nlptoolkit-dependencyparser
Advanced tools
Comparing version 1.0.0 to 1.0.1
@@ -7,3 +7,3 @@ (function (factory) { | ||
else if (typeof define === "function" && define.amd) { | ||
define(["require", "exports", "nlptoolkit-corpus/dist/Corpus", "fs", "./UniversalDependencyTreeBankSentence", "./UniversalDependencyRelation", "./UniversalDependencyTreeBankFeatures", "./UniversalDependencyTreeBankWord", "../ParserEvaluationScore"], factory); | ||
define(["require", "exports", "nlptoolkit-corpus/dist/Corpus", "fs", "./UniversalDependencyTreeBankSentence", "../ParserEvaluationScore"], factory); | ||
} | ||
@@ -17,5 +17,2 @@ })(function (require, exports) { | ||
const UniversalDependencyTreeBankSentence_1 = require("./UniversalDependencyTreeBankSentence"); | ||
const UniversalDependencyRelation_1 = require("./UniversalDependencyRelation"); | ||
const UniversalDependencyTreeBankFeatures_1 = require("./UniversalDependencyTreeBankFeatures"); | ||
const UniversalDependencyTreeBankWord_1 = require("./UniversalDependencyTreeBankWord"); | ||
const ParserEvaluationScore_1 = require("../ParserEvaluationScore"); | ||
@@ -25,3 +22,3 @@ class UniversalDependencyTreeBankCorpus extends Corpus_1.Corpus { | ||
super(); | ||
let sentence = undefined; | ||
let sentence = ""; | ||
let data = fs.readFileSync(fileName, 'utf8'); | ||
@@ -31,44 +28,7 @@ let lines = data.split("\n"); | ||
if (line.length == 0) { | ||
this.addSentence(sentence); | ||
sentence = undefined; | ||
this.addSentence(new UniversalDependencyTreeBankSentence_1.UniversalDependencyTreeBankSentence(sentence)); | ||
sentence = ""; | ||
} | ||
else { | ||
if (line.startsWith("#")) { | ||
if (sentence == null) { | ||
sentence = new UniversalDependencyTreeBankSentence_1.UniversalDependencyTreeBankSentence(); | ||
} | ||
sentence.addComment(line.trim()); | ||
} | ||
else { | ||
let items = line.split("\t"); | ||
if (items.length != 10) { | ||
console.log("Line does not contain 10 items ->" + line); | ||
} | ||
else { | ||
let id = items[0]; | ||
if (id.match("^\\d+$")) { | ||
let surfaceForm = items[1]; | ||
let lemma = items[2]; | ||
let upos = UniversalDependencyRelation_1.UniversalDependencyRelation.getDependencyPosType(items[3]); | ||
if (upos == undefined) { | ||
console.log("Line does not contain universal pos ->" + line); | ||
} | ||
let xpos = items[4]; | ||
let features = new UniversalDependencyTreeBankFeatures_1.UniversalDependencyTreeBankFeatures(items[5]); | ||
let relation; | ||
if (items[6] != "_") { | ||
let to = Number.parseInt(items[6]); | ||
let dependencyType = items[7].toUpperCase(); | ||
relation = new UniversalDependencyRelation_1.UniversalDependencyRelation(to, dependencyType); | ||
} | ||
else { | ||
relation = undefined; | ||
} | ||
let deps = items[8]; | ||
let misc = items[9]; | ||
let word = new UniversalDependencyTreeBankWord_1.UniversalDependencyTreeBankWord(Number.parseInt(id), surfaceForm, lemma, upos, xpos, features, relation, deps, misc); | ||
sentence.addWord(word); | ||
} | ||
} | ||
} | ||
sentence += line + "\n"; | ||
} | ||
@@ -75,0 +35,0 @@ } |
@@ -5,3 +5,3 @@ import { Sentence } from "nlptoolkit-corpus/dist/Sentence"; | ||
private comments; | ||
constructor(); | ||
constructor(sentence?: string); | ||
addComment(comment: string): void; | ||
@@ -8,0 +8,0 @@ toString(): string; |
@@ -7,3 +7,3 @@ (function (factory) { | ||
else if (typeof define === "function" && define.amd) { | ||
define(["require", "exports", "nlptoolkit-corpus/dist/Sentence", "../ParserEvaluationScore"], factory); | ||
define(["require", "exports", "nlptoolkit-corpus/dist/Sentence", "./UniversalDependencyTreeBankWord", "../ParserEvaluationScore", "./UniversalDependencyRelation", "./UniversalDependencyTreeBankFeatures"], factory); | ||
} | ||
@@ -15,7 +15,53 @@ })(function (require, exports) { | ||
const Sentence_1 = require("nlptoolkit-corpus/dist/Sentence"); | ||
const UniversalDependencyTreeBankWord_1 = require("./UniversalDependencyTreeBankWord"); | ||
const ParserEvaluationScore_1 = require("../ParserEvaluationScore"); | ||
const UniversalDependencyRelation_1 = require("./UniversalDependencyRelation"); | ||
const UniversalDependencyTreeBankFeatures_1 = require("./UniversalDependencyTreeBankFeatures"); | ||
class UniversalDependencyTreeBankSentence extends Sentence_1.Sentence { | ||
constructor() { | ||
constructor(sentence) { | ||
super(); | ||
this.comments = new Array(); | ||
if (sentence != undefined) { | ||
let lines = sentence.split("\n"); | ||
for (let line of lines) { | ||
if (line == "") { | ||
continue; | ||
} | ||
if (line.startsWith("#")) { | ||
this.addComment(line.trim()); | ||
} | ||
else { | ||
let items = line.split("\t"); | ||
if (items.length != 10) { | ||
console.log("Line does not contain 10 items ->" + line); | ||
} | ||
else { | ||
let id = items[0]; | ||
if (id.match("^\\d+$")) { | ||
let surfaceForm = items[1]; | ||
let lemma = items[2]; | ||
let upos = UniversalDependencyRelation_1.UniversalDependencyRelation.getDependencyPosType(items[3]); | ||
if (upos == undefined) { | ||
console.log("Line does not contain universal pos ->" + line); | ||
} | ||
let xpos = items[4]; | ||
let features = new UniversalDependencyTreeBankFeatures_1.UniversalDependencyTreeBankFeatures(items[5]); | ||
let relation; | ||
if (items[6] != "_") { | ||
let to = Number.parseInt(items[6]); | ||
let dependencyType = items[7].toUpperCase(); | ||
relation = new UniversalDependencyRelation_1.UniversalDependencyRelation(to, dependencyType); | ||
} | ||
else { | ||
relation = undefined; | ||
} | ||
let deps = items[8]; | ||
let misc = items[9]; | ||
let word = new UniversalDependencyTreeBankWord_1.UniversalDependencyTreeBankWord(Number.parseInt(id), surfaceForm, lemma, upos, xpos, features, relation, deps, misc); | ||
this.addWord(word); | ||
} | ||
} | ||
} | ||
} | ||
} | ||
} | ||
@@ -22,0 +68,0 @@ addComment(comment) { |
{ | ||
"name": "nlptoolkit-dependencyparser", | ||
"version": "1.0.0", | ||
"version": "1.0.1", | ||
"description": "Turkish Dependency Parser Library", | ||
@@ -23,6 +23,6 @@ "main": "index.js", | ||
"mocha": "^9.1.3", | ||
"nlptoolkit-corpus": "^1.0.2", | ||
"nlptoolkit-corpus": "^1.0.5", | ||
"nlptoolkit-datastructure": "^1.0.0", | ||
"nlptoolkit-dictionary": "^1.0.2", | ||
"nlptoolkit-morphologicalanalysis": "^1.0.1", | ||
"nlptoolkit-dictionary": "^1.0.4", | ||
"nlptoolkit-morphologicalanalysis": "^1.0.3", | ||
"nlptoolkit-xmlparser": "^1.0.4", | ||
@@ -29,0 +29,0 @@ "ts-node": "^10.4.0", |
import {Corpus} from "nlptoolkit-corpus/dist/Corpus"; | ||
import * as fs from "fs"; | ||
import {UniversalDependencyTreeBankSentence} from "./UniversalDependencyTreeBankSentence"; | ||
import {UniversalDependencyRelation} from "./UniversalDependencyRelation"; | ||
import {UniversalDependencyTreeBankFeatures} from "./UniversalDependencyTreeBankFeatures"; | ||
import {UniversalDependencyTreeBankWord} from "./UniversalDependencyTreeBankWord"; | ||
import {ParserEvaluationScore} from "../ParserEvaluationScore"; | ||
@@ -13,3 +10,3 @@ | ||
super(); | ||
let sentence = undefined | ||
let sentence = "" | ||
let data = fs.readFileSync(fileName, 'utf8') | ||
@@ -19,41 +16,6 @@ let lines = data.split("\n") | ||
if (line.length == 0){ | ||
this.addSentence(sentence); | ||
sentence = undefined; | ||
this.addSentence(new UniversalDependencyTreeBankSentence(sentence)); | ||
sentence = ""; | ||
} else { | ||
if (line.startsWith("#")){ | ||
if (sentence == null){ | ||
sentence = new UniversalDependencyTreeBankSentence(); | ||
} | ||
sentence.addComment(line.trim()); | ||
} else { | ||
let items = line.split("\t"); | ||
if (items.length != 10){ | ||
console.log("Line does not contain 10 items ->" + line); | ||
} else { | ||
let id = items[0]; | ||
if (id.match("^\\d+$")){ | ||
let surfaceForm = items[1]; | ||
let lemma = items[2]; | ||
let upos = UniversalDependencyRelation.getDependencyPosType(items[3]); | ||
if (upos == undefined){ | ||
console.log("Line does not contain universal pos ->" + line); | ||
} | ||
let xpos = items[4]; | ||
let features = new UniversalDependencyTreeBankFeatures(items[5]); | ||
let relation | ||
if (items[6] != "_"){ | ||
let to = Number.parseInt(items[6]); | ||
let dependencyType = items[7].toUpperCase(); | ||
relation = new UniversalDependencyRelation(to, dependencyType); | ||
} else { | ||
relation = undefined; | ||
} | ||
let deps = items[8]; | ||
let misc = items[9]; | ||
let word = new UniversalDependencyTreeBankWord(Number.parseInt(id), surfaceForm, | ||
lemma, upos, xpos, features, relation, deps, misc); | ||
sentence.addWord(word); | ||
} | ||
} | ||
} | ||
sentence += line + "\n"; | ||
} | ||
@@ -60,0 +22,0 @@ } |
import {Sentence} from "nlptoolkit-corpus/dist/Sentence"; | ||
import {UniversalDependencyTreeBankWord} from "./UniversalDependencyTreeBankWord"; | ||
import {ParserEvaluationScore} from "../ParserEvaluationScore"; | ||
import {UniversalDependencyRelation} from "./UniversalDependencyRelation"; | ||
import {UniversalDependencyTreeBankFeatures} from "./UniversalDependencyTreeBankFeatures"; | ||
@@ -9,4 +11,45 @@ export class UniversalDependencyTreeBankSentence extends Sentence{ | ||
constructor() { | ||
constructor(sentence?: string) { | ||
super(); | ||
if (sentence != undefined){ | ||
let lines = sentence.split("\n") | ||
for (let line of lines){ | ||
if (line == ""){ | ||
continue | ||
} | ||
if (line.startsWith("#")){ | ||
this.addComment(line.trim()); | ||
} else { | ||
let items = line.split("\t"); | ||
if (items.length != 10){ | ||
console.log("Line does not contain 10 items ->" + line); | ||
} else { | ||
let id = items[0]; | ||
if (id.match("^\\d+$")){ | ||
let surfaceForm = items[1]; | ||
let lemma = items[2]; | ||
let upos = UniversalDependencyRelation.getDependencyPosType(items[3]); | ||
if (upos == undefined){ | ||
console.log("Line does not contain universal pos ->" + line); | ||
} | ||
let xpos = items[4]; | ||
let features = new UniversalDependencyTreeBankFeatures(items[5]); | ||
let relation | ||
if (items[6] != "_"){ | ||
let to = Number.parseInt(items[6]); | ||
let dependencyType = items[7].toUpperCase(); | ||
relation = new UniversalDependencyRelation(to, dependencyType); | ||
} else { | ||
relation = undefined; | ||
} | ||
let deps = items[8]; | ||
let misc = items[9]; | ||
let word = new UniversalDependencyTreeBankWord(Number.parseInt(id), surfaceForm, | ||
lemma, upos, xpos, features, relation, deps, misc); | ||
this.addWord(word); | ||
} | ||
} | ||
} | ||
} | ||
} | ||
} | ||
@@ -13,0 +56,0 @@ |
Sorry, the diff of this file is not supported yet
Sorry, the diff of this file is not supported yet
License Policy Violation
LicenseThis package is not allowed per your license policy. Review the package's license to ensure compliance.
Found 1 instance in 1 package
License Policy Violation
LicenseThis package is not allowed per your license policy. Review the package's license to ensure compliance.
Found 1 instance in 1 package
66408063
104
2229