Mirrors/Stirling-PDF
Mirrors/Stirling-PDF
1
0
Fork 0
mirror of https://github.com/Frooodle/Stirling-PDF.git synced 2025-09-08 17:51:20 +02:00
Code Issues Packages Projects Releases Wiki Activity

refactor: clean up code formatting and improve readability in RedactionService

Browse Source 
Signed-off-by: Balázs Szücs <bszucs1209@gmail.com>
This commit is contained in:
Balázs Szücs 2025-09-02 21:13:43 +02:00
parent 338b77de99
commit f00952b856
1 changed files with 29 additions and 41 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

44
app/core/src/main/java/stirling/software/SPDF/service/RedactionService.java
View File

@ -42,6 +42,7 @@ import org.apache.pdfbox.pdmodel.PDResources;
import org.apache.pdfbox.pdmodel.common.PDRectangle;
import org.apache.pdfbox.pdmodel.common.PDStream;
import org.apache.pdfbox.pdmodel.font.PDFont;
import org.apache.pdfbox.pdmodel.font.PDType0Font;
import org.apache.pdfbox.pdmodel.graphics.PDXObject;
import org.apache.pdfbox.pdmodel.graphics.form.PDFormXObject;
import org.apache.pdfbox.pdmodel.graphics.pattern.PDTilingPattern;
@ -2672,16 +2673,14 @@ public class RedactionService {
merger.setDestinationFileName(finalOutputFile.toString());
for (int pageNum = 0; pageNum < pageCount; pageNum++) {
BufferedImage image = pdfRenderer.renderImageWithDPI(pageNum, 600);
File imagePath =
new File(tempImagesDir, "page_" + pageNum + ".png");
File imagePath = new File(tempImagesDir, "page_" + pageNum + ".png");
ImageIO.write(image, "png", imagePath);
List<String> command =
new ArrayList<>(
Arrays.asList(
"tesseract",
imagePath.toString(),
new File(tempOutputDir, "page_" + pageNum)
.toString(),
new File(tempOutputDir, "page_" + pageNum).toString(),
"-l",
buildLanguageOption(request),
"--dpi",
@ -2696,8 +2695,7 @@ public class RedactionService {
throw new IOException(
"Tesseract restoration failed with return code: " + result.getRc());
}
java.io.File pageOutputPath =
new java.io.File(tempOutputDir, "page_" + pageNum + ".pdf");
File pageOutputPath = new File(tempOutputDir, "page_" + pageNum + ".pdf");
merger.addSource(pageOutputPath);
}
merger.mergeDocuments(null);
@ -2736,13 +2734,8 @@ public class RedactionService {
pageText.append(pdfText.getText()).append(" ");
}
}
String extractedText = pageText.toString().trim();
for (String word : targetWords) {
if (extractedText.toLowerCase().contains(word.toLowerCase())) {}
}
} catch (Exception e) {
log.debug("Failed to get direct text from page", e);
}
}
@ -2776,8 +2769,6 @@ public class RedactionService {
}
String completeText = allText.toString().trim();
if (!completeText.isEmpty() && hasProblematicChars) {}
}
List<MatchRange> matches;
@ -2939,8 +2930,7 @@ public class RedactionService {
NORMALIZE_WHITESPACE
}
public interface SemanticScrubber {
}
public interface SemanticScrubber {}
private static class GlyphCoverageProbe {
private final PDFont font;
@ -2956,7 +2946,7 @@ public class RedactionService {
if (font == null) return coverage;
try {
if (font instanceof org.apache.pdfbox.pdmodel.font.PDType0Font) {
if (font instanceof PDType0Font) {
for (int cid = 0; cid < 65536; cid++) {
try {
String unicode = font.toUnicode(cid);
@ -2964,12 +2954,12 @@ public class RedactionService {
coverage.add(cid);
}
} catch (Exception e) {
log.debug("Failed to get unicode for cid {}", cid, e);
}
}
}
} catch (Exception e) {
log.debug("Failed to get glyph coverage for font {}", font, e);
}
return coverage;
}
@ -2995,17 +2985,17 @@ public class RedactionService {
String charStr = new String(Character.toChars(codePoint));
return font.getStringWidth(charStr) / FONT_SCALE_FACTOR * fontSize;
} catch (Exception e) {
log.debug("Failed to get width for codepoint {}", codePoint, e);
}
}
return switch (strategy) {
case EMBED_WIDTH -> getEmbeddedProgramWidth(codePoint, fontSize);
case EMBED_WIDTH -> getEmbeddedProgramWidth(fontSize);
case AVERAGE_WIDTH -> getAverageFontWidth(fontSize);
case LEGACY_SUM -> getLegacySumFallback(codePoint, fontSize);
};
}
private float getEmbeddedProgramWidth(int codePoint, float fontSize) {
private float getEmbeddedProgramWidth(float fontSize) {
try {
if (font.getFontDescriptor() != null) {
float avgWidth = font.getFontDescriptor().getAverageWidth();
@ -3033,7 +3023,7 @@ public class RedactionService {
validChars++;
}
} catch (Exception e) {
log.debug("Failed to get width for char {}", ch, e);
}
}
@ -3091,7 +3081,7 @@ public class RedactionService {
}
}
private void scrubStructureElement(COSDictionary element, Set<ScrubOption> options) {
private static void scrubStructureElement(COSDictionary element, Set<ScrubOption> options) {
if (element == null) return;
if (options.contains(ScrubOption.REMOVE_ACTUALTEXT)) {
@ -3120,12 +3110,11 @@ public class RedactionService {
}
}
private void normalizeWhitespaceInElement(COSDictionary element) {
private static void normalizeWhitespaceInElement(COSDictionary element) {
for (COSName key : List.of(COSName.ACTUAL_TEXT, COSName.ALT, COSName.TU)) {
COSBase value = element.getDictionaryObject(key);
if (value instanceof COSString cosString) {
String text = cosString.getString();
if (text != null) {
String normalized = text.replaceAll("\\s+", " ").trim();
if (normalized.length() > 256) {
normalized = normalized.substring(0, 256);
@ -3134,7 +3123,6 @@ public class RedactionService {
}
}
}
}
private void scrubAnnotations(PDDocument document, Set<ScrubOption> options) {
try {
@ -3157,7 +3145,7 @@ public class RedactionService {
}
}
} catch (Exception e) {
log.debug("Failed to scrub annotations", e);
}
}
}