Re: [xml] libxml2-2.6.22 test suite

[Daniel Veillard <veillard redhat com>]

On Mon, Dec 05, 2005 at 09:55:17PM -0600, Bruce Dubbs wrote:
> Daniel Veillard wrote:
> > On Sun, Dec 04, 2005 at 05:12:57PM -0600, Bruce Dubbs wrote:
> >
> > > I have been validating packages for the Beyond Linux From Scratch
> > > project and have run into a problem with the libxml2 test suite.
> > >
> > > When building with:
> > >
> > > ./configure --prefix=/usr --with-history &&
> > > make
> > >
> > > the process completes without error.  However, when I run `make check`
> > > the process runs until the section:
> > >
> > > ## Catalog regression tests
> > >
> > > at this point, the command lt-xmlcatalog runs without stopping. 
>
> I figured out the problem.  The Makefile is running
>
> ./xmlcatalog --shell ./test/catalogs/whitex.xml \
>   < ./test/catalogs/whitex.script 2>&1 > result.whitex
>
> But when I look at result.whitex, I get:
>
> > /usr/share/xml/docbook/xml/4.1.2/dbpoolx.mod
> > No entry for PUBLIC -//OASIS//ENTITIESDocBook XML Character Entities
> V4.1.2//EN
> > http://www.oasis-open.org/docbook/xml/4.1.2/dbcentx.mod
> > http://www.oasis-open.org/docbook/xml/4.1.2/dbcentx.mod
> > http://www.oasis-open.org/docbook/xml/4.1.2/docbookx.dtd
> > http://www.oasis-open.org/docbook/xml/4.1.2/docbookx.dtd
> Display all 270 possibilities? (y or n)
> .Makefile.swp                 runsuite
> .deps/                        runsuite.c
> .libs/                        runsuite.o
> .... [lines deleted]
> --More--
>
> and hangs there -- evidently in a spin loop at 100% CPU.
>
> I traced this to a readline configuration problem.  I have /etc/inputrc
> that has a statement:
>
> set show-all-if-ambiguous On
>
> If I comment out this statement, the command completes as it should.
> This seems to be triggered by the tab characters embedded in the
> whitex.script.  Could these be changed to spaces?

  No, it's important to test the behaviour when the public identifiers uses
tabs.

> I'd appreciate any insight you could provide.

  Don't compile with history ! It's not by default, and when turned on
readline gives a problem... Or ignore the regression test failure then.

Daniel

-- 
Daniel Veillard      | Red Hat http://redhat.com/
veillard redhat com  | libxml GNOME XML XSLT toolkit  http://xmlsoft.org/
http://veillard.com/ | Rpmfind RPM search engine http://rpmfind.net/