report below this line. ---

Hi; 

As a beginner packager, I find the -I option to lintian very helpful. I would 
like to be able to configure lintian to display-info by default.  Aliases do 
not help because I mostly use svn-buildpackage. 

--- System information. ---
Architecture: i386
Kernel:       Linux 2.6.22

Debian Release: lenny/sid
  500 unstable        localhost 
  500 unstable        debian.netcologne.de 
  500 testing         localhost 

--- Package information. ---
Depends                               (Version) | Installed
===============================================-+-==============
binutils                                        | 2.18.1~cvs20080103-1
diffstat                            (>= 1.27-1) | 1.45-2
dpkg-dev                           (>= 1.13.17) | 1.14.15
file                                            | 4.21-4
gettext                               (>= 0.16) | 0.17-2
intltool-debian                                 | 0.35.0+20060710.1
libdigest-md5-perl                              | 
 OR perl                               (>> 5.8) | 5.8.8-12
libparse-debianchangelog-perl          (>= 0.6) | 1.1.1-1
liburi-perl                                     | 1.35.dfsg.1-1
man-db                            (>= 2.3.20-1) | 2.5.0-4
perl                                            | 5.8.8-12

Message #16 received at 460350@bugs.debian.org (full text, mbox, reply):

From: Niels Thykier <niels@thykier.net>

To: 460350@bugs.debian.org

Subject: Default options in lintianrc

Date: Mon, 30 May 2011 02:11:30 +0200

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

Hi

I have been looking at #460350 (plus clones) and was thinking maybe we
could do something similar to dpkg-source and its
debian/source/{local-,}options files.  That is, we allow options to be
listed in lintianrc (without leading "--" and only in long-form etc.)

Since lintianrc is currently used for setting variables, I suggest we
keep backwards compatibility; any line starting with "LINTIAN_" will be
read as an "old" variable and let it be translated to the relevant
option (e.g. LINTIAN_ROOT -> root).

Format changes of lintianrc aside, we will also need to refactor the
frontend.  Here I propose that we delay interpretation of options (save
for stuff like --cfg); the exact idea I had in mind was to stuff all
options (with value) into a hash (GetOptions can do that for us as I
recall) and as we parse the lintianrc file, we will options from there
to the hash (as appropriate).

Any comments/ideas?

~Niels

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
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=/gkI
-----END PGP SIGNATURE-----

Message #21 received at 460350@bugs.debian.org (full text, mbox, reply):

From: Niels Thykier <niels@thykier.net>

To: 460350-submitter@bugs.debian.org, 460350@bugs.debian.org

Subject: Re: Bug#460350: Default options in lintianrc

Date: Sun, 05 Jun 2011 10:30:34 +0200

[Message part 1 (text/plain, inline)]

On 2011-05-30 02:11, Niels Thykier wrote:

> Hi
> 

Hey,

I have invoked "do-cracy" on this (though I am still option to
suggestions on this) and here is what I got so far.

> I have been looking at #460350 (plus clones) and was thinking maybe we
> could do something similar to dpkg-source and its
> debian/source/{local-,}options files.  That is, we allow options to be
> listed in lintianrc (without leading "--" and only in long-form etc.)
> 

I have been using this approach; the prototype only accepts a very small
(and not necessarily useful) subset of the boolean options.
  Unfortunately --display-info is not one of them, as it is internally
mapped as a -L option.

The options supported by this patch are:
 * info
 * display-experimental
 * no-override
 * show-overrides

There is a little print(f) debug to STDERR that informs you what the
final value of these options are.  I will remove that before merging
this into the master branch.

> Since lintianrc is currently used for setting variables, I suggest we
> keep backwards compatibility; any line starting with "LINTIAN_" will be
> read as an "old" variable and let it be translated to the relevant
> option (e.g. LINTIAN_ROOT -> root).
> 

No changes here (namely because the affected parts are not boolean options).

> Format changes of lintianrc aside, we will also need to refactor the
> frontend.  Here I propose that we delay interpretation of options (save
> for stuff like --cfg); the exact idea I had in mind was to stuff all
> options (with value) into a hash (GetOptions can do that for us as I
> recall) and as we parse the lintianrc file, we will options from there
> to the hash (as appropriate).
> 

So far there not been a need to really delay interpretation; since the
patch only deals with boolean values (all of which defaulted to
false/off); that being said, I do smell complications with the -L.
  As far as I can tell --display-info is the only option that maps to
-L/--display-level; but --display-level is sensitive to the order.  That
being said; I strongly suspect that the amount of people using -L is
very limited.
  Asking them to add a >=wishlish to their display-level option instead
of using display-info in the config seems like a reasonable solution to
the order problem.  Particularly we can disallow the usage of both
display-info and display-level in the config at the same time.

I think it makes sense to ignore display-level and display-info, if
either --display-info or --display-level was passed on command-line.

For the parsing of display-level in the config, I am thinking of using a
space separation, so:

  -L ">=important" -L "+>=normal/possible" -L +minor/certain

becomes

  display-level= >=important +>=normal/possible +minor/certain


I also think we should disallow the usage of dont-check-part,
tags-from-file, ftp-master-rejects, check-part and
suppress-tags{,-from-file} and rely on people to create profiles for that.
  I would also disallow actions (unpack, check etc.) since I do not
really think they make sense to have in the rc file.

~Niels

[0001-Prototype-config-options-for-some-bool-options-only.patch (application/x-wine-extension-patch, attachment)]

Message #29 received at 460350@bugs.debian.org (full text, mbox, reply):

From: Niels Thykier <niels@thykier.net>

To: 460350@bugs.debian.org

Cc: 460350-submitter@bugs.debian.org

Subject: RFR/ITC: Default options in lintianrc

Date: Mon, 06 Jun 2011 23:05:13 +0200

[Message part 1 (text/plain, inline)]

On 2011-06-05 10:30, Niels Thykier wrote:
> On 2011-05-30 02:11, Niels Thykier wrote:
>
> I have invoked "do-cracy" on this (though I am still option to
> suggestions on this) and here is what I got so far.
> 

Hey

More do-cracy[1].

In case you do not know or have not guessed what I meant with RFR and
ITC, then it is "Request For Review" and "Intend To Commit" (respectively).
  There are a TODO marker in the patch that I will do before committing
(possibly in a separate commit); the only other change I can think of
would be requiring a "--no-cfg" (to be used by our test-suite) and
possibly some "--no-X" options.

[1] Okay - I admit I was bored.

>> I have been looking at #460350 (plus clones) and was thinking maybe we
>> could do something similar to dpkg-source and its
>> debian/source/{local-,}options files.  That is, we allow options to be
>> listed in lintianrc (without leading "--" and only in long-form etc.)
>>
> 
> I have been using this approach; the prototype only accepts a very small
> (and not necessarily useful) subset of the boolean options.
>   Unfortunately --display-info is not one of them, as it is internally
> mapped as a -L option.
> 
> The options supported by this patch are:
>  * info
>  * display-experimental
>  * no-override
>  * show-overrides
> 

Extended to also include:
 * color
 * display-info
 * display-level
 * fail-on-warnings
 * pedantic

> There is a little print(f) debug to STDERR that informs you what the
> final value of these options are.  I will remove that before merging
> this into the master branch.
> 

Still there

>> Since lintianrc is currently used for setting variables, I suggest we
>> keep backwards compatibility; any line starting with "LINTIAN_" will be
>> read as an "old" variable and let it be translated to the relevant
>> option (e.g. LINTIAN_ROOT -> root).
>>
> 
> No changes here (namely because the affected parts are not boolean options).
> 

So, I learned that LINTIAN_ROOT cannot be changed in the config file.
For now I have just left the only variables in place and not bothered to
add code to replace them with options.
  Unless there are any request for this I will probably leave it as this
for now (unless we decide to deprecate --checksums anyway - see #629453)

>> Format changes of lintianrc aside, we will also need to refactor the
>> frontend.  Here I propose that we delay interpretation of options (save
>> for stuff like --cfg); the exact idea I had in mind was to stuff all
>> options (with value) into a hash (GetOptions can do that for us as I
>> recall) and as we parse the lintianrc file, we will options from there
>> to the hash (as appropriate).
>>
> 
> So far there not been a need to really delay interpretation; since the
> patch only deals with boolean values (all of which defaulted to
> false/off); that being said, I do smell complications with the -L.


Turns out that there is hardly any validation (that had to be
relocated).  Moving the validation of $color was all so far.

>   As far as I can tell --display-info is the only option that maps to
> -L/--display-level; but --display-level is sensitive to the order.  That
> being said; I strongly suspect that the amount of people using -L is
> very limited.
>   Asking them to add a >=wishlish to their display-level option instead
> of using display-info in the config seems like a reasonable solution to
> the order problem.  Particularly we can disallow the usage of both
> display-info and display-level in the config at the same time.
> 
> I think it makes sense to ignore display-level and display-info, if
> either --display-info or --display-level was passed on command-line.
> 
> For the parsing of display-level in the config, I am thinking of using a
> space separation, so:
> 
>   -L ">=important" -L "+>=normal/possible" -L +minor/certain
> 
> becomes
> 
>   display-level= >=important +>=normal/possible +minor/certain
> 

Done this.

> 
> I also think we should disallow the usage of dont-check-part,
> tags-from-file, ftp-master-rejects, check-part and
> suppress-tags{,-from-file} and rely on people to create profiles for that.
>   I would also disallow actions (unpack, check etc.) since I do not
> really think they make sense to have in the rc file.
> 

Also done (since they would have to be explicitly added to the
"%cfghash" to be allowed).

Unless anyone feels there are any complications I will probably commit
this in two-three days time.

Thank you in advance,
~Niels

[0001-Allow-some-options-to-appear-in-lintianrc.patch (text/x-diff, attachment)]

Message #39 received at 460350-close@bugs.debian.org (full text, mbox, reply):

From: Niels Thykier <niels@thykier.net>

To: 460350-close@bugs.debian.org

Subject: Bug#460350: fixed in lintian 2.5.1

Date: Wed, 15 Jun 2011 22:18:19 +0000

Source: lintian
Source-Version: 2.5.1

We believe that the bug you reported is fixed in the latest version of
lintian, which is due to be installed in the Debian FTP archive:

lintian_2.5.1.dsc
  to main/l/lintian/lintian_2.5.1.dsc
lintian_2.5.1.tar.gz
  to main/l/lintian/lintian_2.5.1.tar.gz
lintian_2.5.1_all.deb
  to main/l/lintian/lintian_2.5.1_all.deb



A summary of the changes between this version and the previous one is
attached.

Thank you for reporting the bug, which will now be closed.  If you
have further comments please address them to 460350@bugs.debian.org,
and the maintainer will reopen the bug report if appropriate.

Debian distribution maintenance software
pp.
Niels Thykier <niels@thykier.net> (supplier of updated lintian package)

(This message was generated automatically at their request; if you
believe that there is a problem with it please contact the archive
administrators by mailing ftpmaster@debian.org)


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

Format: 1.8
Date: Wed, 15 Jun 2011 23:05:28 +0200
Source: lintian
Binary: lintian
Architecture: source all
Version: 2.5.1
Distribution: unstable
Urgency: low
Maintainer: Debian Lintian Maintainers <lintian-maint@debian.org>
Changed-By: Niels Thykier <niels@thykier.net>
Description: 
 lintian    - Debian package checker
Closes: 460350 496902 541978 588831 598412 605012 614876 616493 616553 625886 626476 626587 626597 626775 627082 627431 628673 628754 629056 629189 629193 629288 629453 629648 630164 630560
Changes: 
 lintian (2.5.1) unstable; urgency=low
 .
   * Summary of tag changes:
     + Added:
       - bad-perm-for-file-in-etc-sudoers.d
       - debian-rules-missing-recommended-target
       - dh_pycentral-is-obsolete
       - dh_python-is-obsolete
       - illegal-multi-arch-value
       - non-empty-dependency_libs-in-la-file
     + Removed:
       - uses-dh-python-with-no-pycompat
       - ldconfig-symlink-before-shlib-in-deb
 .
   * checks/binaries:
     + [NT] Applied patch from Paul Wise to detect binaries embeddeding
       glee, glew and libtheora.  (Closes: #629288)
     + [NT] Also consider multiarch dirs when looking for public shared
       libraries.  (Closes: #630164)
   * checks/debhelper{,.desc}:
     + [NT] Recognise that dh_python and dh_pycentral are now obsolete.
       (Closes: #496902, #541978)
   * checks/changes-file:
     + [NT] Always verify the checksums of changes-files.
   * checks/copyright-file:
     + [NT] Extended current dh_make template checks.  Thanks to
       Nelson A. de Oliveira for the patch.  (Closes: #598412)
   * checks/fields{,.desc}:
     + [NT] Allow "all" and "any" to appear together in the Architecture
       field of a .dsc file.  (Closes: #626775)
     + [NT] Added illegal-multi-arch-value.  (Closes: #616493)
   * checks/files:
     + [NT] Fixed two misnamed udeb tags, which lead to an internal
       error if triggered.  Thanks to Guillem Jover for the report.
       (Closes: #628754)
     + [NT] Added bad-perm-for-file-in-etc-sudoers.d tag.
       (Closes: #588831)
     + [CW] Do not issue missing-pre-dependency-on-multiarch-support for
       udebs.
     + [NT] Limit missing-pre-dependency-on-multiarch-support to real
       files only.  Previously it would also check symlinks.
       (Closes: #629648)
   * checks/java{,.desc}:
     + [NT] Sort the jar files by name, so they are checked in the same
       order.
     + [NT] Improved the description of classpath-contains-relative-path.
       (Closes: #629189)
     + [NT] Reduced false-positives of classpath-contains-relative-path.
       Relative classpaths are now okay if they point to jar files in
       /usr/share/java (but not subdirs thereof) and there is a strong
       libX-java dependency.  Alternatively if the classpath points to
       a file or a symlink in the package itself, then it is also okay.
     + [NT] Allow maven plugins to be without classpaths.
       (Closes: #629056)
   * checks/manpages.desc:
     + [ADB] Don't suggest using debmake's output as a reference.  The
       package is not even in oldstable any more.
   * checks/md5sums.desc:
     + [ADB] Drop references to obsolete debstd.  Thanks to Steve Langasek
       for the pointer.
   * checks/rules:
     + [NT] Removed code for handling packages with dh_python.
     + [NT] Fixed the code for determining Build-Depends for python
       packages.  Thanks to Jakub Wilk for the patch.
       (Closes: #614876, #616553)
     + [NT] Applied patch from Roger Leigh to implement the check for
       rules files missing the build-arch and build-indep targets.
       (Closes: #605012)
   * checks/shared-libs{,.desc}:
     + [NT] Added non-empty-dependency_libs-in-la-file.  (Closes: #628673)
     + [NT] Removed ldconfig-symlink-before-shlib-in-deb.
 .
   * collection/java:
     + [NT] Fixed a regex that would accept any file ending with "jar" to
       be a jar file rather than only accepting ".jar".  This fixes a
       number of false-positives java checks.
   * collection/unpacked:
     + [NT] Added support for unpacking source packages using libdpkg-perl,
       instead of dpkg-source, if the latter is not present.
       (Closes: #626476)
 .
   * data/binaries/multiarch-dirs:
     + [NT] New file - maps architectures to multiarch dirs.
   * data/fields/{perl-provides,virtual-packages}:
     + [NT] Refreshed.  (Closes: #627082, #627431)
   * data/output/ftp-master-*:
     + [NT] Refreshed.
   * data/shared-libs/ldconfig-dirs:
     + [NT] Updated ldconfig path for i386 multiarch paths.  Thanks to
       Raphaël Hertzog for the reminder.
   * data/spelling/corrections-multiword:
     + [CW] Remove the "requires to" correction; it has too many common
       false positives, especially involving relative clauses.
       (Closes: #625886)
 .
   * debian/control:
     + [NT] (Build-)Dependency changes.
       - xz-lzma is now preferred to lzma as Build-Dependency
       - Added depends on libdpkg-perl
       - Reduced dpkg-dev to a suggests, we can use libdpkg-perl as
         fallback.
       - Explicitly depends on bzip2 to always support bz2 source packages.
       - Added xz-utils as suggests - needed for .xz source packages.
 .
   * doc/lintian.xml:
     + [NT] Clarify that the package type is not listed in the output for
       binary packages.  (Closes: #629193)
 .
   * frontend/lintian:
     + [NT] Allow some options to be put into the lintianrc file.
       (Closes: #460350)
     + [NT] Added --no-cfg option that can be specified if no config
       file should be read.
     + [NT] Deprecated --checksums (also known as --md5sums and -m).
       Lintian will now always verify the checksums in changes files.
       The options will be accepted with a warning for now, but may
       be removed in a later release.  (Closes: #629453)
     + [NT] Fixed an issue where fail was called before it was
       available.  Thanks to Toby Corkindale for the report.
       (Closes: #630560, LP: #797553)
 .
   * lib/Lintian/Tag/Info.pm:
     + [NT] Output the name of the check and the type of the check
       along with the description of the tag.
   * lib/Lintian/Tags.pm:
     + [NT] Fixed a flaw in logic that gave incorrect results for
       architecture dependent overrides.  Thanks to Andreas Beckmann for
       the report and the patch.  (Closes: #626587)
 .
   * man/lintian.pod.in:
     + [NT] Mentioned that --checksums and aliases are deprecated.
 .
   * unpack/list-srcpkg:
     + [ADB] Don't rely on the order of fields in the Sources file.  The
       files produced by the Debian archive no longer follow the order
       which the code previously assumed, and we shouldn't assume that
       the order won't change again in the future.  (Closes: #626597)
Checksums-Sha1: 
 0b389840ac927017e03373724f1589dcf190025b 2412 lintian_2.5.1.dsc
 f0c1f42f79bbd71a753601495006e77e9df15add 934889 lintian_2.5.1.tar.gz
 b6aae95d8c5f5757f3509f2ce8bfba8fa3c1dbf8 591218 lintian_2.5.1_all.deb
Checksums-Sha256: 
 5a4a013bfd18620e66890d91b35f06a9abe2eedd48a0b04725b6f0bcf72ef4d1 2412 lintian_2.5.1.dsc
 b8812da6a1f31ed462b71d16ab39b491b75291dc7ac40bca629733dffa4726f6 934889 lintian_2.5.1.tar.gz
 b6920d86725fe75f292594356071298fd173b65240159b6a3625bf1fd80174c5 591218 lintian_2.5.1_all.deb
Files: 
 449b2e67a7d792bcbef9203d3dc0f6c7 2412 devel optional lintian_2.5.1.dsc
 8281c27f660386be3cb91dffd58173e6 934889 devel optional lintian_2.5.1.tar.gz
 c34b7b14852cc1abd53d77ee29eb0d26 591218 devel optional lintian_2.5.1_all.deb

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
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=hoYp
-----END PGP SIGNATURE-----

Send a report that this bug log contains spam.