Mailing-List: contact ctakes-dev-help@incubator.apache.org; run by ezmlm
Precedence: bulk
Reply-To: ctakes-dev@incubator.apache.org
Received-SPF: pass (nike.apache.org: domain of steven.bethard@colorado.edu
 designates 128.138.128.235 as permitted sender)
From: Steven Bethard <steven.bethard@Colorado.EDU>
To: "ctakes-dev@incubator.apache.org" <ctakes-dev@incubator.apache.org>
Date: Fri, 2 Nov 2012 02:20:12 -0600
Subject: Re: dep parser descriptor auto-generation
Thread-Topic: dep parser descriptor auto-generation
Thread-Index: Ac240uGzliVIkpVRTwqdS5jDsOsgww==
Message-ID: <C0501707-2A82-4FE8-A53E-150F439360D4@colorado.edu>
References: <5092F04F.50109@childrens.harvard.edu>
In-Reply-To: <5092F04F.50109@childrens.harvard.edu>
Accept-Language: en-US
Content-Language: en-US
acceptlanguage: en-US
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0

On Nov 1, 2012, at 10:57 PM, Tim Miller <timothy.miller@childrens.harvard.e=
du> wrote:=20
> So I've tracked down the issue where POS tags were null in the=20
> dependency parser, but not when the Chunker was added to the pipeline. =20
> The issue was that the POSTagger uses the subiterator method, which=20
> requires priorityList section in the descriptor to work.  The POSTagger=20
> descriptor is missing it but the Chunker descriptor has it, so loading=20
> the Chunker descriptor was allowing the subiterator call in POSTagger to=
=20
> work.
>=20
> With me so far?  I will check in the change to the POS descriptor. But=20
> to fully fix I had to re-generate the dependency parser descriptors=20
> (using the WriteClearParserDescriptors class).  This results in some=20
> additional changes to those which are a result of the type priority=20
> change (see diff below).  Is it safe to check in all of that or just the=
=20
> POS descriptor change and expect others to generate their own dependency=
=20
> descriptors?

Do we have a test that runs the dependency parser from the descriptor? If w=
e do, and the test passes after these changes, I'd say go ahead.

> -                    <import=20
> name=3D"org.apache.ctakes.typesystem.types.TypeSystem"/>
> +                            <import=20
> location=3D"../../../ctakes-type-system/src/main/resources/org/apache/cta=
kes/typesystem/types/TypeSystem.xml"/>

This is the only part that looks suspicious. We want by-name imports, not b=
y-location imports, in general.

Steve