Re: Auto-indent for various languages=pretty print

["chris glur" <crglur gmail com>]

==================
OK, I'll asume that my previous [the 1st one] post was
delivered.   This is what I've discover since then:-

It must be great to be able to build small programs from structures,
instead of chars and have the code auto-pretty-printed.
And the syntax colouring helps to manage the code.

mc provides facilities [currently mostly for C, but hopefully
can be extended to other languages] whereby: F11 shows a
select-structure-menu like:
 ......
 5       IF () ELSEIF ()
 6       CASE (int) OF
 7       for ()
 8       while ()
...etc...
Q1 -where is this [headed as] " User menu" located ?
Searching /home & /usr for "CASE (int)" didn't find it for me!

Apparently the pretty-printing via F19 is driven by
file ~/.mc/cedit/edit.indent.rc which has:--
#! /bin/sh
# *** External Formatter (Indenter) for GNU Midnight Commander.
# arguments:
#    $1 - Name of the file being edited
#    $2 - Name of the file to be processed

exec >/dev/null

case `echo $1 |sed 's/^.*\.//'` in
    c|h)
    # ftp://ftp.gnu.org/pub/gnu/indent/
    # Please add options to your ~/.indent.pro, not here.
        indent "$2"
    ;;...etc. for c++, java, html...etc....

I.e. the script's section says:
"if the file type [via the name] is *.c or *.h
 then use the "indent" utility with the 2nd argument.
I.e. use the "indent" utility on the C-source-file.

'man indent' tells about the indent utility.

Q2 - In the comment of the script: I think I know what
 "the file being edited" is, but what's "the file to
 be processed" ?

Perhaps the many args for indent are passed via a file ?

Q3  - how exactly does:
"`echo $1 |sed 's/^.*\.//'` in
    c|h)", translate to " if the file is *.c OR *.h" ?

Q4 - does anybody use these F11 & F19 mc facilities ?
Q5 - what's the differnce between mcedit & cooledit and which does
mc use ?

Thanks for any feedback,

== Chris Glur.



The selected structure is placed at the cursor.
And F19 pretty-prints the pre-marked block.


On 11/22/07, chris glur <crglur gmail com> wrote:
> As a long time user of NC/mc I'm keen to use the intended facilities
>  of cooledit re. External Formatter.
> The ability to enter structure-templates instead of single chars
> is the right way to go; and the pretty printer facility is essential.
>
> I'm not a C programmer, and I'd like to make versions for Oberon
> [similar to pascal] & soar [an AI system], but it seems that my
> ver. 4.6.0 can't quiet get C right yet ?
>
> I don't want to hack-around trying to decode what
> file: /usr/share/mc/syntax/c.syntax *EXACTLY* means eg. at line:
>   keyword whole linestart \{\s\t\}\[\s\t\]#*\n brightmagenta
>
> I can speculate & confirm eg. by replacing the "{,}" with
> "BEGIN,END", that it concerns the 'Block-brackets'.
> And probably it's important for the pretty-printing.
>
> Looking at the incomplete efforts at the many <language>.syntax files
> shows a waste of effort?
>
> Let's use the wikipedia method: contributors should document their
>  efforts so that other can build-on-top-of-it.
>
> Q1- is the latest version able to give a usable 'structure selector'
>  & pretty-printer, for at least C ?
> It's acceptable if the next-contained-structure must be specifically
>  located - provided this usage-knowledge is documented.
> We don't want each contributor to re-invent the wheel.
>
> Q2- is there any documentation [other than src code] to describe
>  how to write the syntax-file for its 'interpreter' ?
>
> Thank for any feedback,
>
> == Chris Glur.