xml_schemas/examples/morphological_lexicon/gostja.xml

84 lines
3.0 KiB
XML
Raw Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

<lexicon>
<entry>
<head>
<headword>
<lemma orthography_ids="342" dynamic_accentuation_ids="81" pronunciation_ids="32" hyphenation_ids="7|8" tonemic_accentuation_ids="85">gostja</lemma>
</headword>
<lexicalUnit sloleksId="LE_c77074df05d2719216d477bd01d418ad" sloleksKey="S_gostja" type="single">
<lexeme lexical_unit_lexeme_id="83224">gostja</lexeme>
</lexicalUnit>
<grammar>
<category>samostalnik</category>
<subcategory type="pronunciation">slovene</subcategory>
<grammarFeature name="type">common</grammarFeature>
<grammarFeature name="gender">feminine</grammarFeature>
</grammar>
<measureList>
<measure type="frequency" source="Gigafida 2.0">285818</measure>
</measureList>
<relatedEntryList/>
</head>
<body>
<wordFormList>
<wordForm id="333330">
<msd language="sl" system="JOS">Sozei</msd>
<msd language="en" system="UD">NOUN|...</msd>
<grammarFeatureList>
<grammarFeature name="number">singular</grammarFeature>
<grammarFeature name="case">nominative</grammarFeature>
</grammarFeatureList>
<formRepresentations>
<orthographyList>
<orthography id="342" morphologyPatterns="Sz.1.2.0">
<form>gostja</form>
<measureList>
<measure type="frequency" source="Gigafida 2.0">813</measure>
<measure type="frequency" source="GOS 1.0">432</measure>
</measureList>
</orthography>
</orthographyList>
<accentuationList type="dynamic">
<accentuation id="81" orthography_id="342" accentuationPatterns="cjvt:AP.1|zrc:S202b" status="preferred">
<form>gôstja</form>
<measureList>
<measure type="frequency" source="Gigafida 2.0">77</measure>
</measureList>
</accentuation>
</accentuationList>
<accentuationList type="tonemic">
<accentuation id="85" orthography_id="342" accentuationPatterns="zrc:S202b|zrc:S202i" status="preferred">
<form>góstja</form>
<measureList>
<measure type="frequency" source="XXX">7777</measure>
</measureList>
</accentuation>
</accentuationList>
<pronunciationList>
<pronunciation id="32" dynamic_accentuation_id="81" tonemic_accentuation_id="85">
<form script="IPA">ˈɡo:stja</form>
<form script="SAMPA">"gO:stja</form>
<measureList>
<measure type="frequency" source="XXX">7777</measure>
</measureList>
</pronunciation>
</pronunciationList>
<hyphenationList>
<hyphenation id="7" orthography_id="342" status="preferred">
<form>go|stja</form>
<measureList>
<measure type="frequency" source="XXX">7777</measure>
</measureList>
</hyphenation>
<hyphenation id="8" orthography_id="342">
<form>gos|tja</form>
<measureList>
<measure type="frequency" source="XXX">7777</measure>
</measureList>
</hyphenation>
</hyphenationList>
</formRepresentations>
</wordForm>
</wordFormList>
</body>
</entry>
</lexicon>