X-Git-Url: http://git.megacz.com/?p=sbp.git;a=blobdiff_plain;f=tests%2Ftibdoc.g;h=b3b107eb01a52e48a89e53ebdef0b148172a7b1c;hp=918e80cccf12995c8b0e60a55954253a6c0a2dd1;hb=22a2ca7685dcf29e9ddcc6e26443ea0277385fca;hpb=a172395c46dac0c9fe82b6aa236abcd4f422fc08 diff --git a/tests/tibdoc.g b/tests/tibdoc.g index 918e80c..b3b107e 100644 --- a/tests/tibdoc.g +++ b/tests/tibdoc.g @@ -1,3 +1,4 @@ +// interactions between !=> and &~ mean that I need to rethink the chartage // indentation styling... // literal blocks [[need to ignore bracing]] double-colon style? // definition -- by prior line indentation, like headings in the original structured text @@ -5,80 +6,139 @@ // dropcap // output formats: latex, contex, ps, pdf, html, man, txt, rfc -// URGENT: why does swapping [a-zA-Z0-9] for alphanum in "item" cause severe breakage? -// URGENT: why does enabling "url" cause severe breakage? (probably same cause) -// ... something to do with unit productions - -// output formats: latex, contex, ps, pdf, html, man, txt, rfc -// bullet list -// numbered list -// horizontal rule -// dropcap -// smallcap -// strikethrough -// link +// escapification +// comment // math // image -// today's date, last edit date -// blockquote -// superscript -// subscript -// citations/references -// typewriter-text // figures +// "reference-style" links +// +// this[1] is fun +// +// [1] http://... +// + +// nonbreaking text? +// degree: 15^o +// Arrows: <- -> => <= <-> + +// textblocks: +// - attention, caution, danger, error, hint, important, note, tip, warning +// definition +// sidebar +// figure-with-caption +// epigraph (end-of-chapter note) +// compound paragraph (??) +// csv-table? +// table of contents +// header, footer +// #include + +// simple macros (#define) (\define) + +// table representation +// +// \table +// a bbb c +// ddd e +// +// [a] ... +// [b] ... +// [c] ... + // FIXME: these have to go at the top so they have their dropAll bit set before PreSequence.build... -ws !::= w** -w !::= [\r\n\ ] -nw ::= [~\r\n\ ] +w = " " | "\n" | "\r" +ws = "()":: w** +// | "()":: w** "#" (~[\n])* "\n" ws +nw = ~[\r\n\ ] ////////////////////////////////////////////////////////////////////////////// -s ::= {Doc} => top - -Doc ::= Header Body /ws => doc -Header ::= "header" { kv */ ws } /ws => header -Body ::= Section*/ws => body -Section ::= SectionHeader Paragraph*/ws /ws => section -SectionHeader ::= "==" SectionHeaderBody "==" -SectionHeaderBody ::= "=" SectionHeaderBody "=" - > ws text ws - -kv ::= word "=" text /ws => kv1 - -num !::= [0-9]++ -Paragraph ::= { "\"\"" ws text } => "blockquote" - > { "*" " " ws text } => "ul" - > { "#" " " ws text } => "ol" - > { num " " ws text } => "ol" - > { "---" "-"* } => "hr" - > { text } => "p" - -text ::= item */ ws -pre ::= "[verbatim]" { [~]+ } /ws => "verbatim" // FIXME doesn't work -item ::= pre - > email - > structured - > styled - > "\"" text "\"" => quoted - > [a-zA-Z0-9]++ - > symbol - -symbol ::= symbolx & sym++ -symbolx ::= "--" => emdash - | "," - | ":" - | ";" - -styled ::= "**" text "**" => bold - | "__" text "__" => ul - | "~~" text "~~" => it // hard to type - | "((" text "))" => footnote - -structured ::= glyph -// | url - -glyph ::= "(r)" | "(c)" | "(tm)" // euro symbol? +s = Doc + +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! + +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 "\"\"") + +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 = euro:: "(e)" + | r:: "(r)" + | c:: "(c)" + | tm:: "(tm)" + | emdash:: "--" + | ellipses:: "..." + | cent:: "\\cent" + +command = "\\" [a-z]++ // URLs ////////////////////////////////////////////////////////////////////////////// @@ -87,44 +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;/?:&=$\-_.+]++ -password ::= [a-zA-Z0-9;/?:&=$\-_.+]++ -urlchar ::= [a-zA-Z0-9;/?:&=$\-_.+@] - | "%" [0-9] [0-9] => "%" -url ::= "mailto" ":" email - > method "://" url_login? host (":" port)? ("/" urlpath)? => "url" -url_login ::= username (":" password) "@" => "login" -method ::= [+\-.a-z0-9]+ -port ::= [0-9]+ - -domain ::= part +/ "." -part ::= [a-zA-Z0-9\-]++ // interesting use of boolean grammars -// &~ ([\-0-9] [~]* | [~]* [\-0-9]) - -email ::= username "@" host => email -host ::= [0-9]+ "." [0-9]+ "." [0-9]+ "." [0-9]+ => "ip" - | domain +urlpath = urlchar* -> ~urlv // this ~urlv should be handled by url! bug! +username = [a-zA-Z0-9;/?:&=$\-_.+]++ +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 = user:username "@" host:host -> ~[.] +nums = [0-9]++ +host = IP:: nums "." nums "." nums "." nums + | DNS:: domain // Tokens /////////////////////////////////////////////////////////////////// -word ::= alphanum++ +bareword = alphanum++ | quoted -quoted ::= "\"" (([~\"\\] | escaped)+) "\"" - | "\"\"" => "" -escaped ::= "\\n" => "\n" - | "\\r" => "\r" - | "\\" [~nr] +quoted = "\"" ((~[\"\\] | escaped)+) "\"" + | "\"\"":: "\"\"" +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] +alpha = [a-zA-Z] +alphanum = [a-zA-Z0-9] +sym = ~[a-zA-Z0-9\ \r\n=\">] +//sym = [,()]