bansp
commented
3 years ago I have a customization that I was (still am) going to propose as a Feature Request to the TEI Council, via ISO MAF. @TomazErjavec -- if you'd like to use it in the Parla-Clarin ODD, I'd be super happy because that would add to the range of real-life cases that the standard encodes.
<classSpec xmlns="http://www.tei-c.org/ns/1.0" module="tei" type="atts" ident="att.offset" mode="add">
<desc versionDate="2020-02-06" xml:lang="en">provides attributes for specifying the beginning
and end of a linguistic or textual segment, by addressing the character offsets.</desc>
<classes/>
<constraintSpec ident="endPosnotstartPos" scheme="schematron"
xmlns:sch="http://purl.oclc.org/dsdl/schematron">
<!-- paraphrased from span.xml-->
<constraint>
<sch:rule context="*[@endPos]">
<sch:assert test="@startPos">If @endPos is supplied on <sch:name/>, @startPos must
be supplied as well</sch:assert>
</sch:rule>
</constraint>
</constraintSpec>
<constraintSpec ident="startPosnotendPos" scheme="schematron"
xmlns:sch="http://purl.oclc.org/dsdl/schematron">
<!-- paraphrased from span.xml-->
<constraint>
<sch:rule context="*[@startPos]">
<sch:assert test="@endPos">If @startPos is supplied on
<sch:name/>, @endPos must be supplied as well</sch:assert></sch:rule>
</constraint>
</constraintSpec>
<constraintSpec ident="offsetBaseObligatory" scheme="schematron"
xmlns:sch="http://purl.oclc.org/dsdl/schematron">
<constraint>
<sch:rule context="*[@endPos]"><sch:assert test="ancestor-or-self::*[@offsetBase]"
>offset attributes require @offsetBase to be defined in the same element or its
ancestor</sch:assert></sch:rule>
</constraint>
</constraintSpec>
<attList>
<attDef ident="offsetBase" usage="opt">
<desc versionDate="2020-02-06" xml:lang="en">points at the element that forms the
basis for offset calculations in standoff annotations. An element using the
<att>startPos</att> and <att>endPos</att> attributes either has to define
<att>offsetBase</att> as well, or <att>offsetBase</att> should be defined on an
ancestor element.</desc>
<datatype>
<dataRef key="teidata.pointer"/>
</datatype>
</attDef>
<attDef ident="startPos" usage="opt">
<desc versionDate="2020-02-06" xml:lang="en">specifies the starting point of a sequence of characters
or bytes, or of elements that can be pointed at with a URI.</desc>
<datatype minOccurs="1" maxOccurs="1">
<dataRef key="teidata.count"/>
</datatype>
</attDef>
<attDef ident="endPos" usage="opt">
<desc versionDate="2020-02-06" xml:lang="en">specifies the end-point of a sequence of characters
or bytes, or of elements that can be pointed at with a URI.</desc>
<datatype minOccurs="1" maxOccurs="1">
<dataRef key="teidata.count"/>
</datatype>
</attDef>
</attList>
<exemplum xml:lang="en">
<p>The example below comes from a part of the CoMParS (Collection of Multi-lingual Parallel
Sequences) project and presents a fragment of a monolingual subcorpus of German.</p>
<p>The individual sequences (in this case, a sentence) are listed in the <gi>text</gi> part
of the corpus, while the linguistic analysis is performed in the <gi>standOff</gi> part,
which consists, among others, of segmentation information. CoMParS adheres to ISO LAF
principles and uses inter-character points with the indexing starting at 0.</p>
<!-- @valid should become "true" once the modification is incorporated into the Guidelines -->
<egXML xmlns="http://www.tei-c.org/ns/Examples" valid="feasible">
<text xml:lang="de">
<body>
<ab xml:id="deu-ab1" n="1">Ich habe mich im Winter in dich verliebt.</ab>
</body>
</text>
<!--
'I'c'h' 'h'a'b'e' 'm'i'c'h' 'i'm' 'W'i'n't'e'r' 'i'n' 'd'i'c'h' 'v'e'r'l'i'e'b't'.'
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 -->
<standOff xmlns="">
<listAnnotation n="1" offsetBase="#deu-ab1" type="sequence">
<listAnnotation type="segmentation">
<seg startPos="0" endPos="3" xml:id="deu-ab1tok1">Ich</seg>
<seg startPos="4" endPos="8" xml:id="deu-ab1tok2">habe</seg>
<seg startPos="9" endPos="13" xml:id="deu-ab1tok3">mich</seg>
<seg startPos="14" endPos="16" xml:id="deu-ab1tok4">im</seg>
<seg startPos="17" endPos="23" xml:id="deu-ab1tok5">Winter</seg>
<seg startPos="24" endPos="26" xml:id="deu-ab1tok6">in</seg>
<seg startPos="27" endPos="31" xml:id="deu-ab1tok7">dich</seg>
<seg startPos="32" endPos="40" xml:id="deu-ab1tok8">verliebt</seg>
<seg startPos="40" endPos="41" xml:id="deu-ab1tok9">.</seg>
</listAnnotation>
</listAnnotation>
</standOff>
</egXML>
<p>Segmentation information gathered above is subsequently used by all other (numerous) annotation layers.</p>
<p>The CoMParS ODD contains the following statements that include <gi>seg</gi> and
<gi>listAnnotation</gi> into the att.offset class: <egXML
xmlns="http://www.tei-c.org/ns/Examples" valid="true">
<elementSpec ident="seg" module="linking" mode="change">
<classes mode="change">
<memberOf key="att.offset"/>
</classes>
</elementSpec>
<elementSpec ident="listAnnotation" module="spoken" mode="change">
<classes mode="change">
<memberOf key="att.offset"/>
</classes>
</elementSpec>
</egXML>
</p>
</exemplum>
<remarks versionDate="2017-03-21" xml:lang="en">
<p>Two options are possible (and practiced) for the start index. Some systems assume
that indexing starts with 0, some assume that the initial index value is 1. This
decision should be documented in the header, together with other project-specific
encoding decisions. Linguistic analysis in the ISO LAF (Linguistic Annotation
Framework, ISO 24612), MAF (Morphosyntactic Annotation Framework, ISO 24611), as
well as W3C XPointer assume inter-character points and indices starting at 0. W3C
XPath counts characters, beginning at 1.</p>
</remarks>
<listRef>
<ptr target="#STECAT"/>
</listRef>
</classSpec>Aha, that formulation precedes the recent changes in the Guidelines, so some tweaking may be in order. But if the general direction is found interesting for Parla-CL, I can take care of the details and synchronization (in fact, that would be a nice prompt for me to push ISO MAF ahead).
Two words on naming the attributes: A version of this used to be in my att.referring ticket that I withdrew at some point. That ticket overloaded the existing @from and @to attributes, but on Laurent Romary's advice, I resigned from trying to talk the Council into overloading the names. This suggestion doesn't touch the other existing pair, @start and @end, with different definitions, and instead uses the safe pair of @startPos and @endPos, and assumes a straightforward mapping into GrAF and friends.
In MAF, <span> and <seg> are added to it, and it's straightforward to treat <w> likewise, should someone desire to.
TomazErjavec
matyaskopp
I need to create sentence
<s>level synchronization with audio track but I probably found a bug in parla-clarin docs and no suitable solution of my issue...https://github.com/clarin-eric/parla-clarin/blob/261f8cb50c4a7be796e8185f294b216d32cb1561/Schema/parla-clarin-odd.xml#L1618-L1620
According to P5 guidelines
@startand@endare not allowed in<seg>: https://www.tei-c.org/release/doc/tei-p5-doc/en/html/ref-att.timed.html