Fix dangling JSON span by subclassing JFlexXrefSimple

Also:
- Make JsonXref case-sensitive, revising {File}
  to match any case.
- Escape all HTML special characters in JsonXref.
- Use Common.xref in JsonSymbolTokenizer.

Author: idodeclare

src/org/opensolaris/opengrok/analysis/json/Json.lexh

@@ -0,0 +1,30 @@
+/*
+ * CDDL HEADER START
+ *
+ * The contents of this file are subject to the terms of the
+ * Common Development and Distribution License (the "License").  
+ * You may not use this file except in compliance with the License.
+ *
+ * See LICENSE.txt included in this distribution for the specific
+ * language governing permissions and limitations under the License.
+ *
+ * When distributing Covered Code, include this CDDL HEADER in each
+ * file and include the License file at LICENSE.txt.
+ * If applicable, add the following below this CDDL HEADER, with the
+ * fields enclosed by brackets "[]" replaced with your own identifying
+ * information: Portions Copyright [yyyy] [name of copyright owner]
+ *
+ * CDDL HEADER END
+ */
+
+/*
+ * Copyright (c) 2017, Oracle and/or its affiliates. All rights reserved.
+ * Portions Copyright (c) 2017, Chris Fraire <[email protected]>.
+ */
+
+/*
+ * TODO add unicode support
+ */
+Identifier = [a-zA-Z_$] [a-zA-Z0-9_$]*
+
+Number = (0[xX][0-9a-fA-F]+|[0-9]+\.[0-9]+|[0-9]+)(([eE][+-]?[0-9]+)?[ufdlUFDL]*)?

src/org/opensolaris/opengrok/analysis/json/JsonSymbolTokenizer.lex

@@ -44,32 +44,30 @@ super(in);
 %include CommonTokenizer.lexh
 %char
 
-/* TODO : add unicode support */
-Identifier = [a-zA-Z_$] [a-zA-Z0-9_$]*
-
-Number = (0[xX][0-9a-fA-F]+|[0-9]+\.[0-9]+|[0-9]+)(([eE][+-]?[0-9]+)?[ufdlUFDL]*)?
-
 %state STRING
 
+%include Json.lexh
 %%
 
 //TODO improve per http://www.ecma-international.org/publications/files/ECMA-ST/ECMA-404.pdf
 
 <YYINITIAL> {
-{Identifier} {String id = yytext();
+{Identifier} {
+    String id = yytext();
                 if(!Consts.kwd.contains(id)){
                         setAttribs(id, yychar, yychar + yylength());
-                        return yystate(); }
-              }
+                        return yystate();
+		}
+ }
 
-{Number}        {}
+ {Number}        {}
 
  \"     { yybegin(STRING); }
 }
 
 <STRING> {
+ \\[\"\\]    {}
  \"     { yybegin(YYINITIAL); }
-\\\\ | \\\"     {}
 }
 
 <YYINITIAL, STRING> {

src/org/opensolaris/opengrok/analysis/json/JsonXref.lex

@@ -27,18 +27,18 @@
  */
 
 package org.opensolaris.opengrok.analysis.json;
-import org.opensolaris.opengrok.analysis.JFlexXref;
+import org.opensolaris.opengrok.analysis.JFlexXrefSimple;
 import java.io.IOException;
 import java.io.Writer;
 import java.io.Reader;
+import org.opensolaris.opengrok.web.HtmlConsts;
 import org.opensolaris.opengrok.web.Util;
 
 %%
 %public
 %class JsonXref
-%extends JFlexXref
+%extends JFlexXrefSimple
 %unicode
-%ignorecase
 %int
 %include CommonXref.lexh
 %{
@@ -49,18 +49,17 @@ import org.opensolaris.opengrok.web.Util;
   protected void setLineNumber(int x) { yyline = x; }
 %}
 
-//TODO add unicode support
-Identifier = [a-zA-Z_$] [a-zA-Z0-9_$]+
-
-File = [a-zA-Z]{FNameChar}* "." ("js"|"properties"|"props"|"xml"|"conf"|"txt"|"htm"|"html"|"ini"|"diff"|"patch")
-
-Number = (0[xX][0-9a-fA-F]+|[0-9]+\.[0-9]+|[0-9]+)(([eE][+-]?[0-9]+)?[ufdlUFDL]*)?
+File = [a-zA-Z]{FNameChar}* "." ([Jj][Ss] |
+    [Pp][Rr][Oo][Pp][Ee][Rr][Tt][Ii][Ee][Ss] | [Pp][Rr][Oo][Pp][Ss] |
+    [Xx][Mm][Ll] | [Cc][Oo][Nn][Ff] | [Tt][Xx][Tt] | [Hh][Tt][Mm] |
+    [Hh][Tt][Mm][Ll]? | [Ii][Nn][Ii] | [Dd][Ii][Ff][Ff] | [Pp][Aa][Tt][Cc][Hh])
 
 %state  STRING
 
 %include Common.lexh
 %include CommonURI.lexh
 %include CommonPath.lexh
+%include Json.lexh
 %%
 <YYINITIAL>{
 
@@ -89,22 +88,30 @@ Number = (0[xX][0-9a-fA-F]+|[0-9]+\.[0-9]+|[0-9]+)(([eE][+-]?[0-9]+)?[ufdlUFDL]*
         { out.write(Util.breadcrumbPath(urlPrefix+"defs=",yytext(),'.'));}
 */
 
-{Number}        { out.write("<span class=\"n\">"); out.write(yytext()); out.write("</span>"); }
+{Number}        {
+    disjointSpan(HtmlConsts.NUMBER_CLASS);
+    out.write(yytext());
+    disjointSpan(null);
+ }
 
- \"     { yybegin(STRING);out.write("<span class=\"s\">\"");}
+ \"     {
+    pushSpan(STRING, HtmlConsts.STRING_CLASS);
+    out.write(htmlize(yytext()));
+  }
 }
 
 <STRING> {
- \" {WhiteSpace} \"  { out.write(yytext());}
- \"     { yybegin(YYINITIAL); out.write("\"</span>"); }
- \\\\   { out.write("\\\\"); }
- \\\"   { out.write("\\\""); }
+ \\[\"\\] |
+ \" {WhiteSpace} \"  { out.write(htmlize(yytext())); }
+ \"     {
+    out.write(htmlize(yytext()));
+    yypop();
+ }
 }
 
 <YYINITIAL, STRING> {
-"&"     {out.write( "&amp;");}
-"<"     {out.write( "&lt;");}
-">"     {out.write( "&gt;");}
+[&<>\'\"]    { out.write(htmlize(yytext())); }
+
 {WhspChar}*{EOL}      { startNewLine(); }
  {WhiteSpace}   { out.write(yytext()); }
  [!-~]  { out.write(yycharat(0)); }

test/org/opensolaris/opengrok/analysis/json/sample_xref.html

@@ -7,31 +7,31 @@
 <title>sampleFile - OpenGrok cross reference for /sampleFile</title></head><body>
 <script type="text/javascript">/* <![CDATA[ */
 function get_sym_list(){return [];} /* ]]> */</script><a class="l" name="1" href="#1">1</a>{
-<a class="l" name="2" href="#2">2</a>	<span class="s">"lic01"</span> : <span class="s">"CDDL HEADER START"</span>,
-<a class="l" name="3" href="#3">3</a>	<span class="s">"lic02"</span> : <span class="s">""</span>,
-<a class="l" name="4" href="#4">4</a>	<span class="s">"lic03"</span> : <span class="s">"The contents of this file are subject to the terms of the"</span>,
-<a class="l" name="5" href="#5">5</a>	<span class="s">"lic04"</span> : <span class="s">"Common Development and Distribution License (the \"License\")."</span>,
-<a class="l" name="6" href="#6">6</a>	<span class="s">"lic05"</span> : <span class="s">"You may not use this file except in compliance with the License."</span>,
-<a class="l" name="7" href="#7">7</a>	<span class="s">"lic06"</span> : <span class="s">""</span>,
-<a class="l" name="8" href="#8">8</a>	<span class="s">"lic07"</span> : <span class="s">"See <a href="/source/s?path=LICENSE.txt">LICENSE.txt</a> included in this distribution for the specific"</span>,
-<a class="l" name="9" href="#9">9</a>	<span class="s">"lic08"</span> : <span class="s">"language governing permissions and limitations under the License."</span>,
-<a class="hl" name="10" href="#10">10</a>	<span class="s">"lic09"</span> : <span class="s">""</span>,
-<a class="l" name="11" href="#11">11</a>	<span class="s">"lic10"</span> : <span class="s">"When distributing Covered Code, include this CDDL HEADER in each"</span>,
-<a class="l" name="12" href="#12">12</a>	<span class="s">"lic11"</span> : <span class="s">"file and include the License file at <a href="/source/s?path=LICENSE.txt">LICENSE.txt</a>."</span>,
-<a class="l" name="13" href="#13">13</a>	<span class="s">"lic12"</span> : <span class="s">"If applicable, add the following below this CDDL HEADER, with the"</span>,
-<a class="l" name="14" href="#14">14</a>	<span class="s">"lic13"</span> : <span class="s">"fields enclosed by brackets \"[]\" replaced with your own identifying"</span>,
-<a class="l" name="15" href="#15">15</a>	<span class="s">"lic14"</span> : <span class="s">"information: Portions Copyright [yyyy] [name of copyright owner]"</span>,
-<a class="l" name="16" href="#16">16</a>	<span class="s">"lic15"</span> : <span class="s">""</span>,
-<a class="l" name="17" href="#17">17</a>	<span class="s">"lic16"</span> : <span class="s">"CDDL HEADER END"</span>,
-<a class="l" name="18" href="#18">18</a>	<span class="s">"spc01"</span> : <b>null</b>,
-<a class="l" name="19" href="#19">19</a>	<span class="s">"cpy01"</span> : <span class="s">"Copyright (c) 2017, Chris Fraire &lt;[email protected]&gt;."</span>,
-<a class="hl" name="20" href="#20">20</a>	<span class="s">"strings"</span> : [<span class="s">"\u002F"</span>, <span class="s">"\u002f"</span>, <span class="s">"\/"</span>, <span class="s">"/"</span>, <span class="s">"\uD834\uDD1E"</span>, <span class="s">"\""</span>, <span class="s">"\'"</span>,
-<a class="l" name="21" href="#21">21</a>	    <span class="s">"\\"</span>],
-<a class="l" name="22" href="#22">22</a>	<span class="s">"spc02"</span> : <b>null</b>,
-<a class="l" name="23" href="#23">23</a>	<span class="s">"numbers"</span> : [ <span class="n">0</span>, -<span class="n">1</span>, <span class="n">1.2</span>, -<span class="n">2.4</span>, -<span class="n">1e5</span>, <span class="n">1e-6</span>, <span class="n">20.5e97</span>, -<span class="n">33.2e-4</span> ],
-<a class="l" name="24" href="#24">24</a>	<span class="s">"spc03"</span> : <b>null</b>,
-<a class="l" name="25" href="#25">25</a>	<span class="s">"booleans"</span> : [ <b>true</b>, <b>false</b> ],
-<a class="l" name="26" href="#26">26</a>	<span class="s">"END"</span> : <span class="s">""</span>
+<a class="l" name="2" href="#2">2</a>	<span class="s">&quot;lic01&quot;</span> : <span class="s">&quot;CDDL HEADER START&quot;</span>,
+<a class="l" name="3" href="#3">3</a>	<span class="s">&quot;lic02&quot;</span> : <span class="s">&quot;&quot;</span>,
+<a class="l" name="4" href="#4">4</a>	<span class="s">&quot;lic03&quot;</span> : <span class="s">&quot;The contents of this file are subject to the terms of the&quot;</span>,
+<a class="l" name="5" href="#5">5</a>	<span class="s">&quot;lic04&quot;</span> : <span class="s">&quot;Common Development and Distribution License (the \&quot;License\&quot;).&quot;</span>,
+<a class="l" name="6" href="#6">6</a>	<span class="s">&quot;lic05&quot;</span> : <span class="s">&quot;You may not use this file except in compliance with the License.&quot;</span>,
+<a class="l" name="7" href="#7">7</a>	<span class="s">&quot;lic06&quot;</span> : <span class="s">&quot;&quot;</span>,
+<a class="l" name="8" href="#8">8</a>	<span class="s">&quot;lic07&quot;</span> : <span class="s">&quot;See <a href="/source/s?path=LICENSE.txt">LICENSE.txt</a> included in this distribution for the specific&quot;</span>,
+<a class="l" name="9" href="#9">9</a>	<span class="s">&quot;lic08&quot;</span> : <span class="s">&quot;language governing permissions and limitations under the License.&quot;</span>,
+<a class="hl" name="10" href="#10">10</a>	<span class="s">&quot;lic09&quot;</span> : <span class="s">&quot;&quot;</span>,
+<a class="l" name="11" href="#11">11</a>	<span class="s">&quot;lic10&quot;</span> : <span class="s">&quot;When distributing Covered Code, include this CDDL HEADER in each&quot;</span>,
+<a class="l" name="12" href="#12">12</a>	<span class="s">&quot;lic11&quot;</span> : <span class="s">&quot;file and include the License file at <a href="/source/s?path=LICENSE.txt">LICENSE.txt</a>.&quot;</span>,
+<a class="l" name="13" href="#13">13</a>	<span class="s">&quot;lic12&quot;</span> : <span class="s">&quot;If applicable, add the following below this CDDL HEADER, with the&quot;</span>,
+<a class="l" name="14" href="#14">14</a>	<span class="s">&quot;lic13&quot;</span> : <span class="s">&quot;fields enclosed by brackets \&quot;[]\&quot; replaced with your own identifying&quot;</span>,
+<a class="l" name="15" href="#15">15</a>	<span class="s">&quot;lic14&quot;</span> : <span class="s">&quot;information: Portions Copyright [yyyy] [name of copyright owner]&quot;</span>,
+<a class="l" name="16" href="#16">16</a>	<span class="s">&quot;lic15&quot;</span> : <span class="s">&quot;&quot;</span>,
+<a class="l" name="17" href="#17">17</a>	<span class="s">&quot;lic16&quot;</span> : <span class="s">&quot;CDDL HEADER END&quot;</span>,
+<a class="l" name="18" href="#18">18</a>	<span class="s">&quot;spc01&quot;</span> : <b>null</b>,
+<a class="l" name="19" href="#19">19</a>	<span class="s">&quot;cpy01&quot;</span> : <span class="s">&quot;Copyright (c) 2017, Chris Fraire &lt;[email protected]&gt;.&quot;</span>,
+<a class="hl" name="20" href="#20">20</a>	<span class="s">&quot;strings&quot;</span> : [<span class="s">&quot;\u002F&quot;</span>, <span class="s">&quot;\u002f&quot;</span>, <span class="s">&quot;\/&quot;</span>, <span class="s">&quot;/&quot;</span>, <span class="s">&quot;\uD834\uDD1E&quot;</span>, <span class="s">&quot;\&quot;&quot;</span>, <span class="s">&quot;\&apos;&quot;</span>,
+<a class="l" name="21" href="#21">21</a>	    <span class="s">&quot;\\&quot;</span>],
+<a class="l" name="22" href="#22">22</a>	<span class="s">&quot;spc02&quot;</span> : <b>null</b>,
+<a class="l" name="23" href="#23">23</a>	<span class="s">&quot;numbers&quot;</span> : [ <span class="n">0</span>, -<span class="n">1</span>, <span class="n">1.2</span>, -<span class="n">2.4</span>, -<span class="n">1e5</span>, <span class="n">1e-6</span>, <span class="n">20.5e97</span>, -<span class="n">33.2e-4</span> ],
+<a class="l" name="24" href="#24">24</a>	<span class="s">&quot;spc03&quot;</span> : <b>null</b>,
+<a class="l" name="25" href="#25">25</a>	<span class="s">&quot;booleans&quot;</span> : [ <b>true</b>, <b>false</b> ],
+<a class="l" name="26" href="#26">26</a>	<span class="s">&quot;END&quot;</span> : <span class="s">&quot;&quot;</span>
 <a class="l" name="27" href="#27">27</a>}
 <a class="l" name="28" href="#28">28</a></body>
 </html>

test/org/opensolaris/opengrok/analysis/json/truncated_xref.html

@@ -5,5 +5,5 @@
       class="xref">
 <head>
 <title>sampleFile - OpenGrok cross reference for /sampleFile</title></head><body>
-<a class="l" name="1" href="#1">1</a>  <span class="s">"Oops this string is not ter</span></body>
+<a class="l" name="1" href="#1">1</a>  <span class="s">&quot;Oops this string is not ter</span></body>
 </html>