bug#5118: DocView: copy/search text

bug#5118: DocView: copy/search text

[Juri Linkov]

Package: emacs,doc-view
Severity: wishlist

Please add two features that would be easy to implement:

1. Copying the text of the current page to the kill ring.

PDF viewers allow selecting a rectangular area to copy it to the
clipboard.  If this is not possible in DocView, then at least copying
the text of the whole page would be a satisfactory compensation.
Info and Dired bind similar commands to `w'.

2. Search currently is useless in DocView since it doesn't highlight
the current match.

What would be more useful is to implement Occur-like UI.  For instance,
typing `doc-view-occur' would display an *Occur* buffer like:

Page  1:   text text match text text
Page  1:   text text text match text
Page  1:   text text text text match
Page  2:   text match text text text
Page 42:   match text text text text

that allows to navigate to the corresponding page.

-- 
Juri Linkov
http://www.jurta.org/emacs/

bug#5118: DocView: copy/search text

[Kevin Rodgers]

Juri Linkov wrote:
> Package: emacs,doc-view
> Severity: wishlist
> 
> Please add two features that would be easy to implement:
> 
> 1. Copying the text of the current page to the kill ring.

I don't know anything about DocView, but shouldn't that be possible
in any major mode via C-x C-p (mark-page) then M-w (kill-ring-save)?

> PDF viewers allow selecting a rectangular area to copy it to the
> clipboard.  If this is not possible in DocView, then at least copying
> the text of the whole page would be a satisfactory compensation.
> Info and Dired bind similar commands to `w'.

-- 
Kevin Rodgers
Denver, Colorado, USA

bug#5118: DocView: copy/search text

[Juri Linkov]

>> Please add two features that would be easy to implement:
>>
>> 1. Copying the text of the current page to the kill ring.
>
> I don't know anything about DocView, but shouldn't that be possible
> in any major mode via C-x C-p (mark-page) then M-w (kill-ring-save)?

`C-x C-p' and `M-w' will copy binary data, whereas `pdftotext' converts
it to plain text in a hidden buffer that can be displayed with `C-c C-t'
where you can do `C-x C-p' and `M-w'.  But the problem is that it's
difficult to find the current page in the text representation to be able
to copy it.

-- 
Juri Linkov
http://www.jurta.org/emacs/

bug#5118: DocView: copy/search text

[Stefan Monnier]

>>> Please add two features that would be easy to implement:
>>> 1. Copying the text of the current page to the kill ring.
>> I don't know anything about DocView, but shouldn't that be possible
>> in any major mode via C-x C-p (mark-page) then M-w (kill-ring-save)?

> `C-x C-p' and `M-w' will copy binary data, whereas `pdftotext' converts
> it to plain text in a hidden buffer that can be displayed with `C-c C-t'
> where you can do `C-x C-p' and `M-w'.  But the problem is that it's
> difficult to find the current page in the text representation to be able
> to copy it.

doc-view-search-internal seems to just count ^Ls, so assuming
doc-view-search-internal works (which it appears to do), it shouldn't be
that hard.


        Stefan

bug#5118: DocView: copy/search text

[Juri Linkov]

>>>> Please add two features that would be easy to implement:
>>>> 1. Copying the text of the current page to the kill ring.
>>> I don't know anything about DocView, but shouldn't that be possible
>>> in any major mode via C-x C-p (mark-page) then M-w (kill-ring-save)?
>
>> `C-x C-p' and `M-w' will copy binary data, whereas `pdftotext' converts
>> it to plain text in a hidden buffer that can be displayed with `C-c C-t'
>> where you can do `C-x C-p' and `M-w'.  But the problem is that it's
>> difficult to find the current page in the text representation to be able
>> to copy it.
>
> doc-view-search-internal seems to just count ^Ls, so assuming
> doc-view-search-internal works (which it appears to do), it shouldn't be
> that hard.

I meant it's difficult for the user to find the current page (that is
displayed in the image buffer) to find the corresponding page in the
text buffer (with the output from `pdftotext').

I don't mind having a command that would switch from the image buffer
to the text buffer and narrow it to the current page.  So it would be easy
to copy any text from the textual representation of the current page.

-- 
Juri Linkov
http://www.jurta.org/emacs/