diff options
author | daanpeters <d.p.c.h.peters@student.tue.nl> | 2014-05-28 10:43:37 +0200 |
---|---|---|
committer | daanpeters <d.p.c.h.peters@student.tue.nl> | 2014-05-28 10:43:37 +0200 |
commit | 62f2a42f5d5be918561c9996d7fa77b0ef2a9aa9 (patch) | |
tree | 4291993828510f9070a68aaa3810d7c4462d5101 | |
parent | 01cc77792f3ee3fef1d628e98a9a7be7e96b056f (diff) | |
parent | 9969b6a6cbae322680cfcbc27df3d37b0954f00a (diff) | |
download | Goldfarmer-62f2a42f5d5be918561c9996d7fa77b0ef2a9aa9.tar.gz |
Merge origin/master
Conflicts:
src/main/Analyzor.java
src/main/FarmShell.java
-rw-r--r-- | nbproject/configs/such_database.properties | 2 | ||||
-rw-r--r-- | nbproject/project.properties | 162 | ||||
-rw-r--r-- | src/analysis/BrandChecker.java | 32 | ||||
-rw-r--r-- | src/database/NamedPreparedStatement.java | 11 | ||||
-rw-r--r-- | src/database/QueryUtils.java | 8 | ||||
-rw-r--r-- | src/main/Analyzor.java | 736 | ||||
-rw-r--r-- | src/main/FarmShell.java | 429 | ||||
-rw-r--r-- | test/analysis/BrandCheckerTest.java | 4 |
8 files changed, 693 insertions, 691 deletions
diff --git a/nbproject/configs/such_database.properties b/nbproject/configs/such_database.properties index bba41ec..9dffee6 100644 --- a/nbproject/configs/such_database.properties +++ b/nbproject/configs/such_database.properties @@ -1 +1 @@ -$label=such database
+$label=such database diff --git a/nbproject/project.properties b/nbproject/project.properties index b262ab6..ab8ae05 100644 --- a/nbproject/project.properties +++ b/nbproject/project.properties @@ -1,81 +1,81 @@ -annotation.processing.enabled=true
-annotation.processing.enabled.in.editor=false
-annotation.processing.processors.list=
-annotation.processing.run.all.processors=true
-annotation.processing.source.output=${build.generated.sources.dir}/ap-source-output
-application.title=Goldfarmer
-application.vendor=maurice
-build.classes.dir=${build.dir}/classes
-build.classes.excludes=**/*.java,**/*.form
-# This directory is removed when the project is cleaned:
-build.dir=build
-build.generated.dir=${build.dir}/generated
-build.generated.sources.dir=${build.dir}/generated-sources
-# Only compile against the classpath explicitly listed here:
-build.sysclasspath=ignore
-build.test.classes.dir=${build.dir}/test/classes
-build.test.results.dir=${build.dir}/test/results
-# Uncomment to specify the preferred debugger connection transport:
-#debug.transport=dt_socket
-debug.classpath=\
- ${run.classpath}
-debug.test.classpath=\
- ${run.test.classpath}
-# Files in build.classes.dir which should be excluded from distribution jar
-dist.archive.excludes=
-# This directory is removed when the project is cleaned:
-dist.dir=dist
-dist.jar=${dist.dir}/Goldfarmer.jar
-dist.javadoc.dir=${dist.dir}/javadoc
-endorsed.classpath=
-excludes=
-file.reference.joda-time-2.3.jar=lib/joda-time-2.3.jar
-file.reference.postgresql-9.3-1101.jdbc41.jar=lib/postgresql-9.3-1101.jdbc41.jar
-includes=**
-jar.compress=false
-javac.classpath=\
- ${file.reference.joda-time-2.3.jar}:\
- ${file.reference.postgresql-9.3-1101.jdbc41.jar}
-# Space-separated list of extra javac options
-javac.compilerargs=
-javac.deprecation=false
-javac.processorpath=\
- ${javac.classpath}
-javac.source=1.7
-javac.target=1.7
-javac.test.classpath=\
- ${javac.classpath}:\
- ${build.classes.dir}:\
- ${libs.junit_4.classpath}
-javac.test.processorpath=\
- ${javac.test.classpath}
-javadoc.additionalparam=
-javadoc.author=false
-javadoc.encoding=${source.encoding}
-javadoc.noindex=false
-javadoc.nonavbar=false
-javadoc.notree=false
-javadoc.private=false
-javadoc.splitindex=true
-javadoc.use=true
-javadoc.version=false
-javadoc.windowtitle=
-main.class=main.Main
-manifest.file=manifest.mf
-meta.inf.dir=${src.dir}/META-INF
-mkdist.disabled=false
-platform.active=default_platform
-project.licensePath=./nbproject/licenseheader.txt
-run.classpath=\
- ${javac.classpath}:\
- ${build.classes.dir}
-# Space-separated list of JVM arguments used when running the project.
-# You may also define separate properties like run-sys-prop.name=value instead of -Dname=value.
-# To set system properties for unit tests define test-sys-prop.name=value:
-run.jvmargs=
-run.test.classpath=\
- ${javac.test.classpath}:\
- ${build.test.classes.dir}
-source.encoding=UTF-8
-src.dir=src
-test.src.dir=test
+annotation.processing.enabled=true +annotation.processing.enabled.in.editor=false +annotation.processing.processors.list= +annotation.processing.run.all.processors=true +annotation.processing.source.output=${build.generated.sources.dir}/ap-source-output +application.title=Goldfarmer +application.vendor=maurice +build.classes.dir=${build.dir}/classes +build.classes.excludes=**/*.java,**/*.form +# This directory is removed when the project is cleaned: +build.dir=build +build.generated.dir=${build.dir}/generated +build.generated.sources.dir=${build.dir}/generated-sources +# Only compile against the classpath explicitly listed here: +build.sysclasspath=ignore +build.test.classes.dir=${build.dir}/test/classes +build.test.results.dir=${build.dir}/test/results +# Uncomment to specify the preferred debugger connection transport: +#debug.transport=dt_socket +debug.classpath=\ + ${run.classpath} +debug.test.classpath=\ + ${run.test.classpath} +# Files in build.classes.dir which should be excluded from distribution jar +dist.archive.excludes= +# This directory is removed when the project is cleaned: +dist.dir=dist +dist.jar=${dist.dir}/Goldfarmer.jar +dist.javadoc.dir=${dist.dir}/javadoc +endorsed.classpath= +excludes= +file.reference.joda-time-2.3.jar=lib/joda-time-2.3.jar +file.reference.postgresql-9.3-1101.jdbc41.jar=lib/postgresql-9.3-1101.jdbc41.jar +includes=** +jar.compress=false +javac.classpath=\ + ${file.reference.joda-time-2.3.jar}:\ + ${file.reference.postgresql-9.3-1101.jdbc41.jar} +# Space-separated list of extra javac options +javac.compilerargs= +javac.deprecation=false +javac.processorpath=\ + ${javac.classpath} +javac.source=1.7 +javac.target=1.7 +javac.test.classpath=\ + ${javac.classpath}:\ + ${build.classes.dir}:\ + ${libs.junit_4.classpath} +javac.test.processorpath=\ + ${javac.test.classpath} +javadoc.additionalparam= +javadoc.author=false +javadoc.encoding=${source.encoding} +javadoc.noindex=false +javadoc.nonavbar=false +javadoc.notree=false +javadoc.private=false +javadoc.splitindex=true +javadoc.use=true +javadoc.version=false +javadoc.windowtitle= +main.class=main.Main +manifest.file=manifest.mf +meta.inf.dir=${src.dir}/META-INF +mkdist.disabled=false +platform.active=default_platform +project.licensePath=./nbproject/licenseheader.txt +run.classpath=\ + ${javac.classpath}:\ + ${build.classes.dir} +# Space-separated list of JVM arguments used when running the project. +# You may also define separate properties like run-sys-prop.name=value instead of -Dname=value. +# To set system properties for unit tests define test-sys-prop.name=value: +run.jvmargs= +run.test.classpath=\ + ${javac.test.classpath}:\ + ${build.test.classes.dir} +source.encoding=UTF-8 +src.dir=src +test.src.dir=test diff --git a/src/analysis/BrandChecker.java b/src/analysis/BrandChecker.java index 10e22b4..ee9c7b4 100644 --- a/src/analysis/BrandChecker.java +++ b/src/analysis/BrandChecker.java @@ -1,17 +1,3 @@ -/* - * DO WHAT THE FUCK YOU WANT TO PUBLIC LICENSE Version 2, December 2004 - * - * Copyright (C) 2004 Sam Hocevar - * - * Everyone is permitted to copy and distribute verbatim or modified copies - * of this license document, and changing it is allowed as long as the name is - * changed. - * - * DO WHAT THE FUCK YOU WANT TO PUBLIC LICENSE TERMS AND CONDITIONS FOR COPYING, - * DISTRIBUTION AND MODIFICATION - * - * 0. You just DO WHAT THE FUCK YOU WANT TO. - */ package analysis; import java.io.FileInputStream; @@ -93,7 +79,7 @@ public class BrandChecker { if (line.isEmpty()) { return; } - + if (!line.contains("-")) { System.err.println("illformatted rule: " + line + ", missing -"); } else { @@ -110,13 +96,13 @@ public class BrandChecker { // Read the <name> line. String name = parts[0].trim(); - + // Read the positive words. - String positive = parts[1].replaceAll(" ",""); + String positive = parts[1].replaceAll(" ", ""); String[] sequence = positive.split(","); - + if (parts.length == 3) { - String negative = parts[2].replaceAll(" ", ""); + String negative = parts[2].replaceAll(" ", ""); String[] blacklist = negative.split(","); ruleset.add(new BrandRule(name, sequence, blacklist)); } else { @@ -139,7 +125,7 @@ public class BrandChecker { * The words that should be in the text. */ private final HashMap<String, Boolean> names; - + /** * A blacklist of words that are not interesting. */ @@ -164,7 +150,7 @@ public class BrandChecker { } else { this.blacklist = null; } - + for (String name : names) { this.names.put(name, Boolean.FALSE); } @@ -177,7 +163,7 @@ public class BrandChecker { */ public boolean analyze(String[] words) { reset(); - + int found = 0; for (String word : words) { @@ -201,7 +187,7 @@ public class BrandChecker { public String getBrand() { return brand; } - + private void reset() { for (String name : this.names.keySet()) { this.names.put(name, Boolean.FALSE); diff --git a/src/database/NamedPreparedStatement.java b/src/database/NamedPreparedStatement.java index ebb775b..9305d32 100644 --- a/src/database/NamedPreparedStatement.java +++ b/src/database/NamedPreparedStatement.java @@ -2,6 +2,7 @@ package database; import java.sql.Connection; import java.sql.PreparedStatement; +import java.sql.ResultSet; import java.sql.SQLException; import java.sql.Timestamp; import java.sql.Types; @@ -104,4 +105,14 @@ public class NamedPreparedStatement { throw ex; } } + + public ResultSet executeQuery() throws SQLException { + try { + return getStmt().executeQuery(); + } catch (SQLException ex) { + System.err.println("Query error: " + ex.getMessage()); + System.err.println(stmt); + throw ex; + } + } } diff --git a/src/database/QueryUtils.java b/src/database/QueryUtils.java index 2cc6fd6..b95903f 100644 --- a/src/database/QueryUtils.java +++ b/src/database/QueryUtils.java @@ -1,7 +1,6 @@ package database; import java.sql.SQLException; -import java.util.Locale; /** * Utilities to create queries. @@ -9,8 +8,9 @@ import java.util.Locale; * @author Maurice Laveaux */ public class QueryUtils { - public final static String insertRating - = buildQuery("mentionsbrand", new String[]{"tweetid","brand"},"tweetid","brand", "rating"); + + public final static String insertRating + = buildQuery("mentionsbrand", new String[]{"tweetid", "brand"}, "tweetid", "brand", "rating"); public final static String insertProfile = buildQuery("twitteruser", new String[]{"userid"}, "userid", "displayname", "timezone", "tweetcount", "followercount", @@ -96,7 +96,7 @@ public class QueryUtils { statement.setLong("tweetid", tweetid); statement.setInt("rating", rating); statement.setString("brand", brand); - + } public static void setInsertBrandParams(NamedPreparedStatement brandStmt, diff --git a/src/main/Analyzor.java b/src/main/Analyzor.java index 58f7dfc..b896f62 100644 --- a/src/main/Analyzor.java +++ b/src/main/Analyzor.java @@ -1,368 +1,368 @@ -package main; - -import analysis.BrandChecker; -import database.NamedPreparedStatement; -import database.QueryUtils; -import java.io.File; -import java.io.FileInputStream; -import java.io.FileNotFoundException; -import java.io.IOException; -import java.io.InputStream; -import java.io.PrintWriter; -import java.io.UnsupportedEncodingException; -import java.sql.Connection; -import java.sql.PreparedStatement; -import java.sql.ResultSet; -import java.sql.SQLException; -import java.util.List; -import java.util.HashMap; -import java.util.Locale; -import java.util.Map.Entry; -import java.util.Scanner; - -/** - * The sentiment analysis class that rates tweets based on a unigram and bigram - * set of weights. - */ -public class Analyzor { - - /** - * The map that matches single words to their weights. - */ - private final HashMap<String, Double> unimap = new HashMap(); - - /** - * The map that matches word pairs to their weights. - */ - private final HashMap<String, Double> bimap = new HashMap(); - - /** - * The results of a query, maybe return from query(). - */ - private ResultSet data; - - /** - * The persistent connection to the database. - */ - private final Connection connection; - - /** - * @param connection An open connection to the database. - */ - public Analyzor(Connection connection) { - this.connection = connection; - } - - /** - * Read the unigram and bigram lexica. - * - * @throws FileNotFoundException - */ - public void readLexicon() throws FileNotFoundException { - if (!unimap.isEmpty()) { - // data is already read. - return; - } - System.err.println("Trying to read lexicons..."); - // A unigram is in the format (WS = whitespace): - // word <WS> rating <WS> ??? <WS> ?? - // A bigram has an two WS-separated words instead of one. - try (Scanner uniScanner = new Scanner(new File("unigrams-pmilexicon.txt")); - Scanner biScanner = new Scanner(new File("bigrams-pmilexicon.txt"));) { - //Fill the map of unigrams - int lineno = 1; - while (uniScanner.hasNext()) { - - String words = uniScanner.next(); - Double d = Double.valueOf(uniScanner.next()); - unimap.put(words.toLowerCase(), d); - if (uniScanner.hasNextLine()) { - uniScanner.nextLine(); - } - lineno++; - - } - - //fill the map of bigrams - while (biScanner.hasNext()) { - String words = biScanner.next() + " " + biScanner.next(); - bimap.put(words.toLowerCase(), Double.valueOf(biScanner.next())); - if (biScanner.hasNextLine()) { - biScanner.nextLine(); - } - } - } - System.err.println("Lexicons are read."); - } - - /** - * Executes a query that the analyzer can analyze. - * - * @param query The query string to execute. - * @throws SQLException When database connection isn't available. - */ - public void query(String query) throws SQLException { - PreparedStatement statement; - //make a connection to the database and execute the query - statement = connection.prepareStatement(query); - data = statement.executeQuery(); - } - - /** - * Run a sentiment analysis and fill the database with the output. - * - * @param query The sql text for the query. - * @throws SQLException - * @throws IOException - */ - public void sentimentAnalysis(String query) throws SQLException, IOException { - query(query); - - //read the lexicons - readLexicon(); - - //go to the start of te dataset - if (data == null) { - System.err.println("data is empty, try querying first"); - return; - } - - Double value; - String text; - - //for all tuples - while (data.next()) { - //get the text - text = data.getString("text"); - text = splitPunctToWords(text); - // test is the tweet text you are going to analyze - String[] words = text.split("\\s+"); // text splitted into separate words - double positiverate = 0; // positive rating - - // Rate the text with unigrams - for (String word : words) { - value = unimap.get(word); - if (value != null) { - positiverate += unimap.get(word); - } - } - // Rate the text with bigrams - for (int i = 0; i < words.length - 1; i++) { - String pair = words[i] + " " + words[i + 1]; - value = bimap.get(pair); - if (value != null) { - positiverate += bimap.get(pair); - } - } - //insert the rating into the database - NamedPreparedStatement m_insertRating; - m_insertRating = new NamedPreparedStatement(connection, QueryUtils.insertRating); - QueryUtils.setInsertParams(m_insertRating, data.getLong("tweetid"), data.getString("brand"), (int) (positiverate * 10)); - m_insertRating.executeUpdate(); - //don't print the rate - //System.out.println(text + ": " + (int) (positiverate * 10)); - } - } - - /** - * Make a wordcloud of the results of some query. - * - * @param query The sql text for a query. - * @throws SQLException - * @throws FileNotFoundException - * @throws UnsupportedEncodingException - */ - public void makeWordCloud(String query) throws SQLException, FileNotFoundException, UnsupportedEncodingException { - - query(query); - //go to the start of the ResultSet data - if (data == null) { - System.err.println("data is empty, try querying first"); - return; - } - - String text; - String brand; - String[] words; - HashMap<String,HashMap<String, Integer>> wordcloud = new HashMap<>(); - - while (data.next()) { - //get brand - brand=data.getString("brand"); - //make hashmap for each brand - if(!wordcloud.containsKey(brand)){ - wordcloud.put(brand, new HashMap<String,Integer>()); - } - //get the text - text = data.getString("text"); - //remove punctuation, convert to lowercase and split on words - text = removePunct(text); - text = text.toLowerCase(); - words = text.split("\\s+"); - //for all words - for (String word : words) { - //if it is empty, a space or a stripe, skip it - if(word.equals("") || word.equals(" ") || word.equals("-")){ - continue; - } - //if the word is already in the map, increment the amount - if(wordcloud.get(brand).containsKey(word)){ - wordcloud.get(brand).put(word, wordcloud.get(brand).get(word) + 1); - } - //if the word is not already in the map, make an entry with amount = 1 - else{ - wordcloud.get(brand).put(word, 1); - } - } - } - //print the words and their frequency in a csv file - mapToCSV(wordcloud, "wordcloud.csv", "brand,word,count"); - } - - //generate csv for disco from the query - public void disco(String query) throws SQLException, FileNotFoundException, UnsupportedEncodingException { - //do the query - query(query); - PrintWriter writer = new PrintWriter("output.csv", "UTF-8"); - //print the first row - for (int i = 1; i < data.getMetaData().getColumnCount(); i++) { - writer.print(data.getMetaData().getColumnLabel(i) + ", "); - } - writer.println(data.getMetaData().getColumnLabel(data.getMetaData().getColumnCount())); - //print the values - while (data.next()) { - for (int i = 1; i < data.getMetaData().getColumnCount(); i++) { - if (data.getObject(i) == null) { - writer.print(", "); - } else { - writer.print(data.getObject(i).toString().replaceAll("[,\n]", " ") + ", "); - } - } - if (data.getObject(data.getMetaData().getColumnCount()) == null) { - writer.println("0"); - } else { - writer.println(data.getObject(data.getMetaData().getColumnCount()).toString().replace(",", " ")); - } - } - writer.close(); - } - - public void getBrands() throws SQLException { - PreparedStatement statement; - //make a connection to the database and execute the query - statement = connection.prepareStatement("delete from mentionsbrand"); - statement.executeUpdate(); - BrandChecker checker = new BrandChecker("brandonlyrules.txt"); - query("select * from tweet"); - NamedPreparedStatement m_insertBrand = new NamedPreparedStatement(connection, QueryUtils.insertBrand); - while (data.next()) { - List<String> brands = checker.getBrands(data.getString("text")); - if (brands.isEmpty()) { - QueryUtils.setInsertBrandParams(m_insertBrand, data.getLong("tweetid"), "no"); - m_insertBrand.executeUpdate(); - } else { - for (String brand : brands) { - QueryUtils.setInsertBrandParams(m_insertBrand, data.getLong("tweetid"), brand); - m_insertBrand.executeUpdate(); - } - } - } - } - - //gets the amount of users that tweet about a brand in a timezone - //makes a csv file timezone, brand, amount - public void timezone(String query) throws SQLException, FileNotFoundException, UnsupportedEncodingException{ - query(query); - - InputStream inFile = new FileInputStream("timezone.txt"); - Scanner readFile = new Scanner(inFile); - HashMap<String,String> toTimezone = new HashMap<>(); - while (readFile.hasNextLine()) { - String line = readFile.nextLine(); - if(line.split(",").length>1){ - toTimezone.put(line.split(",")[0], line.split(",")[1]); - } - } - - - - //hashmap timezone, brand, amount - HashMap<String, HashMap<String, Integer>> timeMap = new HashMap<>(); - String timezone; - String brand; - - while(data.next()){ - timezone = data.getString("timezone"); - if (toTimezone.containsKey(timezone)){ - timezone=toTimezone.get(timezone); - } else { - timezone="other"; - } - brand = data.getString("brand"); - //if the timezone is already in the map - if(timeMap.containsKey(timezone)){ - //if the brand for that timezone is already in the map - if(timeMap.get(timezone).containsKey(brand)){ - //increment the amount - timeMap.get(timezone).put(brand, timeMap.get(timezone).get(brand) + 1); - } - //if the brand for that timezone is not yet in the map - else{ - //make a new entry for that brand with amount = 1 - timeMap.get(timezone).put(brand, 1); - } - } - //if the timezone is not yet in the map - else{ - //make a new hashmap for this map and fill it with the brand and the amount - timeMap.put(timezone, new HashMap<String, Integer>()); - timeMap.get(timezone).put(brand, 1); - } - } - - - //make the CSV out of the map - mapToCSV(timeMap, "timezone.csv", "timezone,brand,count"); - } - - //replaces punctuation so it will be splitted - //also removes urls - private String splitPunctToWords(String text) { - text = text.replaceAll("https?://\\S*", ""); - text = text.replaceAll("[!?):;\"']", " $0"); - text = text.replaceAll("[.,-](\\s|$)", " $0"); - text = text.replaceAll("\\s[(\"']", "$0 "); - return text; - } - - //removes punctuation - //also removes urls - private String removePunct(String text) { - text = text.replaceAll("https?://\\S*", " "); - text = text.replaceAll("@\\S*", " "); - text = text.replaceAll("[^a-zA-Z0-9#_-]", " "); - return text; - } - - //prints a hashmap into a csv for a html application - //Hashmap<key1, HashMap<key2, value>> becomes key1, key2, value - //only for String, String, Integer - void mapToCSV(HashMap<String, HashMap<String, Integer>> map, String fileName, String firstLine) - throws FileNotFoundException, UnsupportedEncodingException{ - - PrintWriter writer = new PrintWriter(fileName, "UTF-8"); - - writer.println(firstLine); - - //loop over brands - for(Entry en : map.entrySet()){ - //loop over words - for(Entry e : map.get(en.getKey()).entrySet()){ - writer.println(en.getKey() + "," + e.getKey() + "," + e.getValue()); - } - } - - writer.close(); - System.out.println("csv file made, please put it next to html file and run this"); - } -} +package main;
+
+import analysis.BrandChecker;
+import database.NamedPreparedStatement;
+import database.QueryUtils;
+import java.io.File;
+import java.io.FileInputStream;
+import java.io.FileNotFoundException;
+import java.io.IOException;
+import java.io.InputStream;
+import java.io.PrintWriter;
+import java.io.UnsupportedEncodingException;
+import java.sql.Connection;
+import java.sql.PreparedStatement;
+import java.sql.ResultSet;
+import java.sql.SQLException;
+import java.util.List;
+import java.util.HashMap;
+import java.util.Locale;
+import java.util.Map.Entry;
+import java.util.Scanner;
+
+/**
+ * The sentiment analysis class that rates tweets based on a unigram and bigram
+ * set of weights.
+ */
+public class Analyzor {
+
+ /**
+ * The map that matches single words to their weights.
+ */
+ private final HashMap<String, Double> unimap = new HashMap();
+
+ /**
+ * The map that matches word pairs to their weights.
+ */
+ private final HashMap<String, Double> bimap = new HashMap();
+
+ /**
+ * The results of a query, maybe return from query().
+ */
+ private ResultSet data;
+
+ /**
+ * The persistent connection to the database.
+ */
+ private final Connection connection;
+
+ /**
+ * @param connection An open connection to the database.
+ */
+ public Analyzor(Connection connection) {
+ this.connection = connection;
+ }
+
+ /**
+ * Read the unigram and bigram lexica.
+ *
+ * @throws FileNotFoundException
+ */
+ public void readLexicon() throws FileNotFoundException {
+ if (!unimap.isEmpty()) {
+ // data is already read.
+ return;
+ }
+ System.err.println("Trying to read lexicons...");
+ // A unigram is in the format (WS = whitespace):
+ // word <WS> rating <WS> ??? <WS> ??
+ // A bigram has an two WS-separated words instead of one.
+ try (Scanner uniScanner = new Scanner(new File("unigrams-pmilexicon.txt"));
+ Scanner biScanner = new Scanner(new File("bigrams-pmilexicon.txt"));) {
+ //Fill the map of unigrams
+ int lineno = 1;
+ while (uniScanner.hasNext()) {
+
+ String words = uniScanner.next();
+ Double d = Double.valueOf(uniScanner.next());
+ unimap.put(words.toLowerCase(), d);
+ if (uniScanner.hasNextLine()) {
+ uniScanner.nextLine();
+ }
+ lineno++;
+
+ }
+
+ //fill the map of bigrams
+ while (biScanner.hasNext()) {
+ String words = biScanner.next() + " " + biScanner.next();
+ bimap.put(words.toLowerCase(), Double.valueOf(biScanner.next()));
+ if (biScanner.hasNextLine()) {
+ biScanner.nextLine();
+ }
+ }
+ }
+ System.err.println("Lexicons are read.");
+ }
+
+ /**
+ * Executes a query that the analyzer can analyze.
+ *
+ * @param query The query string to execute.
+ * @throws SQLException When database connection isn't available.
+ */
+ public void query(String query) throws SQLException {
+ PreparedStatement statement;
+ //make a connection to the database and execute the query
+ statement = connection.prepareStatement(query);
+ data = statement.executeQuery();
+ }
+
+ /**
+ * Run a sentiment analysis and fill the database with the output.
+ *
+ * @param query The sql text for the query.
+ * @throws SQLException
+ * @throws IOException
+ */
+ public void sentimentAnalysis(String query) throws SQLException, IOException {
+ query(query);
+
+ //read the lexicons
+ readLexicon();
+
+ //go to the start of te dataset
+ if (data == null) {
+ System.err.println("data is empty, try querying first");
+ return;
+ }
+
+ Double value;
+ String text;
+
+ //for all tuples
+ while (data.next()) {
+ //get the text
+ text = data.getString("text");
+ text = splitPunctToWords(text);
+ // test is the tweet text you are going to analyze
+ String[] words = text.split("\\s+"); // text splitted into separate words
+ double positiverate = 0; // positive rating
+
+ // Rate the text with unigrams
+ for (String word : words) {
+ value = unimap.get(word);
+ if (value != null) {
+ positiverate += unimap.get(word);
+ }
+ }
+ // Rate the text with bigrams
+ for (int i = 0; i < words.length - 1; i++) {
+ String pair = words[i] + " " + words[i + 1];
+ value = bimap.get(pair);
+ if (value != null) {
+ positiverate += bimap.get(pair);
+ }
+ }
+ //insert the rating into the database
+ NamedPreparedStatement m_insertRating;
+ m_insertRating = new NamedPreparedStatement(connection, QueryUtils.insertRating);
+ QueryUtils.setInsertParams(m_insertRating, data.getLong("tweetid"), data.getString("brand"), (int) (positiverate * 10));
+ m_insertRating.executeUpdate();
+ //don't print the rate
+ //System.out.println(text + ": " + (int) (positiverate * 10));
+ }
+ }
+
+ /**
+ * Make a wordcloud of the results of some query.
+ *
+ * @param query The sql text for a query.
+ * @throws SQLException
+ * @throws FileNotFoundException
+ * @throws UnsupportedEncodingException
+ */
+ public void makeWordCloud(String query) throws SQLException, FileNotFoundException, UnsupportedEncodingException {
+
+ query(query);
+ //go to the start of the ResultSet data
+ if (data == null) {
+ System.err.println("data is empty, try querying first");
+ return;
+ }
+
+ String text;
+ String brand;
+ String[] words;
+ HashMap<String,HashMap<String, Integer>> wordcloud = new HashMap<>();
+
+ while (data.next()) {
+ //get brand
+ brand=data.getString("brand");
+ //make hashmap for each brand
+ if(!wordcloud.containsKey(brand)){
+ wordcloud.put(brand, new HashMap<String,Integer>());
+ }
+ //get the text
+ text = data.getString("text");
+ //remove punctuation, convert to lowercase and split on words
+ text = removePunct(text);
+ text = text.toLowerCase();
+ words = text.split("\\s+");
+ //for all words
+ for (String word : words) {
+ //if it is empty, a space or a stripe, skip it
+ if(word.equals("") || word.equals(" ") || word.equals("-")){
+ continue;
+ }
+ //if the word is already in the map, increment the amount
+ if(wordcloud.get(brand).containsKey(word)){
+ wordcloud.get(brand).put(word, wordcloud.get(brand).get(word) + 1);
+ }
+ //if the word is not already in the map, make an entry with amount = 1
+ else{
+ wordcloud.get(brand).put(word, 1);
+ }
+ }
+ }
+ //print the words and their frequency in a csv file
+ mapToCSV(wordcloud, "wordcloud.csv", "brand,word,count");
+ }
+
+ //generate csv for disco from the query
+ public void disco(String query) throws SQLException, FileNotFoundException, UnsupportedEncodingException {
+ //do the query
+ query(query);
+ PrintWriter writer = new PrintWriter("output.csv", "UTF-8");
+ //print the first row
+ for (int i = 1; i < data.getMetaData().getColumnCount(); i++) {
+ writer.print(data.getMetaData().getColumnLabel(i) + ", ");
+ }
+ writer.println(data.getMetaData().getColumnLabel(data.getMetaData().getColumnCount()));
+ //print the values
+ while (data.next()) {
+ for (int i = 1; i < data.getMetaData().getColumnCount(); i++) {
+ if (data.getObject(i) == null) {
+ writer.print(", ");
+ } else {
+ writer.print(data.getObject(i).toString().replaceAll("[,\n]", " ") + ", ");
+ }
+ }
+ if (data.getObject(data.getMetaData().getColumnCount()) == null) {
+ writer.println("0");
+ } else {
+ writer.println(data.getObject(data.getMetaData().getColumnCount()).toString().replace(",", " "));
+ }
+ }
+ writer.close();
+ }
+
+ public void getBrands() throws SQLException {
+ PreparedStatement statement;
+ //make a connection to the database and execute the query
+ statement = connection.prepareStatement("delete from mentionsbrand");
+ statement.executeUpdate();
+ BrandChecker checker = new BrandChecker("brandonlyrules.txt");
+ query("select * from tweet");
+ NamedPreparedStatement m_insertBrand = new NamedPreparedStatement(connection, QueryUtils.insertBrand);
+ while (data.next()) {
+ List<String> brands = checker.getBrands(data.getString("text"));
+ if (brands.isEmpty()) {
+ QueryUtils.setInsertBrandParams(m_insertBrand, data.getLong("tweetid"), "no");
+ m_insertBrand.executeUpdate();
+ } else {
+ for (String brand : brands) {
+ QueryUtils.setInsertBrandParams(m_insertBrand, data.getLong("tweetid"), brand);
+ m_insertBrand.executeUpdate();
+ }
+ }
+ }
+ }
+
+ //gets the amount of users that tweet about a brand in a timezone
+ //makes a csv file timezone, brand, amount
+ public void timezone(String query) throws SQLException, FileNotFoundException, UnsupportedEncodingException{
+ query(query);
+
+ InputStream inFile = new FileInputStream("timezone.txt");
+ Scanner readFile = new Scanner(inFile);
+ HashMap<String,String> toTimezone = new HashMap<>();
+ while (readFile.hasNextLine()) {
+ String line = readFile.nextLine();
+ if(line.split(",").length>1){
+ toTimezone.put(line.split(",")[0], line.split(",")[1]);
+ }
+ }
+
+
+
+ //hashmap timezone, brand, amount
+ HashMap<String, HashMap<String, Integer>> timeMap = new HashMap<>();
+ String timezone;
+ String brand;
+
+ while(data.next()){
+ timezone = data.getString("timezone");
+ if (toTimezone.containsKey(timezone)){
+ timezone=toTimezone.get(timezone);
+ } else {
+ timezone="other";
+ }
+ brand = data.getString("brand");
+ //if the timezone is already in the map
+ if(timeMap.containsKey(timezone)){
+ //if the brand for that timezone is already in the map
+ if(timeMap.get(timezone).containsKey(brand)){
+ //increment the amount
+ timeMap.get(timezone).put(brand, timeMap.get(timezone).get(brand) + 1);
+ }
+ //if the brand for that timezone is not yet in the map
+ else{
+ //make a new entry for that brand with amount = 1
+ timeMap.get(timezone).put(brand, 1);
+ }
+ }
+ //if the timezone is not yet in the map
+ else{
+ //make a new hashmap for this map and fill it with the brand and the amount
+ timeMap.put(timezone, new HashMap<String, Integer>());
+ timeMap.get(timezone).put(brand, 1);
+ }
+ }
+
+
+ //make the CSV out of the map
+ mapToCSV(timeMap, "timezone.csv", "timezone,brand,count");
+ }
+
+ //replaces punctuation so it will be splitted
+ //also removes urls
+ private String splitPunctToWords(String text) {
+ text = text.replaceAll("https?://\\S*", "");
+ text = text.replaceAll("[!?):;\"']", " $0");
+ text = text.replaceAll("[.,-](\\s|$)", " $0");
+ text = text.replaceAll("\\s[(\"']", "$0 ");
+ return text;
+ }
+
+ //removes punctuation
+ //also removes urls
+ private String removePunct(String text) {
+ text = text.replaceAll("https?://\\S*", " ");
+ text = text.replaceAll("@\\S*", " ");
+ text = text.replaceAll("[^a-zA-Z0-9#_-]", " ");
+ return text;
+ }
+
+ //prints a hashmap into a csv for a html application
+ //Hashmap<key1, HashMap<key2, value>> becomes key1, key2, value
+ //only for String, String, Integer
+ void mapToCSV(HashMap<String, HashMap<String, Integer>> map, String fileName, String firstLine)
+ throws FileNotFoundException, UnsupportedEncodingException{
+
+ PrintWriter writer = new PrintWriter(fileName, "UTF-8");
+
+ writer.println(firstLine);
+
+ //loop over brands
+ for(Entry en : map.entrySet()){
+ //loop over words
+ for(Entry e : map.get(en.getKey()).entrySet()){
+ writer.println(en.getKey() + "," + e.getKey() + "," + e.getValue());
+ }
+ }
+
+ writer.close();
+ System.out.println("csv file made, please put it next to html file and run this");
+ }
+}
diff --git a/src/main/FarmShell.java b/src/main/FarmShell.java index 8fc515a..6bf350e 100644 --- a/src/main/FarmShell.java +++ b/src/main/FarmShell.java @@ -1,212 +1,217 @@ -package main; - -import database.ConnectionBuilder; -import java.io.IOException; -import java.sql.Connection; -import java.sql.SQLException; -import java.util.Arrays; -import java.util.NoSuchElementException; -import java.util.Scanner; - -/** - * - * @author s123188 - */ -public class FarmShell { - - /** - * A scanner for the stdin. - */ - private final Scanner scanner = new Scanner(System.in); - - private Analyzor cached_analyzor; - private final ConnectionBuilder dbConnectionBuilder; - - FarmShell(ConnectionBuilder dbConnectionBuilder) { - this.dbConnectionBuilder = dbConnectionBuilder; - } - - private void printPrompt() { - System.out.print("$ "); - } - - private Analyzor getAnalyzor() throws SQLException { - if (cached_analyzor == null) { - Connection dbCon = dbConnectionBuilder.create(); - cached_analyzor = new Analyzor(dbCon); - } - return cached_analyzor; - } - - /** - * Processes commands from stdin until the exit command is received or EOF. - */ - public void process_forever() { - System.err.println("Entering interactive shell, type 'help' for help " - + "or 'exit' to leave. '.' repeats the previous interactive " - + "command."); - // print prompt for reading first command - printPrompt(); - String lastLine = ""; - while (scanner.hasNextLine()) { - String line = scanner.nextLine().trim(); - // repeat last command - if (line.equals(".")) { - line = lastLine; - } - if (!execute(line)) { - // requested to terminate - break; - } - if (!line.isEmpty()) { - lastLine = line; - } - // print prompt for reading next line - printPrompt(); - } - } - - /** - * Execute a single commands. - * - * @param cmd A single line of the command. - * @return Whether to continue or exit the application. - */ - public boolean execute(String cmd) { - String[] args = cmd.trim().split("\\s+", 2); - if (!args[0].isEmpty()) { - // non-empty command, let's see whether it makes sense? - return execute(args); - } - return true; - } - - /** - * Executes a command with optional parameters. - * - * @param args An array with the first argument containing the command with - * optional parameters in following arguments. - * @return true if more commands are allowed to be executed, false - * otherwise. - */ - public boolean execute(String[] args) { - try { - Command command = Command.fromString(args[0]); - String[] params = Arrays.copyOfRange(args, 1, args.length); - execute(command, params); - } catch (IllegalArgumentException ex) { - System.err.println(ex.getMessage()); - } catch (IOException ex) { - System.err.println("Command " + args[0] + " failed with " + ex); - ex.printStackTrace(); - } catch (NoSuchElementException ex) { - if ("EXIT NOW".equals(ex.getMessage())) { - // thrown by the "exit" command to signal exit - return false; - } else { - System.err.println("ZOMG SOMETHIGN FAILED: " + ex.getMessage()); - ex.printStackTrace(); - } - } catch (SQLException ex) { - System.err.println("such " + ex); - } - // another satisfied customer, next! - return true; - } - - private void execute(Command command, String[] params) throws SQLException, IOException { - if (params.length < command.getParamCount()) { - throw new IllegalArgumentException("Expected " - + command.getParamCount() + " parameters, got only " - + params.length); - } - switch (command) { - case filterbots: - System.out.println("not yet implemented"); - break; - case sentiment: - getAnalyzor().sentimentAnalysis(params[0]); - break; - case wordcloud: - getAnalyzor().makeWordCloud(params[0]); - break; - case timezone: - getAnalyzor().timezone(params[0]); - case disco: - getAnalyzor().disco(params[0]); - break; - case getBrands: - getAnalyzor().getBrands(); - break; - case help: - for (String line : HELP) { - System.out.println(line); - } - for (Command cmd : Command.values()) { - System.out.printf(" %-10s", cmd.name()); - if (!cmd.getDescription().isEmpty()) { - System.out.print(" " + cmd.getDescription()); - } - if (cmd.getParamCount() == 1) { - System.out.print(" (1 arg)"); - } else if (cmd.getParamCount() > 1) { - System.out.printf(" (%d args)", cmd.getParamCount()); - } - System.out.println(); - } - break; - case exit: - throw new NoSuchElementException("EXIT NOW"); - default: - throw new AssertionError(command.name()); - } - } - - enum Command { - - filterbots("marks all users as bot or not", 1), - sentiment("analyzes all tweets on positivity (about a brand)", 1), - wordcloud("makes a wordcloud of the text of the tweets", 1), - getBrands("fills the database with the brands of a tweet"), - timezone("makes a map per brand for the users", 1), - disco("makes a outputfile for disco", 1), - exit("Returns to shell"), - help("Get help"); - - private final String description; - private final int paramCount; - - Command(String description) { - this.description = description; - this.paramCount = 0; - } - - Command(String description, int paramCount) { - this.description = description; - this.paramCount = paramCount; - } - - public String getDescription() { - return description; - } - - public int getParamCount() { - return paramCount; - } - - public static Command fromString(String command) { - for (Command cmd : values()) { - if (cmd.name().equals(command)) { - return cmd; - } - } - throw new IllegalArgumentException("Unrecognized command. Hint: help"); - } - }; - - private final String[] HELP = new String[]{ - "Interactive TweetShell", - "", - "Available commands:" - }; -} +package main;
+
+import database.ConnectionBuilder;
+import java.io.IOException;
+import java.sql.Connection;
+import java.sql.SQLException;
+import java.util.Arrays;
+import java.util.NoSuchElementException;
+import java.util.Scanner;
+
+/**
+ *
+ * @author s123188
+ */
+public class FarmShell {
+
+ /**
+ * A scanner for the stdin.
+ */
+ private final Scanner scanner = new Scanner(System.in);
+
+ private Analyzor cached_analyzor;
+ private final ConnectionBuilder dbConnectionBuilder;
+
+ FarmShell(ConnectionBuilder dbConnectionBuilder) {
+ this.dbConnectionBuilder = dbConnectionBuilder;
+ }
+
+ private void printPrompt() {
+ System.out.print("$ ");
+ }
+
+ private Analyzor getAnalyzor() throws SQLException {
+ if (cached_analyzor == null) {
+ Connection dbCon = dbConnectionBuilder.create();
+ cached_analyzor = new Analyzor(dbCon);
+ }
+ return cached_analyzor;
+ }
+
+ /**
+ * Processes commands from stdin until the exit command is received or EOF.
+ */
+ public void process_forever() {
+ System.err.println("Entering interactive shell, type 'help' for help "
+ + "or 'exit' to leave. '.' repeats the previous interactive "
+ + "command.");
+ // print prompt for reading first command
+ printPrompt();
+ String lastLine = "";
+ while (scanner.hasNextLine()) {
+ String line = scanner.nextLine().trim();
+ // repeat last command
+ if (line.equals(".")) {
+ line = lastLine;
+ }
+ if (!execute(line)) {
+ // requested to terminate
+ break;
+ }
+ if (!line.isEmpty()) {
+ lastLine = line;
+ }
+ // print prompt for reading next line
+ printPrompt();
+ }
+ }
+
+ /**
+ * Execute a single commands.
+ *
+ * @param cmd A single line of the command.
+ * @return Whether to continue or exit the application.
+ */
+ public boolean execute(String cmd) {
+ String[] args = cmd.trim().split("\\s+", 2);
+ if (!args[0].isEmpty()) {
+ // non-empty command, let's see whether it makes sense?
+ return execute(args);
+ }
+ return true;
+ }
+
+ /**
+ * Executes a command with optional parameters.
+ *
+ * @param args An array with the first argument containing the command with
+ * optional parameters in following arguments.
+ * @return true if more commands are allowed to be executed, false
+ * otherwise.
+ */
+ public boolean execute(String[] args) {
+ try {
+ Command command = Command.fromString(args[0]);
+ String[] params = Arrays.copyOfRange(args, 1, args.length);
+ execute(command, params);
+ } catch (IllegalArgumentException ex) {
+ System.err.println(ex.getMessage());
+ } catch (IOException ex) {
+ System.err.println("Command " + args[0] + " failed with " + ex);
+ ex.printStackTrace();
+ } catch (NoSuchElementException ex) {
+ if ("EXIT NOW".equals(ex.getMessage())) {
+ // thrown by the "exit" command to signal exit
+ return false;
+ } else {
+ System.err.println("ZOMG SOMETHIGN FAILED: " + ex.getMessage());
+ ex.printStackTrace();
+ }
+ } catch (SQLException ex) {
+ System.err.println("such " + ex);
+ }
+ // another satisfied customer, next!
+ return true;
+ }
+
+ private void execute(Command command, String[] params) throws SQLException, IOException {
+ if (params.length < command.getParamCount()) {
+ throw new IllegalArgumentException("Expected "
+ + command.getParamCount() + " parameters, got only "
+ + params.length);
+ }
+ switch (command) {
+ case filterbots:
+ System.out.println("not yet implemented");
+ break;
+ case sentiment:
+ // if there is no query, update all unrated items.
+ if (params.length > 0) {
+ getAnalyzor().sentimentAnalysis(params[0]);
+ } else {
+ getAnalyzor().sentimentAnalysis("");
+ }
+ break;
+ case wordcloud:
+ getAnalyzor().makeWordCloud(params[0]);
+ break;
+ case timezone:
+ getAnalyzor().timezone(params[0]);
+ case disco:
+ getAnalyzor().disco(params[0]);
+ break;
+ case getBrands:
+ getAnalyzor().getBrands();
+ break;
+ case help:
+ for (String line : HELP) {
+ System.out.println(line);
+ }
+ for (Command cmd : Command.values()) {
+ System.out.printf(" %-10s", cmd.name());
+ if (!cmd.getDescription().isEmpty()) {
+ System.out.print(" " + cmd.getDescription());
+ }
+ if (cmd.getParamCount() == 1) {
+ System.out.print(" (1 arg)");
+ } else if (cmd.getParamCount() > 1) {
+ System.out.printf(" (%d args)", cmd.getParamCount());
+ }
+ System.out.println();
+ }
+ break;
+ case exit:
+ throw new NoSuchElementException("EXIT NOW");
+ default:
+ throw new AssertionError(command.name());
+ }
+ }
+
+ enum Command {
+
+ filterbots("marks all users as bot or not", 1),
+ sentiment("analyzes all tweets on brand positivity (optional arg: tweet/brand selection query)"),
+ wordcloud("makes a wordcloud of the text of the tweets", 1),
+ getBrands("fills the database with the brands of a tweet"),
+ timezone("makes a map per brand for the users", 1),
+ disco("makes a outputfile for disco", 1),
+ exit("Returns to shell"),
+ help("Get help");
+
+ private final String description;
+ private final int paramCount;
+
+ Command(String description) {
+ this.description = description;
+ this.paramCount = 0;
+ }
+
+ Command(String description, int paramCount) {
+ this.description = description;
+ this.paramCount = paramCount;
+ }
+
+ public String getDescription() {
+ return description;
+ }
+
+ public int getParamCount() {
+ return paramCount;
+ }
+
+ public static Command fromString(String command) {
+ for (Command cmd : values()) {
+ if (cmd.name().equals(command)) {
+ return cmd;
+ }
+ }
+ throw new IllegalArgumentException("Unrecognized command. Hint: help");
+ }
+ };
+
+ private final String[] HELP = new String[]{
+ "Interactive TweetShell",
+ "",
+ "Available commands:"
+ };
+}
diff --git a/test/analysis/BrandCheckerTest.java b/test/analysis/BrandCheckerTest.java index 23d8445..f55035b 100644 --- a/test/analysis/BrandCheckerTest.java +++ b/test/analysis/BrandCheckerTest.java @@ -82,10 +82,10 @@ public class BrandCheckerTest { public void testBullshit() { doTest("This applepie is delicious", new String[]{}); } - + @Test public void multipleBrands() { - doTest("This tweet contains both iphone 4s,galaxy s5 and iphone", new String[]{"iphone 4s","galaxy s5"}); + doTest("This tweet contains both iphone 4s,galaxy s5 and iphone", new String[]{"iphone 4s", "galaxy s5"}); } } |