Re: [O] make new links show as figures?

[Cook, Malcolm]

>>
 >> So it turns out that this is a REALLY simple patch (two characters):
 >>
 >> +++ b/lisp/org.el
 >> @@ -18853,7 +18853,7 @@ BEG and END default to the buffer boundaries."
 >> (widen)
 >> (setq beg (or beg (point-min)) end (or end (point-max)))
 >> (goto-char beg)
 >> -   (let ((re (concat
 >> "\\[\\[\\(\\(file:\\)\\|\\([./~]\\)\\)\\([^]\n]+?"
 >> +   (let ((re (concat
 >> "\\[.*\\[\\(\\(file:\\)\\|\\([./~]\\)\\)\\([^]\n]+?"
 >> (substring (org-image-file-name-regexp) 0 -2)
 >> "\\)\\]" (if include-linked "" "\\]")))
 >> (case-fold-search t)
 >>
 >> This seems to work in all cases I can think of (see test file below).
 >> Let me know if this works for you and i will apply to master when i
 >> get home and have git push access.
 >>
 >> I tried it on some files with images as links, and it works
 >> great. Thanks a lot!
[Cook, Malcolm] 

Hi - I'm trying to follow this thread but having a challenge.  Can you please 
summarize/repeat the original intended effect of this change....  

~Malcolm


 >
 >So, I've done some testing of the export functions (html and latex
 >only) and found the following behaviors with this patch:
 >
 >- HTML
 >- image in description only displayed if the protocol (file:) is
 >included.
 >- LaTeX
 >- same, required file: protocol in description
 >- additionally, pdflatex \includgraphics barfs on any path w/ a
 >=./= path prefix -- this is true for the current implmenation as
 >well.
 >
 >Changing the regexp for in-buffer image expansion to mimic the export
 >behavior is difficult at best. I think the inconsistency is
 >acceptable, and I will documentment the behavior in the
 >`org-display-inline-images' docstring.
 >
 >The latex breakage on paths beginning with =./= should perhaps be
 >addressed in the latex exporter, but this is orthogonal to the
 >current issue.
 >
 >If there is no objection, I will push a fix tonight or tomorrow.
 >
 >rick
 >