"Sebastien Vauban" <sva-n...@mygooglest.com> writes:
> Hi Eric,
>
> Eric Schulte wrote:
>> aditya siram <aditya.si...@gmail.com> writes:
>>> What's the rationale for having padlines by default in tangled source? It
>>> generates wrong programs for languages where whitespace is significant
>>> (Haskell) and, for me, doesn't noticeably improve the look of the tangled
>>> file in cases where it isn't.
>>
>> It is possible to change the value of default header arguments on a
>> per-language basis because e.g., while (:padlines "yes") may make sense
>> for sh, it probably doesn't for Haskell.
>
> Could it be possible that ":padline yes" does not insert a blank line in front
> of the very first block, only *between* all blocks?
>
> Best regards,
> Seb
I just pushed up a commit which implements this behavior. See the
attached file for an example.
#+Title: Examples with the new padline behavior
#+headers: :tangle pad-yes-with-shebang.sed
#+headers: :shebang "#!/bin/sed -f"
#+headers: :padline "yes"
#+begin_src sed
1 {N;s/\n//1}
#+end_src
#+headers: :tangle pad-no-with-shebang.sed
#+headers: :shebang "#!/bin/sed -f"
#+headers: :padline "no"
#+begin_src sed
1 {N;s/\n//1}
#+end_src
#+headers: :tangle pad-yes-without-shebang.sed
#+headers: :padline "yes"
#+begin_src sed
1 {N;s/\n//1}
#+end_src
#+headers: :tangle pad-no-without-shebang.sed
#+headers: :padline "no"
#+begin_src sed
1 {N;s/\n//1}
#+end_src
1. Tangle the above four blocks with =C-c C-v t=.
2. Execute the following code block to view the contents of the
resulting sed files.
#+begin_src sh :results scalar
head pad*sed
#+end_src
#+RESULTS:
: ==> padline-example.sed <==
: #!/bin/sed -f
:
: 1 {N;s/\n//1}
:
: ==> pad-no-without-shebang.sed <==
: 1 {N;s/\n//1}
:
: ==> pad-no-with-shebang.sed <==
: #!/bin/sed -f
: 1 {N;s/\n//1}
:
: ==> pad-yes-without-shebang.sed <==
: 1 {N;s/\n//1}
:
: ==> pad-yes-with-shebang.sed <==
: #!/bin/sed -f
:
: 1 {N;s/\n//1}
--
Eric Schulte
https://cs.unm.edu/~eschulte
PGP: 0x614CA05D
