first implementation - image/drawing integration

2026-04-04 12:56:56 +02:00
parent fc5b22fba1
commit 5acfdd33c1
42 changed files with 2854 additions and 151 deletions
@@ -0,0 +1,65 @@
+package com.aiteacher.document;
+
+import org.springframework.ai.document.Document;
+import org.springframework.stereotype.Service;
+
+import java.util.ArrayList;
+import java.util.HashMap;
+import java.util.List;
+import java.util.Map;
+import java.util.UUID;
+
+/**
+ * Splits a SectionEntity's full text into overlapping chunks for vector embedding.
+ * Target size: ~1800 characters (~450 tokens); overlap: 200 characters.
+ */
+@Service
+public class TextChunkingService {
+
+    private static final int TARGET_CHARS = 1800;
+    private static final int OVERLAP_CHARS = 200;
+
+    public List<Document> chunk(SectionEntity section, String bookTitle) {
+        String text = section.getFullText();
+        if (text == null || text.isBlank()) return List.of();
+
+        List<String> windows = split(text);
+        List<Document> documents = new ArrayList<>();
+
+        for (int i = 0; i < windows.size(); i++) {
+            String chunkId = UUID.randomUUID().toString();
+            Map<String, Object> metadata = buildMetadata(section, bookTitle, i, windows.size(), chunkId);
+            documents.add(new Document(chunkId, windows.get(i), metadata));
+        }
+        return documents;
+    }
+
+    private List<String> split(String text) {
+        List<String> windows = new ArrayList<>();
+        int start = 0;
+        while (start < text.length()) {
+            int end = Math.min(start + TARGET_CHARS, text.length());
+            windows.add(text.substring(start, end));
+            if (end == text.length()) break;
+            start = end - OVERLAP_CHARS;
+        }
+        return windows;
+    }
+
+    private Map<String, Object> buildMetadata(SectionEntity section, String bookTitle,
+                                               int index, int total, String chunkId) {
+        Map<String, Object> m = new HashMap<>();
+        m.put("type", "TEXT");
+        m.put("book_id", section.getBookId().toString());
+        m.put("book_title", bookTitle);
+        m.put("chapter_id", section.getChapterId());
+        m.put("section_id", section.getId());
+        m.put("section_title", section.getTitle() != null ? section.getTitle() : "");
+        m.put("page_start", section.getPageStart());
+        m.put("page_end", section.getPageEnd());
+        m.put("chunk_index", index);
+        m.put("total_chunks", total);
+        m.put("chunk_id", chunkId);
+        return m;
+    }
+}