IssueID #1130: added support for collocation datasets
This commit is contained in:
Cyprian Laskowski
parent
39b0f6e6b3
commit
41be10e153
|
|
@ -0,0 +1,84 @@
|
||||||
|
<dictionary>
|
||||||
|
<entry>
|
||||||
|
<head>
|
||||||
|
<headword>
|
||||||
|
<lemma>absorbirati</lemma>
|
||||||
|
</headword>
|
||||||
|
<lexicalUnit id="294" type="single">
|
||||||
|
<lexeme>absorbirati</lexeme>
|
||||||
|
</lexicalUnit>
|
||||||
|
<grammar>
|
||||||
|
<category>glagol</category>
|
||||||
|
</grammar>
|
||||||
|
<measureList>
|
||||||
|
<measure type="frequency" source="Gigafida 1.0">3295</measure>
|
||||||
|
</measureList>
|
||||||
|
</head>
|
||||||
|
<body>
|
||||||
|
<senseList>
|
||||||
|
<sense id="1">
|
||||||
|
<syntacticStructureList>
|
||||||
|
<syntacticStructure id="34" name="gbz_pp_sbz2" pp="brez">
|
||||||
|
<exampleContainerList>
|
||||||
|
<exampleContainer>
|
||||||
|
<multiwordExample type="collocation" cluster="1" frequency="5" logDice="2.23">
|
||||||
|
<comp num="1" role="headword" sloleks="LE_a293f9ab871299f116dff2cc1421367a">absorbirati</comp>
|
||||||
|
<comp num="2" role="other" sloleks="LE_4485e634af6a4b16093fe3140632418a">brez</comp>
|
||||||
|
<comp num="3" role="collocate" sloleks="LE_b66fc194167f71cb086b3c0b80820711">težave</comp>
|
||||||
|
</multiwordExample>
|
||||||
|
</exampleContainer>
|
||||||
|
<exampleContainer>
|
||||||
|
<multiwordExample type="collocation" cluster="1" frequency="14" logDice="11.25">
|
||||||
|
<comp num="1" role="headword" sloleks="LE_a293f9ab871299f116dff2cc1421367a">absorbirati</comp>
|
||||||
|
<comp num="2" role="other" sloleks="LE_ba405c6d2b3a4b0cf9f6216af3812c6b">brez</comp>
|
||||||
|
<comp num="3" role="collocate" sloleks="LE_274a6fc240c9e3c4c2dc3e4804a0c11f">zahteve</comp>
|
||||||
|
</multiwordExample>
|
||||||
|
</exampleContainer>
|
||||||
|
</exampleContainerList>
|
||||||
|
</syntacticStructure>
|
||||||
|
<syntacticStructure id="34" name="gbz_pp_sbz2" pp="iz">
|
||||||
|
<exampleContainerList>
|
||||||
|
<exampleContainer>
|
||||||
|
<multiwordExample type="collocation" cluster="2" frequency="14" logDice="11.25">
|
||||||
|
<comp num="1" role="headword" sloleks="LE_a293f9ab871299f116dff2cc1421367a">absorbirati</comp>
|
||||||
|
<comp num="2" role="other" sloleks="LE_ba405c6d2b3a4b0cf9f6216af3812c6b">iz</comp>
|
||||||
|
<comp num="3" role="collocate" sloleks="LE_274a6fc240c9e3c4c2dc3e4804a0c11f">prebavila</comp>
|
||||||
|
</multiwordExample>
|
||||||
|
</exampleContainer>
|
||||||
|
<exampleContainer>
|
||||||
|
<multiwordExample type="collocation" cluster="2" frequency="14" logDice="11.25">
|
||||||
|
<comp num="1" role="headword" sloleks="LE_a293f9ab871299f116dff2cc1421367a">absorbirati</comp>
|
||||||
|
<comp num="2" role="other" sloleks="LE_ba405c6d2b3a4b0cf9f6216af3812c6b">iz</comp>
|
||||||
|
<comp num="3" role="collocate" sloleks="LE_274a6fc240c9e3c4c2dc3e4804a0c11f">želodca</comp>
|
||||||
|
</multiwordExample>
|
||||||
|
</exampleContainer>
|
||||||
|
</exampleContainerList>
|
||||||
|
</syntacticStructure>
|
||||||
|
</syntacticStructureList>
|
||||||
|
</sense>
|
||||||
|
<sense id="2">
|
||||||
|
<syntacticStructureList>
|
||||||
|
<syntacticStructure id="34" name="gbz_pp_sbz2" pp="brez">
|
||||||
|
<exampleContainerList>
|
||||||
|
<exampleContainer>
|
||||||
|
<multiwordExample type="collocation" cluster="1" frequency="5" logDice="2.23">
|
||||||
|
<comp num="1" role="headword" sloleks="LE_a293f9ab871299f116dff2cc1421367a">absorbirati</comp>
|
||||||
|
<comp num="2" role="other" sloleks="LE_4485e634af6a4b16093fe3140632418a">brez</comp>
|
||||||
|
<comp num="3" role="collocate" sloleks="LE_b66fc194167f71cb086b3c0b80820711">težave</comp>
|
||||||
|
</multiwordExample>
|
||||||
|
</exampleContainer>
|
||||||
|
<exampleContainer>
|
||||||
|
<multiwordExample type="collocation" cluster="2" frequency="14" logDice="11.25">
|
||||||
|
<comp num="1" role="headword" sloleks="LE_a293f9ab871299f116dff2cc1421367a">absorbirati</comp>
|
||||||
|
<comp num="2" role="other" sloleks="LE_ba405c6d2b3a4b0cf9f6216af3812c6b">brez</comp>
|
||||||
|
<comp num="3" role="collocate" sloleks="LE_274a6fc240c9e3c4c2dc3e4804a0c11f">prebavila</comp>
|
||||||
|
</multiwordExample>
|
||||||
|
</exampleContainer>
|
||||||
|
</exampleContainerList>
|
||||||
|
</syntacticStructure>
|
||||||
|
</syntacticStructureList>
|
||||||
|
</sense>
|
||||||
|
</senseList>
|
||||||
|
</body>
|
||||||
|
</entry>
|
||||||
|
</dictionary>
|
||||||
52
resources/schema/collocation_dictionary.xsd
Normal file
52
resources/schema/collocation_dictionary.xsd
Normal file
|
|
@ -0,0 +1,52 @@
|
||||||
|
<?xml version="1.0" encoding="UTF-8"?>
|
||||||
|
<xsd:schema xmlns:xsd="http://www.w3.org/2001/XMLSchema">
|
||||||
|
|
||||||
|
<xsd:import namespace="http://www.w3.org/XML/1998/namespace"
|
||||||
|
schemaLocation="http://www.w3.org/2001/xml.xsd" />
|
||||||
|
<xsd:include schemaLocation="inventory.xsd"/>
|
||||||
|
|
||||||
|
<xsd:element name="dictionary">
|
||||||
|
<xsd:complexType>
|
||||||
|
<xsd:sequence>
|
||||||
|
<xsd:element name="entry" minOccurs="0" maxOccurs="unbounded">
|
||||||
|
<xsd:complexType>
|
||||||
|
<xsd:sequence>
|
||||||
|
<xsd:element name="head">
|
||||||
|
<xsd:complexType>
|
||||||
|
<xsd:sequence>
|
||||||
|
<xsd:element name="status" type="statusType" minOccurs="0"/>
|
||||||
|
<xsd:element name="headword" type="headwordType"/>
|
||||||
|
<xsd:element name="lexicalUnit" type="lexicalUnitType" minOccurs="0"/>
|
||||||
|
<xsd:element name="grammar" type="grammarType" minOccurs="0"/>
|
||||||
|
<xsd:element name="measureList" type="measureListType" minOccurs="0"/>
|
||||||
|
</xsd:sequence>
|
||||||
|
</xsd:complexType>
|
||||||
|
</xsd:element>
|
||||||
|
<xsd:element name="body">
|
||||||
|
<xsd:complexType>
|
||||||
|
<xsd:sequence>
|
||||||
|
<xsd:element name="senseList">
|
||||||
|
<xsd:complexType>
|
||||||
|
<xsd:sequence>
|
||||||
|
<xsd:element name="sense" minOccurs="0" maxOccurs="unbounded">
|
||||||
|
<xsd:complexType>
|
||||||
|
<xsd:sequence>
|
||||||
|
<xsd:element name="syntacticStructureList" type="syntacticStructureListType"/>
|
||||||
|
</xsd:sequence>
|
||||||
|
<xsd:attribute name="id" type="xsd:int"/>
|
||||||
|
</xsd:complexType>
|
||||||
|
</xsd:element>
|
||||||
|
</xsd:sequence>
|
||||||
|
</xsd:complexType>
|
||||||
|
</xsd:element>
|
||||||
|
</xsd:sequence>
|
||||||
|
</xsd:complexType>
|
||||||
|
</xsd:element>
|
||||||
|
</xsd:sequence>
|
||||||
|
</xsd:complexType>
|
||||||
|
</xsd:element>
|
||||||
|
</xsd:sequence>
|
||||||
|
</xsd:complexType>
|
||||||
|
</xsd:element>
|
||||||
|
|
||||||
|
</xsd:schema>
|
||||||
|
|
@ -343,9 +343,14 @@
|
||||||
|
|
||||||
<xsd:complexType name="syntacticStructureType">
|
<xsd:complexType name="syntacticStructureType">
|
||||||
<xsd:sequence>
|
<xsd:sequence>
|
||||||
|
<xsd:choice>
|
||||||
<xsd:element name="component" type="componentType" minOccurs="0" maxOccurs="unbounded"/>
|
<xsd:element name="component" type="componentType" minOccurs="0" maxOccurs="unbounded"/>
|
||||||
|
<xsd:element name="exampleContainerList" type="exampleContainerListType" minOccurs="0" maxOccurs="unbounded"/>
|
||||||
|
</xsd:choice>
|
||||||
</xsd:sequence>
|
</xsd:sequence>
|
||||||
<xsd:attribute name="id" type="xsd:int"/>
|
<xsd:attribute name="id" type="xsd:int"/>
|
||||||
|
<xsd:attribute name="name" type="xsd:string"/>
|
||||||
|
<xsd:attribute name="pp" type="xsd:string"/>
|
||||||
</xsd:complexType>
|
</xsd:complexType>
|
||||||
|
|
||||||
<xsd:complexType name="syntacticStructureListType">
|
<xsd:complexType name="syntacticStructureListType">
|
||||||
|
|
|
||||||
Loading…
Reference in New Issue
Block a user