transcript-parser - npm Package Compare versions

Comparing version 0.1.1 to 0.2.0

app.js

 "use strict";
 /***********************
- Module dependencies
+ * Module dependencies
  ***********************/
 // const S = require('string');
 const _ = require('lodash');
+const Promise = require('bluebird');
 
 
 /***********************
- Object creation
+ * Object creation
  ***********************/
@@ -33,4 +34,8 @@ const TranscriptParser = function (options) {
 const proto = TranscriptParser.prototype;
+const tp = this;
 
-proto.parseOne = function(transcript) {
+/***********************
+ * Synchronous parseOne method
+ ***********************/
+proto.parseOneSync = function(transcript) {
   var lines = transcript.split(this.regex.newLine)
@@ -81,6 +86,67 @@ .filter(line => line.length > 0); //Remove blank lines
 
-proto.resolveAliases = function(data) {
-  var aliases = this.settings.aliases;
+/***********************
+ * Asynchronous parseOne method
+ ***********************/
+proto.parseOne = function(transcript, cb) {
+  //Output object
+  const output = {};
+  //Object containing the speakers and their lines
+  output.speaker = {};
+  //List of the speakers, in order
+  output.order = [];
+
+  //Current speaker
+  var speaker = 'none';
+
+  //Remove blank lines
+  return Promise.filter(transcript.split(this.regex.newLine), line => line.length > 0)
+    .then(lines => {
+      if(this.settings.removeActions) {
+        return Promise.map(lines, line => line.split(this.regex.action).join(''))
+      }
+      return Promise.resolve(lines);
+    }).then(lines => {
+      if(this.settings.removeAnnotations) {
+        //Remove annotations
+        return Promise.map(lines, line => line.split(this.regex.annotation).join(''));
+      } else if(this.settings.removeTimestamps) {
+        //Remove timestamps
+        return Promise.map(lines, line => line.split(this.regex.timestamp).join(''));
+      }
+      return Promise.resolve(lines);
+    }).then(lines => {
+      return Promise.each(lines, (line, index) => {
+        if(line.match(this.regex.speaker)) {
+          //Regex match
+          speaker = this.regex.speaker.exec(line)[1];
+          //Remove the speaker from the line
+          line = line.replace(this.regex.speaker, '');
+        }
+        //If the speaker's key doesn't already exist
+        if(!(speaker in output.speaker) &&
+          //And the speaker is defined or the setting to remove undefined speakers is false
+          (speaker !== 'none' || !this.settings.removeUnknownSpeakers)) {
+          //Set the output's speaker key to a new empty array
+          output.speaker[speaker] = [];
+        }
+        //If the speaker is defined or the setting to remove undefined speakers is false
+        if(speaker !== 'none' || !this.settings.removeUnknownSpeakers) {
+          //Add the text to the output speaker's key and speaker name to the order array
+          output.speaker[speaker].push(line);
+          output.order.push(speaker);
+        }
+      });
+    }).then(() => {
+      cb(null, output);
+    })
+    .catch(err => cb(err));
+};
+
+/***********************
+ * Synchronous resolveAliases method
+ ***********************/
+proto.resolveAliasesSync = function(data) {
+  const aliases = this.settings.aliases;
   if(_.isEmpty(aliases)) return data;
-  var speakers = data.speaker;
+  const speakers = data.speaker;
 
@@ -122,3 +188,42 @@ for(var speaker in speakers) {
 
+/***********************
+ * Asynchronous resolveAliases method
+ ***********************/
+proto.resolveAliases = function(data, cb) {
+  const aliases = this.settings.aliases;
+  if(_.isEmpty(aliases)) return cb(null, data);
+  const speakers = data.speaker;
 
+  return Promise.all(_.keys(speakers).map(speakerName => {
+    return Promise.all(_.keys(aliases).map(trueName => {
+      return Promise.each(aliases[trueName], regex => {
+        //If the regex matches
+        if(regex.test(speakerName)) {
+          //Add the lines from the regex-matched speaker
+          //to the new speaker if the new speaker exists
+          speakers[trueName] = speakers[trueName] ?
+            _.concat(speakers[trueName], speakers[speakerName]) :
+            //Otherwise, make a new list 
+            speakers[trueName] = speakers[speakerName];
+          //Delete the old key
+          delete speakers[speakerName];
+          return;
+        }
+      })
+    }))
+  })).then(() => {
+    return Promise.each(data.order, (speaker, speakerIndex) => {
+      return Promise.all(_.map(aliases, (alias, trueName) => {
+        return Promise.all(_.map(alias, (regex, regexIndex) => {
+          if(speaker.search(regex) !== -1) {
+            return data.order[speakerIndex] = trueName;
+          }
+        }));
+      }));
+    });
+  }).then(() => {
+    cb(null, data);
+  }).catch(err => cb(err));
+};
+
 module.exports = TranscriptParser;

package.json

 {
   "name": "transcript-parser",
-  "version": "0.1.1",
+  "version": "0.2.0",
   "description": "Parses plaintext speech/debate/radio transcripts into JavaScript objects.",
@@ -5,0 +5,0 @@ "main": "app.js",

README.md

@@ -18,4 +18,4 @@ transcript-parser
     
-    //Do not use readFileSync in production
-    const output = tp.parseOne(fs.readFileSync('transcript.txt', {encoding: 'UTF-8'}));
+    //Do not use fs.readFileSync in production
+    const output = tp.parseOneSync(fs.readFileSync('transcript.txt', {encoding: 'UTF-8'}));
     console.log(output);
@@ -52,2 +52,16 @@
 
+### .parseOneSync()
+
+The `parseOneSync()` method parses a string and returns an object representing it.
+
+#### Syntax
+
+`tp.parseOneSync(transcript)`
+
+##### Parameters
+
+- `transcript`
+    + The transcript, as a `string`.
+
+
 ### .parseOne()
@@ -59,3 +73,3 @@
 
-`tp.parseOne(transcript)`
+`tp.parseOne(transcript, callback)`
 
@@ -65,5 +79,23 @@ ##### Parameters
 - `transcript`
-    - The transcript, as a `string`.
+    + The transcript, as a `string`.
+- `callback(err, data)`
+    + A callback to be exectuted on function completion.
 
 
+### .resolveAliasesSync()
+
+The `resolveAliasesSync()` method resolves all aliases specified in the configuration passed to the `TranscriptParser`'s constructor (see above).
+
+Renames the names in the `order` list to match the new names in the transcript. Note that there is a signifigant performance penalty, so don't use this method unless you need it.
+
+#### Syntax
+
+`tp.resolveAliasesSync(data)`
+
+##### Parameters
+
+- `data`
+    + The transcript object after being parsed.
+ 
+
 ### .resolveAliases()
@@ -77,3 +109,3 @@
 
-`tp.resolveAliases(data)`
+`tp.resolveAliases(data, callback)`
 
@@ -83,2 +115,5 @@ ##### Parameters
 - `data`
-    - The transcript object after being parsed.
+    + The transcript object after being parsed.
+- `callback(err, resolved)`
+    + A callback to be executed on function completion.
+

test/parser.js

@@ -15,3 +15,7 @@ "use strict";
 
-  describe('#parseOne()', function(){
+  /*
+   * For the synchronous parseOne method
+   *
+   */
+  describe('#parseOneSync()', function(){
     const tp = new TranscriptParser();
@@ -21,3 +25,3 @@
       const parser = new TranscriptParser();
-      var result = parser.parseOne('PERSON A: Hello, (PAUSES) (DRINKS WATER) my name is Bob.(APPLAUSE)');
+      var result = parser.parseOneSync('PERSON A: Hello, (PAUSES) (DRINKS WATER) my name is Bob.(APPLAUSE)');
       result.speaker.should.eql({
@@ -32,3 +36,3 @@ 'PERSON A': [
       const parser = new TranscriptParser({removeActions: false});
-      var result = parser.parseOne('PERSON A: Hello, (PAUSES) (DRINKS WATER) my name is Bob.(APPLAUSE)');
+      var result = parser.parseOneSync('PERSON A: Hello, (PAUSES) (DRINKS WATER) my name is Bob.(APPLAUSE)');
       result.speaker.should.eql({
@@ -43,3 +47,3 @@ 'PERSON A': [
       const parser = new TranscriptParser({removeAnnotations: false, removeTimestamps: false});
-      var result = parser.parseOne('[20:20:34] BERMAN: [2:1:41] The...');
+      var result = parser.parseOneSync('[20:20:34] BERMAN: [2:1:41] The...');
       result.speaker.should.eql({
@@ -54,3 +58,3 @@ '[20:20:34] BERMAN': [
       const parser = new TranscriptParser({removeAnnotations: false, removeTimestamps: true});
-      var result = parser.parseOne('[20:20:34] BERMAN: [2:1:41] The [first] name...');
+      var result = parser.parseOneSync('[20:20:34] BERMAN: [2:1:41] The [first] name...');
       result.speaker.should.eql({
@@ -65,3 +69,3 @@ 'BERMAN': [
       const parser = new TranscriptParser({removeUnknownSpeakers: true});
-      var result = parser.parseOne('The quick [brown] fox jumps over the (lazy) dog.');
+      var result = parser.parseOneSync('The quick [brown] fox jumps over the (lazy) dog.');
       result.should.eql({
@@ -77,3 +81,3 @@ speaker: {},
         .then(info => {
-          this.result = tp.parseOne(info);
+          this.result = tp.parseOneSync(info);
           return readExpected(1);
@@ -86,6 +90,108 @@ }).then(expected => {
     });
+
   });
 
-  describe('#resolveAliases()', function () {
 
+  /*
+   * For the asynchronous parseOne method
+   *
+   */
+  describe('#parseOne()', function(){
+    const tp = new TranscriptParser();
+
+    it('should remove actions by default', function(done) {
+      const parser = new TranscriptParser();
+      parser.parseOne('PERSON A: Hello, (PAUSES) (DRINKS WATER) my name is Bob.(APPLAUSE)',
+      function(err, result) {
+        if(err) return done(err);
+        result.speaker.should.eql({
+          'PERSON A': [
+            'Hello, my name is Bob.'
+          ]
+        });
+        done();
+      });
+    });
+
+    it('should respect the removeActions setting', function(done) {
+      const parser = new TranscriptParser({removeActions: false});
+      var result = parser.parseOne('PERSON A: Hello, (PAUSES) (DRINKS WATER) my name is Bob.(APPLAUSE)',
+      function(err, result) {
+        if(err) return done(err);
+        result.speaker.should.eql({
+          'PERSON A': [
+            'Hello, (PAUSES) (DRINKS WATER) my name is Bob.(APPLAUSE)'
+          ]
+        });
+        done();
+      });
+    });
+
+    it('should respect the removeTimestamps setting', function(done) {
+      const parser = new TranscriptParser({removeAnnotations: false, removeTimestamps: false});
+      var result = parser.parseOne('[20:20:34] BERMAN: [2:1:41] The...',
+      function(err, result) {
+        if(err) return done(err);
+        result.speaker.should.eql({
+          '[20:20:34] BERMAN': [
+            '[2:1:41] The...'
+          ]
+        });
+        done();
+      });
+    });
+
+    it('should be able to remove timestamps without removing annotations', function(done) {
+      const parser = new TranscriptParser({removeAnnotations: false, removeTimestamps: true});
+      var result = parser.parseOne('[20:20:34] BERMAN: [2:1:41] The [first] name...',
+      function(err, result) {
+        if(err) return done(err);
+        result.speaker.should.eql({
+          'BERMAN': [
+            'The [first] name...'
+          ]
+        });
+        done();
+      });
+    });
+
+    it('should respect the remove unknown speakers setting', function(done) {
+      const parser = new TranscriptParser({removeUnknownSpeakers: true});
+      var result = parser.parseOne('The quick [brown] fox jumps over the (lazy) dog.',
+      function(err, result) {
+        if(err) return done(err);
+        result.should.eql({
+          speaker: {},
+          order: []
+        });
+        done();
+      });
+    });
+
+    it('should parse a transcript correctly', function(done) {
+      readSample(1)
+        .bind({})
+        .then(info => {
+          return Promise.fromCallback(cb => {
+            tp.parseOne(info, cb);
+          });
+        })
+        .then(result => {
+          this.result = result;
+          return readExpected(1);
+        }).then(expected => {
+          this.result.should.be.eql(JSON.parse(expected));
+          done();
+        })
+        .catch(e => done(e));
+    });
+
+  });
+
+  /*
+   * For the synchronous resolveAliases method
+   *
+   */
+  describe('#resolveAliasesSync()', function () {
+
     it('should resolve aliases correctly', function(done) {
@@ -98,4 +204,4 @@ const tp = new TranscriptParser({
         .then(info => {
-          this.result = tp.parseOne(info);
-          this.result = tp.resolveAliases(this.result);
+          this.result = tp.parseOneSync(info);
+          this.result = tp.resolveAliasesSync(this.result);
           return readExpected(2);
@@ -113,4 +219,4 @@ }).then(expected => {
         .then(info => {
-          var parsed = tp.parseOne(info);
-          var resolved = tp.resolveAliases(parsed);
+          var parsed = tp.parseOneSync(info);
+          var resolved = tp.resolveAliasesSync(parsed);
           parsed.should.equal(resolved);
@@ -121,4 +227,47 @@ done();
     });
+  });
 
+  /*
+   * For the asynchronous resolveAliases method
+   *
+   */
+  describe('#resolveAliases()', function () {
+
+    it('should resolve aliases correctly', function(done) {
+      const tp = new TranscriptParser({
+        aliases: { "DONALD TRUMP": [ /.*TRUMP.*/ ] }
+      });
+      readSample(2)
+        .bind({})
+        .then(info => {
+          return Promise.fromCallback(cb => tp.parseOne(info, cb));
+        }).then(result => {
+          return Promise.fromCallback(cb => tp.resolveAliases(result, cb));
+        }).then(result => {
+          this.result = result;
+          return readExpected(2);
+        }).then(expected => {
+          this.result.should.eql(JSON.parse(expected));
+          done();
+        })
+        .catch(e => done(e));
+    });
+
+    it('should return unchanged data if aliases are not set', function(done) {
+      const tp = new TranscriptParser({aliases: {}});
+      readSample(2)
+      .bind({})
+        .then(info => {
+          return Promise.fromCallback(cb => tp.parseOne(info, cb));
+        }).then(parsed => {
+          this.parsed = parsed;
+          return Promise.fromCallback(cb => tp.resolveAliases(parsed, cb));
+        }).then(resolved => {
+          this.parsed.should.equal(resolved);
+          done();
+        })
+        .catch(e => done(e));
+    });
   });
+
 });
@@ -125,0 +274,0 @@

No alert changes

Improved metrics

Total package byte prevSize

157352

increased by6.3%

Lines of code

496

increased by91.51%

Number of lines in readme file

114

increased by46.15%

No dependency changes