| author | wenzelm | 
| Sun, 25 Feb 2018 15:44:46 +0100 | |
| changeset 67721 | 5348bea4accd | 
| parent 64370 | 865b39487b5d | 
| child 67820 | e30d6368c7c8 | 
| permissions | -rw-r--r-- | 
| 44698 | 1 | /* Title: Pure/PIDE/yxml.scala | 
| 27930 | 2 | Author: Makarius | 
| 3 | ||
| 44698 | 4 | Efficient text representation of XML trees. Suitable for direct | 
| 5 | inlining into plain text. | |
| 27930 | 6 | */ | 
| 7 | ||
| 8 | package isabelle | |
| 9 | ||
| 10 | ||
| 40453 | 11 | import scala.collection.mutable | 
| 36684 | 12 | |
| 13 | ||
| 32450 | 14 | object YXML | 
| 15 | {
 | |
| 27943 
f34ff5e7728f
replaced Pattern.split by chunks iterator (more efficient, resembles ML version more closely);
 wenzelm parents: 
27930diff
changeset | 16 | /* chunk markers */ | 
| 27930 | 17 | |
| 56661 
ef623f6f036b
avoid octal escape literals -- deprecated in scala-2.11.0;
 wenzelm parents: 
56600diff
changeset | 18 | val X = '\u0005' | 
| 
ef623f6f036b
avoid octal escape literals -- deprecated in scala-2.11.0;
 wenzelm parents: 
56600diff
changeset | 19 | val Y = '\u0006' | 
| 56600 | 20 | |
| 21 | val is_X = (c: Char) => c == X | |
| 22 | val is_Y = (c: Char) => c == Y | |
| 23 | ||
| 43774 
6dfdb70496fe
added Outer_Syntax.quote_string, which is conceptually a bit different from Token.unparse;
 wenzelm parents: 
43746diff
changeset | 24 | val X_string = X.toString | 
| 
6dfdb70496fe
added Outer_Syntax.quote_string, which is conceptually a bit different from Token.unparse;
 wenzelm parents: 
43746diff
changeset | 25 | val Y_string = Y.toString | 
| 27930 | 26 | |
| 43782 | 27 | def detect(s: String): Boolean = s.exists(c => c == X || c == Y) | 
| 28 | ||
| 27930 | 29 | |
| 38268 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 30 | /* string representation */ // FIXME byte array version with pseudo-utf-8 (!?) | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 31 | |
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 32 | def string_of_body(body: XML.Body): String = | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 33 |   {
 | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 34 | val s = new StringBuilder | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 35 |     def attrib(p: (String, String)) { s += Y; s ++= p._1; s += '='; s ++= p._2 }
 | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 36 | def tree(t: XML.Tree): Unit = | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 37 |       t match {
 | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 38 | case XML.Elem(Markup(name, atts), ts) => | 
| 55554 | 39 | s += X; s += Y; s ++= name; atts.foreach(attrib); s += X | 
| 38268 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 40 | ts.foreach(tree) | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 41 | s += X; s += Y; s += X | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 42 | case XML.Text(text) => s ++= text | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 43 | } | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 44 | body.foreach(tree) | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 45 | s.toString | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 46 | } | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 47 | |
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 48 | def string_of_tree(tree: XML.Tree): String = string_of_body(List(tree)) | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 49 | |
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 50 | |
| 27930 | 51 | /* parsing */ | 
| 52 | ||
| 27993 
6dd90ef9f927
simplified exceptions: use plain error function / RuntimeException;
 wenzelm parents: 
27971diff
changeset | 53 |   private def err(msg: String) = error("Malformed YXML: " + msg)
 | 
| 27930 | 54 |   private def err_attribute() = err("bad attribute")
 | 
| 55 |   private def err_element() = err("bad element")
 | |
| 56 | private def err_unbalanced(name: String) = | |
| 57 |     if (name == "") err("unbalanced element")
 | |
| 44181 | 58 |     else err("unbalanced element " + quote(name))
 | 
| 27930 | 59 | |
| 27944 
2bf3f30558ed
parse_attrib: more efficient due to indexOf('=');
 wenzelm parents: 
27943diff
changeset | 60 |   private def parse_attrib(source: CharSequence) = {
 | 
| 
2bf3f30558ed
parse_attrib: more efficient due to indexOf('=');
 wenzelm parents: 
27943diff
changeset | 61 | val s = source.toString | 
| 
2bf3f30558ed
parse_attrib: more efficient due to indexOf('=');
 wenzelm parents: 
27943diff
changeset | 62 |     val i = s.indexOf('=')
 | 
| 
2bf3f30558ed
parse_attrib: more efficient due to indexOf('=');
 wenzelm parents: 
27943diff
changeset | 63 | if (i <= 0) err_attribute() | 
| 38234 
4b610fbb2d83
YXML.parse: refrain from interning, let XML.Cache do it (partially);
 wenzelm parents: 
38231diff
changeset | 64 | (s.substring(0, i), s.substring(i + 1)) | 
| 27944 
2bf3f30558ed
parse_attrib: more efficient due to indexOf('=');
 wenzelm parents: 
27943diff
changeset | 65 | } | 
| 27930 | 66 | |
| 67 | ||
| 38267 
e50c283dd125
type XML.Body as basic data representation language (Scala version);
 wenzelm parents: 
38234diff
changeset | 68 | def parse_body(source: CharSequence): XML.Body = | 
| 32450 | 69 |   {
 | 
| 27930 | 70 | /* stack operations */ | 
| 71 | ||
| 40453 | 72 | def buffer(): mutable.ListBuffer[XML.Tree] = new mutable.ListBuffer[XML.Tree] | 
| 73 | var stack: List[(Markup, mutable.ListBuffer[XML.Tree])] = List((Markup.Empty, buffer())) | |
| 27930 | 74 | |
| 36684 | 75 | def add(x: XML.Tree) | 
| 76 |     {
 | |
| 77 |       (stack: @unchecked) match { case ((_, body) :: _) => body += x }
 | |
| 27930 | 78 | } | 
| 79 | ||
| 36684 | 80 | def push(name: String, atts: XML.Attributes) | 
| 81 |     {
 | |
| 27930 | 82 | if (name == "") err_element() | 
| 38230 
ed147003de4b
simplified type XML.Tree: embed Markup directly, avoid slightly odd triple;
 wenzelm parents: 
36685diff
changeset | 83 | else stack = (Markup(name, atts), buffer()) :: stack | 
| 36684 | 84 | } | 
| 27930 | 85 | |
| 36684 | 86 | def pop() | 
| 87 |     {
 | |
| 88 |       (stack: @unchecked) match {
 | |
| 38475 | 89 |         case ((Markup.Empty, _) :: _) => err_unbalanced("")
 | 
| 38230 
ed147003de4b
simplified type XML.Tree: embed Markup directly, avoid slightly odd triple;
 wenzelm parents: 
36685diff
changeset | 90 | case ((markup, body) :: pending) => | 
| 
ed147003de4b
simplified type XML.Tree: embed Markup directly, avoid slightly odd triple;
 wenzelm parents: 
36685diff
changeset | 91 | stack = pending | 
| 
ed147003de4b
simplified type XML.Tree: embed Markup directly, avoid slightly odd triple;
 wenzelm parents: 
36685diff
changeset | 92 | add(XML.Elem(markup, body.toList)) | 
| 36684 | 93 | } | 
| 27930 | 94 | } | 
| 95 | ||
| 96 | ||
| 97 | /* parse chunks */ | |
| 98 | ||
| 56600 | 99 |     for (chunk <- Library.separated_chunks(is_X, source) if chunk.length != 0) {
 | 
| 34099 
2541de190d92
added decode_chars, with raw character view on byte buffer and adhoc decoding via toString;
 wenzelm parents: 
32450diff
changeset | 100 | if (chunk.length == 1 && chunk.charAt(0) == Y) pop() | 
| 27945 
d2dc5a1903e8
tuned parse performance: avoid splitting terminal Y chunk;
 wenzelm parents: 
27944diff
changeset | 101 |       else {
 | 
| 56600 | 102 |         Library.separated_chunks(is_Y, chunk).toList match {
 | 
| 34099 
2541de190d92
added decode_chars, with raw character view on byte buffer and adhoc decoding via toString;
 wenzelm parents: 
32450diff
changeset | 103 | case ch :: name :: atts if ch.length == 0 => | 
| 38234 
4b610fbb2d83
YXML.parse: refrain from interning, let XML.Cache do it (partially);
 wenzelm parents: 
38231diff
changeset | 104 | push(name.toString, atts.map(parse_attrib)) | 
| 27945 
d2dc5a1903e8
tuned parse performance: avoid splitting terminal Y chunk;
 wenzelm parents: 
27944diff
changeset | 105 | case txts => for (txt <- txts) add(XML.Text(txt.toString)) | 
| 
d2dc5a1903e8
tuned parse performance: avoid splitting terminal Y chunk;
 wenzelm parents: 
27944diff
changeset | 106 | } | 
| 27930 | 107 | } | 
| 108 | } | |
| 42719 | 109 |     (stack: @unchecked) match {
 | 
| 38475 | 110 | case List((Markup.Empty, body)) => body.toList | 
| 38230 
ed147003de4b
simplified type XML.Tree: embed Markup directly, avoid slightly odd triple;
 wenzelm parents: 
36685diff
changeset | 111 | case (Markup(name, _), _) :: _ => err_unbalanced(name) | 
| 27930 | 112 | } | 
| 113 | } | |
| 114 | ||
| 32450 | 115 | def parse(source: CharSequence): XML.Tree = | 
| 27930 | 116 |     parse_body(source) match {
 | 
| 117 | case List(result) => result | |
| 118 |       case Nil => XML.Text("")
 | |
| 119 |       case _ => err("multiple results")
 | |
| 120 | } | |
| 27960 | 121 | |
| 29521 | 122 | |
| 123 | /* failsafe parsing */ | |
| 124 | ||
| 45666 | 125 | private def markup_broken(source: CharSequence) = | 
| 55551 | 126 | XML.Elem(Markup.Broken, List(XML.Text(source.toString))) | 
| 29521 | 127 | |
| 38267 
e50c283dd125
type XML.Body as basic data representation language (Scala version);
 wenzelm parents: 
38234diff
changeset | 128 | def parse_body_failsafe(source: CharSequence): XML.Body = | 
| 32450 | 129 |   {
 | 
| 29521 | 130 |     try { parse_body(source) }
 | 
| 45666 | 131 |     catch { case ERROR(_) => List(markup_broken(source)) }
 | 
| 29521 | 132 | } | 
| 133 | ||
| 32450 | 134 | def parse_failsafe(source: CharSequence): XML.Tree = | 
| 135 |   {
 | |
| 27960 | 136 |     try { parse(source) }
 | 
| 45666 | 137 |     catch { case ERROR(_) => markup_broken(source) }
 | 
| 27960 | 138 | } | 
| 27930 | 139 | } |