update metagrammar
[sbp.git] / tests / tibdoc.g
index 442e203..b3b107e 100644 (file)
@@ -1,4 +1,3 @@
-x  ="x"
 // interactions between !=> and &~ mean that I need to rethink the chartage
 // indentation styling...
 // literal blocks [[need to ignore bracing]] double-colon style?
@@ -20,10 +19,7 @@ x  ="x"
 //   [1] http://...
 //
 
-// consider ++bold++ and **italic**?
-// \br
 // nonbreaking text?
-// ellipsis detection (...)
 // degree: 15^o
 // Arrows: <- -> => <= <->
 
@@ -40,7 +36,6 @@ x  ="x"
 // #include
 
 // simple macros (#define) (\define)
-// today's date
 
 // table representation
 //  
@@ -53,82 +48,97 @@ x  ="x"
 //     [c] ...
 
 // FIXME: these have to go at the top so they have their dropAll bit set before PreSequence.build...
-ws      !  = w**
-w       !  =  [\r\n\ ]
+w          = " " | "\n" | "\r"
+ws         =  "()":: w**
+//           |  "()":: w** "#" (~[\n])* "\n" ws
 nw         = ~[\r\n\ ]
 
 //////////////////////////////////////////////////////////////////////////////
 
-s                   = Doc => top
+s                   = Doc
 
-Doc                 = {Header} Body                   /ws => doc
-Header              = "header" { kv */ ws }           /ws => header
-Body                = Section*/ws                         => body
-Section             = { SectionHeader Paragraph* /ws => section }
+Doc                 = head:Header ws! body:Body
+Header              = { "\\header" { KeyVal */ ws } /ws }
+Body                = { Section } */ws
+Section             = SectionHeader ws! Paragraph*
 SectionHeader       = "==" SectionHeaderBody "=="
 SectionHeaderBody   =  "=" SectionHeaderBody "="
-                    >      ws text ws
-
-sp    !  = " "**
-blank    = sp "\n" sp "\n" ws
-
-kv           = word "=" text /ws => kv1
-wp !  = w++
-num !  = [0-9]++
-Paragraph    = { "\"\"" ws  text }  => "blockquote"
-             > { "---" "-"*      }  => "hr"
-             > { text }             => "p"
-
-onums     !  = nums (". "|") ")
-any       !  = ~[]*
-
-uli          = "* "         (ws text &~ any (oli|uli)) => "li"
-oli          = ("# "|onums) (ws text &~ any (oli|uli)) => "li"
-
-text         = Item => text
-Itemx        = ws Item
-             | ()
-Item         = blockquote
-             > { uli+/ws => "ul" }       Itemx  => []
-             | { oli+/ws => "ol" }       Itemx  => []
-             > pre                       Itemx  => []
-             > structured                Itemx  => []
-             > structuredx               Itemx  => []
-             > styled                    Itemx  => []
-             > qtext                     Itemx  => []
-             > (alphanum++ => stringify) Itemx  => []
-             > symbol                    Itemx  => []
-             > (sym++ => stringify)      Itemx  => []
-             > Paragraph                 Itemx  => []
-
-blockquote   = "\"\"" text "\"\""        => "blockquote"
-             | "\"\"" block              => "blockquote"
+                    >      ws! text ws!
+
+sp       = " "**
+blank    = sp! "\n" sp! "\n" ws!
+
+KeyVal       = key:bareword "=" val:text /ws
+wp           = w++
+num          = [0-9]++
+
+Paragraph   = { Blockquote:: "\"\" "    text }
+            > HR::         { "---" "-"*      }
+            >         { OL:: (text &~ (text! ws! {oli}+)) ws! {oli}+  }
+            > P::          { text          }
+
+onums        = nums (". "|") ")!
+any          = ~[]*
+
+uli          =  "* "         (ws! text &~ any (oli|uli)!)
+oli          = onums!        (ws! text &~ any (oli|uli)!)
+
+text         = Item
+
+Item*/ws     =
+               blockquote
+//             > ^"#" ws! { ~[]* }
+             > Verbatim
+             > InlineGrammar
+             > link                       
+             > structured                 
+             > styled                     
+             > (Chars:: alphanum++)       
+             > Quotes:: "\"" text "\""         
+             > (Symbol:: sym++)           
+             > { Block:: text }
+
+word = Chars:: bareword
+
+blockquote = "adsfafewag"
+//blockquote   = Blockquote:: "\"\"" (block | text "\"\"")
              
-qtext        = "\"" text "\""            => "quoted"
-pre          = "[verbatim]" { ~[]+ } /ws => "verbatim"   // FIXME doesn't work
-
-styled       = "__" text "__"      => underline
-             | "((" text "))"      => footnote
-             | ( "[[" text "]]"    => tt
-               >  "[" word "]"     => citation
-               )
-             | "!!" text "!!"      => strikethrough
-             | "^^" (word|block)   => superscript
-             | ",," (word|block)   => subscript
-             | "\\sc" block        => smallcap
-             | "**" text "**"      => bold
-             | "!" (word|block)    => keyword
-             >  "*" text "*"       => it
-
-block   = { text }
-structured   = { text } "->" (url|email) => link
-              //> alphanum++ "->" (url|email) => link
-structuredx   = glyph
+Verbatim      = "[verbatim]" ws! { (~[])++ }
+
+#import meta.g as meta
+InlineGrammar = "\grammar"   ws! { meta.Grammar }
+
+styled       = Underline::     "__" text "__"      
+             | Footnote::      "((" text "))"      
+             | TT::            "[[" text "]]"
+             | Citation::       "[" word "]"
+             | Strikethrough:: "!!" text "!!"      
+             | Superscript::   "^^" (word|block)   
+             | Subscript::     ",," (word|block)   
+             | Smallcap::      "\\sc" block        
+             | Bold::           "++" text "++"
+             | Keyword::         "!" (word|block)    
+             | Italic::         "**" text "**"
+
+block         = { text }
+
+link          = text:({ text }|word)  "->" ws! href:href
+href          = url | email | {href}
+
+structured    = command & "\\" ([a-zA-Z0-9]++)! block?
+              > glyph
               > email
               > url
 
-glyph        = "(r)" | "(c)" | "(tm)" | "--"  // euro symbol?
-             | "\\today" -> ~[a-z] => today
+glyph        = euro::     "(e)"
+             | r::        "(r)"
+             | c::        "(c)"
+             | tm::       "(tm)"
+             | emdash::   "--"
+             | ellipses:: "..."
+             | cent::     "\\cent"
+
+command      = "\\" [a-z]++
 
 
 // URLs //////////////////////////////////////////////////////////////////////////////
@@ -137,44 +147,54 @@ glyph        = "(r)" | "(c)" | "(tm)" | "--"  // euro symbol?
 // subtypes of url (ftp, etc) as conjunctions, but the "master pattern"
 // only gets parsed once
 
-urlpath      = urlchar*
-username     = [a-zA-Z0-9;/?:&=$\-_.+]++ => stringify
-password     = [a-zA-Z0-9;/?:&=$\-_.+]++ => stringify
-urlchar      = [a-zA-Z0-9;/?:&=$\-_.+@]
-             | "%" [0-9] [0-9]       => "%"
-url          = "mailto" ":"   email
-             > method "://" url_login? host (":" nums)? ("/" urlpath)?     => "url"
-url_login    = username (":" password) "@"                                 => "login"
-method       = [+\-.a-z0-9]+ => stringify
-domain       = (part +/ ".") -> ~"." => domain
-part         = [a-zA-Z0-9\-]++  => stringify
+urlpath      = urlchar* -> ~urlv           // this ~urlv should be handled by url! bug!
+username     = [a-zA-Z0-9;/?:&=$\-_.+]++
+password     = [a-zA-Z0-9;/?:&=$\-_.+]++
+urlc         = [a-zA-Z0-9;/?:&=$\-_.+@]
+urlv         = urlc | [%]
+urlchar      = urlc
+             | urlescape:: "%" [0-9] [0-9]
+url          = Mailto:: "mailto" ":"   email -> ~urlv
+             > URL::
+                  method:method 
+                  "://"
+                  login:url_login?
+                  host:host
+                  port:(":" nums)?
+                  path:("/" urlpath)?
+                  ref:("#" urlpath)?
+                     -> ~(urlv|[\#])
+url_login    = Login:: username:username password:(":" password) "@"
+method       = [+\-.a-z0-9]+
+domain       = (part +/ ".") -> ~"."
+part         = [a-zA-Z0-9\-]++
 // interesting use of boolean grammars
 //            &~ ([\-0-9] ~[]* | ~[]* [\-0-9])
 
-email        = username "@" host -> ~[.] => emailaddr
-nums         = [0-9]++ => stringify
-host         = nums "." nums "." nums "." nums => "ip"
-             | domain
+email        = user:username "@" host:host -> ~[.]
+nums         = [0-9]++
+host         = IP::  nums "." nums "." nums "." nums
+             | DNS:: domain
 
 
 
 // Tokens ///////////////////////////////////////////////////////////////////
 
-word       = alphanum++ => stringify
+bareword   = alphanum++
            | quoted
 
 quoted     = "\"" ((~[\"\\] | escaped)+) "\""
-           | "\"\"" => ""
-escaped    = "\\n" => "\n"
-           | "\\r" => "\r"
+           | "\"\"":: "\"\""
+escaped    = lf:: "\\n"
+           | cr:: "\\r"
            | "\\" ~[nr]
 
 
 // Chars ///////////////////////////////////////////////////////////////
 
 alpha      = [a-zA-Z]
-//num        = [0-9]
 alphanum   = [a-zA-Z0-9]
 sym        = ~[a-zA-Z0-9\ \r\n=\">]
+//sym        = [,()]