runassudo / gfx2gfx-pdftext Goto Github PK

Images are not positioned properly in latest commit (2a84c79), while they were in upstream (c40abbf).

Offending swf: swf.zip

As a result of the manual positioning of the letters, contiguous blocks of text are not recognised as such by PDF readers. As a result, copy-pasting does not work well, and nor does in-text search.

This may be related to the issue that text is interpreted by programs like qpdfview and evince back-to-front, possibly related to the topdown setting of PDFlib.

Hi, I've succesfully compiled the gfx2gfx-pdftext v0.9.2 build 8d5a70b code (and the same is true for SWFTools v0.9.2) under Mac OS X 10.11.6 but I had to fix the jpeg.c file (as I described in the thread https://github.com/matthiaskramm/swftools/issues/37) given that I had some compiling errors.

In particular I've compiled gfx2gfx without errors but if I use it with the command:
gfx2gfx test.swf -o test.pdf
I get the following errors (swf file zipped and attached test.swf.zip):
Error: ID 142 unknown
Error: ID 145 unknown
Error: ID 148 unknown
Error: ID 151 unknown
Error: ID 154 unknown
Error: ID 157 unknown
Error: ID 160 unknown
Error: ID 163 unknown
Error: ID 166 unknown
Error: ID 169 unknown
Error: ID 172 unknown
Error: ID 175 unknown
Error: ID 178 unknown
Error: ID 181 unknown
Error: ID 184 unknown
Error: ID 187 unknown
Error: ID 190 unknown
Error: ID 193 unknown
Error: ID 196 unknown
Error: ID 199 unknown
Error: ID 202 unknown
Error: ID 205 unknown
Error: ID 208 unknown
Error: ID 211 unknown
Error: ID 214 unknown
Error: ID 217 unknown
Error: ID 220 unknown
Error: ID 223 unknown
Error: ID 226 unknown
Error: ID 229 unknown
Error: ID 232 unknown
Error: ID 235 unknown
Error: ID 238 unknown
Error: ID 241 unknown
Error: ID 244 unknown
Error: ID 247 unknown
Error: ID 250 unknown
Error: ID 253 unknown
Error: ID 256 unknown
Error: ID 259 unknown
Error: ID 262 unknown
Error: ID 265 unknown
Error: ID 268 unknown
Error: ID 271 unknown
Error: ID 274 unknown
Error: ID 277 unknown
Error: ID 280 unknown
Error: ID 283 unknown
Error: ID 286 unknown
Error: ID 289 unknown
Error: ID 292 unknown
Error: ID 295 unknown
Error: ID 298 unknown

Do you have any idea what this is? How can I fix it?

I would like to be able to exclude specific strings from the generated pdf.

Is there any file I could modify to do so?
I can't find the responsible method in the text.c or the gfx2gfx.c file.

Thanks for your help.

As reported by @sonst-was.

e.g. page_3.zip

In pdf.js, the letters are huge and reversed. Adobe Reader (acroread) on Linux displays some glyphs correctly, but raises a bunch of errors and fails to display most.

• Unusual glyph positioning (kerning??) not respected

e.g. U+239f, which is erroneously placed at the left of the bounding box rather than the right; italic f, which is placed too close to the next letter.

• Some glyphs don't render

e.g. U+239b, U+23aa

gfx2gfx-pdf2text - part of swftools 0.9.2 (build )

missing build is 8d5a70b

Compiled under ubuntu 17.04, typically flawless conversion of swf and gau files. This particular page (originally 694.gau) converted with no errors under -r0 option, but caused acrobat 8.1 to crash right here when combining pages 654-714. Garbled graphics in the output, but all 693 prior pages fine. Single page conversion with -r0 and -r300 attached.

swf2pdf.zip

Using gfx2gfx compiled on Windows with mingw. Unknown if this is related. Issue also occurs with upstream non-pdftext version.

Example of offending SWF is on file with @RunasSudo

thanks for your fancy project, i'm not sure whether you can read Chinese or not(i got problems while using this tool converting some swf in Chinese)，actually, i came across 2 problems:

version: gfx2gfx-pdf2text - part of swftools 0.9.2 (build 8d5a70b)

• some character convertion is not right
  demo file address
• fatal exception: [1106] PDF_encoding_set_char: Integer parameter 'slot' has bad value 256
  demo file address

the 2nd problem is easy to fix(i don't know whether this fix is right or not), i just add 2 lines code after

        if(gt7bits>=128)
            gt7bits=0;

but there still exists the 1st problem, some character is missing after convertion

This seems to be a compatibility issue between viewers. In some viewers, including Adobe Reader, MuPDF, xpdf and Firefox's pdf.js, text does not display. qpdf, evince, okular and Google Drive view the PDFs correctly.

A temporary fix appears to be to post-process the PDF with Ghostscript or Poppler:

gs -o output.pdf -sDEVICE=pdfwrite -dPDFSETTINGS=/prepress input.pdf

or

pdftocairo -pdf input.pdf output.pdf

This produces a PDF which appears to be readable in all the above applications.

1.swf.zip
however above is test file.
after convert into pdf result is different 😄

output

After compiling some pdf's, I ran into an error with text copying. Before the invisible text -fix, text was still copiable and would paste properly. Currently, trying to highlight text from the generated pdf's doesn't work properly.

I only highlighted the word 'dream' but as the picture shows, more words are highlighted. Pasting this results in a mess of incoherency:

Additional details:
I used parameter -r 300
Tested in Adobe Reader and Google Drive