|
| 1 | +package org.itmo; |
| 2 | + |
| 3 | +import org.apache.kafka.clients.producer.KafkaProducer; |
| 4 | + |
| 5 | +import java.io.FileNotFoundException; |
| 6 | +import java.io.IOException; |
| 7 | +import java.nio.charset.StandardCharsets; |
| 8 | +import java.nio.file.Files; |
| 9 | +import java.nio.file.InvalidPathException; |
| 10 | +import java.nio.file.Path; |
| 11 | +import java.nio.file.Paths; |
| 12 | +import java.util.ArrayList; |
| 13 | +import java.util.List; |
| 14 | + |
| 15 | +public class FileProducer extends KafkaProducerBase { |
| 16 | + private final String filename; |
| 17 | + private final int topN; |
| 18 | + public static final String[] TASK_TYPES = { |
| 19 | + "WORD_COUNT", |
| 20 | + "TOP_N", |
| 21 | + "SENTIMENT", |
| 22 | + "NAME_REPLACE", |
| 23 | + "SORT_SENTENCES" |
| 24 | + }; |
| 25 | + // public static final String[] TASK_TOPICS = { |
| 26 | + // "task-topic-word-count", |
| 27 | + // "task-topic-top-n", |
| 28 | + // "task-topic-sentiment", |
| 29 | + // "task-topic-name-replace", |
| 30 | + // "task-topic-sort-sentences" |
| 31 | + // }; |
| 32 | + |
| 33 | + public FileProducer(KafkaProducer<String, byte[]> sharedProducer, String filename) { |
| 34 | + super(sharedProducer); |
| 35 | + this.filename = filename; |
| 36 | + this.topN = 10; |
| 37 | + } |
| 38 | + |
| 39 | + public FileProducer(KafkaProducer<String, byte[]> sharedProducer, String filename, int topN) { |
| 40 | + super(sharedProducer); |
| 41 | + this.filename = filename; |
| 42 | + this.topN = topN; |
| 43 | + } |
| 44 | + |
| 45 | + /** |
| 46 | + * Разбивает файл на секции (по параграфам) и отправляет задачи в Kafka topics |
| 47 | + * @return количество созданных секций |
| 48 | + */ |
| 49 | + public int prepareTextAndProduce() throws IOException { |
| 50 | + Path file; |
| 51 | + |
| 52 | + try { |
| 53 | + file = Paths.get(this.filename); |
| 54 | + if (!Files.isRegularFile(file)) { |
| 55 | + throw new FileNotFoundException(); |
| 56 | + } |
| 57 | + } catch (InvalidPathException e) { |
| 58 | + throw new IOException("Failed to open file " + filename + "\n" + e.getMessage()); |
| 59 | + } catch (FileNotFoundException e) { |
| 60 | + throw new IOException("File " + filename + " is not present or a directory"); |
| 61 | + } |
| 62 | + |
| 63 | + // Читаем весь файл |
| 64 | + String fileContent = new String(Files.readAllBytes(file), StandardCharsets.UTF_8); |
| 65 | + |
| 66 | + // Разбиваем на секции по параграфам (двойной перенос строки) или предложениям |
| 67 | + List<String> sections = splitIntoSections(fileContent); |
| 68 | + |
| 69 | + System.out.println("[Producer] Split file into " + sections.size() + " sections"); |
| 70 | + |
| 71 | + // Создаем и отправляем задачи для каждой секции |
| 72 | + try { |
| 73 | + int taskCounter = 0; |
| 74 | + for (int sectionIndex = 0; sectionIndex < sections.size(); sectionIndex++) { |
| 75 | + String section = sections.get(sectionIndex); |
| 76 | + // System.out.println("[Producer] Processing section " + sectionIndex + " (length: " + section.length() + " chars)"); |
| 77 | + |
| 78 | + // Отправляем задачи всех типов для каждой секции |
| 79 | + for (int taskTypeIndex = 0; taskTypeIndex < TASK_TYPES.length; taskTypeIndex++) { |
| 80 | + String taskType = TASK_TYPES[taskTypeIndex]; |
| 81 | + String taskTopic = TASK_TYPES[taskTypeIndex]; |
| 82 | + |
| 83 | + TaskMessage task; |
| 84 | + if ("TOP_N".equals(taskType)) { |
| 85 | + task = new TaskMessage(taskType, section, sectionIndex, topN); |
| 86 | + } else { |
| 87 | + task = new TaskMessage(taskType, section, sectionIndex); |
| 88 | + } |
| 89 | + |
| 90 | + byte[] messageBytes = MessageSerializer.serialize(task); |
| 91 | + // Use taskId as key for partition distribution |
| 92 | + String key = task.getTaskId(); |
| 93 | + taskCounter++; |
| 94 | + // System.out.println("[Producer] Sending task #" + (taskCounter) + ": " + taskType + |
| 95 | + // " for section " + sectionIndex + " (message size: " + messageBytes.length + " bytes)"); |
| 96 | + super.produce(taskTopic, key, messageBytes); |
| 97 | + } |
| 98 | + } |
| 99 | + |
| 100 | + // Ждем завершения всех отправок |
| 101 | + super.producer.flush(); |
| 102 | + System.out.println("[Producer] All tasks sent to Kafka topics (" + (sections.size() * TASK_TYPES.length) + " tasks total)"); |
| 103 | + |
| 104 | + } catch (Exception e) { |
| 105 | + throw new IOException("Error while sending messages to Kafka: " + e.getMessage(), e); |
| 106 | + } |
| 107 | + |
| 108 | + return sections.size(); |
| 109 | + } |
| 110 | + |
| 111 | + /** |
| 112 | + * Подсчитывает количество секций в файле без разбиения |
| 113 | + */ |
| 114 | + public int countSections() throws IOException { |
| 115 | + Path file; |
| 116 | + |
| 117 | + try { |
| 118 | + file = Paths.get(this.filename); |
| 119 | + if (!Files.isRegularFile(file)) { |
| 120 | + throw new FileNotFoundException(); |
| 121 | + } |
| 122 | + } catch (InvalidPathException e) { |
| 123 | + throw new IOException("Failed to open file " + filename + "\n" + e.getMessage()); |
| 124 | + } catch (FileNotFoundException e) { |
| 125 | + throw new IOException("File " + filename + " is not present or a directory"); |
| 126 | + } |
| 127 | + |
| 128 | + // Читаем весь файл |
| 129 | + String fileContent = new String(Files.readAllBytes(file), StandardCharsets.UTF_8); |
| 130 | + |
| 131 | + // Разбиваем на секции по параграфам (двойной перенос строки) или предложениям |
| 132 | + List<String> sections = splitIntoSections(fileContent); |
| 133 | + |
| 134 | + return sections.size(); |
| 135 | + } |
| 136 | + |
| 137 | + /** |
| 138 | + * Разбивает текст на секции по параграфам |
| 139 | + */ |
| 140 | + private List<String> splitIntoSections(String text) { |
| 141 | + List<String> sections = new ArrayList<>(); |
| 142 | + |
| 143 | + // Разбиваем по двойному переносу строки (параграфы) |
| 144 | + String[] paragraphs = text.split("\\n\\s*\\n+"); |
| 145 | + |
| 146 | + // Если получилось слишком мало параграфов, разбиваем по предложениям |
| 147 | + if (paragraphs.length < 2) { |
| 148 | + paragraphs = text.split("[.!?]+\\s+"); |
| 149 | + } |
| 150 | + |
| 151 | + // Объединяем очень короткие секции |
| 152 | + StringBuilder currentSection = new StringBuilder(); |
| 153 | + int minSectionLength = 100; // минимальная длина секции |
| 154 | + |
| 155 | + for (String paragraph : paragraphs) { |
| 156 | + String trimmed = paragraph.trim(); |
| 157 | + if (trimmed.isEmpty()) { |
| 158 | + continue; |
| 159 | + } |
| 160 | + |
| 161 | + if (currentSection.length() > 0 && |
| 162 | + currentSection.length() + trimmed.length() < minSectionLength) { |
| 163 | + // Добавляем к текущей секции |
| 164 | + currentSection.append(" ").append(trimmed); |
| 165 | + } else { |
| 166 | + // Сохраняем предыдущую секцию и начинаем новую |
| 167 | + if (currentSection.length() > 0) { |
| 168 | + sections.add(currentSection.toString()); |
| 169 | + } |
| 170 | + currentSection = new StringBuilder(trimmed); |
| 171 | + } |
| 172 | + } |
| 173 | + |
| 174 | + // Добавляем последнюю секцию |
| 175 | + if (currentSection.length() > 0) { |
| 176 | + sections.add(currentSection.toString()); |
| 177 | + } |
| 178 | + |
| 179 | + // Если всё ещё пусто, используем весь текст как одну секцию |
| 180 | + if (sections.isEmpty()) { |
| 181 | + sections.add(text.trim()); |
| 182 | + } |
| 183 | + |
| 184 | + return sections; |
| 185 | + } |
| 186 | +} |
0 commit comments