| author | wenzelm | 
| Thu, 28 Mar 2013 15:00:27 +0100 | |
| changeset 51569 | 4e084727faae | 
| parent 48996 | a8bad1369ada | 
| child 55551 | 4a5f65df29fa | 
| 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 | |
| 43774 
6dfdb70496fe
added Outer_Syntax.quote_string, which is conceptually a bit different from Token.unparse;
 wenzelm parents: 
43746diff
changeset | 19 | val X = '\5' | 
| 
6dfdb70496fe
added Outer_Syntax.quote_string, which is conceptually a bit different from Token.unparse;
 wenzelm parents: 
43746diff
changeset | 20 | val Y = '\6' | 
| 
6dfdb70496fe
added Outer_Syntax.quote_string, which is conceptually a bit different from Token.unparse;
 wenzelm parents: 
43746diff
changeset | 21 | val X_string = X.toString | 
| 
6dfdb70496fe
added Outer_Syntax.quote_string, which is conceptually a bit different from Token.unparse;
 wenzelm parents: 
43746diff
changeset | 22 | val Y_string = Y.toString | 
| 27930 | 23 | |
| 43782 | 24 | def detect(s: String): Boolean = s.exists(c => c == X || c == Y) | 
| 25 | ||
| 27930 | 26 | |
| 38268 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 27 | /* 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 | 28 | |
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 29 | def string_of_body(body: XML.Body): String = | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 30 |   {
 | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 31 | val s = new StringBuilder | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 32 |     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 | 33 | def tree(t: XML.Tree): Unit = | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 34 |       t match {
 | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 35 | case XML.Elem(Markup(name, atts), ts) => | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 36 | s += X; s += Y; s++= name; atts.foreach(attrib); s += X | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 37 | ts.foreach(tree) | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 38 | s += X; s += Y; s += X | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 39 | case XML.Text(text) => s ++= text | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 40 | } | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 41 | body.foreach(tree) | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 42 | s.toString | 
| 
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 | |
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 45 | 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 | 46 | |
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 47 | |
| 27930 | 48 | /* parsing */ | 
| 49 | ||
| 27993 
6dd90ef9f927
simplified exceptions: use plain error function / RuntimeException;
 wenzelm parents: 
27971diff
changeset | 50 |   private def err(msg: String) = error("Malformed YXML: " + msg)
 | 
| 27930 | 51 |   private def err_attribute() = err("bad attribute")
 | 
| 52 |   private def err_element() = err("bad element")
 | |
| 53 | private def err_unbalanced(name: String) = | |
| 54 |     if (name == "") err("unbalanced element")
 | |
| 44181 | 55 |     else err("unbalanced element " + quote(name))
 | 
| 27930 | 56 | |
| 27944 
2bf3f30558ed
parse_attrib: more efficient due to indexOf('=');
 wenzelm parents: 
27943diff
changeset | 57 |   private def parse_attrib(source: CharSequence) = {
 | 
| 
2bf3f30558ed
parse_attrib: more efficient due to indexOf('=');
 wenzelm parents: 
27943diff
changeset | 58 | val s = source.toString | 
| 
2bf3f30558ed
parse_attrib: more efficient due to indexOf('=');
 wenzelm parents: 
27943diff
changeset | 59 |     val i = s.indexOf('=')
 | 
| 
2bf3f30558ed
parse_attrib: more efficient due to indexOf('=');
 wenzelm parents: 
27943diff
changeset | 60 | if (i <= 0) err_attribute() | 
| 38234 
4b610fbb2d83
YXML.parse: refrain from interning, let XML.Cache do it (partially);
 wenzelm parents: 
38231diff
changeset | 61 | (s.substring(0, i), s.substring(i + 1)) | 
| 27944 
2bf3f30558ed
parse_attrib: more efficient due to indexOf('=');
 wenzelm parents: 
27943diff
changeset | 62 | } | 
| 27930 | 63 | |
| 64 | ||
| 38267 
e50c283dd125
type XML.Body as basic data representation language (Scala version);
 wenzelm parents: 
38234diff
changeset | 65 | def parse_body(source: CharSequence): XML.Body = | 
| 32450 | 66 |   {
 | 
| 27930 | 67 | /* stack operations */ | 
| 68 | ||
| 40453 | 69 | def buffer(): mutable.ListBuffer[XML.Tree] = new mutable.ListBuffer[XML.Tree] | 
| 70 | var stack: List[(Markup, mutable.ListBuffer[XML.Tree])] = List((Markup.Empty, buffer())) | |
| 27930 | 71 | |
| 36684 | 72 | def add(x: XML.Tree) | 
| 73 |     {
 | |
| 74 |       (stack: @unchecked) match { case ((_, body) :: _) => body += x }
 | |
| 27930 | 75 | } | 
| 76 | ||
| 36684 | 77 | def push(name: String, atts: XML.Attributes) | 
| 78 |     {
 | |
| 27930 | 79 | if (name == "") err_element() | 
| 38230 
ed147003de4b
simplified type XML.Tree: embed Markup directly, avoid slightly odd triple;
 wenzelm parents: 
36685diff
changeset | 80 | else stack = (Markup(name, atts), buffer()) :: stack | 
| 36684 | 81 | } | 
| 27930 | 82 | |
| 36684 | 83 | def pop() | 
| 84 |     {
 | |
| 85 |       (stack: @unchecked) match {
 | |
| 38475 | 86 |         case ((Markup.Empty, _) :: _) => err_unbalanced("")
 | 
| 38230 
ed147003de4b
simplified type XML.Tree: embed Markup directly, avoid slightly odd triple;
 wenzelm parents: 
36685diff
changeset | 87 | case ((markup, body) :: pending) => | 
| 
ed147003de4b
simplified type XML.Tree: embed Markup directly, avoid slightly odd triple;
 wenzelm parents: 
36685diff
changeset | 88 | stack = pending | 
| 
ed147003de4b
simplified type XML.Tree: embed Markup directly, avoid slightly odd triple;
 wenzelm parents: 
36685diff
changeset | 89 | add(XML.Elem(markup, body.toList)) | 
| 36684 | 90 | } | 
| 27930 | 91 | } | 
| 92 | ||
| 93 | ||
| 94 | /* parse chunks */ | |
| 95 | ||
| 48996 | 96 |     for (chunk <- Library.separated_chunks(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 | 97 | if (chunk.length == 1 && chunk.charAt(0) == Y) pop() | 
| 27945 
d2dc5a1903e8
tuned parse performance: avoid splitting terminal Y chunk;
 wenzelm parents: 
27944diff
changeset | 98 |       else {
 | 
| 48996 | 99 |         Library.separated_chunks(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 | 100 | 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 | 101 | push(name.toString, atts.map(parse_attrib)) | 
| 27945 
d2dc5a1903e8
tuned parse performance: avoid splitting terminal Y chunk;
 wenzelm parents: 
27944diff
changeset | 102 | case txts => for (txt <- txts) add(XML.Text(txt.toString)) | 
| 
d2dc5a1903e8
tuned parse performance: avoid splitting terminal Y chunk;
 wenzelm parents: 
27944diff
changeset | 103 | } | 
| 27930 | 104 | } | 
| 105 | } | |
| 42719 | 106 |     (stack: @unchecked) match {
 | 
| 38475 | 107 | case List((Markup.Empty, body)) => body.toList | 
| 38230 
ed147003de4b
simplified type XML.Tree: embed Markup directly, avoid slightly odd triple;
 wenzelm parents: 
36685diff
changeset | 108 | case (Markup(name, _), _) :: _ => err_unbalanced(name) | 
| 27930 | 109 | } | 
| 110 | } | |
| 111 | ||
| 32450 | 112 | def parse(source: CharSequence): XML.Tree = | 
| 27930 | 113 |     parse_body(source) match {
 | 
| 114 | case List(result) => result | |
| 115 |       case Nil => XML.Text("")
 | |
| 116 |       case _ => err("multiple results")
 | |
| 117 | } | |
| 27960 | 118 | |
| 29521 | 119 | |
| 120 | /* failsafe parsing */ | |
| 121 | ||
| 45666 | 122 | private def markup_broken(source: CharSequence) = | 
| 123 | XML.elem(Markup.Broken.name, List(XML.Text(source.toString))) | |
| 29521 | 124 | |
| 38267 
e50c283dd125
type XML.Body as basic data representation language (Scala version);
 wenzelm parents: 
38234diff
changeset | 125 | def parse_body_failsafe(source: CharSequence): XML.Body = | 
| 32450 | 126 |   {
 | 
| 29521 | 127 |     try { parse_body(source) }
 | 
| 45666 | 128 |     catch { case ERROR(_) => List(markup_broken(source)) }
 | 
| 29521 | 129 | } | 
| 130 | ||
| 32450 | 131 | def parse_failsafe(source: CharSequence): XML.Tree = | 
| 132 |   {
 | |
| 27960 | 133 |     try { parse(source) }
 | 
| 45666 | 134 |     catch { case ERROR(_) => markup_broken(source) }
 | 
| 27960 | 135 | } | 
| 27930 | 136 | } |