| author | wenzelm | 
| Wed, 27 Feb 2019 16:28:46 +0100 | |
| changeset 69844 | b21ddfa7042b | 
| parent 67820 | e30d6368c7c8 | 
| child 71534 | f10bffaa2bae | 
| 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 | 
| 67820 
e30d6368c7c8
clarified argument formats: explicit Unit, allow XML.Elem as well;
 wenzelm parents: 
64370diff
changeset | 26 | val XY_string = X_string + Y_string | 
| 27930 | 27 | |
| 43782 | 28 | def detect(s: String): Boolean = s.exists(c => c == X || c == Y) | 
| 67820 
e30d6368c7c8
clarified argument formats: explicit Unit, allow XML.Elem as well;
 wenzelm parents: 
64370diff
changeset | 29 | def detect_elem(s: String): Boolean = s.startsWith(XY_string) | 
| 43782 | 30 | |
| 27930 | 31 | |
| 38268 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 32 | /* 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 | 33 | |
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 34 | def string_of_body(body: XML.Body): String = | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 35 |   {
 | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 36 | val s = new StringBuilder | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 37 |     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 | 38 | def tree(t: XML.Tree): Unit = | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 39 |       t match {
 | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 40 | case XML.Elem(Markup(name, atts), ts) => | 
| 55554 | 41 | 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 | 42 | ts.foreach(tree) | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 43 | s += X; s += Y; s += X | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 44 | case XML.Text(text) => s ++= text | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 45 | } | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 46 | body.foreach(tree) | 
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 47 | s.toString | 
| 
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 | |
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 50 | 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 | 51 | |
| 
beb86b805590
more uniform XML/YXML string_of_body/string_of_tree;
 wenzelm parents: 
38267diff
changeset | 52 | |
| 27930 | 53 | /* parsing */ | 
| 54 | ||
| 27993 
6dd90ef9f927
simplified exceptions: use plain error function / RuntimeException;
 wenzelm parents: 
27971diff
changeset | 55 |   private def err(msg: String) = error("Malformed YXML: " + msg)
 | 
| 27930 | 56 |   private def err_attribute() = err("bad attribute")
 | 
| 57 |   private def err_element() = err("bad element")
 | |
| 58 | private def err_unbalanced(name: String) = | |
| 59 |     if (name == "") err("unbalanced element")
 | |
| 44181 | 60 |     else err("unbalanced element " + quote(name))
 | 
| 27930 | 61 | |
| 27944 
2bf3f30558ed
parse_attrib: more efficient due to indexOf('=');
 wenzelm parents: 
27943diff
changeset | 62 |   private def parse_attrib(source: CharSequence) = {
 | 
| 
2bf3f30558ed
parse_attrib: more efficient due to indexOf('=');
 wenzelm parents: 
27943diff
changeset | 63 | val s = source.toString | 
| 
2bf3f30558ed
parse_attrib: more efficient due to indexOf('=');
 wenzelm parents: 
27943diff
changeset | 64 |     val i = s.indexOf('=')
 | 
| 
2bf3f30558ed
parse_attrib: more efficient due to indexOf('=');
 wenzelm parents: 
27943diff
changeset | 65 | if (i <= 0) err_attribute() | 
| 38234 
4b610fbb2d83
YXML.parse: refrain from interning, let XML.Cache do it (partially);
 wenzelm parents: 
38231diff
changeset | 66 | (s.substring(0, i), s.substring(i + 1)) | 
| 27944 
2bf3f30558ed
parse_attrib: more efficient due to indexOf('=');
 wenzelm parents: 
27943diff
changeset | 67 | } | 
| 27930 | 68 | |
| 69 | ||
| 38267 
e50c283dd125
type XML.Body as basic data representation language (Scala version);
 wenzelm parents: 
38234diff
changeset | 70 | def parse_body(source: CharSequence): XML.Body = | 
| 32450 | 71 |   {
 | 
| 27930 | 72 | /* stack operations */ | 
| 73 | ||
| 40453 | 74 | def buffer(): mutable.ListBuffer[XML.Tree] = new mutable.ListBuffer[XML.Tree] | 
| 75 | var stack: List[(Markup, mutable.ListBuffer[XML.Tree])] = List((Markup.Empty, buffer())) | |
| 27930 | 76 | |
| 36684 | 77 | def add(x: XML.Tree) | 
| 78 |     {
 | |
| 79 |       (stack: @unchecked) match { case ((_, body) :: _) => body += x }
 | |
| 27930 | 80 | } | 
| 81 | ||
| 36684 | 82 | def push(name: String, atts: XML.Attributes) | 
| 83 |     {
 | |
| 27930 | 84 | if (name == "") err_element() | 
| 38230 
ed147003de4b
simplified type XML.Tree: embed Markup directly, avoid slightly odd triple;
 wenzelm parents: 
36685diff
changeset | 85 | else stack = (Markup(name, atts), buffer()) :: stack | 
| 36684 | 86 | } | 
| 27930 | 87 | |
| 36684 | 88 | def pop() | 
| 89 |     {
 | |
| 90 |       (stack: @unchecked) match {
 | |
| 38475 | 91 |         case ((Markup.Empty, _) :: _) => err_unbalanced("")
 | 
| 38230 
ed147003de4b
simplified type XML.Tree: embed Markup directly, avoid slightly odd triple;
 wenzelm parents: 
36685diff
changeset | 92 | case ((markup, body) :: pending) => | 
| 
ed147003de4b
simplified type XML.Tree: embed Markup directly, avoid slightly odd triple;
 wenzelm parents: 
36685diff
changeset | 93 | stack = pending | 
| 
ed147003de4b
simplified type XML.Tree: embed Markup directly, avoid slightly odd triple;
 wenzelm parents: 
36685diff
changeset | 94 | add(XML.Elem(markup, body.toList)) | 
| 36684 | 95 | } | 
| 27930 | 96 | } | 
| 97 | ||
| 98 | ||
| 99 | /* parse chunks */ | |
| 100 | ||
| 56600 | 101 |     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 | 102 | if (chunk.length == 1 && chunk.charAt(0) == Y) pop() | 
| 27945 
d2dc5a1903e8
tuned parse performance: avoid splitting terminal Y chunk;
 wenzelm parents: 
27944diff
changeset | 103 |       else {
 | 
| 56600 | 104 |         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 | 105 | 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 | 106 | push(name.toString, atts.map(parse_attrib)) | 
| 27945 
d2dc5a1903e8
tuned parse performance: avoid splitting terminal Y chunk;
 wenzelm parents: 
27944diff
changeset | 107 | case txts => for (txt <- txts) add(XML.Text(txt.toString)) | 
| 
d2dc5a1903e8
tuned parse performance: avoid splitting terminal Y chunk;
 wenzelm parents: 
27944diff
changeset | 108 | } | 
| 27930 | 109 | } | 
| 110 | } | |
| 42719 | 111 |     (stack: @unchecked) match {
 | 
| 38475 | 112 | case List((Markup.Empty, body)) => body.toList | 
| 38230 
ed147003de4b
simplified type XML.Tree: embed Markup directly, avoid slightly odd triple;
 wenzelm parents: 
36685diff
changeset | 113 | case (Markup(name, _), _) :: _ => err_unbalanced(name) | 
| 27930 | 114 | } | 
| 115 | } | |
| 116 | ||
| 32450 | 117 | def parse(source: CharSequence): XML.Tree = | 
| 27930 | 118 |     parse_body(source) match {
 | 
| 119 | case List(result) => result | |
| 120 |       case Nil => XML.Text("")
 | |
| 67820 
e30d6368c7c8
clarified argument formats: explicit Unit, allow XML.Elem as well;
 wenzelm parents: 
64370diff
changeset | 121 |       case _ => err("multiple XML trees")
 | 
| 
e30d6368c7c8
clarified argument formats: explicit Unit, allow XML.Elem as well;
 wenzelm parents: 
64370diff
changeset | 122 | } | 
| 
e30d6368c7c8
clarified argument formats: explicit Unit, allow XML.Elem as well;
 wenzelm parents: 
64370diff
changeset | 123 | |
| 
e30d6368c7c8
clarified argument formats: explicit Unit, allow XML.Elem as well;
 wenzelm parents: 
64370diff
changeset | 124 | def parse_elem(source: CharSequence): XML.Tree = | 
| 
e30d6368c7c8
clarified argument formats: explicit Unit, allow XML.Elem as well;
 wenzelm parents: 
64370diff
changeset | 125 |     parse_body(source) match {
 | 
| 
e30d6368c7c8
clarified argument formats: explicit Unit, allow XML.Elem as well;
 wenzelm parents: 
64370diff
changeset | 126 | case List(elem: XML.Elem) => elem | 
| 
e30d6368c7c8
clarified argument formats: explicit Unit, allow XML.Elem as well;
 wenzelm parents: 
64370diff
changeset | 127 |       case _ => err("single XML element expected")
 | 
| 27930 | 128 | } | 
| 27960 | 129 | |
| 29521 | 130 | |
| 131 | /* failsafe parsing */ | |
| 132 | ||
| 45666 | 133 | private def markup_broken(source: CharSequence) = | 
| 55551 | 134 | XML.Elem(Markup.Broken, List(XML.Text(source.toString))) | 
| 29521 | 135 | |
| 38267 
e50c283dd125
type XML.Body as basic data representation language (Scala version);
 wenzelm parents: 
38234diff
changeset | 136 | def parse_body_failsafe(source: CharSequence): XML.Body = | 
| 32450 | 137 |   {
 | 
| 29521 | 138 |     try { parse_body(source) }
 | 
| 45666 | 139 |     catch { case ERROR(_) => List(markup_broken(source)) }
 | 
| 29521 | 140 | } | 
| 141 | ||
| 32450 | 142 | def parse_failsafe(source: CharSequence): XML.Tree = | 
| 143 |   {
 | |
| 27960 | 144 |     try { parse(source) }
 | 
| 45666 | 145 |     catch { case ERROR(_) => markup_broken(source) }
 | 
| 27960 | 146 | } | 
| 27930 | 147 | } |