schemas.text.json Maven / Gradle / Ivy

Go to download
{
    "$schema": "https://json-schema.org/draft/2020-12/schema#",
    "$id": "https://raw.githubusercontent.com/ArDoCo/Core/main/framework/text-provider-json/src/main/resources/schemas/text.json",
    "title": "Text",
    "description": "A definition of a text",
    "type": "object",
    "required": [
        "sentences"
    ],
    "properties": {
        "sentences": {
            "description": "the words that are contained in this sentence",
            "type": "array",
            "uniqueItems": true,
            "items": {
                "description": "Sentence in a text",
                "type": "object",
                "required": [
                    "sentenceNo",
                    "text",
                    "constituencyTree",
                    "words"
                ],
                "properties": {
                    "sentenceNo": {
                        "description": "index of the sentence",
                        "type": "integer",
                        "minimum": 0
                    },
                    "text": {
                        "description": "the text of the sentence",
                        "type": "string"
                    },
                    "constituencyTree": {
                        "description": "the constituency tree of the sentence in bracket notation",
                        "type": "string"
                    },
                    "words": {
                        "description": "the words that are contained in this sentence",
                        "type": "array",
                        "uniqueItems": true,
                        "items": {
                            "description": "Definition of a word",
                            "type": "object",
                            "required": [
                                "sentenceNo",
                                "id",
                                "text",
                                "lemma",
                                "posTag",
                                "outgoingDependencies",
                                "incomingDependencies"
                            ],
                            "properties": {
                                "sentenceNo": {
                                    "description": "index of the sentence the word is contained in",
                                    "type": "integer",
                                    "minimum": 0
                                },
                                "id": {
                                    "description": "The id of the word. Should be ascending from 1 for the first word in the text.",
                                    "type": "integer",
                                    "minimum": 1
                                },
                                "text": {
                                    "description": "the text of the word",
                                    "type": "string"
                                },
                                "lemma": {
                                    "description": "the lemma of the word",
                                    "type": "string"
                                },
                                "posTag": {
                                    "$ref": "#/$defs/posTags"
                                },
                                "outgoingDependencies": {
                                    "description": "the outgoing dependencies",
                                    "type": "array",
                                    "uniqueItems": false,
                                    "items": {
                                        "type": "object",
                                        "required": [
                                            "targetWordId",
                                            "dependencyType"
                                        ],
                                        "properties": {
                                            "targetWordId": {
                                                "description": "The id of the word the dependency points to.",
                                                "type": "integer"
                                            },
                                            "dependencyType": {
                                                "$refs": "#/$defs/dependencyTypes"
                                            }
                                        }
                                    }
                                },
                                "incomingDependencies": {
                                    "description": "the incoming dependencies",
                                    "type": "array",
                                    "uniqueItems": false,
                                    "items": {
                                        "type": "object",
                                        "required": [
                                            "sourceWordId",
                                            "dependencyType"
                                        ],
                                        "properties": {
                                            "sourceWordId": {
                                                "description": "The id of the word the dependency originates from.",
                                                "type": "integer"
                                            },
                                            "dependencyType": {
                                                "$refs": "#/$defs/dependencyTypes"
                                            }
                                        }
                                    }
                                }
                            }
                        }
                    }
                }
            }
        }
    },
    "$defs": {
        "posTags": {
            "description": "the lemma of the word",
            "type": "string",
            "enum": [
                "JJ",
                "JJR",
                "JJS",
                "RB",
                "RBR",
                "RBS",
                "WRB",
                "CC",
                "IN",
                "CD",
                "DT",
                "WDT",
                "EX",
                "FW",
                "HYPH",
                "LS",
                "NN",
                "NNS",
                "NNP",
                "NNPS",
                "PDT",
                "POS",
                "PRP",
                "PRP$",
                "WP$",
                "WP",
                "RP",
                "SYM",
                "TO",
                "UH",
                "VB",
                "VBD",
                "VBG",
                "VBN",
                "VBP",
                "VBZ",
                "MD",
                ".",
                ",",
                ":",
                "-LRB-",
                "-RRB-",
                "-NONE-",
                "``",
                "''",
                "$",
                "#",
                "HYPH",
                "NFP",
                "ADD",
                "AFX",
                "GW",
                "XX"
            ]
        },
        "dependencyTypes": {
            "description": "The valid dependency tags",
            "type": "string",
            "enum": [
                "APPOS",
                "NSUBJ",
                "POSS",
                "OBJ",
                "IOBJ",
                "NMOD",
                "NSUBJPASS",
                "POBJ",
                "AGENT",
                "NUM",
                "PREDET",
                "RCMOD",
                "CSUBJ",
                "CCOMP",
                "XCOMP",
                "OBL",
                "VOCATIVE",
                "EXPL",
                "DISLOCATED",
                "ADVCL",
                "ADVMOD",
                "DISCOURSE",
                "AUXILIARY",
                "COP",
                "MARK",
                "ACL",
                "AMOD",
                "DET",
                "CLF",
                "CASE",
                "CONJ",
                "CC",
                "FIXED",
                "FLAT",
                "COMPOUND",
                "LIST",
                "PARATAXIS",
                "ORPHAN",
                "GOES_WITH",
                "REPARANDUM",
                "PUNCT",
                "CSUBJ_PASS",
                "ACL_RELCL",
                "COMPOUND_PRT",
                "NMOD_POSS",
                "REF",
                "NSUBJ_XSUBJ",
                "NSUBJ_PASS_XSUBJ",
                "NSUBJ_RELSUBJ",
                "NSUBJ_PASS_RELSUBJ",
                "OBJ_RELOBJ"
            ]
        }
    }
}