CTAN Sumbission -- cmap package: create searchable PDF files

[Robin Fairbairns]

Vladimir Volovich writes:

> I've uploaded the file cmap.tar.gz into the incoming directory of
> ftp.tex.ac.uk. It is a new (version 1.0a) "cmap" package.
> The license is LPPL.
>
> The cmap package is to be used with pdf(e)latex and provides CMap
> files for various LaTeX font encodings (currently, T1, T2A, T2B, T2C,
> T5; this list will grow in the future) and a latex package which adds
> the /ToUnicode entries into the font dictionaries of fonts used in the
> document on the fly, making the resulting PDF files "searchable" and
> "cut-and-pasteable" in Acrobat Reader and other PDF previewers.
>
> Thus, when the document uses Type 1 fonts like cm-super or other fonts
> in standard font encodings, teh PDF files will contain the additional
> information in the form of CMap entries which will define the mappings
> from LaTeX font encodings to Unicode, making functions like search in
> PDF and cut-and-paste from PDF to other applications work properly.
>
> Please install it in the CTAN:macros/latex/contrib/supported/cmap
> directory.
>
> Contributions and/or requests for other font encodings support, as
> well as user feedback/bug reports are welcome.
>
> P.S. currently, this package is mirrored to
> macros/latex/contrib/supported/t2/etc/cmap; i think that a more
> generic location in macros/latex/contrib/supported/cmap is better,
> since the package is of general use. i'll remove it from the T2
> package once it will be installed on
> macros/latex/contrib/supported/cmap

i've installed the package as requested: thanks for the upload.

Robin Fairbairns

For the CTAN team

[Bjørn-Helge Mevik]

Robin.Fa...@cl.cam.ac.uk (Robin Fairbairns) writes:

> Vladimir Volovich writes:
>
>> The cmap package is to be used with pdf(e)latex and provides CMap
>> files for various LaTeX font encodings (currently, T1, T2A, T2B, T2C,
>> T5; this list will grow in the future) and a latex package which adds
>> the /ToUnicode entries into the font dictionaries of fonts used in the
>> document on the fly, making the resulting PDF files "searchable" and
>> "cut-and-pasteable" in Acrobat Reader and other PDF previewers.

I use pdflatex and cm-super, and the generated pdf files are
searchable and selectable wether I use cmap or not. (I've tried with
Acrobat Reader on Linux and Windows, as well as xpdf on Linux.) I'm
curious: what is it that doesn't work without cmap?

--
Bjørn-Helge Mevik

[Heiko Oberdiek]

use...@mevik.net (Bjørn-Helge Mevik) wrote:

For example, some ligatures are not detected without cmap:

\documentclass{article}
%\usepackage{cmap}
\usepackage[T1]{fontenc}
\begin{document}
fl
ffl
fi
ffi
\end{document}

I see a difference with AR5/Linux: ffl and ffi are not detected.
But AR4/Linux here works fine without cmap, too. (?)

Yours sincerely
Heiko <ober...@uni-freiburg.de>

[Vladimir Volovich]

"BM" == Bjørn-Helge Mevik writes:

BM> I use pdflatex and cm-super, and the generated pdf files are
BM> searchable and selectable wether I use cmap or not. (I've tried
BM> with Acrobat Reader on Linux and Windows, as well as xpdf on
BM> Linux.) I'm curious: what is it that doesn't work without cmap?

there were reports that if one uses the germandbls.alt in
cm-super-t1.enc to get the original CM-like shape of the sharp s, then
this letter is not searchable in acrobat reader; with the cmap
package, it becomes searchable; if some T1-encoded fonts based on
type1 fonts used in the document are using non-standard glyph names,
the cmap package will make them searchable too; also, cyrillic,
vietnamese, greek, etc fonts require the cmap to be searchable.

Best,
v.