pmh_grammar.leg

%{
/* PEG Markdown Highlight
 * Copyright 2011 Ali Rantakari -- http://hasseg.org
 * Licensed under the GPL2+ and MIT licenses (see LICENSE for more info).
 * 
 * pmh_grammar.leg
 * 
 * This is a slightly adapted version of the PEG grammar from John MacFarlane's
 * peg-markdown compiler.
 */

/// header_code_here

#define elem(x)     mk_element((parser_data *)G->data, x, thunk->begin, thunk->end)
#define elem_s(x)   mk_element((parser_data *)G->data, x, s->pos, thunk->end)
#define mk_sep      mk_element((parser_data *)G->data, pmh_SEPARATOR, 0,0)
#define mk_notype   mk_element((parser_data *)G->data, pmh_NO_TYPE, 0,0)
#define etext(x)    mk_etext((parser_data *)G->data, x)
#define ADD(x)      add((parser_data *)G->data, x)
#define EXT(x)      extension((parser_data *)G->data, x)
#define REF_EXISTS(x) reference_exists((parser_data *)G->data, x)
#define GET_REF(x)  get_reference((parser_data *)G->data, x)
#define PARSING_REFERENCES ((parser_data *)G->data)->parsing_only_references
#define FREE_LABEL(l) { free(l->label); l->label = NULL; }
#define FREE_ADDRESS(l) { free(l->address); l->address = NULL; }

// This gives us the text matched with < > as it appears in the original input:
#define COPY_YYTEXT_ORIG() copy_input_span((parser_data *)G->data, thunk->begin, thunk->end)

%}

Doc =       ( Block )*

# placeholder for marking locations
LocMarker = &. { $$ = elem(pmh_NO_TYPE); }


Block =     BlankLine*
            ( BlockQuote
            | Verbatim
            | Note
            | Reference
            | HorizontalRule
            | Heading
            | OrderedList
            | BulletList
            | HtmlBlock
            | StyleBlock
            | Para
            | Plain )

Para =      NonindentSpace Inlines BlankLine+

Plain =     Inlines

AtxInline = !Newline !(Sp? '#'* Sp Newline) Inline

AtxStart =  < ( "######" | "#####" | "####" | "###" | "##" | "#" ) >
            { $$ = elem((pmh_element_type)(pmh_H1 + (strlen(yytext) - 1))); }

AtxHeading = < s:AtxStart Sp? ( AtxInline )+ (Sp? '#'* Sp)? Newline >
            { ADD(elem_s(s->type)); }

SetextHeading = SetextHeading1 | SetextHeading2

SetextBottom1 = "===" '='* Newline

SetextBottom2 = "---" '-'* Newline

SetextHeading1 =  &(RawLine SetextBottom1)
                  s:LocMarker
                  < ( !Endline Inline )+ Sp? Newline
                  SetextBottom1 >
                  { ADD(elem_s(pmh_H1)); }

SetextHeading2 =  &(RawLine SetextBottom2)
                  s:LocMarker
                  < ( !Endline Inline )+ Sp? Newline
                  SetextBottom2 >
                  { ADD(elem_s(pmh_H2)); }

Heading = SetextHeading | AtxHeading

BlockQuote = a:BlockQuoteRaw
            { pmh_realelement *rawlist = mk_element((parser_data *)G->data, pmh_RAW_LIST, 0,0);
              rawlist->children = reverse(a);
              ADD(rawlist);
            }

BlockQuoteRaw =  a:StartList
                 (( < '>' ' '? > { ADD(elem(pmh_BLOCKQUOTE)); } Line { a = cons($$, a); } )
                  ( !'>' !BlankLine Line { a = cons($$, a); } )*
                  ( < BlankLine > { a = cons(etext("\n"), a); } )*
                 )+
                 { $$ = a; }

NonblankIndentedLine = !BlankLine IndentedLine

VerbatimChunk = ( BlankLine )*
                ( NonblankIndentedLine )+

Verbatim =     < s:LocMarker
                 ( VerbatimChunk )+ >
                 { ADD(elem_s(pmh_VERBATIM)); }

HorizontalRule = < NonindentSpace
                 ( '*' Sp '*' Sp '*' (Sp '*')*
                 | '-' Sp '-' Sp '-' (Sp '-')*
                 | '_' Sp '_' Sp '_' (Sp '_')*)
                 Sp Newline > BlankLine+
                 { ADD(elem(pmh_HRULE)); }

Bullet = !HorizontalRule NonindentSpace < ('+' | '*' | '-') > Spacechar+
         { ADD(elem(pmh_LIST_BULLET)); }

BulletList = &Bullet (ListTight | ListLoose)

ListTight = a:StartList
            ( ListItemTight
              { pmh_realelement *el = mk_notype;
                el->children = $$;
                a = cons(el, a);
              } )+
            BlankLine* !(Bullet | Enumerator)
            { pmh_realelement *cur = a;
              while (cur != NULL) {
                  pmh_realelement *rawlist = mk_element((parser_data *)G->data, pmh_RAW_LIST, 0,0);
                  rawlist->children = reverse(cur->children);
                  ADD(rawlist);
                  cur = cur->next;
              }
            }

ListLoose = a:StartList
            ( b:ListItem BlankLine*
              { b = cons(etext("\n\n"), b); /* In loose list, \n\n added to end of each element */
                pmh_realelement *el = mk_notype;
                el->children = b;
                a = cons(el, a);
              } )+
            { pmh_realelement *cur = a;
              while (cur != NULL) {
                  pmh_realelement *rawlist = mk_element((parser_data *)G->data, pmh_RAW_LIST, 0,0);
                  rawlist->children = reverse(cur->children);
                  ADD(rawlist);
                  cur = cur->next;
              }
            }

ListItem =  ( Bullet | Enumerator )
            a:StartList
            ListBlock { a = cons($$, a); }
            ( ListContinuationBlock { a = cons($$, a); } )*
            { $$ = a; }

ListItemTight =
            ( Bullet | Enumerator )
            a:StartList
            ListBlock { a = cons($$, a); }
            ( !BlankLine
              ListContinuationBlock { a = cons($$, a); } )*
            !ListContinuationBlock
            { $$ = a; }

ListBlock = a:StartList
            !BlankLine Line { a = cons($$, a); }
            ( ListBlockLine { a = cons(elem(pmh_RAW), a); } )*
            { $$ = a; }

ListContinuationBlock = a:StartList
                        ( < BlankLine* >
                          { if (*yytext == '\0') /* if strlen(yytext) == 0 */
                                a = cons(elem(pmh_SEPARATOR), a);
                            else
                                a = cons(elem(pmh_RAW), a);
                          } )
                        ( Indent ListBlock { a = cons($$, a); } )+
                        { $$ = a; }

Enumerator = NonindentSpace < [0-9]+ '.' > Spacechar+
             { ADD(elem(pmh_LIST_ENUMERATOR)); }

OrderedList = &Enumerator (ListTight | ListLoose)

ListBlockLine = !BlankLine
                !( Indent? (Bullet | Enumerator) )
                !HorizontalRule
                OptionallyIndentedLine

# Parsers for different kinds of block-level HTML content.
# This is repetitive due to constraints of PEG grammar.

HtmlBlockOpenAddress = '<' Spnl ("address" | "ADDRESS") Spnl HtmlAttribute* '>'
HtmlBlockCloseAddress = '<' Spnl '/' ("address" | "ADDRESS") Spnl '>'
HtmlBlockAddress = HtmlBlockOpenAddress (HtmlBlockAddress | !HtmlBlockCloseAddress .)* HtmlBlockCloseAddress

HtmlBlockOpenBlockquote = '<' Spnl ("blockquote" | "BLOCKQUOTE") Spnl HtmlAttribute* '>'
HtmlBlockCloseBlockquote = '<' Spnl '/' ("blockquote" | "BLOCKQUOTE") Spnl '>'
HtmlBlockBlockquote = HtmlBlockOpenBlockquote (HtmlBlockBlockquote | !HtmlBlockCloseBlockquote .)* HtmlBlockCloseBlockquote

HtmlBlockOpenCenter = '<' Spnl ("center" | "CENTER") Spnl HtmlAttribute* '>'
HtmlBlockCloseCenter = '<' Spnl '/' ("center" | "CENTER") Spnl '>'
HtmlBlockCenter = HtmlBlockOpenCenter (HtmlBlockCenter | !HtmlBlockCloseCenter .)* HtmlBlockCloseCenter

HtmlBlockOpenDir = '<' Spnl ("dir" | "DIR") Spnl HtmlAttribute* '>'
HtmlBlockCloseDir = '<' Spnl '/' ("dir" | "DIR") Spnl '>'
HtmlBlockDir = HtmlBlockOpenDir (HtmlBlockDir | !HtmlBlockCloseDir .)* HtmlBlockCloseDir

HtmlBlockOpenDiv = '<' Spnl ("div" | "DIV") Spnl HtmlAttribute* '>'
HtmlBlockCloseDiv = '<' Spnl '/' ("div" | "DIV") Spnl '>'
HtmlBlockDiv = HtmlBlockOpenDiv (HtmlBlockDiv | !HtmlBlockCloseDiv .)* HtmlBlockCloseDiv

HtmlBlockOpenDl = '<' Spnl ("dl" | "DL") Spnl HtmlAttribute* '>'
HtmlBlockCloseDl = '<' Spnl '/' ("dl" | "DL") Spnl '>'
HtmlBlockDl = HtmlBlockOpenDl (HtmlBlockDl | !HtmlBlockCloseDl .)* HtmlBlockCloseDl

HtmlBlockOpenFieldset = '<' Spnl ("fieldset" | "FIELDSET") Spnl HtmlAttribute* '>'
HtmlBlockCloseFieldset = '<' Spnl '/' ("fieldset" | "FIELDSET") Spnl '>'
HtmlBlockFieldset = HtmlBlockOpenFieldset (HtmlBlockFieldset | !HtmlBlockCloseFieldset .)* HtmlBlockCloseFieldset

HtmlBlockOpenForm = '<' Spnl ("form" | "FORM") Spnl HtmlAttribute* '>'
HtmlBlockCloseForm = '<' Spnl '/' ("form" | "FORM") Spnl '>'
HtmlBlockForm = HtmlBlockOpenForm (HtmlBlockForm | !HtmlBlockCloseForm .)* HtmlBlockCloseForm

HtmlBlockOpenH1 = '<' Spnl ("h1" | "H1") Spnl HtmlAttribute* '>'
HtmlBlockCloseH1 = '<' Spnl '/' ("h1" | "H1") Spnl '>'
HtmlBlockH1 = < s:LocMarker HtmlBlockOpenH1 (HtmlBlockH1 | !HtmlBlockCloseH1 .)* HtmlBlockCloseH1 >
                { ADD(elem_s(pmh_H1)); }

HtmlBlockOpenH2 = '<' Spnl ("h2" | "H2") Spnl HtmlAttribute* '>'
HtmlBlockCloseH2 = '<' Spnl '/' ("h2" | "H2") Spnl '>'
HtmlBlockH2 = < s:LocMarker HtmlBlockOpenH2 (HtmlBlockH2 | !HtmlBlockCloseH2 .)* HtmlBlockCloseH2 >
                { ADD(elem_s(pmh_H2)); }

HtmlBlockOpenH3 = '<' Spnl ("h3" | "H3") Spnl HtmlAttribute* '>'
HtmlBlockCloseH3 = '<' Spnl '/' ("h3" | "H3") Spnl '>'
HtmlBlockH3 = < s:LocMarker HtmlBlockOpenH3 (HtmlBlockH3 | !HtmlBlockCloseH3 .)* HtmlBlockCloseH3 >
                { ADD(elem_s(pmh_H3)); }

HtmlBlockOpenH4 = '<' Spnl ("h4" | "H4") Spnl HtmlAttribute* '>'
HtmlBlockCloseH4 = '<' Spnl '/' ("h4" | "H4") Spnl '>'
HtmlBlockH4 = < s:LocMarker HtmlBlockOpenH4 (HtmlBlockH4 | !HtmlBlockCloseH4 .)* HtmlBlockCloseH4 >
                { ADD(elem_s(pmh_H4)); }

HtmlBlockOpenH5 = '<' Spnl ("h5" | "H5") Spnl HtmlAttribute* '>'
HtmlBlockCloseH5 = '<' Spnl '/' ("h5" | "H5") Spnl '>'
HtmlBlockH5 = < s:LocMarker HtmlBlockOpenH5 (HtmlBlockH5 | !HtmlBlockCloseH5 .)* HtmlBlockCloseH5 >
                { ADD(elem_s(pmh_H5)); }

HtmlBlockOpenH6 = '<' Spnl ("h6" | "H6") Spnl HtmlAttribute* '>'
HtmlBlockCloseH6 = '<' Spnl '/' ("h6" | "H6") Spnl '>'
HtmlBlockH6 = < s:LocMarker HtmlBlockOpenH6 (HtmlBlockH6 | !HtmlBlockCloseH6 .)* HtmlBlockCloseH6 >
                { ADD(elem_s(pmh_H6)); }

HtmlBlockOpenMenu = '<' Spnl ("menu" | "MENU") Spnl HtmlAttribute* '>'
HtmlBlockCloseMenu = '<' Spnl '/' ("menu" | "MENU") Spnl '>'
HtmlBlockMenu = HtmlBlockOpenMenu (HtmlBlockMenu | !HtmlBlockCloseMenu .)* HtmlBlockCloseMenu

HtmlBlockOpenNoframes = '<' Spnl ("noframes" | "NOFRAMES") Spnl HtmlAttribute* '>'
HtmlBlockCloseNoframes = '<' Spnl '/' ("noframes" | "NOFRAMES") Spnl '>'
HtmlBlockNoframes = HtmlBlockOpenNoframes (HtmlBlockNoframes | !HtmlBlockCloseNoframes .)* HtmlBlockCloseNoframes

HtmlBlockOpenNoscript = '<' Spnl ("noscript" | "NOSCRIPT") Spnl HtmlAttribute* '>'
HtmlBlockCloseNoscript = '<' Spnl '/' ("noscript" | "NOSCRIPT") Spnl '>'
HtmlBlockNoscript = HtmlBlockOpenNoscript (HtmlBlockNoscript | !HtmlBlockCloseNoscript .)* HtmlBlockCloseNoscript

HtmlBlockOpenOl = '<' Spnl ("ol" | "OL") Spnl HtmlAttribute* '>'
HtmlBlockCloseOl = '<' Spnl '/' ("ol" | "OL") Spnl '>'
HtmlBlockOl = HtmlBlockOpenOl (HtmlBlockOl | !HtmlBlockCloseOl .)* HtmlBlockCloseOl

HtmlBlockOpenP = '<' Spnl ("p" | "P") Spnl HtmlAttribute* '>'
HtmlBlockCloseP = '<' Spnl '/' ("p" | "P") Spnl '>'
HtmlBlockP = HtmlBlockOpenP (HtmlBlockP | !HtmlBlockCloseP .)* HtmlBlockCloseP

HtmlBlockOpenPre = '<' Spnl ("pre" | "PRE") Spnl HtmlAttribute* '>'
HtmlBlockClosePre = '<' Spnl '/' ("pre" | "PRE") Spnl '>'
HtmlBlockPre = HtmlBlockOpenPre (HtmlBlockPre | !HtmlBlockClosePre .)* HtmlBlockClosePre

HtmlBlockOpenTable = '<' Spnl ("table" | "TABLE") Spnl HtmlAttribute* '>'
HtmlBlockCloseTable = '<' Spnl '/' ("table" | "TABLE") Spnl '>'
HtmlBlockTable = HtmlBlockOpenTable (HtmlBlockTable | !HtmlBlockCloseTable .)* HtmlBlockCloseTable

HtmlBlockOpenUl = '<' Spnl ("ul" | "UL") Spnl HtmlAttribute* '>'
HtmlBlockCloseUl = '<' Spnl '/' ("ul" | "UL") Spnl '>'
HtmlBlockUl = HtmlBlockOpenUl (HtmlBlockUl | !HtmlBlockCloseUl .)* HtmlBlockCloseUl

HtmlBlockOpenDd = '<' Spnl ("dd" | "DD") Spnl HtmlAttribute* '>'
HtmlBlockCloseDd = '<' Spnl '/' ("dd" | "DD") Spnl '>'
HtmlBlockDd = HtmlBlockOpenDd (HtmlBlockDd | !HtmlBlockCloseDd .)* HtmlBlockCloseDd

HtmlBlockOpenDt = '<' Spnl ("dt" | "DT") Spnl HtmlAttribute* '>'
HtmlBlockCloseDt = '<' Spnl '/' ("dt" | "DT") Spnl '>'
HtmlBlockDt = HtmlBlockOpenDt (HtmlBlockDt | !HtmlBlockCloseDt .)* HtmlBlockCloseDt

HtmlBlockOpenFrameset = '<' Spnl ("frameset" | "FRAMESET") Spnl HtmlAttribute* '>'
HtmlBlockCloseFrameset = '<' Spnl '/' ("frameset" | "FRAMESET") Spnl '>'
HtmlBlockFrameset = HtmlBlockOpenFrameset (HtmlBlockFrameset | !HtmlBlockCloseFrameset .)* HtmlBlockCloseFrameset

HtmlBlockOpenLi = '<' Spnl ("li" | "LI") Spnl HtmlAttribute* '>'
HtmlBlockCloseLi = '<' Spnl '/' ("li" | "LI") Spnl '>'
HtmlBlockLi = HtmlBlockOpenLi (HtmlBlockLi | !HtmlBlockCloseLi .)* HtmlBlockCloseLi

HtmlBlockOpenTbody = '<' Spnl ("tbody" | "TBODY") Spnl HtmlAttribute* '>'
HtmlBlockCloseTbody = '<' Spnl '/' ("tbody" | "TBODY") Spnl '>'
HtmlBlockTbody = HtmlBlockOpenTbody (HtmlBlockTbody | !HtmlBlockCloseTbody .)* HtmlBlockCloseTbody

HtmlBlockOpenTd = '<' Spnl ("td" | "TD") Spnl HtmlAttribute* '>'
HtmlBlockCloseTd = '<' Spnl '/' ("td" | "TD") Spnl '>'
HtmlBlockTd = HtmlBlockOpenTd (HtmlBlockTd | !HtmlBlockCloseTd .)* HtmlBlockCloseTd

HtmlBlockOpenTfoot = '<' Spnl ("tfoot" | "TFOOT") Spnl HtmlAttribute* '>'
HtmlBlockCloseTfoot = '<' Spnl '/' ("tfoot" | "TFOOT") Spnl '>'
HtmlBlockTfoot = HtmlBlockOpenTfoot (HtmlBlockTfoot | !HtmlBlockCloseTfoot .)* HtmlBlockCloseTfoot

HtmlBlockOpenTh = '<' Spnl ("th" | "TH") Spnl HtmlAttribute* '>'
HtmlBlockCloseTh = '<' Spnl '/' ("th" | "TH") Spnl '>'
HtmlBlockTh = HtmlBlockOpenTh (HtmlBlockTh | !HtmlBlockCloseTh .)* HtmlBlockCloseTh

HtmlBlockOpenThead = '<' Spnl ("thead" | "THEAD") Spnl HtmlAttribute* '>'
HtmlBlockCloseThead = '<' Spnl '/' ("thead" | "THEAD") Spnl '>'
HtmlBlockThead = HtmlBlockOpenThead (HtmlBlockThead | !HtmlBlockCloseThead .)* HtmlBlockCloseThead

HtmlBlockOpenTr = '<' Spnl ("tr" | "TR") Spnl HtmlAttribute* '>'
HtmlBlockCloseTr = '<' Spnl '/' ("tr" | "TR") Spnl '>'
HtmlBlockTr = HtmlBlockOpenTr (HtmlBlockTr | !HtmlBlockCloseTr .)* HtmlBlockCloseTr

HtmlBlockOpenScript = '<' Spnl ("script" | "SCRIPT") Spnl HtmlAttribute* '>'
HtmlBlockCloseScript = '<' Spnl '/' ("script" | "SCRIPT") Spnl '>'
HtmlBlockScript = HtmlBlockOpenScript (!HtmlBlockCloseScript .)* HtmlBlockCloseScript

HtmlBlockInTags = HtmlBlockAddress
                | HtmlBlockBlockquote
                | HtmlBlockCenter
                | HtmlBlockDir
                | HtmlBlockDiv
                | HtmlBlockDl
                | HtmlBlockFieldset
                | HtmlBlockForm
                | HtmlBlockH1
                | HtmlBlockH2
                | HtmlBlockH3
                | HtmlBlockH4
                | HtmlBlockH5
                | HtmlBlockH6
                | HtmlBlockMenu
                | HtmlBlockNoframes
                | HtmlBlockNoscript
                | HtmlBlockOl
                | HtmlBlockP
                | HtmlBlockPre
                | HtmlBlockTable
                | HtmlBlockUl
                | HtmlBlockDd
                | HtmlBlockDt
                | HtmlBlockFrameset
                | HtmlBlockLi
                | HtmlBlockTbody
                | HtmlBlockTd
                | HtmlBlockTfoot
                | HtmlBlockTh
                | HtmlBlockThead
                | HtmlBlockTr
                | HtmlBlockScript

HtmlBlock = < s:LocMarker
              ( HtmlBlockInTags | HtmlComment | HtmlBlockSelfClosing ) >
            BlankLine+
            { ADD(elem_s(pmh_HTMLBLOCK)); }

HtmlBlockSelfClosing = '<' Spnl HtmlBlockType Spnl HtmlAttribute* '/' Spnl '>'

HtmlBlockType = "address" | "blockquote" | "center" | "dir" | "div" | "dl" | "fieldset" | "form" | "h1" | "h2" | "h3" |
                "h4" | "h5" | "h6" | "hr" | "isindex" | "menu" | "noframes" | "noscript" | "ol" | "p" | "pre" | "table" |
                "ul" | "dd" | "dt" | "frameset" | "li" | "tbody" | "td" | "tfoot" | "th" | "thead" | "tr" | "script" |
                "ADDRESS" | "BLOCKQUOTE" | "CENTER" | "DIR" | "DIV" | "DL" | "FIELDSET" | "FORM" | "H1" | "H2" | "H3" |
                "H4" | "H5" | "H6" | "HR" | "ISINDEX" | "MENU" | "NOFRAMES" | "NOSCRIPT" | "OL" | "P" | "PRE" | "TABLE" |
                "UL" | "DD" | "DT" | "FRAMESET" | "LI" | "TBODY" | "TD" | "TFOOT" | "TH" | "THEAD" | "TR" | "SCRIPT"

StyleOpen =     '<' Spnl ("style" | "STYLE") Spnl HtmlAttribute* '>'
StyleClose =    '<' Spnl '/' ("style" | "STYLE") Spnl '>'
InStyleTags =   StyleOpen (!StyleClose .)* StyleClose
StyleBlock =    < s:LocMarker InStyleTags >
                BlankLine*
                { ADD(elem_s(pmh_HTMLBLOCK)); }

Inlines  =  ( !Endline Inline
              | Endline &Inline )+ Endline?

Inline  = Str
        | Endline
        | UlOrStarLine
        | Space
        | Strong
        | Emph
        | Image
        | Link
        | NoteReference
        | InlineNote
        | Code
        | RawHtml
        | Entity
        | EscapedChar
        | Symbol

Space = Spacechar+

Str = NormalChar (NormalChar | '_'+ &Alphanumeric)*

EscapedChar =   '\\' !Newline [-\\`|*_{}[\]()#+.!><]

Entity =    < s:LocMarker
            ( HexEntity | DecEntity | CharEntity ) >
            { ADD(elem_s(pmh_HTML_ENTITY)); }

Endline =   LineBreak | TerminalEndline | NormalEndline

NormalEndline =   Sp Newline !BlankLine !'>' !AtxStart
                  !(Line ("===" '='* | "---" '-'*) Newline)

TerminalEndline = Sp Newline Eof

LineBreak = "  " NormalEndline

Symbol =    SpecialChar

# This keeps the parser from getting bogged down on long strings of '*' or '_',
# or strings of '*' or '_' with space on each side:
UlOrStarLine =  (UlLine | StarLine)
StarLine =      "****" '*'* | Spacechar '*'+ &Spacechar
UlLine   =      "____" '_'* | Spacechar '_'+ &Spacechar

Emph =      EmphStar | EmphUl

OneStarOpen  =  !StarLine < '*' > !Spacechar !Newline { $$ = elem(pmh_NO_TYPE); }
OneStarClose =  !Spacechar !Newline Inline !StrongStar < '*' > { $$ = elem(pmh_NO_TYPE); }

EmphStar =  s:OneStarOpen
            ( !OneStarClose Inline )*
            OneStarClose
            { ADD(elem_s(pmh_EMPH)); }

OneUlOpen  =  !UlLine < '_' > !Spacechar !Newline { $$ = elem(pmh_NO_TYPE); }
OneUlClose =  !Spacechar !Newline Inline !StrongUl < '_' > !Alphanumeric { $$ = elem(pmh_NO_TYPE); }

EmphUl =    s:OneUlOpen
            ( !OneUlClose Inline )*
            OneUlClose
            { ADD(elem_s(pmh_EMPH)); }

Strong = StrongStar | StrongUl

TwoStarOpen =   !StarLine < "**" > !Spacechar !Newline { $$ = elem(pmh_NO_TYPE); }
TwoStarClose =  !Spacechar !Newline Inline < "**" > { $$ = elem(pmh_NO_TYPE); }

StrongStar =    s:TwoStarOpen
                ( !TwoStarClose Inline )*
                TwoStarClose
                { ADD(elem_s(pmh_STRONG)); }

TwoUlOpen =     !UlLine < "__" > !Spacechar !Newline { $$ = elem(pmh_NO_TYPE); }
TwoUlClose =    !Spacechar !Newline Inline < "__" > !Alphanumeric { $$ = elem(pmh_NO_TYPE); }

StrongUl =  s:TwoUlOpen
            ( !TwoUlClose Inline )*
            TwoUlClose
            { ADD(elem_s(pmh_STRONG)); }

Image = '!' ( ExplicitLink | ReferenceLink )
        {
            if ($$ != NULL) {
                $$->type = pmh_IMAGE;
                $$->pos -= 1;
                ADD($$);
            }
        }

Link =  ( ExplicitLink | ReferenceLink | AutoLink )
        { if ($$) ADD($$); } # AutoLink does not return $$

ReferenceLink = ReferenceLinkDouble | ReferenceLinkSingle

ReferenceLinkDouble =  < s:Label Spnl !"[]" l:Label >
                        {
                        	pmh_realelement *reference = GET_REF(l->label);
                            if (reference) {
                                $$ = elem_s(pmh_LINK);
                                $$->label = strdup(l->label);
                                $$->address = strdup(reference->address);
                            } else
                                $$ = NULL;
                            FREE_LABEL(s);
                            FREE_LABEL(l);
                        }

ReferenceLinkSingle =  < s:Label (Spnl "[]")? >
                        {
                        	pmh_realelement *reference = GET_REF(s->label);
                            if (reference) {
                                $$ = elem_s(pmh_LINK);
                                $$->label = strdup(s->label);
                                $$->address = strdup(reference->address);
                            } else
                                $$ = NULL;
                            FREE_LABEL(s);
                        }

ExplicitLink =  < s:Label Spnl '(' Sp l:Source Spnl Title Sp ')' >
                {
                    $$ = elem_s(pmh_LINK);
                    if (l->address != NULL)
                        $$->address = strdup(l->address);
                    FREE_LABEL(s);
                    FREE_ADDRESS(l);
                }

Source  = { $$ = mk_notype; }
          ( '<' < SourceContents > { $$->address = COPY_YYTEXT_ORIG(); } '>'
          | < SourceContents > { $$->address = COPY_YYTEXT_ORIG(); } )

SourceContents = ( ( !'(' !')' !'>' Nonspacechar )+ | '(' SourceContents ')')*

Title = ( TitleSingle | TitleDouble | "" )

TitleSingle = '\'' ( !( '\'' Sp ( ')' | Newline ) ) . )* '\''

TitleDouble = '"' ( !( '"' Sp ( ')' | Newline ) ) . )* '"'

AutoLink = AutoLinkUrl | AutoLinkEmail

AutoLinkUrl =  < s:LocMarker { s->type = pmh_AUTO_LINK_URL; }
               '<'
                 < [A-Za-z]+ "://" ( !Newline !'>' . )+ >
                 { s->address = COPY_YYTEXT_ORIG(); }
               '>' >
               {
                s->end = thunk->end;
                ADD(s);
               }

AutoLinkEmail = < s:LocMarker { s->type = pmh_AUTO_LINK_EMAIL; }
                '<'
                  ( "mailto:" )? < [-A-Za-z0-9+_./!%~$]+ '@' ( !Newline !'>' . )+ >
                  { s->address = COPY_YYTEXT_ORIG(); }
                '>' >
               {
                s->end = thunk->end;
                ADD(s);
               }

Reference = < s:LocMarker
              NonindentSpace !"[]" l:Label ':' Spnl r:RefSrc RefTitle > BlankLine+
              {
                pmh_realelement *el = elem_s(pmh_REFERENCE);
                el->label = strdup(l->label);
                el->address = strdup(r->address);
                ADD(el);
                FREE_LABEL(l);
                FREE_ADDRESS(r);
              }

Label = < s:LocMarker
        '[' ( !'^' &{ EXT(pmh_EXT_NOTES) } | &. &{ !EXT(pmh_EXT_NOTES) } )
        < ( !']' Inline )* >
        { s->label = COPY_YYTEXT_ORIG(); }
        ']' >
        {
            s->pos = s->pos;
            s->end = thunk->end;
            $$ = s;
        }

RefSrc = < Nonspacechar+ >
		 { $$ = mk_notype; $$->address = COPY_YYTEXT_ORIG(); }

RefTitle =  ( RefTitleSingle | RefTitleDouble | RefTitleParens | EmptyTitle )

EmptyTitle = ""

RefTitleSingle = Spnl '\'' ( !('\'' Sp Newline | Newline ) . )* '\''

RefTitleDouble = Spnl '"' ( !('"' Sp Newline | Newline) . )* '"'

RefTitleParens = Spnl '(' ( !(')' Sp Newline | Newline) . )* ')'

# Starting point for parsing only references:
References = ( Reference | SkipBlock )*

Ticks1 = < "`" > !'`' { $$ = elem(pmh_NO_TYPE); }
Ticks2 = < "``" > !'`' { $$ = elem(pmh_NO_TYPE); }
Ticks3 = < "```" > !'`' { $$ = elem(pmh_NO_TYPE); }
Ticks4 = < "````" > !'`' { $$ = elem(pmh_NO_TYPE); }
Ticks5 = < "`````" > !'`' { $$ = elem(pmh_NO_TYPE); }

Code = < ( s:Ticks1 Sp ( ( !'`' Nonspacechar )+ | !Ticks1 '`'+ | !( Sp Ticks1 ) ( Spacechar | Newline !BlankLine ) )+ Sp Ticks1
       | s:Ticks2 Sp ( ( !'`' Nonspacechar )+ | !Ticks2 '`'+ | !( Sp Ticks2 ) ( Spacechar | Newline !BlankLine ) )+ Sp Ticks2
       | s:Ticks3 Sp ( ( !'`' Nonspacechar )+ | !Ticks3 '`'+ | !( Sp Ticks3 ) ( Spacechar | Newline !BlankLine ) )+ Sp Ticks3
       | s:Ticks4 Sp ( ( !'`' Nonspacechar )+ | !Ticks4 '`'+ | !( Sp Ticks4 ) ( Spacechar | Newline !BlankLine ) )+ Sp Ticks4
       | s:Ticks5 Sp ( ( !'`' Nonspacechar )+ | !Ticks5 '`'+ | !( Sp Ticks5 ) ( Spacechar | Newline !BlankLine ) )+ Sp Ticks5
       ) >
       { ADD(elem_s(pmh_CODE)); }

RawHtml =   < s:LocMarker (HtmlComment | HtmlBlockScript | HtmlTag) >
            { ADD(elem_s(pmh_HTML)); }

BlankLine =     Sp Newline

Quoted =        '"' (!'"' .)* '"' | '\'' (!'\'' .)* '\''
HtmlAttribute = (AlphanumericAscii | '-')+ Spnl ('=' Spnl (Quoted | (!'>' Nonspacechar)+))? Spnl
HtmlComment =   < s:LocMarker "<!--" (!"-->" .)* "-->" >
                { ADD(elem_s(pmh_COMMENT)); }
HtmlTag =       '<' Spnl '/'? AlphanumericAscii+ Spnl HtmlAttribute* '/'? Spnl '>'
Eof =           !.
Spacechar =     ' ' | '\t'
Nonspacechar =  !Spacechar !Newline .
Newline =       '\n' | '\r' '\n'?
Sp =            Spacechar*
Spnl =          Sp (Newline Sp)?
SpecialChar =   '*' | '_' | '`' | '&' | '[' | ']' | '(' | ')' | '<' | '!' | '#' | '\\' | '\'' | '"' | ExtendedSpecialChar
NormalChar =    !( SpecialChar | Spacechar | Newline ) .
# Not used anywhere in grammar:
#NonAlphanumeric = [\000-\057\072-\100\133-\140\173-\177]
Alphanumeric = [0-9A-Za-z] | '\200' | '\201' | '\202' | '\203' | '\204' | '\205' | '\206' | '\207' | '\210' | '\211' | '\212' | '\213' | '\214' | '\215' | '\216' | '\217' | '\220' | '\221' | '\222' | '\223' | '\224' | '\225' | '\226' | '\227' | '\230' | '\231' | '\232' | '\233' | '\234' | '\235' | '\236' | '\237' | '\240' | '\241' | '\242' | '\243' | '\244' | '\245' | '\246' | '\247' | '\250' | '\251' | '\252' | '\253' | '\254' | '\255' | '\256' | '\257' | '\260' | '\261' | '\262' | '\263' | '\264' | '\265' | '\266' | '\267' | '\270' | '\271' | '\272' | '\273' | '\274' | '\275' | '\276' | '\277' | '\300' | '\301' | '\302' | '\303' | '\304' | '\305' | '\306' | '\307' | '\310' | '\311' | '\312' | '\313' | '\314' | '\315' | '\316' | '\317' | '\320' | '\321' | '\322' | '\323' | '\324' | '\325' | '\326' | '\327' | '\330' | '\331' | '\332' | '\333' | '\334' | '\335' | '\336' | '\337' | '\340' | '\341' | '\342' | '\343' | '\344' | '\345' | '\346' | '\347' | '\350' | '\351' | '\352' | '\353' | '\354' | '\355' | '\356' | '\357' | '\360' | '\361' | '\362' | '\363' | '\364' | '\365' | '\366' | '\367' | '\370' | '\371' | '\372' | '\373' | '\374' | '\375' | '\376' | '\377'
AlphanumericAscii = [A-Za-z0-9]

HexEntity =     '&' '#' [Xx] [0-9a-fA-F]+ ';'
DecEntity =     '&' '#' [0-9]+ ';'
CharEntity =    '&' [A-Za-z0-9]+ ';'

NonindentSpace =    "   " | "  " | " " | ""
Indent =            "\t" | "    "
IndentedLine =      Indent Line
OptionallyIndentedLine = Indent? Line

# StartList starts a list data structure that can be added to with cons:
StartList = &.
            { $$ = NULL; }

Line =  RawLine
       { $$ = mk_element((parser_data *)G->data, pmh_RAW, $$->pos, $$->end); }

RawLine = ( < (!'\r' !'\n' .)* Newline > | < .+ > Eof )
          { $$ = elem(pmh_RAW); }

SkipBlock = ( !BlankLine RawLine )+ BlankLine*
          | BlankLine+

# Syntax extensions

ExtendedSpecialChar = &{ EXT(pmh_EXT_NOTES) } ( '^' )

NoteReference = &{ EXT(pmh_EXT_NOTES) }
                RawNoteReference

RawNoteReference = "[^" ( !Newline !']' . )+ ']'

Note =          &{ EXT(pmh_EXT_NOTES) }
                NonindentSpace RawNoteReference ':' Sp
                ( RawNoteBlock )
                ( &Indent RawNoteBlock )*

InlineNote =    &{ EXT(pmh_EXT_NOTES) }
                "^["
                ( !']' Inline )+
                ']'

# Not used anywhere in grammar:
#Notes =         ( Note | SkipBlock )*

RawNoteBlock =  ( !BlankLine OptionallyIndentedLine )+
                ( BlankLine* )

%%