src/Pure/Thy/markdown.ML
author wenzelm
Thu, 15 Oct 2015 12:43:02 +0200
changeset 61445 31aadb15eda5
parent 61444 1fcdfc1a7e50
child 61446 9b09acfb7e06
permissions -rw-r--r--
more document structure; tuned signature;
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
61441
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
     1
(*  Title:      Pure/Thy/markdown.ML
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
     2
    Author:     Makarius
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
     3
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
     4
Minimal support for Markdown documents (see also http://commonmark.org).
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
     5
*)
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
     6
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
     7
signature MARKDOWN =
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
     8
sig
61443
78bbfadd1034 more document structure;
wenzelm
parents: 61442
diff changeset
     9
  datatype kind = Itemize | Enumerate | Description
61445
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    10
  type marker = {indent: int, kind: kind}
61441
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    11
  type line
61443
78bbfadd1034 more document structure;
wenzelm
parents: 61442
diff changeset
    12
  val line_content: line -> Antiquote.text_antiquote list
61445
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    13
  val make_line: Antiquote.text_antiquote list -> line
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    14
  val empty_line: line
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    15
  datatype block = Paragraph of line list | List of marker * block list
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    16
  val read_lines: line list -> block list
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    17
  val read: Input.source -> block list
61441
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    18
end;
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    19
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    20
structure Markdown: MARKDOWN =
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    21
struct
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    22
61445
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    23
(* document lines *)
61441
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    24
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    25
datatype kind = Itemize | Enumerate | Description;
61445
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    26
type marker = {indent: int, kind: kind};
61441
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    27
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    28
datatype line =
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    29
  Line of
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    30
   {content: Antiquote.text_antiquote list,
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    31
    is_empty: bool,
61445
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    32
    marker: (marker * Position.T) option};
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    33
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    34
val eof_line =
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    35
  Line {content = [Antiquote.Text [(Symbol.eof, Position.none)]],
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    36
    is_empty = false, marker = NONE};
61441
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    37
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    38
fun line_content (Line {content, ...}) = content;
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    39
fun line_is_empty (Line {is_empty, ...}) = is_empty;
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    40
fun line_marker (Line {marker, ...}) = marker;
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    41
61443
78bbfadd1034 more document structure;
wenzelm
parents: 61442
diff changeset
    42
78bbfadd1034 more document structure;
wenzelm
parents: 61442
diff changeset
    43
(* make line *)
78bbfadd1034 more document structure;
wenzelm
parents: 61442
diff changeset
    44
61441
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    45
local
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    46
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    47
fun bad_blank ((s, _): Symbol_Pos.T) = Symbol.is_ascii_blank s andalso s <> Symbol.space;
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    48
val bad_blanks = maps (fn Antiquote.Text ss => filter bad_blank ss | _ => []);
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    49
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    50
fun check_blanks content =
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    51
  (case bad_blanks content of
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    52
    [] => ()
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    53
  | (c, pos) :: _ =>
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    54
      error ("Bad blank character " ^ quote (ML_Syntax.print_char c) ^ Position.here pos));
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    55
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    56
fun is_space ((s, _): Symbol_Pos.T) = s = Symbol.space;
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    57
val is_empty = forall (fn Antiquote.Text ss => forall is_space ss | _ => false);
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    58
61444
1fcdfc1a7e50 more document structure;
wenzelm
parents: 61443
diff changeset
    59
val scan_marker =
1fcdfc1a7e50 more document structure;
wenzelm
parents: 61443
diff changeset
    60
  Scan.many is_space --
61445
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    61
  (Symbol_Pos.$$ "\<^item>" >> K Itemize ||
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    62
   Symbol_Pos.$$ "\<^enum>" >> K Enumerate ||
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    63
   Symbol_Pos.$$ "\<^descr>" >> K Description) >> (fn (a, b) => {indent = length a, kind = b});
61442
467ebb937294 clarified;
wenzelm
parents: 61441
diff changeset
    64
61444
1fcdfc1a7e50 more document structure;
wenzelm
parents: 61443
diff changeset
    65
fun read_marker (Antiquote.Text ss :: _) =
61445
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    66
      #1 (Scan.finite Symbol_Pos.stopper (Scan.option (scan_marker -- Symbol_Pos.scan_pos)) ss)
61444
1fcdfc1a7e50 more document structure;
wenzelm
parents: 61443
diff changeset
    67
  | read_marker _ = NONE;
61441
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    68
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    69
in
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    70
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    71
fun make_line content =
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    72
  let
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    73
    val _ = check_blanks content;
61444
1fcdfc1a7e50 more document structure;
wenzelm
parents: 61443
diff changeset
    74
    val marker = read_marker content;
1fcdfc1a7e50 more document structure;
wenzelm
parents: 61443
diff changeset
    75
  in Line {content = content, is_empty = is_empty content, marker = marker} end;
61441
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    76
61445
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    77
val empty_line = make_line [];
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    78
61441
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    79
end;
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    80
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    81
61445
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    82
(* document blocks *)
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    83
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    84
datatype block = Paragraph of line list | List of marker * block list;
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    85
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    86
fun add_span (opt_marker, body) document =
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    87
  (case (opt_marker, document) of
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    88
    (SOME marker, List (list_marker, list_body) :: rest) =>
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    89
      if marker = list_marker then List (list_marker, body @ list_body) :: rest
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    90
      else List (marker, body) :: document
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    91
  | (SOME marker, _) => List (marker, body) :: document
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    92
  | (NONE, _) => body @ document);
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    93
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
    94
61443
78bbfadd1034 more document structure;
wenzelm
parents: 61442
diff changeset
    95
(* read document *)
61441
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    96
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    97
local
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
    98
61442
467ebb937294 clarified;
wenzelm
parents: 61441
diff changeset
    99
fun plain_line line =
61445
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
   100
  not (line_is_empty line) andalso is_none (line_marker line) andalso line <> eof_line;
61441
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
   101
61444
1fcdfc1a7e50 more document structure;
wenzelm
parents: 61443
diff changeset
   102
val parse_paragraph = Scan.many1 plain_line >> Paragraph;
1fcdfc1a7e50 more document structure;
wenzelm
parents: 61443
diff changeset
   103
61441
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
   104
val parse_span =
61444
1fcdfc1a7e50 more document structure;
wenzelm
parents: 61443
diff changeset
   105
  parse_paragraph >> (fn par => (NONE, [par])) ||
1fcdfc1a7e50 more document structure;
wenzelm
parents: 61443
diff changeset
   106
  Scan.one (is_some o line_marker) -- Scan.many plain_line --
1fcdfc1a7e50 more document structure;
wenzelm
parents: 61443
diff changeset
   107
    Scan.repeat (Scan.one line_is_empty |-- parse_paragraph) >>
61445
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
   108
      (fn ((line, lines), pars) =>
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
   109
        (Option.map #1 (line_marker line), Paragraph (line :: lines) :: pars));
61443
78bbfadd1034 more document structure;
wenzelm
parents: 61442
diff changeset
   110
78bbfadd1034 more document structure;
wenzelm
parents: 61442
diff changeset
   111
val parse_document =
61445
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
   112
  parse_span ::: Scan.repeat (Scan.option (Scan.one line_is_empty) |-- parse_span)
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
   113
    >> (fn spans => fold_rev add_span spans []);
61441
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
   114
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
   115
in
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
   116
61445
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
   117
val read_lines =
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
   118
  Scan.read (Scan.stopper (K eof_line) (fn line => line = eof_line))
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
   119
    (Scan.repeat (Scan.many line_is_empty |-- parse_document) --| Scan.many line_is_empty) #>
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
   120
  the_default [] #> flat;
61441
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
   121
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
   122
end;
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
   123
61445
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
   124
val read = Antiquote.read #> Antiquote.split_lines #> map make_line #> read_lines;
31aadb15eda5 more document structure;
wenzelm
parents: 61444
diff changeset
   125
61441
20ff1d5c74e1 minimal support for Markdown documents;
wenzelm
parents:
diff changeset
   126
end;