mirror of
https://github.com/arsvendg/Stirling-PDF.git
synced 2026-07-16 11:23:10 +02:00
smart redaction (#6195)
Co-authored-by: James Brunton <[email protected]>
This commit is contained in:
EthanHealy01
co-authored by
James Brunton
James Brunton
parent
7f3ca7ea70
commit
35a712a278
+201
-1
@@ -8,13 +8,18 @@ import java.util.List;
|
||||
import java.util.Map;
|
||||
import java.util.Optional;
|
||||
import java.util.Set;
|
||||
import java.util.regex.Matcher;
|
||||
import java.util.regex.Pattern;
|
||||
import java.util.stream.Collectors;
|
||||
|
||||
import org.apache.commons.csv.CSVFormat;
|
||||
import org.apache.commons.csv.CSVPrinter;
|
||||
import org.apache.commons.csv.QuoteMode;
|
||||
import org.apache.pdfbox.pdmodel.PDDocument;
|
||||
import org.apache.pdfbox.pdmodel.PDPage;
|
||||
import org.apache.pdfbox.pdmodel.common.PDRectangle;
|
||||
import org.apache.pdfbox.text.PDFTextStripper;
|
||||
import org.apache.pdfbox.text.TextPosition;
|
||||
import org.springframework.stereotype.Service;
|
||||
|
||||
import com.fasterxml.jackson.annotation.JsonIgnore;
|
||||
@@ -24,6 +29,7 @@ import lombok.Data;
|
||||
import lombok.RequiredArgsConstructor;
|
||||
import lombok.extern.slf4j.Slf4j;
|
||||
|
||||
import stirling.software.SPDF.pdf.parser.PageImageLocator;
|
||||
import stirling.software.SPDF.pdf.parser.PdfIngester;
|
||||
import stirling.software.SPDF.pdf.parser.PdfModels.ParsedPage;
|
||||
import stirling.software.SPDF.pdf.parser.PdfModels.RawLine;
|
||||
@@ -32,6 +38,7 @@ import stirling.software.SPDF.pdf.parser.PdfModels.TextFragment;
|
||||
import stirling.software.SPDF.pdf.parser.TabulaTableParser;
|
||||
import stirling.software.common.util.ExceptionUtils;
|
||||
import stirling.software.common.util.PdfUtils;
|
||||
import stirling.software.common.util.RegexPatternUtils;
|
||||
import stirling.software.proprietary.model.api.ai.AiPdfContentType;
|
||||
import stirling.software.proprietary.model.api.ai.AiWorkflowFileRequest;
|
||||
import stirling.software.proprietary.model.api.ai.AiWorkflowTextSelection;
|
||||
@@ -122,6 +129,7 @@ public class PdfContentExtractor {
|
||||
}
|
||||
csvStrings.add(sw.toString());
|
||||
}
|
||||
|
||||
return csvStrings;
|
||||
}
|
||||
|
||||
@@ -295,7 +303,6 @@ public class PdfContentExtractor {
|
||||
private List<AiWorkflowTextSelection> extractPageText(
|
||||
PDDocument document, List<Integer> selectedPages, int maxCharacters)
|
||||
throws IOException {
|
||||
PDFTextStripper textStripper = new PDFTextStripper();
|
||||
List<AiWorkflowTextSelection> pages = new ArrayList<>();
|
||||
int remainingCharacters = maxCharacters;
|
||||
|
||||
@@ -304,10 +311,29 @@ public class PdfContentExtractor {
|
||||
break;
|
||||
}
|
||||
|
||||
PDFTextStripper textStripper = new PDFTextStripper();
|
||||
textStripper.setSortByPosition(true);
|
||||
textStripper.setStartPage(pageNumber);
|
||||
textStripper.setEndPage(pageNumber);
|
||||
|
||||
String pageText = textStripper.getText(document).trim();
|
||||
|
||||
// Prepend page dimensions so the AI agent can reason about absolute coordinates.
|
||||
PDPage page = document.getPage(pageNumber - 1);
|
||||
PDRectangle bbox = page.getBBox();
|
||||
String dimensionHeader =
|
||||
String.format(
|
||||
"--- Page dimensions: %.0fx%.0f pts"
|
||||
+ " (PDF user-space: origin bottom-left, Y up) ---\n",
|
||||
bbox.getWidth(), bbox.getHeight());
|
||||
pageText = dimensionHeader + pageText;
|
||||
|
||||
// Append image metadata so the AI agent can reason about images spatially.
|
||||
String imageAnnotation = buildImageAnnotation(document, pageNumber - 1);
|
||||
if (!imageAnnotation.isEmpty()) {
|
||||
pageText = pageText + imageAnnotation;
|
||||
}
|
||||
|
||||
if (pageText.isBlank()) {
|
||||
continue;
|
||||
}
|
||||
@@ -327,6 +353,56 @@ public class PdfContentExtractor {
|
||||
return pages;
|
||||
}
|
||||
|
||||
/**
|
||||
* Builds a human-readable description of all images on a page to append to page text. Uses PDF
|
||||
* user-space coordinates (origin bottom-left, Y up) so the AI can reference exact bounding
|
||||
* boxes when requesting image redaction.
|
||||
*/
|
||||
private String buildImageAnnotation(PDDocument document, int pageIndex) {
|
||||
try {
|
||||
List<ImageBlock> images = extractImagePositions(document, pageIndex);
|
||||
if (images.isEmpty()) {
|
||||
return "";
|
||||
}
|
||||
PDPage page = document.getPage(pageIndex);
|
||||
PDRectangle bbox = page.getBBox();
|
||||
float pageWidth = bbox.getWidth();
|
||||
float pageHeight = bbox.getHeight();
|
||||
|
||||
StringBuilder sb = new StringBuilder("\n\n--- Images on this page ---");
|
||||
for (int i = 0; i < images.size(); i++) {
|
||||
ImageBlock img = images.get(i);
|
||||
String position = spatialLabel(img, pageWidth, pageHeight);
|
||||
float w = img.x2() - img.x1();
|
||||
float h = img.y2() - img.y1();
|
||||
sb.append(
|
||||
String.format(
|
||||
"\nImage %d: position=%s, size=%.0fx%.0f pts,"
|
||||
+ " bounds=(x1=%.0f, y1=%.0f, x2=%.0f, y2=%.0f)",
|
||||
i + 1, position, w, h, img.x1(), img.y1(), img.x2(), img.y2()));
|
||||
}
|
||||
return sb.toString();
|
||||
} catch (Exception e) {
|
||||
log.debug(
|
||||
"Failed to extract image positions for page {}: {}", pageIndex, e.getMessage());
|
||||
return "";
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* Returns a human-readable spatial label (e.g. "top-left", "center") for an image based on its
|
||||
* centre relative to the page dimensions. Coordinates are in PDF user-space (Y up).
|
||||
*/
|
||||
private static String spatialLabel(ImageBlock img, float pageWidth, float pageHeight) {
|
||||
float cx = (img.x1() + img.x2()) / 2f;
|
||||
float cy = (img.y1() + img.y2()) / 2f;
|
||||
|
||||
String horiz = cx < pageWidth / 3f ? "left" : cx < 2 * pageWidth / 3f ? "center" : "right";
|
||||
// PDF Y increases upward, so higher Y = higher on the page = "top"
|
||||
String vert = cy > 2 * pageHeight / 3f ? "top" : cy > pageHeight / 3f ? "middle" : "bottom";
|
||||
return vert + "-" + horiz;
|
||||
}
|
||||
|
||||
private ExtractedFileText buildExtractedFileText(
|
||||
String fileName, List<AiWorkflowTextSelection> pages) {
|
||||
ExtractedFileText fileText = new ExtractedFileText();
|
||||
@@ -347,6 +423,130 @@ public class PdfContentExtractor {
|
||||
return text.substring(0, end);
|
||||
}
|
||||
|
||||
// -----------------------------------------------------------------------
|
||||
// Text position finding
|
||||
// -----------------------------------------------------------------------
|
||||
|
||||
/**
|
||||
* A located text match inside a PDF: 0-based page index and bounding box in PDFBox coordinates
|
||||
* (origin bottom-left).
|
||||
*/
|
||||
public record TextBlock(int pageIndex, float x1, float y1, float x2, float y2) {}
|
||||
|
||||
/**
|
||||
* An image found on a PDF page: 0-based page index and bounding box in PDF user-space
|
||||
* coordinates (origin bottom-left, Y increases upward).
|
||||
*/
|
||||
public record ImageBlock(int pageIndex, float x1, float y1, float x2, float y2) {}
|
||||
|
||||
/**
|
||||
* Extract the bounding boxes of all raster/vector images on the given (0-based) page.
|
||||
*
|
||||
* @param document the open PDF
|
||||
* @param pageIndex 0-based page index
|
||||
* @return list of located images in document order
|
||||
*/
|
||||
public List<ImageBlock> extractImagePositions(PDDocument document, int pageIndex)
|
||||
throws IOException {
|
||||
PDPage page = document.getPage(pageIndex);
|
||||
PageImageLocator locator = new PageImageLocator(page, pageIndex);
|
||||
locator.processPage(page);
|
||||
return locator.getImageBoxes().stream()
|
||||
.map(b -> new ImageBlock(b.pageIndex(), b.x1(), b.y1(), b.x2(), b.y2()))
|
||||
.toList();
|
||||
}
|
||||
|
||||
/**
|
||||
* Find all occurrences of {@code pattern} in {@code document} and return their bounding boxes.
|
||||
*
|
||||
* @param document the open PDF
|
||||
* @param pattern the search string or regex
|
||||
* @param useRegex {@code true} to treat {@code pattern} as a regular expression
|
||||
* @return list of located matches, in page order
|
||||
*/
|
||||
public List<TextBlock> findTextPositions(PDDocument document, String pattern, boolean useRegex)
|
||||
throws IOException {
|
||||
LocalTextFinder finder = new LocalTextFinder(pattern, useRegex);
|
||||
finder.getText(document);
|
||||
return finder.found;
|
||||
}
|
||||
|
||||
private static final class LocalTextFinder extends PDFTextStripper {
|
||||
|
||||
private final String searchTerm;
|
||||
private final boolean useRegex;
|
||||
final List<TextBlock> found = new ArrayList<>();
|
||||
|
||||
private final List<TextPosition> pagePositions = new ArrayList<>();
|
||||
private final StringBuilder pageText = new StringBuilder();
|
||||
|
||||
LocalTextFinder(String searchTerm, boolean useRegex) throws IOException {
|
||||
this.searchTerm = searchTerm;
|
||||
this.useRegex = useRegex;
|
||||
setWordSeparator(" ");
|
||||
setLineSeparator("\n");
|
||||
}
|
||||
|
||||
@Override
|
||||
protected void startPage(PDPage page) throws IOException {
|
||||
super.startPage(page);
|
||||
pagePositions.clear();
|
||||
pageText.setLength(0);
|
||||
}
|
||||
|
||||
@Override
|
||||
protected void writeString(String text, List<TextPosition> positions) {
|
||||
pageText.append(text);
|
||||
pagePositions.addAll(positions);
|
||||
}
|
||||
|
||||
@Override
|
||||
protected void writeWordSeparator() {
|
||||
pageText.append(getWordSeparator());
|
||||
pagePositions.add(null);
|
||||
}
|
||||
|
||||
@Override
|
||||
protected void writeLineSeparator() {
|
||||
pageText.append(getLineSeparator());
|
||||
pagePositions.add(null);
|
||||
}
|
||||
|
||||
@Override
|
||||
protected void endPage(PDPage page) throws IOException {
|
||||
String text = pageText.toString();
|
||||
if (!text.isEmpty() && searchTerm != null && !searchTerm.isBlank()) {
|
||||
String term = searchTerm.trim();
|
||||
String regex = useRegex ? term : "\\Q" + term + "\\E";
|
||||
Pattern pat = RegexPatternUtils.getInstance().createSearchPattern(regex, true);
|
||||
Matcher matcher = pat.matcher(text);
|
||||
while (matcher.find()) {
|
||||
float minX = Float.MAX_VALUE;
|
||||
float minY = Float.MAX_VALUE;
|
||||
float maxX = -Float.MAX_VALUE;
|
||||
float maxY = -Float.MAX_VALUE;
|
||||
boolean hit = false;
|
||||
for (int i = matcher.start(); i < matcher.end(); i++) {
|
||||
if (i < pagePositions.size()) {
|
||||
TextPosition tp = pagePositions.get(i);
|
||||
if (tp != null) {
|
||||
hit = true;
|
||||
minX = Math.min(minX, tp.getX());
|
||||
maxX = Math.max(maxX, tp.getX() + tp.getWidth());
|
||||
minY = Math.min(minY, tp.getY() - tp.getHeight());
|
||||
maxY = Math.max(maxY, tp.getY());
|
||||
}
|
||||
}
|
||||
}
|
||||
if (hit) {
|
||||
found.add(new TextBlock(getCurrentPageNo() - 1, minX, minY, maxX, maxY));
|
||||
}
|
||||
}
|
||||
}
|
||||
super.endPage(page);
|
||||
}
|
||||
}
|
||||
|
||||
// --- Types shared with AiWorkflowService (package-private) ---
|
||||
|
||||
interface PdfContentResult {
|
||||
|
||||
Reference in New Issue
Block a user