On Fri, May 20, 2016 at 12:11 PM, Eric S Fraga <e.fr...@ucl.ac.uk> wrote: > On Friday, 20 May 2016 at 17:06, Ken Mankoff wrote: > > [...] > >> Thanks for clarifying your workflow. That seems to work for a small >> example. But what would you do if you had a manuscript with 100 code >> blocks in it? And in general you're exporting it a lot without >> changing anything in code, just editing the non-code text. But >> occasionally you want to modify a code block (and its results) too? > > So, if I understand correctly, you have org-export-babel-evaluate set to > nil? If I set that, I no longer get asked whether I want anything > evaluated but the (existing) results do get exported. I just tried it > to confirm.
Using this block, #+begin_src R :exports results print("hello") #+end_src I get the following (with *no* pre-existing #+results block manually created with C-c C-c): | *o-e-b-e* | *o-c-b-e* | *buffer* | *doc* | |-----------+-----------+-----------+------------------| | nil | t | nothing | =print("hello")= | | t | t | nothing | =hello= | | t | nil | nothing | =hello= | | nil | nil | nothing | =print("hello")= | |-----------+-----------+-----------+------------------| Variables org-export-babel-evaluate and org-confirm-babel-evaluate abbreviated above. I find two things bizarre: - these options (well, just o-e-b-e it seems) would change what's exported? - why without C-c C-c which initiates the #+results block, exporting which causes babel to execute code doesn't insert the results block. Is this intended or for a reason? I've not really looked into these variables and just use :eval yes/no to toggle things. I find it easy enough to create all my blocks with :eval no, and do a quick change to =yes= as I'm iteratively exporting. Finally, I can do a replace-string to catch any missed :eval yes's before a final export. > >> In the above scenario, I used to edit the code, and =C-c C-c= it >> manually to get the updates into the results. When I exported the >> document (C-e l l), it would export in ~1 second (even a >100 page >> document) and not prompt me about executing code blocks. > > Yes, I do this as well in such a case. > > -- > : Eric S Fraga (0xFFFCF67D), Emacs 25.0.92.1, Org release_8.3.4-775-g3308a5