Is it possible to Extract Text from the specific area exactly？

[barryy70424910]

I been trying use Adobe Acrobat SDK to extract text for many days ...

it is OK for me to get the whole text of the page , But those text basically is not ordered.

Normally content comes like top content , footer content , Main content

We don't get those text like we see those words orderly.

Another reason why I am asking is , we have used another 3rd party tools like PDFbox

With their tools , giving the specific area , it return text successfully . And unfortunately , this tools doesn't read pdf successfully.

And , adobe acrobat SDK read all pdf files well .

Now this is what I plainly to do 

Giving a specific area , and return the text . Just Like we read pdf files , we select it and copy it.

Firtst . Is it possible to do that ?

Second . I used pdfDoc.CreateTextSelect(pageNumber, pdfRect);

This function return text which is not I want when those texts are in form or image .

I was giving the smaller pdfRect to CreateTextSelect function ,  but it finally return its own BoundingRect like the bigger one.

And Also , function return texts like : DQM0 L VDDQDQ8VSSVDDDQ7VSSQ M VSSQDQ10DQ9DQ6DQ5VDDQ N VSSQDQ12DQ14DQ1DQ3VDDQ P ...

Correct me if I am wrong , Is it possible to do that ? or Am I using the wrong method?

[Test Screen Name]

Try JavaScript GetPageNthWord and GetPageNthWordQuads, though “order” is a slippery and uncertain concept for PDF text.

[barryy70424910]

Hi , thank you for replying .

this is what I get contexts by using AcroPDTextSelect

this , is by using GetPageNthWord and GetPageNthWordQuads

I still don't get those content in from split .

and even I get this contents , I can't get where these word were suppose to be.

so this is why I am wondering if there is a function , or a way to get contents in a specific area

[Test Screen Name]

No but you can check the quads to exclude text outside the area.

[barryy70424910]

Thanks for replying .

Yes , now I get the 8 point of the rectangle information.

And I use an Array to store the information and adjust these points. Thank you so much .

How do I use these adjusted points to send back to the function , and get the text that I want?

[Test Screen Name]

You cannot send the points to the function. Instead you will see all words but you can ignore the words which are outside your target area.

[barryy70424910]

now I have already process Image to get the circle sharp , where text is in it.

In the picture up there , like I want to get VDDQ

and I use rectangle to select the range , before the function goes , I set it all up.

But it return string like "VDDQDQ8VSSVDDDQ7VSSQ" , I don't get to know the text in advance

This is why Im asking if there is another way to get context in specific area .

Thank you still !

[Test Screen Name]

You must do what I said: get each word and check its quad. why Does this not work?

[barryy70424910]

ok , maybe it  wasn't clear.

https://drive.google.com/open?id=1T9_1hdp1e1DPU4VmLADdW36qGW-8k7qn

this is the pdf file.

And of course I did some implements by GetPageNthWord  and GetPageNthWordQuads with it

totally , It return (numWords 😃 24 on that page by GetPageNthWord .

I expected these word return like 1 2 3 ... 7 8 9

A DQ26 DQ24 ....

but the face isn't like that , as the output shows ,  it return VDDQDQ31DQ28DQ27DQ23DQ24DQ16… which is not proper for me.

I found that those coordinate combine as a rectangle which seems relative to the content.

And TextStripper in apache PDFBox tool is able to extract the contexts in circle shape with proper range.

I wanna know what  is wrong with it , why can't I implement  3rd party tools' function with Acrobat SDK?

[lrosenth]

The reason you have limited functionality is you are using the limited functionality APIs…

If you want more power and capabilities, use the Plugin APIs instead of the IAC APIs.

[Bernd Alheit]

On the third page the function getPageNumWords returns the value 129

So there are 129 words on this page.

[Thom Parker]

I get the same thing number of words as Bernd and the pin names in the ovals extract individually.

What version of Acrobat are you using?

[barryy70424910]

Really?  After license , I have installed Adobe Acrobat  Pro DC.

And I add Adobe + Adobe Acrobat 10.0 Type Library to references in my project

Here is my code​ and suppose I have input the right pageNumber

It return 24 words only .

[Bernd Alheit]

Execute this.getPageNumWords(2); in the Acrobat console and you will get 129.

[barryy70424910]

sorry guys , I really didn't get the same words like you guys did.

I did my project with C#

I'm really new to this area , I know there are many layers in pdf.

Am I really on the wrong direction just like lrosenth​said?

[Test Screen Name]

The PDF link is not public. So I have no idea what you expect to see.

[barryy70424910]

AS4C4M32S.pdf - Google Drive

sorry , one more try?

[Test Screen Name]

1. We cannot tell from fragments of code, what do you set pageNumber to?

2. Please run the one line in the console suggested above, what do you get?