package service;

import extraction.DimensionIdentifier;
import extraction.PositionedTextStripper;
import extraction.TextGrouper;
import lombok.extern.slf4j.Slf4j;
import model.DimensionResult;
import model.TextElement;
import model.TextGroup;
import org.apache.pdfbox.Loader;
import org.apache.pdfbox.pdmodel.PDDocument;

import java.io.ByteArrayOutputStream;
import java.io.File;
import java.io.IOException;
import java.io.InputStream;
import java.io.StringWriter;
import java.net.MalformedURLException;
import java.net.URL;
import java.nio.file.Path;
import java.util.List;
import java.util.stream.Collectors;

@Slf4j
public class PdfExtractionService {

    private final TextGrouper textGrouper;
    private final DimensionIdentifier dimensionIdentifier;

    public PdfExtractionService(TextGrouper textGrouper,
                                DimensionIdentifier dimensionIdentifier) {
        this.textGrouper = textGrouper;
        this.dimensionIdentifier = dimensionIdentifier;
    }

    public List<DimensionResult> extractDimensions(String pdfUrl) throws IOException {
        try (PDDocument document = Loader.loadPDF(getPdfBytes(pdfUrl))) {
//            int totalPages = document.getNumberOfPages();
            PositionedTextStripper stripper = new PositionedTextStripper();
            stripper.setSortByPosition(true);
            StringWriter writer = new StringWriter();
            stripper.writeText(document, writer);
            List<TextElement> allElements = stripper.getElements();

            List<TextGroup> groups = textGrouper.groupTextElements(allElements);
            return dimensionIdentifier.identifyDimensions(groups);
        }
    }

    private byte[] getPdfBytes(String pdfUrl) throws IOException {
        try (InputStream in = new URL(pdfUrl).openStream();
             ByteArrayOutputStream buffer = new ByteArrayOutputStream()) {
            byte[] chunk = new byte[8192];
            int len;
            while ((len = in.read(chunk)) != -1) {
                buffer.write(chunk, 0, len);
            }
            return buffer.toByteArray();
        }
    }

    /**
     * 区域提取：直接返回区域内的原始文本组数据（经过乱码修正），
     * 不经过 DimensionIdentifier 的模式匹配和符号推断逻辑。
     */
    public List<DimensionResult> extractRawTextForRegion(String pdfUrl,
                                                         double rx, double ry, double rw, double rh) throws IOException {
        try (PDDocument document = Loader.loadPDF(getPdfBytes(pdfUrl))) {
            PositionedTextStripper stripper = new PositionedTextStripper();
            stripper.setSortByPosition(true);
            StringWriter writer = new StringWriter();
            stripper.writeText(document, writer);
            List<TextElement> allElements = stripper.getElements();
            List<TextGroup> groups = textGrouper.groupTextElements(allElements);

            // 筛选落在指定区域内的 TextGroup，直接返回其文本内容
            List<DimensionResult> results = new java.util.ArrayList<>();
            for (TextGroup g : groups) {
                String text = g.getText().trim();
                if (text.isEmpty()) continue;

                // 对旋转文本和水平文本分别计算边界框
                double gx, gy, gw, gh;
                if (g.isRotated()) {
                    // 旋转文本: X=列位, Width=Y跨度, Height=字符厚度
                    gx = g.getX();
                    gy = g.getY();
                    gw = g.getHeight(); // X方向宽度 = 字符厚度
                    gh = g.getWidth();  // Y方向高度 = Y跨度
                } else {
                    gx = g.getX();
                    gy = g.getY() - g.getHeight();
                    gw = g.getWidth();
                    gh = g.getHeight();
                }
                if (gx + gw < rx || gx > rx + rw || gy + gh < ry || gy > ry + rh) continue;

                // 中心点Y检查：防止大字号符号（如Φ fs=24）的边界框向上延伸过多导致误入区域。
                // 仅对非旋转文本应用，旋转文本的Y跨度由文本长度决定，不受此问题影响。
                if (!g.isRotated()) {
                    double centerY = gy + gh / 2.0;
                    if (gh > 0 && (centerY < ry || centerY > ry + rh)) continue;
                }

                // 超宽边界框检查：CAD导出PDF中文本字距极大时，边界框右边缘远超实际可视范围
                // 当文本宽度异常（超过字符数×字号×2）且起始位置在区域左侧外时，跳过
                if (!g.isRotated()) {
                    double normalWidth = text.length() * g.getFontSize();
                    if (gw > normalWidth * 2 && gx < rx - g.getFontSize()) {
                        continue;
                    }
                }

                log.debug("Region match: text=[{}] rotated={} gx={} gy={} gw={} gh={} origX={} origY={} origW={} origH={}",
                        text, g.isRotated(), gx, gy, gw, gh, g.getX(), g.getY(), g.getWidth(), g.getHeight());

                DimensionResult r = new DimensionResult();
                r.setDimension(text);
                r.setTolerance(null);
                r.setType("text");
                r.setX(Math.round(gx * 100.0) / 100.0);
                r.setY(Math.round(gy * 100.0) / 100.0);
                r.setWidth(Math.round(gw * 100.0) / 100.0);
                r.setHeight(Math.round(gh * 100.0) / 100.0);
                r.setPage(g.getPageNum());
                results.add(r);
            }
            return results;
        }
    }
}