fix: switch to pdftohtml for pdf to html conversions (#998)

* fix: switch to pdftohtml for pdf to html conversions

* build: include poppler-utils in dockerfile for pdftohtml
This commit is contained in:
Eric
2024-03-29 17:02:33 -04:00
committed by GitHub
parent 27bbf7a513
commit dfb8c64f5a
37 changed files with 101 additions and 58 deletions

View File

@@ -244,6 +244,6 @@ public class EndpointConfiguration {
}
}
}
private static final String REMOVE_BLANKS = "remove-blanks";
}

View File

@@ -291,6 +291,6 @@ public class UserController {
}
return ResponseEntity.ok(apiKey);
}
private static final String LOGIN_MESSAGETYPE_CREDSUPDATED = "/login?messageType=credsUpdated";
}

View File

@@ -29,18 +29,6 @@ import stirling.software.SPDF.utils.WebResponseUtils;
@Tag(name = "Convert", description = "Convert APIs")
public class ConvertPDFToOffice {
@PostMapping(consumes = "multipart/form-data", value = "/pdf/html")
@Operation(
summary = "Convert PDF to HTML",
description =
"This endpoint converts a PDF file to HTML format. Input:PDF Output:HTML Type:SISO")
public ResponseEntity<byte[]> processPdfToHTML(@ModelAttribute PDFFile request)
throws Exception {
MultipartFile inputFile = request.getFileInput();
PDFToFile pdfToFile = new PDFToFile();
return pdfToFile.processPdfToOfficeFormat(inputFile, "html", "writer_pdf_import");
}
@PostMapping(consumes = "multipart/form-data", value = "/pdf/presentation")
@Operation(
summary = "Convert PDF to Presentation format",

View File

@@ -6,8 +6,6 @@ import java.nio.file.Path;
import java.util.ArrayList;
import java.util.List;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.beans.factory.annotation.Qualifier;
import org.springframework.http.ResponseEntity;
import org.springframework.web.bind.annotation.ModelAttribute;
import org.springframework.web.bind.annotation.PostMapping;

View File

@@ -219,6 +219,6 @@ public class ExtractImageScansController {
});
}
}
private static final String REPLACEFIRST = "[.][^.]+$";
}

View File

@@ -26,7 +26,6 @@ import org.springframework.stereotype.Service;
import com.fasterxml.jackson.databind.ObjectMapper;
import stirling.software.SPDF.model.ApplicationProperties;
import stirling.software.SPDF.model.PipelineConfig;
import stirling.software.SPDF.model.PipelineOperation;

View File

@@ -3,8 +3,6 @@ package stirling.software.SPDF.repository;
import java.util.Optional;
import org.springframework.data.jpa.repository.JpaRepository;
import org.springframework.data.jpa.repository.Query;
import org.springframework.data.repository.query.Param;
import stirling.software.SPDF.model.User;

View File

@@ -25,6 +25,71 @@ import io.github.pixee.security.Filenames;
import stirling.software.SPDF.utils.ProcessExecutor.ProcessExecutorResult;
public class PDFToFile {
public ResponseEntity<byte[]> processPdfToHtml(MultipartFile inputFile)
throws IOException, InterruptedException {
if (!"application/pdf".equals(inputFile.getContentType())) {
return new ResponseEntity<>(HttpStatus.BAD_REQUEST);
}
// Get the original PDF file name without the extension
String originalPdfFileName = Filenames.toSimpleFileName(inputFile.getOriginalFilename());
String pdfBaseName = originalPdfFileName.substring(0, originalPdfFileName.lastIndexOf('.'));
Path tempInputFile = null;
Path tempOutputDir = null;
byte[] fileBytes;
String fileName = "temp.file";
try {
// Save the uploaded file to a temporary location
tempInputFile = Files.createTempFile("input_", ".pdf");
Files.copy(
inputFile.getInputStream(), tempInputFile, StandardCopyOption.REPLACE_EXISTING);
// Prepare the output directory
tempOutputDir = Files.createTempDirectory("output_");
// Run the pdftohtml command with complex output
List<String> command =
new ArrayList<>(
Arrays.asList(
"pdftohtml", "-c", tempInputFile.toString(), pdfBaseName));
ProcessExecutorResult returnCode =
ProcessExecutor.getInstance(ProcessExecutor.Processes.PDFTOHTML)
.runCommandWithOutputHandling(command, tempOutputDir.toFile());
// Get output files
List<File> outputFiles = Arrays.asList(tempOutputDir.toFile().listFiles());
// Return output files in a ZIP archive
fileName = pdfBaseName + "ToHtml.zip";
ByteArrayOutputStream byteArrayOutputStream = new ByteArrayOutputStream();
ZipOutputStream zipOutputStream = new ZipOutputStream(byteArrayOutputStream);
for (File outputFile : outputFiles) {
ZipEntry entry = new ZipEntry(outputFile.getName());
zipOutputStream.putNextEntry(entry);
FileInputStream fis = new FileInputStream(outputFile);
IOUtils.copy(fis, zipOutputStream);
fis.close();
zipOutputStream.closeEntry();
}
zipOutputStream.close();
fileBytes = byteArrayOutputStream.toByteArray();
} finally {
// Clean up the temporary files
if (tempInputFile != null) Files.delete(tempInputFile);
if (tempOutputDir != null) FileUtils.deleteDirectory(tempOutputDir.toFile());
}
return WebResponseUtils.bytesToWebResponse(
fileBytes, fileName, MediaType.APPLICATION_OCTET_STREAM);
}
public ResponseEntity<byte[]> processPdfToOfficeFormat(
MultipartFile inputFile, String outputFormat, String libreOfficeFilter)
throws IOException, InterruptedException {
@@ -39,17 +104,7 @@ public class PDFToFile {
// Validate output format
List<String> allowedFormats =
Arrays.asList(
"doc",
"docx",
"odt",
"ppt",
"pptx",
"odp",
"rtf",
"html",
"xml",
"txt:Text");
Arrays.asList("doc", "docx", "odt", "ppt", "pptx", "odp", "rtf", "xml", "txt:Text");
if (!allowedFormats.contains(outputFormat)) {
return new ResponseEntity<>(HttpStatus.BAD_REQUEST);
}

View File

@@ -24,6 +24,7 @@ public class ProcessExecutor {
public enum Processes {
LIBRE_OFFICE,
PDFTOHTML,
OCR_MY_PDF,
PYTHON_OPENCV,
GHOSTSCRIPT,
@@ -45,6 +46,7 @@ public class ProcessExecutor {
int semaphoreLimit =
switch (key) {
case LIBRE_OFFICE -> 1;
case PDFTOHTML -> 1;
case OCR_MY_PDF -> 2;
case PYTHON_OPENCV -> 8;
case GHOSTSCRIPT -> 16;
@@ -56,6 +58,7 @@ public class ProcessExecutor {
long timeoutMinutes =
switch (key) {
case LIBRE_OFFICE -> 30;
case PDFTOHTML -> 5;
case OCR_MY_PDF -> 30;
case PYTHON_OPENCV -> 30;
case GHOSTSCRIPT -> 5;