[Bug] Extra Spaces in Copied Text

[lbdyck]

When I attempt to copy some text (code example) I am seeing extra spaces when I paste the info. See the attached zip file for the sample page from the pdf and a text file showing the results of pasting.

[quant]

This is in fact happening in quite many pdf's. I had similar problems with another pdf software. The developer's reply was sth like ... sometimes there is no space between words (in the pdf file structure), so the program has to guess as to how many spaces should be there in the extracted text, and whether there should be any space at all ... not easy

[Podhorny]

It looks correct, in PDF there always 2 spaces, try to copy also text above - it has only one space between words:

Code: Select all

You must update the PROFILE EXEC for any user ID that will be running
GOMMAIN. The default user ID is OPMGRM1. The PROFILE EXEC for these user
IDs should include the following statements:
/*  Sample  lines  to  include  in  OPMGRM1  PROFILE  EXEC  */
’CP  SET  RUN  ON’
’ACCESS  194  D’

[Ivan - Tracker Software]

To be honest, into your PDF sample text contains two spaces between words -- try to use Text selection tool and you will see this.

[quant]

To be honest, into your PDF sample text contains two spaces between words -- try to use Text selection tool and you will see this.

That is exactly the point. How do you know that there are two spaces? Visually it looks like, but that line could just be made to "fill the line" (as opposed to be left aligned), so there is still one space between the words, it's just a bit bigger.

See the example I provide. The comparison of extracted text from Adobe and pdf-xchange:

... no comparison, because this editor removes double spaces, haha, this is funny. OK, extract this sentence:

"Recently, the characterization of q-optimal equivalent martingale mea-
sures in market models with jumps has been studied in several papers."

You will see double space extracted in pdf-xchange, but there are no double spaces in the original text, neither in Adobe, it's just that the line is filled.

[Ivan - Tracker Software]

There are no spaces at all into the text you send. This text specified into the PDF into the following way:

Code: Select all

[(Recen)27(tly)84(,)-495(the)-464(c)27(haracterization)-462(of)]TJ
/F2 10.91 Tf
164.08 0 TD
[(q)]TJ
/F7 10.91 Tf
5.26 0 TD
[(-optimal)-462(equiv)54(alen)28(t)-462(martingale)-464(mea-)]TJ

please note numbers like -464, -462, etc. - they means distance between pieces of text. Viewer analyzes this and convert to spaces. Number of spaces depends of font's metric.

[quant]

please note numbers like -464, -462, etc. - they means distance between pieces of text. Viewer analyzes this and convert to spaces. Number of spaces depends of font's metric.

OK, but you see yourself that

"Viewer analyzes this and convert to spaces. Number of spaces depends of font's metric."

is probably not the best way to go about this. Clearly, the original author didn't put 2 spaces between words, it's just that the line was filled. I would think that "intelligent viewer/text extractor" would take this into account, and not merely apply the hardcoded formula

number of spaces = distance / (font metric)

[lbdyck]

Just comparing Viewer to Acrobat Reader the Reader does correctly handle the spaces (*but fails with the new line).

[John - Tracker Supp]

Hi,

We are looking at adding an optional feature that will handle both ways - so as not create an issue if we handle the space duplication that will impact other situations - which it very easily could.

HTH