| author | wenzelm | 
| Mon, 21 Mar 2016 20:22:07 +0100 | |
| changeset 62681 | 45b8dd2d3827 | 
| parent 56661 | ef623f6f036b | 
| child 64370 | 865b39487b5d | 
| permissions | -rw-r--r-- | 
| 44698 | 1 | /* Title: Pure/PIDE/yxml.scala | 
| 45673 
cd41e3903fbf
separate compilation of PIDE vs. Pure sources, which enables independent Scala library;
 wenzelm parents: 
45667diff
changeset | 2 | Module: PIDE | 
| 27930 | 3 | Author: Makarius | 
| 4 | ||
| 44698 | 5 | Efficient text representation of XML trees. Suitable for direct | 
| 6 | inlining into plain text. | |
| 27930 | 7 | */ | 
| 8 | ||
| 9 | package isabelle | |
| 10 | ||
| 11 | ||
| 40453 | 12 | import scala.collection.mutable | 
| 36684 | 13 | |
| 14 | ||
| 32450 | 15 | object YXML | 
| 16 | {
 | |
| 27943 
f34ff5e7728f
replaced Pattern.split by chunks iterator (more efficient, resembles ML version more closely);
 wenzelm parents: 
27930diff
changeset | 17 | /* chunk markers */ | 
| 27930 | 18 | |
| 56661 
ef623f6f036b
avoid octal escape literals -- deprecated in scala-2.11.0;
 wenzelm parents: 
56600diff
changeset | 19 | val X = '\u0005' | 
| 
ef623f6f036b
avoid octal escape literals -- deprecated in scala-2.11.0;
 wenzelm parents: 
56600diff
changeset | 20 | val Y = '\u0006' | 
| 56600 | 21 | |
| 22 | val is_X = (c: Char) => c == X | |
| 23 | val is_Y = (c: Char) => c == Y | |
| 24 | ||
| 43774 
6dfdb70496fe
added Outer_Syntax.quote_string, which is conceptually a bit different from Token.unparse;
 wenzelm parents: 
43746diff
changeset | 25 | val X_string = X.toString | 
| 
6dfdb70496fe
added Outer_Syntax.quote_string, which is conceptually a bit different from Token.unparse;
 wenzelm parents: 
43746diff
changeset | 26 | val Y_string = Y.toString | 
| 27930 | 27 | |
| 43782 | 28 | def detect(s: String): Boolean = s.exists(c => c == X || c == Y) | 
| 29 | ||
| 27930 | 30 | |
| 38268 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 31 | /* 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 | 32 | |
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 33 | def string_of_body(body: XML.Body): String = | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 34 |   {
 | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 35 | val s = new StringBuilder | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 36 |     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 | 37 | def tree(t: XML.Tree): Unit = | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 38 |       t match {
 | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 39 | case XML.Elem(Markup(name, atts), ts) => | 
| 55554 | 40 | 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 | 41 | ts.foreach(tree) | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 42 | s += X; s += Y; s += X | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 43 | case XML.Text(text) => s ++= text | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 44 | } | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 45 | body.foreach(tree) | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 46 | s.toString | 
| 
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 | |
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 49 | 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 | 50 | |
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 51 | |
| 27930 | 52 | /* parsing */ | 
| 53 | ||
| 27993 
6dd90ef9f927
simplified exceptions: use plain error function / RuntimeException;
 wenzelm parents: 
27971diff
changeset | 54 |   private def err(msg: String) = error("Malformed YXML: " + msg)
 | 
| 27930 | 55 |   private def err_attribute() = err("bad attribute")
 | 
| 56 |   private def err_element() = err("bad element")
 | |
| 57 | private def err_unbalanced(name: String) = | |
| 58 |     if (name == "") err("unbalanced element")
 | |
| 44181 | 59 |     else err("unbalanced element " + quote(name))
 | 
| 27930 | 60 | |
| 27944 
2bf3f30558ed
parse_attrib: more efficient due to indexOf('=');
 wenzelm parents: 
27943diff
changeset | 61 |   private def parse_attrib(source: CharSequence) = {
 | 
| 
2bf3f30558ed
parse_attrib: more efficient due to indexOf('=');
 wenzelm parents: 
27943diff
changeset | 62 | val s = source.toString | 
| 
2bf3f30558ed
parse_attrib: more efficient due to indexOf('=');
 wenzelm parents: 
27943diff
changeset | 63 |     val i = s.indexOf('=')
 | 
| 
2bf3f30558ed
parse_attrib: more efficient due to indexOf('=');
 wenzelm parents: 
27943diff
changeset | 64 | if (i <= 0) err_attribute() | 
| 38234 
4b610fbb2d83
YXML.parse: refrain from interning, let XML.Cache do it (partially);
 wenzelm parents: 
38231diff
changeset | 65 | (s.substring(0, i), s.substring(i + 1)) | 
| 27944 
2bf3f30558ed
parse_attrib: more efficient due to indexOf('=');
 wenzelm parents: 
27943diff
changeset | 66 | } | 
| 27930 | 67 | |
| 68 | ||
| 38267 
e50c283dd125
type XML.Body as basic data representation language (Scala version);
 wenzelm parents: 
38234diff
changeset | 69 | def parse_body(source: CharSequence): XML.Body = | 
| 32450 | 70 |   {
 | 
| 27930 | 71 | /* stack operations */ | 
| 72 | ||
| 40453 | 73 | def buffer(): mutable.ListBuffer[XML.Tree] = new mutable.ListBuffer[XML.Tree] | 
| 74 | var stack: List[(Markup, mutable.ListBuffer[XML.Tree])] = List((Markup.Empty, buffer())) | |
| 27930 | 75 | |
| 36684 | 76 | def add(x: XML.Tree) | 
| 77 |     {
 | |
| 78 |       (stack: @unchecked) match { case ((_, body) :: _) => body += x }
 | |
| 27930 | 79 | } | 
| 80 | ||
| 36684 | 81 | def push(name: String, atts: XML.Attributes) | 
| 82 |     {
 | |
| 27930 | 83 | if (name == "") err_element() | 
| 38230 
ed147003de4b
simplified type XML.Tree: embed Markup directly, avoid slightly odd triple;
 wenzelm parents: 
36685diff
changeset | 84 | else stack = (Markup(name, atts), buffer()) :: stack | 
| 36684 | 85 | } | 
| 27930 | 86 | |
| 36684 | 87 | def pop() | 
| 88 |     {
 | |
| 89 |       (stack: @unchecked) match {
 | |
| 38475 | 90 |         case ((Markup.Empty, _) :: _) => err_unbalanced("")
 | 
| 38230 
ed147003de4b
simplified type XML.Tree: embed Markup directly, avoid slightly odd triple;
 wenzelm parents: 
36685diff
changeset | 91 | case ((markup, body) :: pending) => | 
| 
ed147003de4b
simplified type XML.Tree: embed Markup directly, avoid slightly odd triple;
 wenzelm parents: 
36685diff
changeset | 92 | stack = pending | 
| 
ed147003de4b
simplified type XML.Tree: embed Markup directly, avoid slightly odd triple;
 wenzelm parents: 
36685diff
changeset | 93 | add(XML.Elem(markup, body.toList)) | 
| 36684 | 94 | } | 
| 27930 | 95 | } | 
| 96 | ||
| 97 | ||
| 98 | /* parse chunks */ | |
| 99 | ||
| 56600 | 100 |     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 | 101 | if (chunk.length == 1 && chunk.charAt(0) == Y) pop() | 
| 27945 
d2dc5a1903e8
tuned parse performance: avoid splitting terminal Y chunk;
 wenzelm parents: 
27944diff
changeset | 102 |       else {
 | 
| 56600 | 103 |         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 | 104 | 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 | 105 | push(name.toString, atts.map(parse_attrib)) | 
| 27945 
d2dc5a1903e8
tuned parse performance: avoid splitting terminal Y chunk;
 wenzelm parents: 
27944diff
changeset | 106 | case txts => for (txt <- txts) add(XML.Text(txt.toString)) | 
| 
d2dc5a1903e8
tuned parse performance: avoid splitting terminal Y chunk;
 wenzelm parents: 
27944diff
changeset | 107 | } | 
| 27930 | 108 | } | 
| 109 | } | |
| 42719 | 110 |     (stack: @unchecked) match {
 | 
| 38475 | 111 | case List((Markup.Empty, body)) => body.toList | 
| 38230 
ed147003de4b
simplified type XML.Tree: embed Markup directly, avoid slightly odd triple;
 wenzelm parents: 
36685diff
changeset | 112 | case (Markup(name, _), _) :: _ => err_unbalanced(name) | 
| 27930 | 113 | } | 
| 114 | } | |
| 115 | ||
| 32450 | 116 | def parse(source: CharSequence): XML.Tree = | 
| 27930 | 117 |     parse_body(source) match {
 | 
| 118 | case List(result) => result | |
| 119 |       case Nil => XML.Text("")
 | |
| 120 |       case _ => err("multiple results")
 | |
| 121 | } | |
| 27960 | 122 | |
| 29521 | 123 | |
| 124 | /* failsafe parsing */ | |
| 125 | ||
| 45666 | 126 | private def markup_broken(source: CharSequence) = | 
| 55551 | 127 | XML.Elem(Markup.Broken, List(XML.Text(source.toString))) | 
| 29521 | 128 | |
| 38267 
e50c283dd125
type XML.Body as basic data representation language (Scala version);
 wenzelm parents: 
38234diff
changeset | 129 | def parse_body_failsafe(source: CharSequence): XML.Body = | 
| 32450 | 130 |   {
 | 
| 29521 | 131 |     try { parse_body(source) }
 | 
| 45666 | 132 |     catch { case ERROR(_) => List(markup_broken(source)) }
 | 
| 29521 | 133 | } | 
| 134 | ||
| 32450 | 135 | def parse_failsafe(source: CharSequence): XML.Tree = | 
| 136 |   {
 | |
| 27960 | 137 |     try { parse(source) }
 | 
| 45666 | 138 |     catch { case ERROR(_) => markup_broken(source) }
 | 
| 27960 | 139 | } | 
| 27930 | 140 | } |