Markdown ultra-lite

[Alex Tweedly]

I wanted to find a LC implementation of markdown, so I could easily use 
markdown files on a website, and have them be converted to html for 
output. I really wanted it completely in LC, so I could just include it 
without worrying about installing, or checking for, any other languages, 
scripts, etc.

But I could find one, so I thought I might try implementing it myself.

Didn't take long to find a current version in Python (easier for me to 
convert than C or Perl :-) - but heavens there's a lot of code there !! 
And indeed when you read a description (try 
http://spec.commonmark.org/0.28/ for example) you can easily see why 
it's complicated to do  full implementation.

And, sadly, many of those complications could have been avoided, but 
weren't due to historical misfortune :-)     So I have done an 
ultra-light version  ....

Restrictions and limitations:
-----------------------

Links are only 'inline', not 'reference'.
   - the whole inline link must be on a single input line.
   - no autolink, HTML Tags or codespans to confuse them :-)
   - no checks on brackets within the link text - simple cases should 
just work, others won't

Images are inline only, not referenced.
  - whole image spec must be on a single input line
  - doesn't do anything to deal with links within image description
          (mostly because I couldn't figure out what this part of the 
spec was trying to allow :-)

NB - because images are done before links, it is possible to handle the 
common case where the inline image is withing the linktext - i.e. you 
can click on the image to go to the other URL, e.g.

    Here is a [picture ![some alt text](x.png) used within 
link](/images/bigx.png)

Only one style of header is handled - using multiple '#'s rather than 
pseudo-underline.

Emphasis, and everything else, isn't done - and may or may not be 
depending on whether I decide I car about it.

-------------------------

Additions and oddities :

1. If the last char in the URL is a '*' then it becomes a "new tab" link 
(i.e. TARGET='_blank').

2. I added special handling for facebook links - see the code


In case anyone else finds it useful, here's the code .... I'll figure 
out Github etc. and put it up there some day - but for now it's short 
enough I'm just going to include it here. It uses a couple of utility 
functions which are included.

-- Alex.

local sSubstitutions

function markdownToHTML pMD
    put empty into sSubstitutions
    repeat for each line L in pMD
       -- first extract any image specs
       put L into LL
       put 1 into N
       repeat forever
          if decompose2(LL, "![", "](", p1, p2, p3) then
             if decompose1(p3, ")", p3a, p4) then
                put "<img src='" & p3a & "' alt='" & p2 & "'>" into 
sSubstitutions[N]
                put p1 & numtochar(N) & p4 into LL
                add 1 to N
                next repeat
             end if
          end if
          exit repeat
       end repeat

       -- and then do any links
       repeat forever
          if decompose2(LL, "[", "](", p1, p2, p3) then
             if decompose1(p3, ")", p3a, p4) then
                if char -1 of p3a = "*" then
                   put "' TARGET='_blank" into char -1 of p3a
                end if
                if p2 = "fb" then
                   put p1 & "<a href='http://facebook.com/" & p3a & "' 
class='icon fa-facebook' TARGET='_blank'><span class='label'>" & "</a>" 
& p4 into LL
                else
                   put p1 & "<a href=" & quote & p3a & quote & ">" & p2 
& "</a>" & p4 into LL
                end if

                next repeat
             end if
          end if
          exit repeat
       end repeat

       repeat with i = 1 to N
          replace numtochar(i) with sSubstitutions[i] in LL
       end repeat
       if LL is empty then put L into LL

       if LL is empty then
          put CR after tResult
       else
          put word 1 of LL into W

          replace "#" with empty in W
          if W is empty then
             put the number of chars in word 1 of LL into N
             put "<H" & N & ">" & word 2 to -1 of LL & "</H" & N & ">" 
into LL
          end if

          put LL &CR after tResult
       end if
       --      if NOT (tResult ends with CR) then exit repeat
       --      delete char -1 of tResult
    end repeat
    replace (numtochar(13) & numtochar(10)) with CR in tResult
    replace (CR & CR &CR) with (CR & "<p><br><p>" & CR) in tResult
    replace (CR & CR) with (CR & "<p>") in tResult
    return tResult
end markdownToHTML

function decompose1 pIn, pSep1, @p1, @p2
    -- if the input string (pIn) contains the delimiter string (pSep1) then
    --    return TRUE and set the result variables to the parts 'before' 
and 'after' the delimiter
    -- else
    --    return FALSE, and leave p1, p2 unchanged
   put offset(pSep1, pIn) into t1
   if t1 > 0 then
     put char 1 to t1-1 of pIn into p1
     put char (t1+the number of chars in pSep1) to -1 of pIn into p2
     return TRUE
   end if
   return FALSE
end decompose1

function decompose2 pIn, pSep1, pSep2, @p1, @p2, @p3
    -- if the input string (pIn) contains the (non-overlapping) 
delimiter strings (pSep1, pSep2) then
    --    return TRUE and set the result variables to the parts 
'before', 'between' and 'after' the delimiters
    -- else
    --    return FALSE, and leave p1, p2, p3 unchanged
    put offset(pSep1, pIn) into t1
    if t1 > 0 then
       put char t1+the number of chars in pSep1 to -1 of pIn into tIn
       put offset(pSep2, tIn) into t2
       if t2 > 0 then
          put char 1 to t1-1 of pIn into p1
          put char 1 to t2-1 of tIn into p2
          put char t2+the number of chars in pSep2 to -1 of tIn into p3
          return TRUE
       end if
    end if
    return FALSE
end decompose2