Huge News!Announcing our $40M Series B led by Abstract Ventures.Learn More â
ner-node - npm Package Compare versions
Comparing version 0.0.2 to 0.0.3
| { | ||
| "name": "ner-node", | ||
| "version": "0.0.2", | ||
| "version": "0.0.3", | ||
| "description": "Node library for spawning NER server and pinging it to get entities out of raw text.", | ||
@@ -5,0 +5,0 @@ "main": "SocketNER.js", |
@@ -14,3 +14,3 @@ <h1>NER-Node</h2> | ||
| var socketNER = require("SocketNER") | ||
| var socketNER = require("ner-node") | ||
| socketNER(port, classifierFileName, pathToNER, function (obj) { | ||
@@ -26,2 +26,6 @@ // you can define your own function to parse tagged text | ||
| ``` | ||
| <p> | ||
| <strong>Note:-</strong> | ||
| The npm package doesn't come with StanfordNER folder. You can download the latest from Stanford NER website OR get it from the git hub repository. | ||
| </p> | ||
@@ -28,0 +32,0 @@ <h4>Issues & Suggestions</h4> |
157
SocketNER.js
| var spawn = require("child_process").spawn | ||
| var socket = require("net").Socket() | ||
| var deasync = require("deasync") | ||
| // The purpose of deasync is to make the getEntities function synchronous | ||
@@ -9,91 +10,89 @@ // without guessing the time it would take to finish. Previously, I was using | ||
| // by taking the code out of getEntities function and by using a global variable jsonEntities. | ||
| var deasync = require("deasync") | ||
| function SocketNER(port, classifierFileName, pathToNER, callback) { | ||
| //defining defaults if arguments is a false value | ||
| port = port || 1234 | ||
| classifierFileName = classifierFileName || "english.all.3class.distsim.crf.ser.gz" | ||
| pathToNER = pathToNER || "/" | ||
| // using jsonEntities as a list is a clever technique of mine | ||
| // to empty using .pop() while returning the value in getEntities function** | ||
| var jsonEntities = [] | ||
| var taggedText | ||
| var requiredEntityG | ||
| var rawTextG | ||
| //defining defaults if arguments is a false value | ||
| port = port || 1234 | ||
| classifierFileName = classifierFileName || "english.all.3class.distsim.crf.ser.gz" | ||
| pathToNER = pathToNER || "/" | ||
| // using jsonEntities as a list is a clever technique of mine | ||
| // to empty using .pop() while returning the value in getEntities function** | ||
| var jsonEntities = [] | ||
| var taggedText | ||
| var requiredEntityG | ||
| var rawTextG | ||
| // starting server as a seperate process | ||
| var server = spawn( | ||
| "java",[ | ||
| "-mx750m", "-cp", | ||
| pathToNER + "stanford-ner.jar", | ||
| "edu.stanford.nlp.ie.NERServer", | ||
| "-loadClassifier", pathToNER + classifierFileName, | ||
| "-port", port, "-outputFormat", "inlineXML" | ||
| ] | ||
| ) | ||
| // starting server as a seperate process | ||
| var server = spawn( | ||
| "java",[ | ||
| "-mx750m", "-cp", | ||
| pathToNER + "stanford-ner.jar", | ||
| "edu.stanford.nlp.ie.NERServer", | ||
| "-loadClassifier", pathToNER + classifierFileName, | ||
| "-port", port, "-outputFormat", "inlineXML" | ||
| ] | ||
| ) | ||
| // Setup a Socket Connection after Server loads the Classifier | ||
| // I don't know why server's stderr stream gets all output and why stdout don't | ||
| server.stderr.on("data", function (data) { | ||
| // Server would finish loading, when it flushes out 'done [x secs]' | ||
| if (data.toString().search("done") !== -1) { callback(socketNER) } | ||
| }) | ||
| // Setup a Socket Connection after Server loads the Classifier | ||
| // I don't know why server's stderr stream gets all output and why stdout don't | ||
| server.stderr.on("data", function (data) { | ||
| // Server would finish loading, when it flushes out 'done [x secs]' | ||
| if (data.toString().search("done") !== -1) { callback(socketNER) } | ||
| }) | ||
| // Adding event handlers to the socket | ||
| socket.on("connect", function () { socket.write(rawTextG) }) | ||
| socket.on("error", function (err) { console.log(err.toString()) }) | ||
| socket.on("data", function (data) { taggedText = data.toString() }) | ||
| // The NER server (ie. the other end) sends FIN signal after each write request, thereby closes | ||
| // the socket at its end. So I am using the 'end' event to process the returned taggedText | ||
| socket.on("end", function (err) { | ||
| jsonEntities.push(socketNER.parser(taggedText, requiredEntityG)) | ||
| }) | ||
| // Adding event handlers to the socket | ||
| socket.on("connect", function () { socket.write(rawTextG) }) | ||
| socket.on("error", function (err) { console.log(err.toString()) }) | ||
| socket.on("data", function (data) { taggedText = data.toString() }) | ||
| // The NER server (ie. the other end) sends FIN signal after each write request, thereby closes | ||
| // the socket at its end. So I am using the 'end' event to process the returned taggedText | ||
| socket.on("end", function (err) { | ||
| jsonEntities.push(socketNER.parser(taggedText, requiredEntityG)) | ||
| }) | ||
| var socketNER = {} | ||
| socketNER.getEntities = function (rawText, requiredEntity) { | ||
| // taking a copy of requiredEntity preference | ||
| requiredEntityG = requiredEntity | ||
| // replacing line breaks with spaces and adding two line breaks at the end | ||
| // for an unknown reason. May be, it relates to how a request should be sent through socket. | ||
| rawTextG = rawText.replace(/[\r\n\f\t\v]/g, " ") + "\n\n" | ||
| // Reconnecting on each getEntities function call, that's neccessary as the NER sever closes | ||
| // the socket after each such call. | ||
| socket.connect(port) | ||
| // deasync would won't halt the Event Loop ie. it would also next events to be met | ||
| // but at the same time, it won't exit this function untill jsonEntities comes back. | ||
| deasync.loopWhile(function () { return (jsonEntities.length !== 1) }) | ||
| return jsonEntities.pop() // ** or jsonEntities.shift() | ||
| } | ||
| var socketNER = {} | ||
| socketNER.getEntities = function (rawText, requiredEntity) { | ||
| // taking a copy of requiredEntity preference | ||
| requiredEntityG = requiredEntity | ||
| // replacing line breaks with spaces and adding two line breaks at the end | ||
| // for an unknown reason. May be, it relates to how a request should be sent through socket. | ||
| rawTextG = rawText.replace(/[\r\n\f\t\v]/g, " ") + "\n\n" | ||
| // Reconnecting on each getEntities function call, that's neccessary as the NER sever closes | ||
| // the socket after each such call. | ||
| socket.connect(port) | ||
| // deasync would won't halt the Event Loop ie. it would also next events to be met | ||
| // but at the same time, it won't exit this function untill jsonEntities comes back. | ||
| deasync.loopWhile(function () { return (jsonEntities.length !== 1) }) | ||
| return jsonEntities.pop() // ** or jsonEntities.shift() | ||
| } | ||
| // Closes the socket and kills the server process | ||
| socketNER.close = function () { | ||
| socket.end() | ||
| server.kill() | ||
| } | ||
| // Closes the socket and kills the server process | ||
| socketNER.close = function () { | ||
| socket.end() | ||
| server.kill() | ||
| } | ||
| // Passing in 'the parser' to the socketNER return object, | ||
| // so that user could be able to define his own parser later on | ||
| socketNER.parser = function (taggedText, requiredEntity) { | ||
| var matches | ||
| var entities = {} //return value of parser function | ||
| // Change the regex scope according to user's Entitry requirements | ||
| // Please always pass the requiredEntity in Upper case as NER uses upper cased Tags | ||
| var re = requiredEntity ? new RegExp(["<(",requiredEntity,"?)>(.*?)<\/",requiredEntity,"?>"].join(""), "g") | ||
| : /<([A-Z]+?)>(.*?)<\/[A-Z]+?>/g | ||
| while((matches = re.exec(taggedText)) !== null) { | ||
| if (entities[matches[1]]) { | ||
| // if tagName is present, then pushing in the tagValue Array | ||
| entities[matches[1]].push(matches[2]) | ||
| } | ||
| else { | ||
| // otherwise adding the tagName with a new tagValue Array | ||
| entities[matches[1]] = [matches[2]] | ||
| } | ||
| } | ||
| return entities | ||
| } | ||
| // Passing in 'the parser' to the socketNER return object, | ||
| // so that user could be able to define his own parser later on | ||
| socketNER.parser = function (taggedText, requiredEntity) { | ||
| var matches, entities = {} //return value of parser function | ||
| // Change the regex scope according to user's Entitry requirements | ||
| // Please always pass the requiredEntity in Upper case as NER uses upper cased Tags | ||
| var re = requiredEntity ? new RegExp(["<(",requiredEntity,"?)>(.*?)<\/",requiredEntity,"?>"].join(""), "g") | ||
| : /<([A-Z]+?)>(.*?)<\/[A-Z]+?>/g | ||
| while((matches = re.exec(taggedText)) !== null) { | ||
| if (entities[matches[1]]) { | ||
| // if tagName is present, then pushing in the tagValue Array | ||
| entities[matches[1]].push(matches[2]) | ||
| } | ||
| else { | ||
| // otherwise adding the tagName with a new tagValue Array | ||
| entities[matches[1]] = [matches[2]] | ||
| } | ||
| } | ||
| return entities | ||
| } | ||
@@ -100,0 +99,0 @@ } |
@@ -29,3 +29,3 @@ // Terminal command to run the test file :- node test.js | ||
| var nerServer = require("./SocketNER") | ||
| var nerServer = require("ner-node") | ||
| nerServer(8080, null , "./StanfordNER/", function (ner) { | ||
@@ -32,0 +32,0 @@ var jsonEntities = ner.getEntities(rawText, "") |
@@ -33,3 +33,3 @@ // Terminal command to run the test file :- node test2.js | ||
| var nerServer = require("./SocketNER") | ||
| var nerServer = require("ner-node") | ||
@@ -36,0 +36,0 @@ nerServer(8080, null , "./StanfordNER/", function (ner) { |
New alerts
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
Fixed alerts
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
Improved metrics
- Total package byte prevSize
- increased by2.81%
15199
- Number of lines in readme file
- increased by12.5%
36
Worsened metrics
- Lines of code
- decreased by-0.38%
262
No dependency changes