Feature: Support manual redaction

src/main/java/stirling/software/SPDF/config/EndpointConfiguration.java

@@ -135,6 +135,7 @@ public void init() {
         addEndpointToGroup("Security", "remove-cert-sign");
         addEndpointToGroup("Security", "sanitize-pdf");
         addEndpointToGroup("Security", "auto-redact");
+        addEndpointToGroup("Security", "redact");
 
         // Adding endpoints to "Other" group
         addEndpointToGroup("Other", "ocr-pdf");
@@ -234,6 +235,7 @@ public void init() {
         addEndpointToGroup("Java", "markdown-to-pdf");
         addEndpointToGroup("Java", "show-javascript");
         addEndpointToGroup("Java", "auto-redact");
+        addEndpointToGroup("Java", "redact");
         addEndpointToGroup("Java", "pdf-to-csv");
         addEndpointToGroup("Java", "split-by-size-or-count");
         addEndpointToGroup("Java", "overlay-pdf");

src/main/java/stirling/software/SPDF/controller/api/security/RedactController.java

@@ -3,13 +3,18 @@
 import java.awt.*;
 import java.io.ByteArrayOutputStream;
 import java.io.IOException;
+import java.util.Collections;
 import java.util.List;
 
 import org.apache.pdfbox.pdmodel.PDDocument;
+import org.apache.pdfbox.pdmodel.PDPage;
 import org.apache.pdfbox.pdmodel.PDPageContentStream;
+import org.apache.pdfbox.pdmodel.PDPageTree;
 import org.apache.pdfbox.pdmodel.common.PDRectangle;
 import org.springframework.beans.factory.annotation.Autowired;
 import org.springframework.http.ResponseEntity;
+import org.springframework.web.bind.WebDataBinder;
+import org.springframework.web.bind.annotation.InitBinder;
 import org.springframework.web.bind.annotation.ModelAttribute;
 import org.springframework.web.bind.annotation.PostMapping;
 import org.springframework.web.bind.annotation.RequestMapping;
@@ -21,12 +26,17 @@
 import io.swagger.v3.oas.annotations.tags.Tag;
 
 import lombok.extern.slf4j.Slf4j;
+
 import stirling.software.SPDF.model.PDFText;
+import stirling.software.SPDF.model.api.security.ManualRedactPdfRequest;
 import stirling.software.SPDF.model.api.security.RedactPdfRequest;
+import stirling.software.SPDF.model.api.security.RedactionArea;
 import stirling.software.SPDF.pdf.TextFinder;
 import stirling.software.SPDF.service.CustomPDDocumentFactory;
+import stirling.software.SPDF.utils.GeneralUtils;
 import stirling.software.SPDF.utils.PdfUtils;
 import stirling.software.SPDF.utils.WebResponseUtils;
+import stirling.software.SPDF.utils.propertyeditor.StringToArrayListPropertyEditor;
 
 @RestController
 @RequestMapping("/api/v1/security")
@@ -41,6 +51,107 @@ public RedactController(CustomPDDocumentFactory pdfDocumentFactory) {
         this.pdfDocumentFactory = pdfDocumentFactory;
     }
 
+    @InitBinder
+    public void initBinder(WebDataBinder binder) {
+        binder.registerCustomEditor(List.class, "redactions", new StringToArrayListPropertyEditor());
+    }
+
+    @PostMapping(value = "/redact", consumes = "multipart/form-data")
+    @Operation(summary = "Redacts areas and pages in a PDF document", description = "This operation takes an input PDF file with a list of areas, page number(s)/range(s)/function(s) to redact. Input:PDF, Output:PDF, Type:SISO")
+    public ResponseEntity<byte[]> redactPDF(@ModelAttribute ManualRedactPdfRequest request) throws IOException {
+        MultipartFile file = request.getFileInput();
+        List<RedactionArea> redactionAreas = request.getRedactions();
+
+        PDDocument document = pdfDocumentFactory.load(file);
+
+        PDPageTree allPages = document.getDocumentCatalog().getPages();
+
+        redactPages(request, document, allPages);
+        redactAreas(redactionAreas, document, allPages);
+
+        if (request.isConvertPDFToImage()) {
+            PDDocument convertedPdf = PdfUtils.convertPdfToPdfImage(document);
+            document.close();
+            document = convertedPdf;
+        }
+
+        ByteArrayOutputStream baos = new ByteArrayOutputStream();
+        document.save(baos);
+        document.close();
+
+        byte[] pdfContent = baos.toByteArray();
+        return WebResponseUtils.bytesToWebResponse(
+                pdfContent,
+                Filenames.toSimpleFileName(file.getOriginalFilename()).replaceFirst("[.][^.]+$", "")
+                        + "_redacted.pdf");
+    }
+
+    private void redactAreas(List<RedactionArea> redactionAreas, PDDocument document, PDPageTree allPages)
+            throws IOException {
+        Color redactColor = null;
+        for (RedactionArea redactionArea : redactionAreas) {
+            if (redactionArea.getPage() == null || redactionArea.getPage() <= 0
+                    || redactionArea.getHeight() == null || redactionArea.getHeight() <= 0.0D
+                    || redactionArea.getWidth() == null || redactionArea.getWidth() <= 0.0D)
+                continue;
+            PDPage page = allPages.get(redactionArea.getPage() - 1);
+
+            PDPageContentStream contentStream = new PDPageContentStream(
+                    document, page, PDPageContentStream.AppendMode.APPEND, true, true);
+            redactColor = decodeOrDefault(redactionArea.getColor(), Color.BLACK);
+            contentStream.setNonStrokingColor(redactColor);
+
+            float x = redactionArea.getX().floatValue();
+            float y = redactionArea.getY().floatValue();
+            float width = redactionArea.getWidth().floatValue();
+            float height = redactionArea.getHeight().floatValue();
+
+            PDRectangle box = page.getBBox();
+
+            contentStream.addRect(x, box.getHeight() - y - height, width, height);
+            contentStream.fill();
+            contentStream.close();
+        }
+    }
+
+    private void redactPages(ManualRedactPdfRequest request, PDDocument document, PDPageTree allPages)
+            throws IOException {
+        Color redactColor = decodeOrDefault(request.getPageRedactionColor(), Color.BLACK);
+        List<Integer> pageNumbers = getPageNumbers(request, allPages.getCount());
+        for (Integer pageNumber : pageNumbers) {
+            PDPage page = allPages.get(pageNumber);
+
+            PDPageContentStream contentStream = new PDPageContentStream(
+                    document, page, PDPageContentStream.AppendMode.APPEND, true, true);
+            contentStream.setNonStrokingColor(redactColor);
+
+            PDRectangle box = page.getBBox();
+
+            contentStream.addRect(0, 0, box.getWidth(), box.getHeight());
+            contentStream.fill();
+            contentStream.close();
+        }
+    }
+
+    private Color decodeOrDefault(String hex, Color defaultColor) {
+        Color color = null;
+        try {
+            color = Color.decode(hex);
+        } catch (Exception e) {
+            color = defaultColor;
+        }
+
+        return color;
+    }
+
+    private List<Integer> getPageNumbers(ManualRedactPdfRequest request, int pagesCount) {
+        String pageNumbersInput = request.getPageNumbers();
+        String[] parsedPageNumbers = pageNumbersInput != null ? pageNumbersInput.split(",") : new String[0];
+        List<Integer> pageNumbers = GeneralUtils.parsePageList(parsedPageNumbers, pagesCount, false);
+        Collections.sort(pageNumbers);
+        return pageNumbers;
+    }
+
     @PostMapping(value = "/auto-redact", consumes = "multipart/form-data")
     @Operation(
             summary = "Redacts listOfText in a PDF document",

src/main/java/stirling/software/SPDF/controller/web/SecurityWebController.java

@@ -18,6 +18,12 @@ public String autoRedactForm(Model model) {
         return "security/auto-redact";
     }
 
+    @GetMapping("/redact")
+    public String redactForm(Model model) {
+        model.addAttribute("currentPage", "redact");
+        return "security/redact";
+    }
+
     @GetMapping("/add-password")
     @Hidden
     public String addPasswordForm(Model model) {

src/main/java/stirling/software/SPDF/model/api/security/ManualRedactPdfRequest.java

@@ -0,0 +1,21 @@
+package stirling.software.SPDF.model.api.security;
+
+import java.util.List;
+
+import io.swagger.v3.oas.annotations.media.Schema;
+import lombok.Data;
+import lombok.EqualsAndHashCode;
+import stirling.software.SPDF.model.api.PDFWithPageNums;
+
+@Data
+@EqualsAndHashCode(callSuper = true)
+public class ManualRedactPdfRequest extends PDFWithPageNums {
+    @Schema(description = "A list of areas that should be redacted")
+    private List<RedactionArea> redactions;
+
+    @Schema(description = "Convert the redacted PDF to an image", defaultValue = "false")
+    private boolean convertPDFToImage;
+
+    @Schema(description = "The color used to fully redact certain pages")
+    private String pageRedactionColor;
+}

src/main/java/stirling/software/SPDF/model/api/security/RedactionArea.java

@@ -0,0 +1,23 @@
+package stirling.software.SPDF.model.api.security;
+
+import io.swagger.v3.oas.annotations.media.Schema;
+import lombok.Data;
+
+@Data
+public class RedactionArea {
+    @Schema(description = "The left edge point of the area to be redacted.")
+    private Double x;
+    @Schema(description = "The top edge point of the area to be redacted.")
+    private Double y;
+
+    @Schema(description = "The height of the area to be redacted.")
+    private Double height;
+    @Schema(description = "The width of the area to be redacted.")
+    private Double width;
+
+    @Schema(description = "The page on which the area should be redacted.")
+    private Integer page;
+
+    @Schema(description = "The color used to redact the specified area.")
+    private String color;
+}

src/main/java/stirling/software/SPDF/utils/propertyeditor/StringToArrayListPropertyEditor.java

@@ -0,0 +1,37 @@
+package stirling.software.SPDF.utils.propertyeditor;
+
+import java.beans.PropertyEditorSupport;
+import java.util.ArrayList;
+import java.util.List;
+
+import com.fasterxml.jackson.core.type.TypeReference;
+import com.fasterxml.jackson.databind.DeserializationFeature;
+import com.fasterxml.jackson.databind.ObjectMapper;
+
+import lombok.extern.slf4j.Slf4j;
+import stirling.software.SPDF.model.api.security.RedactionArea;
+
+@Slf4j
+public class StringToArrayListPropertyEditor extends PropertyEditorSupport {
+
+    private final ObjectMapper objectMapper = new ObjectMapper();
+
+    @Override
+    public void setAsText(String text) throws IllegalArgumentException {
+        if (text == null || text.trim().isEmpty()) {
+            setValue(new ArrayList<>());
+            return;
+        }
+        try {
+            objectMapper.configure(DeserializationFeature.ACCEPT_SINGLE_VALUE_AS_ARRAY, true);
+            TypeReference<ArrayList<RedactionArea>> typeRef = new TypeReference<ArrayList<RedactionArea>>() {
+            };
+            List<RedactionArea> list = objectMapper.readValue(text, typeRef);
+            setValue(list);
+        } catch (Exception e) {
+            log.error("Exception while converting {}", e);
+            throw new IllegalArgumentException(
+                    "Failed to convert java.lang.String to java.util.List");
+        }
+    }
+}

src/main/resources/messages_en_GB.properties

@@ -82,6 +82,7 @@ pages=Pages
 loading=Loading...
 addToDoc=Add to Document
 reset=Reset
+apply=Apply
 
 legal.privacy=Privacy Policy
 legal.terms=Terms and Conditions
@@ -474,6 +475,10 @@ home.autoRedact.title=Auto Redact
 home.autoRedact.desc=Auto Redacts(Blacks out) text in a PDF based on input text
 autoRedact.tags=Redact,Hide,black out,black,marker,hidden
 
+home.redact.title=Manual Redaction
+home.redact.desc=Redacts a PDF based on selected text, drawn shapes and/or selected page(s)
+redact.tags=Redact,Hide,black out,black,marker,hidden,manual
+
 home.tableExtraxt.title=PDF to CSV
 home.tableExtraxt.desc=Extracts Tables from a PDF converting it to CSV
 tableExtraxt.tags=CSV,Table Extraction,extract,convert
@@ -579,6 +584,15 @@ autoRedact.customPaddingLabel=Custom Extra Padding
 autoRedact.convertPDFToImageLabel=Convert PDF to PDF-Image (Used to remove text behind the box)
 autoRedact.submitButton=Submit
 
+#redact
+redact.title=Manual Redaction
+redact.header=Manual Redaction
+redact.submit=Redact
+redact.pageBasedRedaction=Page-based Redaction
+redact.convertPDFToImageLabel=Convert PDF to PDF-Image (Used to remove text behind the box)
+redact.pageRedactionNumbers.title=Pages
+redact.pageRedactionNumbers.placeholder=(e.g. 1,2,8 or 4,7,12-16 or 2n-1)
+redact.redactionColor.title=Redaction Color
 
 #showJS
 showJS.title=Show Javascript