mirror of
https://github.com/arsvendg/Stirling-PDF.git
synced 2026-07-14 10:34:06 +02:00
Feature/pdf ingestion jpdfium (#6525)
This commit is contained in:
EthanHealy01
+27
-5
@@ -1,11 +1,13 @@
|
||||
package stirling.software.SPDF.model.api.converters;
|
||||
|
||||
import org.springframework.core.io.Resource;
|
||||
import java.nio.charset.StandardCharsets;
|
||||
|
||||
import org.springframework.http.MediaType;
|
||||
import org.springframework.http.ResponseEntity;
|
||||
import org.springframework.web.bind.annotation.ModelAttribute;
|
||||
import org.springframework.web.multipart.MultipartFile;
|
||||
|
||||
import io.github.pixee.security.Filenames;
|
||||
import io.swagger.v3.oas.annotations.Operation;
|
||||
|
||||
import lombok.RequiredArgsConstructor;
|
||||
@@ -15,8 +17,11 @@ import stirling.software.common.annotations.AutoJobPostMapping;
|
||||
import stirling.software.common.annotations.api.ConvertApi;
|
||||
import stirling.software.common.enumeration.ResourceWeight;
|
||||
import stirling.software.common.model.api.PDFFile;
|
||||
import stirling.software.common.util.PDFToFile;
|
||||
import stirling.software.common.pdf.PdfMarkdownConverter;
|
||||
import stirling.software.common.util.TempFile;
|
||||
import stirling.software.common.util.TempFileManager;
|
||||
import stirling.software.common.util.WebResponseUtils;
|
||||
import stirling.software.jpdfium.PdfDocument;
|
||||
|
||||
@ConvertApi
|
||||
@RequiredArgsConstructor
|
||||
@@ -33,10 +38,27 @@ public class ConvertPDFToMarkdown {
|
||||
summary = "Convert PDF to Markdown",
|
||||
description =
|
||||
"This endpoint converts a PDF file to Markdown format. Input:PDF Output:Markdown Type:SISO")
|
||||
public ResponseEntity<Resource> processPdfToMarkdown(@ModelAttribute PDFFile file)
|
||||
public ResponseEntity<byte[]> processPdfToMarkdown(@ModelAttribute PDFFile file)
|
||||
throws Exception {
|
||||
MultipartFile inputFile = file.getFileInput();
|
||||
PDFToFile pdfToFile = new PDFToFile(tempFileManager);
|
||||
return pdfToFile.processPdfToMarkdown(inputFile);
|
||||
|
||||
String originalName = Filenames.toSimpleFileName(inputFile.getOriginalFilename());
|
||||
String baseName =
|
||||
originalName.contains(".")
|
||||
? originalName.substring(0, originalName.lastIndexOf('.'))
|
||||
: originalName;
|
||||
|
||||
String markdown;
|
||||
try (TempFile tempInput = new TempFile(tempFileManager, ".pdf")) {
|
||||
inputFile.transferTo(tempInput.getFile());
|
||||
try (PdfDocument doc = PdfDocument.open(tempInput.getPath())) {
|
||||
markdown = new PdfMarkdownConverter().convert(doc);
|
||||
}
|
||||
}
|
||||
|
||||
return WebResponseUtils.bytesToWebResponse(
|
||||
markdown.getBytes(StandardCharsets.UTF_8),
|
||||
baseName + ".md",
|
||||
MediaType.valueOf("text/markdown"));
|
||||
}
|
||||
}
|
||||
|
||||
+48
-46
@@ -1,16 +1,17 @@
|
||||
package stirling.software.SPDF.model.api.converters;
|
||||
|
||||
import static org.junit.jupiter.api.Assertions.assertEquals;
|
||||
import static org.mockito.ArgumentMatchers.any;
|
||||
import static org.mockito.Mockito.*;
|
||||
import static org.springframework.test.web.servlet.request.MockMvcRequestBuilders.multipart;
|
||||
import static org.springframework.test.web.servlet.result.MockMvcResultMatchers.*;
|
||||
|
||||
import java.io.File;
|
||||
import java.nio.charset.StandardCharsets;
|
||||
import java.nio.file.Path;
|
||||
|
||||
import org.junit.jupiter.api.Test;
|
||||
import org.mockito.ArgumentCaptor;
|
||||
import org.mockito.MockedConstruction;
|
||||
import org.mockito.MockedStatic;
|
||||
import org.mockito.Mockito;
|
||||
import org.springframework.core.io.ByteArrayResource;
|
||||
import org.springframework.core.io.Resource;
|
||||
@@ -21,9 +22,10 @@ import org.springframework.test.web.servlet.MockMvc;
|
||||
import org.springframework.test.web.servlet.setup.MockMvcBuilders;
|
||||
import org.springframework.web.bind.annotation.ExceptionHandler;
|
||||
import org.springframework.web.bind.annotation.RestControllerAdvice;
|
||||
import org.springframework.web.multipart.MultipartFile;
|
||||
|
||||
import stirling.software.common.util.PDFToFile;
|
||||
import stirling.software.common.pdf.PdfMarkdownConverter;
|
||||
import stirling.software.common.util.TempFile;
|
||||
import stirling.software.jpdfium.PdfDocument;
|
||||
|
||||
class ConvertPDFToMarkdownTest {
|
||||
|
||||
@@ -47,68 +49,68 @@ class ConvertPDFToMarkdownTest {
|
||||
@Test
|
||||
void pdfToMarkdownReturnsMarkdownBytes() throws Exception {
|
||||
byte[] md = "# heading\n\ncontent\n".getBytes(StandardCharsets.UTF_8);
|
||||
String expectedMd = "# heading\n\ncontent\n";
|
||||
|
||||
try (MockedConstruction<PDFToFile> construction =
|
||||
Mockito.mockConstruction(
|
||||
PDFToFile.class,
|
||||
(mock, ctx) -> {
|
||||
when(mock.processPdfToMarkdown(any(MultipartFile.class)))
|
||||
.thenAnswer(
|
||||
inv ->
|
||||
ResponseEntity.ok()
|
||||
.header("Content-Type", "text/markdown")
|
||||
.body(new ByteArrayResource(md)));
|
||||
})) {
|
||||
File tmpFile = File.createTempFile("test", ".pdf");
|
||||
tmpFile.deleteOnExit();
|
||||
|
||||
MockMvc mvc = mockMvc();
|
||||
try (MockedConstruction<TempFile> tempMock =
|
||||
Mockito.mockConstruction(
|
||||
TempFile.class,
|
||||
(mock, ctx) -> {
|
||||
when(mock.getFile()).thenReturn(tmpFile);
|
||||
when(mock.getPath()).thenReturn(tmpFile.toPath());
|
||||
});
|
||||
MockedStatic<PdfDocument> docStatic = Mockito.mockStatic(PdfDocument.class);
|
||||
MockedConstruction<PdfMarkdownConverter> converterMock =
|
||||
Mockito.mockConstruction(
|
||||
PdfMarkdownConverter.class,
|
||||
(mock, ctx) -> when(mock.convert(any())).thenReturn(expectedMd))) {
|
||||
|
||||
PdfDocument mockDoc = Mockito.mock(PdfDocument.class);
|
||||
docStatic.when(() -> PdfDocument.open(any(Path.class))).thenReturn(mockDoc);
|
||||
|
||||
MockMultipartFile file =
|
||||
new MockMultipartFile(
|
||||
"fileInput", // must match the field name in PDFFile
|
||||
"input.pdf",
|
||||
"application/pdf",
|
||||
new byte[] {1, 2, 3});
|
||||
"fileInput", "input.pdf", "application/pdf", new byte[] {1, 2, 3});
|
||||
|
||||
// ResponseEntity<Resource> is written synchronously on the request thread,
|
||||
// so there is no async dispatch to wait for (unlike the old StreamingResponseBody
|
||||
// path).
|
||||
mvc.perform(multipart("/api/v1/convert/pdf/markdown").file(file))
|
||||
mockMvc()
|
||||
.perform(multipart("/api/v1/convert/pdf/markdown").file(file))
|
||||
.andExpect(status().isOk())
|
||||
.andExpect(header().string("Content-Type", "text/markdown"))
|
||||
.andExpect(content().bytes(md));
|
||||
|
||||
// Verify that exactly one instance was created
|
||||
assert construction.constructed().size() == 1;
|
||||
|
||||
// And that the uploaded file was passed to processPdfToMarkdown()
|
||||
PDFToFile created = construction.constructed().get(0);
|
||||
ArgumentCaptor<MultipartFile> captor = ArgumentCaptor.forClass(MultipartFile.class);
|
||||
verify(created, times(1)).processPdfToMarkdown(captor.capture());
|
||||
MultipartFile passed = captor.getValue();
|
||||
|
||||
// Minimal plausibility checks
|
||||
assertEquals("input.pdf", passed.getOriginalFilename());
|
||||
assertEquals("application/pdf", passed.getContentType());
|
||||
}
|
||||
}
|
||||
|
||||
@Test
|
||||
void pdfToMarkdownWhenServiceThrowsReturns500() throws Exception {
|
||||
try (MockedConstruction<PDFToFile> ignored =
|
||||
Mockito.mockConstruction(
|
||||
PDFToFile.class,
|
||||
(mock, ctx) -> {
|
||||
when(mock.processPdfToMarkdown(any(MultipartFile.class)))
|
||||
.thenThrow(new RuntimeException("boom"));
|
||||
})) {
|
||||
File tmpFile = File.createTempFile("test", ".pdf");
|
||||
tmpFile.deleteOnExit();
|
||||
|
||||
MockMvc mvc = mockMvc();
|
||||
try (MockedConstruction<TempFile> tempMock =
|
||||
Mockito.mockConstruction(
|
||||
TempFile.class,
|
||||
(mock, ctx) -> {
|
||||
when(mock.getFile()).thenReturn(tmpFile);
|
||||
when(mock.getPath()).thenReturn(tmpFile.toPath());
|
||||
});
|
||||
MockedStatic<PdfDocument> docStatic = Mockito.mockStatic(PdfDocument.class);
|
||||
MockedConstruction<PdfMarkdownConverter> converterMock =
|
||||
Mockito.mockConstruction(
|
||||
PdfMarkdownConverter.class,
|
||||
(mock, ctx) ->
|
||||
when(mock.convert(any()))
|
||||
.thenThrow(new RuntimeException("boom")))) {
|
||||
|
||||
PdfDocument mockDoc = Mockito.mock(PdfDocument.class);
|
||||
docStatic.when(() -> PdfDocument.open(any(Path.class))).thenReturn(mockDoc);
|
||||
|
||||
MockMultipartFile file =
|
||||
new MockMultipartFile(
|
||||
"fileInput", "x.pdf", "application/pdf", new byte[] {0x01});
|
||||
|
||||
mvc.perform(multipart("/api/v1/convert/pdf/markdown").file(file))
|
||||
mockMvc()
|
||||
.perform(multipart("/api/v1/convert/pdf/markdown").file(file))
|
||||
.andExpect(status().isInternalServerError());
|
||||
}
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user