[jira] [Comment Edited] (PDFBOX-4738) Pages do not have objects after splitting

["Maruan Sahyoun (Jira)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/PDFBOX-4738?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17013154#comment-17013154 ] 

Maruan Sahyoun edited comment on PDFBOX-4738 at 1/10/20 7:21 PM:
-----------------------------------------------------------------

Understood that you would liek to get that information - but why is the number of objects of particular importance? What is the use case behind that? As an example some information of PDF might be stored 'inline' in another object or as a separate object which is referenced. At the end it's still the same information.

In addition object cunt per page will not tell you anything about the page being more complex than another. And objects you are referring too are a means of the PDF serialization but not of the in memory representation.


was (Author: msahyoun):
Understood that you would liek to get that information - but why is the number of objects of particular importance? What is the use case behind that? As an example some information of PDF might be stored 'inline' in another object or as a separate object which is referenced. At the end it's still the same information.

> Pages do not have objects after splitting
> -----------------------------------------
>
>                 Key: PDFBOX-4738
>                 URL: https://issues.apache.org/jira/browse/PDFBOX-4738
>             Project: PDFBox
>          Issue Type: Bug
>            Reporter: Yuguang Huang
>            Priority: Major
>
>  
> Hi PDFBOX community, we want to get objs count on pages instead of the whole document. 
> Our way to do it is splitting the whole document into multiple documents containing only one page. But it seems then it returns documents/pages without objects, meaning getDocument().getObjects() returns an empty list. 
> But if we save each page into bytes then load them into PDDocument, we are able to get the object counts. 
>  
> Is there any way we can get the page objs count without involving so much IO? Thanks! 
>  
> Output of the below code with a three-page PDF document:
>  
> Page objects count from splitted pages:
> page [1] num of objs [0]
> page [2] num of objs [0]
> page [3] num of objs [0]
> Page objects count from pages generated from bytes:
> page [1] num of objs [20]
> page [2] num of objs [51]
> page [3] num of objs [20]
>  
> {code:java}
> private static void printNumObjects(String pdfFilename) throws IOException {
>  byte[] fileContent = Files.readAllBytes((new File(pdfFilename)).toPath());
>  PDDocument document = PDDocument.load(fileContent);
>  List<PDDocument> pages = new Splitter().split(document);
>  List<byte[]> pageBytes = pages.stream().map(page -> {
>  try (ByteArrayOutputStream baos = new ByteArrayOutputStream()) {
>  page.save(baos);
>  page.close();
>  return baos.toByteArray();
>  } catch (IOException e) {
>  LOG.error("Failed to get bytes from page.", e);
>  return new byte[0];
>  }
>  }).collect(Collectors.toList());
>  System.out.println("Page objects count from splitted pages:");
>  IntStream.range(0, pages.size()).forEach(i -> System.out.println(String.format("page [%d] num of objs [%d]", i + 1, pages.get(i).getDocument().getObjects().size())));
>  System.out.println("Page objects count from pages generated from bytes:");
>  IntStream.range(0, pageBytes.size()).forEach(i -> {
>  try {
>  System.out.println(String.format("page [%d] num of objs [%d]", i + 1, PDDocument.load(pageBytes.get(i)).getDocument().getObjects().size()));
>  } catch (IOException e) {
>  LOG.error("Failed to load page.", e);
>  }
>  });
> }{code}
>  
>  



--
This message was sent by Atlassian Jira
(v8.3.4#803005)

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@pdfbox.apache.org
For additional commands, e-mail: dev-help@pdfbox.apache.org