i have a pdf that has text rotated 90 degrees, but trying to search for matches within that text region fails. using adobe's acroread works. Reproducible: Always Steps to Reproduce: 1. download http://distfiles.gentoo.org/distfiles/pms-4_p20110118.pdf 2. load it in okular 3. search for "ECLASSDIR" Actual Results: okular reports no matches Expected Results: okular would find the listing of ECLASSDIR on page 47 (Section 12.1 under The Ebuild Environment -> Defined Variables) using KDE 4.8.2 with Okular 0.14.2 the original source can be found here: git://git.overlays.gentoo.org/proj/pms.git
Created attachment 75823 [details] pdf from libreoffice with horizontal and vertical text
Created attachment 75824 [details] source odt for pdf with horizontal and vertical text
Noticed the same with Okular 0.15.4 KDE 4.8.2. Tried with pdfs created with pdflatex and libreoffice writer (the latter attached). Funny enough if one selects the vertical text it is recognized as text, but the string is distorted. In the case of the libreoffice pdf instead of "Hidden" it is "n eddiH" which is also found when searching for it. On a side note opened the odt file (attached as well) from which I created the pdf with okular and the vertical text was not displayed.
FWIW, this is still a problem as of KDE Applications 17.11.80 and Poppler 0.61. See also Bug 338563, which concerns the selection/copying problem. However, note that simply fixing this bug won't be enough, we would also need to strip a varying amount of whitespace/linebreaks. This can be seen by creating a PDF in Inkscape that contains text fragments, each rotated in different multiples of 10°, and then pasting the copied text.
*** This bug has been marked as a duplicate of bug 407133 ***