Re: [Caml-list] ANN: CamlPDF 1.7

[John Whitington <john@coherentgraphics.co.uk>]

Hi,

Armaël Guéneau wrote:
> Hi,
>
> Le 15/08/2013 13:21, John Whitington a écrit :
>> The first new release of the CamlPDF library for a while is here:
>>
>> http://www.github.com/johnwhitington/camlpdf
>>
>> (Or, shortly, via OPAM.)
> Thanks!
>
> I have been playing with CamlPDF a bit, trying to do text extraction.
> I'm a total novice about the PDF format, so i might be doing it wrong,
> but I was wondering if there were facilities, in CamlPDF, to handle
> diacritics and ligatures.
>
> For example, when reading the PDF operators for "Université", I get
>
> Pdfops_TJ (Pdf.Array [Pdf.String "Universit"; Pdf.String "\019";
> Pdf.Real 486.; Pdf.String "e"])

So 0o019 looks like a floating acute in that encoding, followed by a 
kern of 486/1000 of a point to shift leftward, followed by an 'e'. So, 
this is an accented character built by composition of glyphs.

> For "efficient", with "ffi" being ligated, I get
>
> Pdfops_TJ (Pdf.Array [Pdf.String "e\014cient"])

In the font in use here, character 0o014 appears to be a single glyph 
for the ffi ligature.

> How can I convert these back, especially the ligature? I tried to use the
> conversion functions of Pdftext, like codepoints_of_text followed by
> utf8_of_codepoints, but that didn't seem to work. It's highly possible
> that I'm also doing it wrong here.

Drop me a note off-list with the example file and I'll take a look.

Text extraction is almost always possible with modern PDFs. With older 
PDFs, text extraction sometimes isn't possible. You can test this by 
trying to copy & paste the text in Adobe Reader -- if it comes out 
correct, CamlPDF will be able to extract the text too -- if not, it won't.

With Thanks,

-- 
John Whitington
Director, Coherent Graphics Ltd
http://www.coherentpdf.com/