version 3.6.92

* NEWS: Record release date.
style: avoid strncpy
2026-08-09 10:45:15 +00:00 · 2020-07-19 09:24:57 +02:00 · 2020-07-19 09:23:42 +02:00 · 2020-07-18 07:54:02 +02:00 · 2020-07-16 07:31:25 +02:00 · 2020-07-16 07:31:24 +02:00
52 changed files with 2457 additions and 695 deletions
@@ -1 +1 @@
-3.6.90
+3.6.91
@@ -1,5 +1,21 @@
 GNU Bison NEWS

+* Noteworthy changes in release 3.6.92 (2020-07-19) [beta]
+
+Changes in the display of counterexamples.
+
+** Documentation
+
+*** Examples
+
+  The bistromathic demonstrates %param and how to quote sources in the error
+  messages:
+
+    > 123 456
+    1.5-7: syntax error: expected end of file or + or - or * or / or ^ before number
+        1 | 123 456
+          |     ^~~
+
 * Noteworthy changes in release 3.6.91 (2020-07-09) [beta]

 ** Bug fixes
@@ -30,9 +46,17 @@ GNU Bison NEWS
  strings in the grammar which can be parsed in two ways due to the
  conflict.  For example:

-    Example              exp '+' exp • '/' exp
-    First derivation     exp ::=[ exp ::=[ exp '+' exp • ] '/' exp ]
-    Second derivation    exp ::=[ exp '+' exp ::=[ exp • '/' exp ] ]
+    Shift/reduce conflict on token "/":
+      Example              exp "+" exp • "/" exp
+      Shift derivation
+        exp
+        ↳ exp "+" exp
+                  ↳ exp • "/" exp
+      Example              exp "+" exp • "/" exp
+      Reduce derivation
+        exp
+        ↳ exp             "/" exp
+          ↳ exp "+" exp •

  When Bison is installed with text styling enabled, the example is actually
  shown twice, with colors highlighting the ambiguity.
@@ -42,15 +66,24 @@ GNU Bison NEWS
  bison cannot find an example that can be derived in two ways, it instead
  generates two examples that are the same up until the dot:

-    First example        expr • ID $end
-    First derivation     $accept ::=[ s ::=[ a ::=[ expr • ] ID ] $end ]
-    Second example       expr • ID ',' ID $end
-    Second derivation    $accept ::=[ s ::=[ a ::=[ expr ::=[ expr • ID ',' ] ] ID ] $end ]
+    First example        expr • ID ',' ID $end
+    Shift derivation
+      $accept
+      ↳ s                      $end
+        ↳ a                 ID
+          ↳ expr
+            ↳ expr • ID ','
+    Second example       expr • ID $end
+    Reduce derivation
+      $accept
+      ↳ s             $end
+        ↳ a        ID
+          ↳ expr •

  In these cases, the parser usually doesn't have enough lookahead to
  differentiate the two given examples.

-  The counterexamples are "focused": in two different ways.  First, they do
+  The counterexamples are "focused" in two different ways.  First, they do
  not clutter the output with all the derivations from the start symbol,
  rather they start on the "conflicted nonterminal". They go straight to the
  point.  Second, they don't "expand" nonterminal symbols uselessly.
@@ -444,7 +444,7 @@ added the `[-Wother]` part to all the warnings).  Part of the update can be
 done with a crude tool: `build-aux/update-test`.

 Once you ran the test suite, and therefore have many `testsuite.log` files,
-run, from the source tree:
+run `make update-tests`.  Or, by hand, from the *source* tree:

    $ ./build-aux/update-test $build/tests/testsuite.dir/*/testsuite.log

@@ -45,6 +45,7 @@ Csaba Raduly              [email protected]
 Dagobert Michelsen        [email protected]
 Daniel Frużyński          [email protected]
 Daniel Galloway           [email protected]
+Daniela Becker            [email protected]
 Daniel Hagerty            [email protected]
 David Barto               [email protected]
 David J. MacKenzie        [email protected]
@@ -36,8 +36,6 @@ Unless we play it dumb (little structure).
 - How about not evaluating incomplete lines when the text is not finished
  (as shells do).

- Caret diagnostics
-
 ** Questions
 *** Java
 - Should i18n be part of the Lexer?  Currently it's a static method of
@@ -49,6 +49,10 @@
 .cex-5   { color: orange; }
 .cex-6   { color: brown; }
 .cex-7   { color: mauve; }
+.cex-8   { color: #013220; } /* Dark green. */
+.cex-9   { color: #e75480; } /* Dark pink. */
+.cex-10  { color: cyan; }
+.cex-11  { color: orange; }

 /* Cex: derivation rewriting steps.  */
 .cex-step { font-style: italic; }
@@ -52,7 +52,7 @@
  <xsl:if test="nonterminal[@usefulness='useless-in-grammar']">
    <xsl:text>Nonterminals useless in grammar&#10;&#10;</xsl:text>
    <xsl:for-each select="nonterminal[@usefulness='useless-in-grammar']">
-      <xsl:text>   </xsl:text>
+      <xsl:text>    </xsl:text>
      <xsl:value-of select="@name"/>
      <xsl:text>&#10;</xsl:text>
    </xsl:for-each>
@@ -65,7 +65,7 @@
    <xsl:text>Terminals unused in grammar&#10;&#10;</xsl:text>
    <xsl:for-each select="terminal[@usefulness='unused-in-grammar']">
      <xsl:sort select="@symbol-number" data-type="number"/>
-      <xsl:text>   </xsl:text>
+      <xsl:text>    </xsl:text>
      <xsl:value-of select="@name"/>
      <xsl:text>&#10;</xsl:text>
    </xsl:for-each>
@@ -136,6 +136,7 @@
 </xsl:template>

 <xsl:template match="terminal">
+  <xsl:text>    </xsl:text>
  <xsl:value-of select="@name"/>
  <xsl:call-template name="line-wrap">
    <xsl:with-param name="first-line-length">
@@ -148,6 +149,9 @@
    </xsl:with-param>
    <xsl:with-param name="line-length" select="66" />
    <xsl:with-param name="text">
+      <xsl:if test="string-length(@type) != 0">
+        <xsl:value-of select="concat(' &lt;', @type, '&gt;')"/>
+      </xsl:if>
      <xsl:value-of select="concat(' (', @token-number, ')')"/>
      <xsl:for-each select="key('bison:ruleByRhs', @name)">
        <xsl:value-of select="concat(' ', @number)"/>
@@ -157,14 +161,18 @@
 </xsl:template>

 <xsl:template match="nonterminal">
+  <xsl:text>    </xsl:text>
  <xsl:value-of select="@name"/>
+  <xsl:if test="string-length(@type) != 0">
+    <xsl:value-of select="concat(' &lt;', @type, '&gt;')"/>
+  </xsl:if>
  <xsl:value-of select="concat(' (', @symbol-number, ')')"/>
  <xsl:text>&#10;</xsl:text>
  <xsl:variable name="output">
    <xsl:call-template name="line-wrap">
      <xsl:with-param name="line-length" select="66" />
      <xsl:with-param name="text">
-        <xsl:text>    </xsl:text>
+        <xsl:text>        </xsl:text>
        <xsl:if test="key('bison:ruleByLhs', @name)">
          <xsl:text>on@left:</xsl:text>
          <xsl:for-each select="key('bison:ruleByLhs', @name)">
@@ -173,7 +181,7 @@
        </xsl:if>
        <xsl:if test="key('bison:ruleByRhs', @name)">
          <xsl:if test="key('bison:ruleByLhs', @name)">
-            <xsl:text>, </xsl:text>
+            <xsl:text>&#10;        </xsl:text>
          </xsl:if>
          <xsl:text>on@right:</xsl:text>
          <xsl:for-each select="key('bison:ruleByRhs', @name)">
@@ -348,11 +356,11 @@
  <!-- RHS -->
  <xsl:for-each select="rhs/*">
    <xsl:if test="position() = $dot + 1">
-      <xsl:text> .</xsl:text>
+      <xsl:text> •</xsl:text>
    </xsl:if>
    <xsl:apply-templates select="."/>
    <xsl:if test="position() = last() and position() = $dot">
-      <xsl:text> .</xsl:text>
+      <xsl:text> •</xsl:text>
    </xsl:if>
  </xsl:for-each>
  <xsl:if test="$lookaheads">
@@ -227,6 +227,7 @@
  <xsl:text>&#10;</xsl:text>
  <p class="pre">
    <xsl:call-template name="style-rule-set">
+      <xsl:with-param name="anchor" select="'true'" />
      <xsl:with-param
        name="rule-set" select="rules/rule[@usefulness!='useless-in-grammar']"
      />
@@ -238,9 +239,11 @@
 </xsl:template>

 <xsl:template name="style-rule-set">
+  <xsl:param name="anchor"/>
  <xsl:param name="rule-set"/>
  <xsl:for-each select="$rule-set">
    <xsl:apply-templates select=".">
+      <xsl:with-param name="anchor" select="$anchor"/>
      <xsl:with-param name="pad" select="'3'"/>
      <xsl:with-param name="prev-lhs">
        <xsl:if test="position()>1">
@@ -306,9 +309,10 @@
    <xsl:text> Terminals, with rules where they appear</xsl:text>
  </h3>
  <xsl:text>&#10;&#10;</xsl:text>
-  <p class="pre">
+  <ul>
+    <xsl:text>&#10;</xsl:text>
    <xsl:apply-templates select="terminal"/>
-  </p>
+  </ul>
  <xsl:text>&#10;&#10;</xsl:text>
 </xsl:template>

@@ -318,41 +322,64 @@
    <xsl:text> Nonterminals, with rules where they appear</xsl:text>
  </h3>
  <xsl:text>&#10;&#10;</xsl:text>
-  <p class="pre">
+  <ul>
+    <xsl:text>&#10;</xsl:text>
    <xsl:apply-templates
      select="nonterminal[@usefulness!='useless-in-grammar']"
    />
-  </p>
+  </ul>
 </xsl:template>

 <xsl:template match="terminal">
-  <b><xsl:value-of select="@name"/></b>
-  <xsl:value-of select="concat(' (', @token-number, ')')"/>
-  <xsl:for-each select="key('bison:ruleByRhs', @name)">
-    <xsl:apply-templates select="." mode="number-link"/>
-  </xsl:for-each>
+  <xsl:text>  </xsl:text>
+  <li>
+    <b><xsl:value-of select="@name"/></b>
+    <xsl:if test="string-length(@type) != 0">
+      <xsl:value-of select="concat(' &lt;', @type, '&gt;')"/>
+    </xsl:if>
+    <xsl:value-of select="concat(' (', @token-number, ')')"/>
+    <xsl:for-each select="key('bison:ruleByRhs', @name)">
+      <xsl:apply-templates select="." mode="number-link"/>
+    </xsl:for-each>
+  </li>
  <xsl:text>&#10;</xsl:text>
 </xsl:template>

 <xsl:template match="nonterminal">
-  <b><xsl:value-of select="@name"/></b>
-  <xsl:value-of select="concat(' (', @symbol-number, ')')"/>
-  <xsl:text>&#10;    </xsl:text>
-  <xsl:if test="key('bison:ruleByLhs', @name)">
-    <xsl:text>on left:</xsl:text>
-    <xsl:for-each select="key('bison:ruleByLhs', @name)">
-      <xsl:apply-templates select="." mode="number-link"/>
-    </xsl:for-each>
-  </xsl:if>
-  <xsl:if test="key('bison:ruleByRhs', @name)">
-    <xsl:if test="key('bison:ruleByLhs', @name)">
-      <xsl:text>&#10;    </xsl:text>
+  <xsl:text>  </xsl:text>
+  <li>
+    <b><xsl:value-of select="@name"/></b>
+    <xsl:if test="string-length(@type) != 0">
+      <xsl:value-of select="concat(' &lt;', @type, '&gt;')"/>
    </xsl:if>
-    <xsl:text>on right:</xsl:text>
-    <xsl:for-each select="key('bison:ruleByRhs', @name)">
-      <xsl:apply-templates select="." mode="number-link"/>
-    </xsl:for-each>
-  </xsl:if>
+    <xsl:value-of select="concat(' (', @symbol-number, ')')"/>
+    <xsl:text>&#10;    </xsl:text>
+    <ul>
+      <xsl:text>&#10;</xsl:text>
+      <xsl:if test="key('bison:ruleByLhs', @name)">
+        <xsl:text>      </xsl:text>
+        <li>
+          <xsl:text>on left:</xsl:text>
+          <xsl:for-each select="key('bison:ruleByLhs', @name)">
+            <xsl:apply-templates select="." mode="number-link"/>
+          </xsl:for-each>
+        </li>
+        <xsl:text>&#10;</xsl:text>
+      </xsl:if>
+      <xsl:if test="key('bison:ruleByRhs', @name)">
+        <xsl:text>      </xsl:text>
+        <li>
+          <xsl:text>on right:</xsl:text>
+          <xsl:for-each select="key('bison:ruleByRhs', @name)">
+            <xsl:apply-templates select="." mode="number-link"/>
+          </xsl:for-each>
+        </li>
+        <xsl:text>&#10;</xsl:text>
+      </xsl:if>
+    <xsl:text>    </xsl:text>
+    </ul>
+    <xsl:text>&#10;  </xsl:text>
+  </li>
  <xsl:text>&#10;</xsl:text>
 </xsl:template>

@@ -385,7 +412,7 @@
        <xsl:value-of select="concat('state_', @number)"/>
      </xsl:attribute>
    </a>
-    <xsl:text>state </xsl:text>
+    <xsl:text>State </xsl:text>
    <xsl:value-of select="@number"/>
  </h3>
  <xsl:text>&#10;&#10;</xsl:text>
@@ -464,7 +491,12 @@
  </xsl:apply-templates>
 </xsl:template>

+<!--
+anchor = 'true': define as an <a> anchor.
+itemset = 'true': show the items.
+ -->
 <xsl:template match="rule">
+  <xsl:param name="anchor"/>
  <xsl:param name="itemset"/>
  <xsl:param name="pad"/>
  <xsl:param name="prev-lhs"/>
@@ -475,17 +507,21 @@
    <xsl:text>&#10;</xsl:text>
  </xsl:if>

-  <xsl:if test="$itemset != 'true'">
-    <a>
-      <xsl:attribute name="name">
-        <xsl:value-of select="concat('rule_', @number)"/>
-      </xsl:attribute>
-    </a>
-  </xsl:if>
  <xsl:text>  </xsl:text>

  <xsl:choose>
-    <xsl:when test="$itemset = 'true'">
+    <xsl:when test="$anchor = 'true'">
+      <a>
+        <xsl:attribute name="name">
+          <xsl:value-of select="concat('rule_', @number)"/>
+        </xsl:attribute>
+        <xsl:call-template name="lpad">
+          <xsl:with-param name="str" select="string(@number)"/>
+          <xsl:with-param name="pad" select="number($pad)"/>
+        </xsl:call-template>
+      </a>
+    </xsl:when>
+    <xsl:otherwise>
      <a>
        <xsl:attribute name="href">
          <xsl:value-of select="concat('#rule_', @number)"/>
@@ -495,25 +531,13 @@
          <xsl:with-param name="pad" select="number($pad)"/>
        </xsl:call-template>
      </a>
-    </xsl:when>
-    <xsl:otherwise>
-      <xsl:call-template name="lpad">
-        <xsl:with-param name="str" select="string(@number)"/>
-        <xsl:with-param name="pad" select="number($pad)"/>
-      </xsl:call-template>
    </xsl:otherwise>
  </xsl:choose>
  <xsl:text> </xsl:text>

  <!-- LHS -->
  <xsl:choose>
-    <xsl:when test="$itemset != 'true' and $prev-lhs = lhs[text()]">
-      <xsl:call-template name="lpad">
-        <xsl:with-param name="str" select="'|'"/>
-        <xsl:with-param name="pad" select="number(string-length(lhs[text()])) + 2"/>
-      </xsl:call-template>
-    </xsl:when>
-    <xsl:when test="$itemset = 'true' and $prev-lhs = lhs[text()]">
+    <xsl:when test="$prev-lhs = lhs[text()]">
      <xsl:call-template name="lpad">
        <xsl:with-param name="str" select="'|'"/>
        <xsl:with-param name="pad" select="number(string-length(lhs[text()])) + 2"/>
@@ -263,7 +263,6 @@ Writing GLR Parsers
 * Merging GLR Parses::     Using GLR parsers to resolve ambiguities.
 * GLR Semantic Actions::   Considerations for semantic values and deferred actions.
 * Semantic Predicates::    Controlling a parse with arbitrary computations.
-* Compiler Requirements for GLR::  GLR parsers require a modern C compiler.

 Examples

@@ -943,7 +942,6 @@ on the resulting values to produce an arbitrary merged result.
 * Merging GLR Parses::     Using GLR parsers to resolve ambiguities.
 * GLR Semantic Actions::   Considerations for semantic values and deferred actions.
 * Semantic Predicates::    Controlling a parse with arbitrary computations.
-* Compiler Requirements for GLR::  GLR parsers require a modern C compiler.
@end menu

@node Simple GLR Parsers
@@ -1372,14 +1370,14 @@ widget:
@noindent
 is one way to allow the same parser to handle two different syntaxes for
 widgets.  The clause preceded by @code{%?} is treated like an ordinary
-action, except that its text is treated as an expression and is always
+midrule action, except that its text is handled as an expression and is always
 evaluated immediately (even when in nondeterministic mode).  If the
 expression yields 0 (false), the clause is treated as a syntax error,
 which, in a nondeterministic parser, causes the stack in which it is reduced
-to die.  In a deterministic parser, it acts like YYERROR.
+to die.  In a deterministic parser, it acts like @code{YYERROR}.

 As the example shows, predicates otherwise look like semantic actions, and
-therefore you must be take them into account when determining the numbers
+therefore you must take them into account when determining the numbers
 to use for denoting the semantic values of right-hand side symbols.
 Predicate actions, however, have no defined value, and may not be given
 labels.
@@ -1393,7 +1391,7 @@ widget:
  @{ if (!new_syntax) YYERROR; @}
    "widget" id new_args  @{ $$ = f($3, $4); @}
 |  @{ if (new_syntax) YYERROR; @}
-    "widget" id old_args   @{ $$ = f($3, $4); @}
+    "widget" id old_args  @{ $$ = f($3, $4); @}
 ;
@end example

@@ -1411,36 +1409,6 @@ reports an error.
 Finally, be careful in writing predicates: deferred actions have not been
 evaluated, so that using them in a predicate will have undefined effects.

-@node Compiler Requirements for GLR
-@subsection Considerations when Compiling GLR Parsers
-@cindex @code{inline}
-@cindex GLR parsers and @code{inline}
-
-The GLR parsers require a compiler for ISO C89 or
-later.  In addition, they use the @code{inline} keyword, which is not
-C89, but is C99 and is a common extension in pre-C99 compilers.  It is
-up to the user of these parsers to handle
-portability issues.  For instance, if using Autoconf and the Autoconf
-macro @code{AC_C_INLINE}, a mere
-
-@example
-%@{
-  #include <config.h>
-%@}
-@end example
-
-@noindent
-will suffice.  Otherwise, we suggest
-
-@example
-%@{
-  #if (__STDC_VERSION__ < 199901 && ! defined __GNUC__ \
-       && ! defined inline)
-  # define inline
-  #endif
-%@}
-@end example
-
@node Locations
@section Locations
@cindex location
@@ -9936,10 +9904,10 @@ output is actually in color)}:
@example
 Shift/reduce conflict on token "else":
@group
-  Example            @yellow{"if" expr "then"} @blue{"if" expr "then" stmt} @red{•} @yellow{"else" stmt}
-  First derivation   @yellow{if_stmt ::=[ "if" expr "then"} @green{stmt ::=[} @blue{if_stmt ::=[ "if" expr "then" stmt} @red{•} @blue{]} @green{]} @yellow{"else" stmt ]}
  Example            @yellow{"if" expr "then"} @blue{"if" expr "then" stmt} @red{•} @blue{"else" stmt}
-  Second derivation  @yellow{if_stmt ::=[ "if" expr "then"} @green{stmt ::=[} @blue{if_stmt ::=[ "if" expr "then" stmt} @red{•} @blue{"else" stmt ]} @green{]} @yellow{]}
+  Shift derivation   @yellow{if_stmt @arrow{} [ "if" expr "then"} @green{stmt @arrow{} [} @blue{if_stmt @arrow{} [ "if" expr "then" stmt} @red{•} @blue{"else" stmt ]} @green{]} @yellow{]}
+  Example            @yellow{"if" expr "then"} @blue{"if" expr "then" stmt} @red{•} @yellow{"else" stmt}
+  Reduce derivation  @yellow{if_stmt @arrow{} [ "if" expr "then"} @green{stmt @arrow{} [} @blue{if_stmt @arrow{} [ "if" expr "then" stmt} @red{•} @blue{]} @green{]} @yellow{"else" stmt ]}
@end group
@end example
@end ifhtml
@@ -9947,14 +9915,14 @@ Shift/reduce conflict on token "else":
@smallexample
 Shift/reduce conflict on token "else":
@group
-  Example
-    @yellow{"if" expr "then"} @blue{"if" expr "then" stmt} @red{•} @yellow{"else" stmt}
-  First derivation
-    @yellow{if_stmt ::=[ "if" expr "then"} @green{stmt ::=[} @blue{if_stmt ::=[ "if" expr "then" stmt} @red{•} @blue{]} @green{]} @yellow{"else" stmt ]}
  Example
    @yellow{"if" expr "then"} @blue{"if" expr "then" stmt} @red{•} @blue{"else" stmt}
-  Second derivation
-    @yellow{if_stmt ::=[ "if" expr "then"} @green{stmt ::=[} @blue{if_stmt ::=[ "if" expr "then" stmt} @red{•} @blue{"else" stmt ]} @green{]} @yellow{]}
+  Shift derivation
+    @yellow{if_stmt @arrow{} [ "if" expr "then"} @green{stmt @arrow{} [} @blue{if_stmt @arrow{} [ "if" expr "then" stmt} @red{•} @blue{"else" stmt ]} @green{]} @yellow{]}
+  Example
+    @yellow{"if" expr "then"} @blue{"if" expr "then" stmt} @red{•} @yellow{"else" stmt}
+  Reduce derivation
+    @yellow{if_stmt @arrow{} [ "if" expr "then"} @green{stmt @arrow{} [} @blue{if_stmt @arrow{} [ "if" expr "then" stmt} @red{•} @blue{]} @green{]} @yellow{"else" stmt ]}
@end group
@end smallexample
@end ifnothtml
@@ -9987,22 +9955,22 @@ $ @kbd{bison -Wcex sequence.y}
 sequence.y: @dwarning{warning}: 1 shift/reduce conflict [@dwarning{-Wconflicts-sr}]
 sequence.y: @dwarning{warning}: 2 reduce/reduce conflicts [@dwarning{-Wconflicts-rr}]
 Shift/reduce conflict on token "word":
-  Example              @red{•} @yellow{"word"}
-  First derivation     @yellow{sequence ::=[} @green{sequence ::=[} @red{•} @green{]} @yellow{"word" ]}
  Example              @red{•} @green{"word"}
-  Second derivation    @yellow{sequence ::=[} @green{maybeword ::=[} @red{•} @green{"word" ]} @yellow{]}
+  Shift derivation     @yellow{sequence @arrow{} [} @green{maybeword @arrow{} [} @red{•} @green{"word" ]} @yellow{]}
+  Example              @red{•} @yellow{"word"}
+  Reduce derivation    @yellow{sequence @arrow{} [} @green{sequence @arrow{} [} @red{•} @green{]} @yellow{"word" ]}

 Reduce/reduce conflict on tokens $end, "word":
  Example              @red{•}
-  First derivation     @yellow{sequence ::=[} @red{•} @yellow{]}
+  First derivation     @yellow{sequence @arrow{} [} @red{•} @yellow{]}
  Example              @red{•}
-  Second derivation    @yellow{sequence ::=[} @green{maybeword ::=[} @red{•} @green{]} @yellow{]}
+  Second derivation    @yellow{sequence @arrow{} [} @green{maybeword @arrow{} [} @red{•} @green{]} @yellow{]}

 Shift/reduce conflict on token "word":
-  Example              @red{•} @yellow{"word"}
-  First derivation     @yellow{sequence ::=[} @green{sequence ::=[} @blue{maybeword ::=[} @red{•} @blue{]} @green{]} @yellow{"word" ]}
  Example              @red{•} @green{"word"}
-  Second derivation    @yellow{sequence ::=[} @green{maybeword ::=[} @red{•} @green{"word" ]} @yellow{]}
+  Shift derivation     @yellow{sequence @arrow{} [} @green{maybeword @arrow{} [} @red{•} @green{"word" ]} @yellow{]}
+  Example              @red{•} @yellow{"word"}
+  Reduce derivation    @yellow{sequence @arrow{} [} @green{sequence @arrow{} [} @blue{maybeword @arrow{} [} @red{•} @blue{]} @green{]} @yellow{"word" ]}

 sequence.y:8.3-45: @dwarning{warning}: rule useless in parser due to conflicts [@dwarning{-Wother}]
    8 |   @dwarning{%empty    @{ printf ("empty maybeword\n"); @}}
@@ -10033,10 +10001,10 @@ expr: %empty | expr ID ','

@smallexample
 Shift/reduce conflict on token ID:
-  First example        @blue{expr} @red{•} @green{ID} @yellow{$end}
-  First derivation     @yellow{$accept ::=[} @green{s ::=[} @blue{a ::=[ expr} @red{•} @blue{]} @green{ID ]} @yellow{$end ]}
-  Second example       @purple{expr} @red{•} @purple{ID ','} @green{ID} @yellow{$end}
-  Second derivation    @yellow{$accept ::=[} @green{s ::=[} @blue{a ::=[} @purple{expr ::=[ expr} @red{•} @purple{ID ',' ]} @blue{]} @green{ID ]} @yellow{$end ]}
+  First example        @purple{expr} @red{•} @purple{ID ','} @green{ID} @yellow{$end}
+  Shift derivation     @yellow{$accept @arrow{} [} @green{s @arrow{} [} @blue{a @arrow{} [} @purple{expr @arrow{} [ expr} @red{•} @purple{ID ',' ]} @blue{]} @green{ID ]} @yellow{$end ]}
+  Second example       @blue{expr} @red{•} @green{ID} @yellow{$end}
+  Reduce derivation    @yellow{$accept @arrow{} [} @green{s @arrow{} [} @blue{a @arrow{} [ expr} @red{•} @blue{]} @green{ID ]} @yellow{$end ]}
@end smallexample

 This conflict is caused by the parser not having enough information to know
@@ -10432,10 +10400,10 @@ counterexamples within the report, augmented with the corresponding items
 Shift/reduce conflict on token '/':
    1 exp: exp '+' exp •
    4 exp: exp • '/' exp
-  Example              @green{exp '+' exp} @red{•} @yellow{'/' exp}
-  First derivation     @yellow{exp ::=[} @green{exp ::=[ exp '+' exp} @red{•} @green{]} @yellow{'/' exp ]}
  Example              @yellow{exp '+'} @green{exp} @red{•} @green{'/' exp}
-  Second derivation    @yellow{exp ::=[ exp '+'} @green{exp ::=[ exp} @red{•} @green{'/' exp ]} @yellow{]}
+  Shift derivation     @yellow{exp @arrow{} [ exp '+'} @green{exp @arrow{} [ exp} @red{•} @green{'/' exp ]} @yellow{]}
+  Example              @green{exp '+' exp} @red{•} @yellow{'/' exp}
+  Reduce derivation    @yellow{exp @arrow{} [} @green{exp @arrow{} [ exp '+' exp} @red{•} @green{]} @yellow{'/' exp ]}
@end example

 This shows two separate derivations in the grammar for the same @code{exp}:
@@ -14824,13 +14792,13 @@ Bison.  See the file @file{ABOUT-NLS} for more information.
 I can't build Bison because my C compiler is too old.
@end quotation

-Except for GLR parsers (@pxref{Compiler Requirements for GLR}), the C
-code that Bison generates requires only C89 or later.  However, Bison
-itself requires common C99 features such as declarations after
-statements.  Bison's @code{configure} script attempts to enable C99 (or
-later) support on compilers that default to pre-C99.  If your compiler
-lacks these C99 features entirely, GCC may well be a better choice; or
-you can try upgrading to your compiler's latest version.
+Except for GLR parsers (which require C99), the C code that Bison generates
+requires only C89 or later.  However, Bison itself requires common C99
+features such as declarations after statements.  Bison's @code{configure}
+script attempts to enable C99 (or later) support on compilers that default
+to pre-C99.  If your compiler lacks these C99 features entirely, GCC may
+well be a better choice; or you can try upgrading to your compiler's latest
+version.

@node Where can I find help?
@section Where can I find help?
@@ -53,7 +53,9 @@ push-parser model.
 This example demonstrates best practices when using Bison.
 - Its hand-written scanner tracks locations.
 - Its interface is pure.
- It uses the `error` token to get error recovery.
+- It uses %params to pass user information to the parser and scanner.
+- Its scanner uses the `error` token to signal lexical errors and enter
+  error recovery.
 - Its interface is "incremental", well suited for interaction: it uses the
  push-parser API to feed the parser with the incoming tokens.
 - It features an interactive command line with completion based on the
@@ -62,6 +64,13 @@ This example demonstrates best practices when using Bison.
  messages.
 - It uses a custom syntax error with location, lookahead correction and
  token internationalization.
+- Error messages quote the source with squiggles that underline the error:
+```
+> 123 456
+1.5-7: syntax error: expected end of file or + or - or * or / or ^ before number
+    1 | 123 456
+      |     ^~~
+```
 - It supports debug traces with semantic values.
 - It uses named references instead of the traditional $1, $2, etc.

@@ -2,7 +2,9 @@
 This example demonstrates best practices when using Bison.
 - Its hand-written scanner tracks locations.
 - Its interface is pure.
- It uses the `error` token to get error recovery.
+- It uses %params to pass user information to the parser and scanner.
+- Its scanner uses the `error` token to signal lexical errors and enter
+  error recovery.
 - Its interface is "incremental", well suited for interaction: it uses the
  push-parser API to feed the parser with the incoming tokens.
 - It features an interactive command line with completion based on the
@@ -11,6 +13,13 @@ This example demonstrates best practices when using Bison.
  messages.
 - It uses a custom syntax error with location, lookahead correction and
  token internationalization.
+- Error messages quote the source with squiggles that underline the error:
+```
+> 123 456
+1.5-7: syntax error: expected end of file or + or - or * or / or ^ before number
+    1 | 123 456
+      |     ^~~
+```
 - It supports debug traces with semantic values.
 - It uses named references instead of the traditional $1, $2, etc.

@@ -101,14 +101,28 @@ cat >input <<EOF
 EOF
 run 0 '> *
 > ''
-err: 1.1: syntax error: expected end of file or - or ( or exit or number or function etc., before *'
+err: 1.1: syntax error: expected end of file or - or ( or exit or number or function etc., before *
+err:     1 | *
+err:       | ^'
+
+# Underline long errors.
+cat >input <<EOF
+123 123456
+EOF
+run 0 '> 123 123456
+> ''
+err: 1.5-10: syntax error: expected end of file or + or - or * or / or ^ before number
+err:     1 | 123 123456
+err:       |     ^~~~~~'

 cat >input <<EOF
 1 + 2 * * 3
 EOF
 run 0 '> 1 + 2 * * 3
 > ''
-err: 1.9: syntax error: expected - or ( or number or function or variable before *'
+err: 1.9: syntax error: expected - or ( or number or function or variable before *
+err:     1 | 1 + 2 * * 3
+err:       |         ^'

 cat >input <<EOF
 1 / 0
@@ -132,8 +146,14 @@ run 0 '> ((1 ++ 2) ** 3)
 1332
 > ''
 err: 1.6: syntax error: expected - or ( or number or function or variable before +
+err:     1 | ((1 ++ 2) ** 3)
+err:       |      ^
 err: 2.5: syntax error: expected - or ( or number or function or variable before +
-err: 2.16: syntax error: expected - or ( or number or function or variable before *'
+err:     2 | (1 ++ 2) + (3 ** 4)
+err:       |     ^
+err: 2.16: syntax error: expected - or ( or number or function or variable before *
+err:     2 | (1 ++ 2) + (3 ** 4)
+err:       |                ^'

 # The rule "( error )" should work even if there are no tokens between "(" and ")".
 cat >input <<EOF
@@ -142,7 +162,9 @@ EOF
 run 0 '> ()
 666
 > ''
-err: 1.2: syntax error: expected - or ( or number or function or variable before )'
+err: 1.2: syntax error: expected - or ( or number or function or variable before )
+err:     1 | ()
+err:       |  ^'


 cat >input <<EOF
@@ -189,6 +211,8 @@ err: LAC: checking lookahead function: S5
 err: LAC: checking lookahead variable: S6
 err: LAC: checking lookahead NEG: Err
 err: 1.2: syntax error: expected - or ( or number or function or variable before +
+err:     1 | (+_)
+err:       |  ^
 err: LAC: initial context discarded due to error recovery
 err: Shifting token error (1.2: )
 err: Entering state 10
@@ -286,7 +310,9 @@ run 0 '> (1+
 (       -       atan    cos     exp     ln      number  sin     sqrt
 > (1+
 > ''
-err: 1.4: syntax error: expected - or ( or number or function or variable before end of file'
+err: 1.4: syntax error: expected - or ( or number or function or variable before end of file
+err:     1 | (1+
+err:       |    ^'

 # Check the completion of a word.
 sed -e 's/\\t/	/g' >input <<EOF
@@ -294,7 +320,9 @@ sed -e 's/\\t/	/g' >input <<EOF
 EOF
 run 0 '> (atan ( ''
 > ''
-err: 1.9: syntax error: expected - or ( or number or function or variable before end of file'
+err: 1.9: syntax error: expected - or ( or number or function or variable before end of file
+err:     1 | (atan ( ''
+err:       |         ^'

 # Check the completion at the very beginning.
 sed -e 's/\\t/	/g' >input <<EOF
@@ -313,8 +341,9 @@ sed -e 's/\\t/	/g' >input <<EOF
 EOF
 run -n 0 '> 1++ ''
 > ''
-err: 1.1: syntax error: expected - or ( or number or function or variable before +
 err: 1.3: syntax error: expected - or ( or number or function or variable before +
+err:     1 | 1++ ''
+err:       |   ^
 '

 # And even when the error was recovered from.
@@ -323,8 +352,10 @@ sed -e 's/\\t/	/g' >input <<EOF
 EOF
 run -n 0 '> (1++2) + 3 +  ''
 > ''
-err: 1.1: syntax error: expected - or ( or number or function or variable before +
-err: 1.1: syntax error: expected - or ( or number or function or variable before +
 err: 1.4: syntax error: expected - or ( or number or function or variable before +
+err:     1 | (1++2) + 3 +  ''
+err:       |    ^
 err: 1.15: syntax error: expected - or ( or number or function or variable before end of file
+err:     1 | (1++2) + 3 +  ''
+err:       |               ^
 '
@@ -65,6 +65,15 @@

  symrec *putsym (char const *name, int sym_type);
  symrec *getsym (char const *name);
+
+  // Exchanging information with the parser.
+  typedef struct
+  {
+    // Whether to not emit error messages.
+    int silent;
+    // The current input line.
+    const char *line;
+  } user_context;
 }

 // Emitted in the header file, after the definition of YYSTYPE.
@@ -74,9 +83,13 @@
 #   define __attribute__(Spec) /* empty */
 #  endif
 # endif
-  int yylex (const char **line, YYSTYPE *yylval, YYLTYPE *yylloc);
-  void yyerror (YYLTYPE *loc, char const *format, ...)
-    __attribute__ ((__format__ (__printf__, 2, 3)));
+
+  yytoken_kind_t
+  yylex (const char **line, YYSTYPE *yylval, YYLTYPE *yylloc,
+         const user_context *uctx);
+  void yyerror (YYLTYPE *loc, const user_context *uctx,
+                char const *format, ...)
+    __attribute__ ((__format__ (__printf__, 3, 4)));
 }

 // Emitted in the implementation file.
@@ -116,6 +129,9 @@
 // Generate the parser description file (calc.output).
 %verbose

+// User information exchanged with the parser and scanner.
+%param {const user_context *uctx}
+
 // Generate YYSTYPE from the types assigned to symbols.
 %define api.value.type union
 %token
@@ -171,7 +187,7 @@ exp:
  {
    if ($r == 0)
      {
-        yyerror (&@$, "error: division by zero");
+        yyerror (&@$, uctx, "error: division by zero");
        YYERROR;
      }
    else
@@ -257,8 +273,9 @@ symbol_count (void)
 | Scanner.  |
 `----------*/

-int
-yylex (const char **line, YYSTYPE *yylval, YYLTYPE *yylloc)
+yytoken_kind_t
+yylex (const char **line, YYSTYPE *yylval, YYLTYPE *yylloc,
+       const user_context *uctx)
 {
  int c;

@@ -328,7 +345,7 @@ yylex (const char **line, YYSTYPE *yylval, YYLTYPE *yylloc)

      // Stray characters.
    default:
-      yyerror (yylloc, "syntax error: invalid character: %c", c);
+      yyerror (yylloc, uctx, "syntax error: invalid character: %c", c);
      return TOK_YYerror;
    }
 }
@@ -366,8 +383,11 @@ error_format_string (int argc)


 int
-yyreport_syntax_error (const yypcontext_t *ctx)
+yyreport_syntax_error (const yypcontext_t *ctx, const user_context *uctx)
 {
+  if (uctx->silent)
+    return 0;
+
  enum { ARGS_MAX = 6 };
  yysymbol_kind_t arg[ARGS_MAX];
  int argsize = yypcontext_expected_tokens (ctx, arg, ARGS_MAX);
@@ -378,11 +398,12 @@ yyreport_syntax_error (const yypcontext_t *ctx)
    argsize = ARGS_MAX;
  const char *format = error_format_string (1 + argsize + too_many_expected_tokens);

+  const YYLTYPE *loc = yypcontext_location (ctx);
  while (*format)
    // %@: location.
    if (format[0] == '%' && format[1] == '@')
      {
-        YY_LOCATION_PRINT (stderr, *yypcontext_location (ctx));
+        YY_LOCATION_PRINT (stderr, *loc);
        format += 2;
      }
    // %u: unexpected token.
@@ -407,13 +428,25 @@ yyreport_syntax_error (const yypcontext_t *ctx)
        ++format;
      }
  fputc ('\n', stderr);
+
+  // Quote the source line.
+  {
+    fprintf (stderr, "%5d | %s\n", loc->first_line, uctx->line);
+    fprintf (stderr, "%5s | %*s", "", loc->first_column, "^");
+    for (int i = loc->last_column - loc->first_column - 1; 0 < i; --i)
+      putc ('~', stderr);
+    putc ('\n', stderr);
+  }
  return 0;
 }


 // Called by yyparse on error.
-void yyerror (YYLTYPE *loc, char const *format, ...)
+void yyerror (YYLTYPE *loc, const user_context *uctx, char const *format, ...)
 {
+  if (uctx->silent)
+    return;
+
  YY_LOCATION_PRINT (stderr, *loc);
  fputs (": ", stderr);
  va_list args;
@@ -449,11 +482,13 @@ xstrndup (const char *string, size_t n)
 static int
 process_line (YYLTYPE *lloc, const char *line)
 {
+  user_context uctx = {0, line};
  yypstate *ps = yypstate_new ();
  int status = 0;
  do {
    YYSTYPE lval;
-    status = yypush_parse (ps, yylex (&line, &lval, lloc), &lval, lloc);
+    yytoken_kind_t token = yylex (&line, &lval, lloc, &uctx);
+    status = yypush_parse (ps, token, &lval, lloc, &uctx);
  } while (status == YYPUSH_MORE);
  yypstate_delete (ps);
  lloc->last_line++;
@@ -468,18 +503,19 @@ expected_tokens (const char *input,
                 int *tokens, int ntokens)
 {
  YYDPRINTF ((stderr, "expected_tokens (\"%s\")", input));
+  user_context uctx = {1, input};

  // Parse the current state of the line.
  yypstate *ps = yypstate_new ();
  int status = 0;
+  YYLTYPE lloc = { 1, 1, 1, 1 };
  do {
-    YYLTYPE lloc = { 1, 1, 1, 1 };
    YYSTYPE lval;
-    int token = yylex (&input, &lval, &lloc);
+    yytoken_kind_t token = yylex (&input, &lval, &lloc, &uctx);
    // Don't let the parse know when we reach the end of input.
-    if (!token)
+    if (token == TOK_YYEOF)
      break;
-    status = yypush_parse (ps, token, &lval, &lloc);
+    status = yypush_parse (ps, token, &lval, &lloc, &uctx);
  } while (status == YYPUSH_MORE);

  int res = 0;
@@ -313,6 +313,7 @@
 /stdlib.h
 /stdlib.in.h
 /stpcpy.c
+/stpncpy.c
 /strchrnul.c
 /strchrnul.valgrind
 /strdup.c
@@ -175,6 +175,7 @@
 /stdio_h.m4
 /stdlib_h.m4
 /stpcpy.m4
+/stpncpy.m4
 /strchrnul.m4
 /strdup.m4
 /strerror.m4
@@ -169,10 +169,10 @@ AnnotationList__compute_conflicted_tokens (bitset shift_tokens,
  bitset_copy (tokens, shift_tokens);
  for (int i = 0; i < reds->num; ++i)
    {
-      bitset_and (conflicted_tokens_rule, tokens, reds->lookahead_tokens[i]);
+      bitset_and (conflicted_tokens_rule, tokens, reds->lookaheads[i]);
      bitset_or (conflicted_tokens,
                 conflicted_tokens, conflicted_tokens_rule);
-      bitset_or (tokens, tokens, reds->lookahead_tokens[i]);
+      bitset_or (tokens, tokens, reds->lookaheads[i]);
      /* Check that rules are sorted on rule number or the next step in
         AnnotationList__compute_from_inadequacies will misbehave.  */
      aver (i == 0 || reds->rules[i-1] < reds->rules[i]);
@@ -401,7 +401,7 @@ AnnotationList__compute_from_inadequacies (
  struct obstack *annotations_obstackp,
  InadequacyListNodeCount *inadequacy_list_node_count)
 {
-  /* Return an empty list if s->lookahead_tokens = NULL.  */
+  /* Return an empty list if s->lookaheads = NULL.  */
  if (s->consistent)
    return;

@@ -422,7 +422,7 @@ AnnotationList__compute_from_inadequacies (
      /* Allocate the annotation node.  */
      {
        for (int rule_i = 0; rule_i < s->reductions->num; ++rule_i)
-          if (bitset_test (s->reductions->lookahead_tokens[rule_i],
+          if (bitset_test (s->reductions->lookaheads[rule_i],
                           conflicted_token))
            ++contribution_count;
        if (bitset_test (shift_tokens, conflicted_token))
@@ -445,7 +445,7 @@ AnnotationList__compute_from_inadequacies (
        for (int rule_i = 0; rule_i < s->reductions->num; ++rule_i)
          {
            rule *the_rule = s->reductions->rules[rule_i];
-            if (bitset_test (s->reductions->lookahead_tokens[rule_i],
+            if (bitset_test (s->reductions->lookaheads[rule_i],
                             conflicted_token))
              {
                bitset_set (actions, rule_i);
@@ -250,9 +250,9 @@ flush_shift (state *s, int token)
 `--------------------------------------------------------------------*/

 static void
-flush_reduce (bitset lookahead_tokens, int token)
+flush_reduce (bitset lookaheads, int token)
 {
-  bitset_reset (lookahead_tokens, token);
+  bitset_reset (lookaheads, token);
 }


@@ -275,10 +275,10 @@ resolve_sr_conflict (state *s, int ruleno, symbol **errors, int *nerrs)
  /* Find the rule to reduce by to get precedence of reduction.  */
  rule *redrule = reds->rules[ruleno];
  int redprec = redrule->prec->prec;
-  bitset lookahead_tokens = reds->lookahead_tokens[ruleno];
+  bitset lookaheads = reds->lookaheads[ruleno];

  for (symbol_number i = 0; i < ntokens; ++i)
-    if (bitset_test (lookahead_tokens, i)
+    if (bitset_test (lookaheads, i)
        && bitset_test (lookahead_set, i)
        && symbols[i]->content->prec)
      {
@@ -295,7 +295,7 @@ resolve_sr_conflict (state *s, int ruleno, symbol **errors, int *nerrs)
          {
            register_precedence (i, redrule->prec->number);
            log_resolution (redrule, i, shift_resolution);
-            flush_reduce (lookahead_tokens, i);
+            flush_reduce (lookaheads, i);
          }
        else
          /* Matching precedence levels.
@@ -316,7 +316,7 @@ resolve_sr_conflict (state *s, int ruleno, symbol **errors, int *nerrs)
            case right_assoc:
              register_assoc (i, redrule->prec->number);
              log_resolution (redrule, i, right_resolution);
-              flush_reduce (lookahead_tokens, i);
+              flush_reduce (lookaheads, i);
              break;

            case left_assoc:
@@ -329,7 +329,7 @@ resolve_sr_conflict (state *s, int ruleno, symbol **errors, int *nerrs)
              register_assoc (i, redrule->prec->number);
              log_resolution (redrule, i, nonassoc_resolution);
              flush_shift (s, i);
-              flush_reduce (lookahead_tokens, i);
+              flush_reduce (lookaheads, i);
              /* Record an explicit error for this token.  */
              errors[(*nerrs)++] = symbols[i];
              break;
@@ -369,7 +369,7 @@ set_conflicts (state *s, symbol **errors)
  for (int i = 0; i < reds->num; ++i)
    if (reds->rules[i]->prec
        && reds->rules[i]->prec->prec
-        && !bitset_disjoint_p (reds->lookahead_tokens[i], lookahead_set))
+        && !bitset_disjoint_p (reds->lookaheads[i], lookahead_set))
      resolve_sr_conflict (s, i, errors, &nerrs);

  if (nerrs)
@@ -385,13 +385,13 @@ set_conflicts (state *s, symbol **errors)
  /* Loop over all rules which require lookahead in this state.  Check
     for conflicts not resolved above.

-     reds->lookahead_tokens can be NULL if the LR type is LR(0).  */
-  if (reds->lookahead_tokens)
+     reds->lookaheads can be NULL if the LR type is LR(0).  */
+  if (reds->lookaheads)
    for (int i = 0; i < reds->num; ++i)
      {
-        if (!bitset_disjoint_p (reds->lookahead_tokens[i], lookahead_set))
+        if (!bitset_disjoint_p (reds->lookaheads[i], lookahead_set))
          conflicts[s->number] = true;
-        bitset_or (lookahead_set, lookahead_set, reds->lookahead_tokens[i]);
+        bitset_or (lookahead_set, lookahead_set, reds->lookaheads[i]);
      }
 }

@@ -460,7 +460,7 @@ count_state_sr_conflicts (const state *s)
  }

  for (int i = 0; i < reds->num; ++i)
-    bitset_or (lookahead_set, lookahead_set, reds->lookahead_tokens[i]);
+    bitset_or (lookahead_set, lookahead_set, reds->lookaheads[i]);

  bitset_and (lookahead_set, lookahead_set, shift_set);

@@ -499,7 +499,7 @@ count_state_rr_conflicts (const state *s)
    {
      int count = 0;
      for (int j = 0; j < reds->num; ++j)
-        count += bitset_test (reds->lookahead_tokens[j], i);
+        count += bitset_test (reds->lookaheads[j], i);
      if (2 <= count)
        res += count-1;
    }
@@ -534,7 +534,7 @@ count_rule_state_sr_conflicts (rule *r, state *s)
  for (int i = 0; i < reds->num; ++i)
    if (reds->rules[i] == r)
      {
-        bitset lookaheads = reds->lookahead_tokens[i];
+        bitset lookaheads = reds->lookaheads[i];
        int j;
        FOR_EACH_SHIFT (trans, j)
          res += bitset_test (lookaheads, TRANSITION_SYMBOL (trans, j));
@@ -576,8 +576,8 @@ count_rule_state_rr_conflicts (rule *r, state *s)
        if (reds->rules[j] != r)
          {
            bitset_and (lookaheads,
-                        reds->lookahead_tokens[i],
-                        reds->lookahead_tokens[j]);
+                        reds->lookaheads[i],
+                        reds->lookaheads[j]);
            res += bitset_count (lookaheads);
          }
  bitset_free (lookaheads);
@@ -76,17 +76,29 @@ typedef struct
 {
  derivation *d1;
  derivation *d2;
+  bool shift_reduce;
  bool unifying;
  bool timeout;
 } counterexample;

 static counterexample *
 new_counterexample (derivation *d1, derivation *d2,
+                    bool shift_reduce,
                    bool u, bool t)
 {
-  counterexample *res = xmalloc (sizeof (counterexample));
-  res->d1 = d1;
-  res->d2 = d2;
+  counterexample *res = xmalloc (sizeof *res);
+  res->shift_reduce = shift_reduce;
+  if (shift_reduce)
+    {
+      // Display the shift first.
+      res->d1 = d2;
+      res->d2 = d1;
+    }
+  else
+    {
+      res->d1 = d1;
+      res->d2 = d2;
+    }
  res->unifying = u;
  res->timeout = t;
  return res;
@@ -101,13 +113,14 @@ free_counterexample (counterexample *cex)
 }

 static void
-print_counterexample (counterexample *cex, FILE *out, const char *prefix)
+print_counterexample (const counterexample *cex, FILE *out, const char *prefix)
 {
+  const bool flat = getenv ("YYFLAT");
  fprintf (out, "  %s%-20s ",
           prefix, cex->unifying ? _("Example") : _("First example"));
  derivation_print_leaves (cex->d1, out, prefix);
-  fprintf (out, "  %s%-20s ",
-           prefix, _("First derivation"));
+  fprintf (out, flat ? "  %s%-20s " : "  %s%s",
+           prefix, cex->shift_reduce ? _("Shift derivation") : _("First derivation"));
  derivation_print (cex->d1, out, prefix);

  // If we output to the terminal (via stderr) and we have color
@@ -119,8 +132,8 @@ print_counterexample (counterexample *cex, FILE *out, const char *prefix)
               prefix, cex->unifying ? _("Example") : _("Second example"));
      derivation_print_leaves (cex->d2, out, prefix);
    }
-  fprintf (out, "  %s%-20s ",
-           prefix, _("Second derivation"));
+  fprintf (out, flat ? "  %s%-20s " : "  %s%s",
+           prefix, cex->shift_reduce ? _("Reduce derivation") : _("Second derivation"));
  derivation_print (cex->d2, out, prefix);

  fputc ('\n', out);
@@ -144,7 +157,7 @@ typedef struct si_bfs_node
 static si_bfs_node *
 si_bfs_new (state_item_number si, si_bfs_node *parent)
 {
-  si_bfs_node *res = xcalloc (1, sizeof (si_bfs_node));
+  si_bfs_node *res = xcalloc (1, sizeof *res);
  res->si = si;
  res->parent = parent;
  res->reference_count = 1;
@@ -175,6 +188,8 @@ si_bfs_free (si_bfs_node *n)
    }
 }

+typedef gl_list_t si_bfs_node_list;
+
 /**
 * start is a state_item such that conflict_sym is an element of FIRSTS of the
 * nonterminal after the dot in start. Because of this, we should be able to
@@ -188,15 +203,16 @@ expand_to_conflict (state_item_number start, symbol_number conflict_sym)
 {
  si_bfs_node *init = si_bfs_new (start, NULL);

-  gl_list_t queue = gl_list_create (GL_LINKED_LIST, NULL, NULL,
-                                    (gl_listelement_dispose_fn) si_bfs_free,
-                                    true, 1, (const void **) &init);
+  si_bfs_node_list queue
+    = gl_list_create (GL_LINKED_LIST, NULL, NULL,
+                      (gl_listelement_dispose_fn) si_bfs_free,
+                      true, 1, (const void **) &init);
  si_bfs_node *node = NULL;
  // breadth-first search for a path of productions to the conflict symbol
  while (gl_list_size (queue) > 0)
    {
      node = (si_bfs_node *) gl_list_get_at (queue, 0);
-      state_item *silast = state_items + node->si;
+      state_item *silast = &state_items[node->si];
      symbol_number sym = item_number_as_symbol_number (*silast->item);
      if (sym == conflict_sym)
        break;
@@ -238,7 +254,7 @@ expand_to_conflict (state_item_number start, symbol_number conflict_sym)

  for (si_bfs_node *n = node; n != NULL; n = n->parent)
    {
-      state_item *si = state_items + n->si;
+      state_item *si = &state_items[n->si];
      item_number *pos = si->item;
      if (SI_PRODUCTION (si))
        {
@@ -274,7 +290,7 @@ expand_to_conflict (state_item_number start, symbol_number conflict_sym)
 */
 static derivation *
 complete_diverging_example (symbol_number conflict_sym,
-                            gl_list_t path, derivation_list derivs)
+                            state_item_list path, derivation_list derivs)
 {
  // The idea is to transfer each pending symbol on the productions
  // associated with the given StateItems to the resulting derivation.
@@ -395,15 +411,15 @@ complete_diverging_example (symbol_number conflict_sym,
 /* Iterate backwards through the shifts of the path in the reduce
   conflict, and find a path of shifts from the shift conflict that
   goes through the same states. */
-static gl_list_t
-nonunifying_shift_path (gl_list_t reduce_path, state_item *shift_conflict)
+static state_item_list
+nonunifying_shift_path (state_item_list reduce_path, state_item *shift_conflict)
 {
  gl_list_node_t tmp = gl_list_add_last (reduce_path, NULL);
  gl_list_node_t next_node = gl_list_previous_node (reduce_path, tmp);
  gl_list_node_t node = gl_list_previous_node (reduce_path, next_node);
  gl_list_remove_node (reduce_path, tmp);
  state_item *si = shift_conflict;
-  gl_list_t result =
+  state_item_list result =
    gl_list_create_empty (GL_LINKED_LIST, NULL, NULL, NULL, true);
  // FIXME: bool paths_merged;
  for (; node != NULL; next_node = node,
@@ -425,10 +441,10 @@ nonunifying_shift_path (gl_list_t reduce_path, state_item *shift_conflict)

      // bfs to find a shift to the right state
      si_bfs_node *init = si_bfs_new (si - state_items, NULL);
-      gl_list_t queue =
-        gl_list_create (GL_LINKED_LIST, NULL, NULL,
-                        (gl_listelement_dispose_fn) si_bfs_free,
-                        true, 1, (const void **) &init);
+      si_bfs_node_list queue
+        = gl_list_create (GL_LINKED_LIST, NULL, NULL,
+                          (gl_listelement_dispose_fn) si_bfs_free,
+                          true, 1, (const void **) &init);
      si_bfs_node *sis = NULL;
      state_item *prevsi = NULL;
      while (gl_list_size (queue) > 0)
@@ -438,7 +454,7 @@ nonunifying_shift_path (gl_list_t reduce_path, state_item *shift_conflict)
          if (sis->si == 0)
            break;

-          state_item *search_si = state_items + sis->si;
+          state_item *search_si = &state_items[sis->si];
          // if the current state-item is a production item,
          // its reverse production items get added to the queue.
          // Otherwise, look for a reverse transition to the target state.
@@ -447,7 +463,7 @@ nonunifying_shift_path (gl_list_t reduce_path, state_item *shift_conflict)
          state_item_number sin;
          BITSET_FOR_EACH (biter, rsi, sin, 0)
            {
-              prevsi = state_items + sin;
+              prevsi = &state_items[sin];
              if (SI_TRANSITION (search_si))
                {
                  if (prevsi->state == refsi->state)
@@ -465,9 +481,9 @@ nonunifying_shift_path (gl_list_t reduce_path, state_item *shift_conflict)
      // prepend path to shift we found
      if (sis)
        {
-          gl_list_node_t ln = gl_list_add_first (result, state_items + sis->si);
+          gl_list_node_t ln = gl_list_add_first (result, &state_items[sis->si]);
          for (si_bfs_node *n = sis->parent; n; n = n->parent)
-            ln = gl_list_add_after (result, ln, state_items + n->si);
+            ln = gl_list_add_after (result, ln, &state_items[n->si]);

        }
      si = prevsi;
@@ -493,17 +509,17 @@ nonunifying_shift_path (gl_list_t reduce_path, state_item *shift_conflict)
 static counterexample *
 example_from_path (bool shift_reduce,
                   state_item_number itm2,
-                   gl_list_t shortest_path, symbol_number next_sym)
+                   state_item_list shortest_path, symbol_number next_sym)
 {
  derivation *deriv1 =
    complete_diverging_example (next_sym, shortest_path, NULL);
-  gl_list_t path_2
+  state_item_list path_2
    = shift_reduce
    ? nonunifying_shift_path (shortest_path, &state_items [itm2])
    : shortest_path_from_start (itm2, next_sym);
  derivation *deriv2 = complete_diverging_example (next_sym, path_2, NULL);
  gl_list_free (path_2);
-  return new_counterexample (deriv1, deriv2, false, true);
+  return new_counterexample (deriv1, deriv2, shift_reduce, false, true);
 }

 /*
@@ -525,7 +541,7 @@ typedef struct
 static search_state *
 initial_search_state (state_item *conflict1, state_item *conflict2)
 {
-  search_state *res = xmalloc (sizeof (search_state));
+  search_state *res = xmalloc (sizeof *res);
  res->states[0] = new_parse_state (conflict1);
  res->states[1] = new_parse_state (conflict2);
  parse_state_retain (res->states[0]);
@@ -537,7 +553,7 @@ initial_search_state (state_item *conflict1, state_item *conflict2)
 static search_state *
 new_search_state (parse_state *ps1, parse_state *ps2, int complexity)
 {
-  search_state *res = xmalloc (sizeof (search_state));
+  search_state *res = xmalloc (sizeof *res);
  res->states[0] = ps1;
  res->states[1] = ps2;
  parse_state_retain (res->states[0]);
@@ -549,8 +565,8 @@ new_search_state (parse_state *ps1, parse_state *ps2, int complexity)
 static search_state *
 copy_search_state (search_state *parent)
 {
-  search_state *res = xmalloc (sizeof (search_state));
-  memcpy (res, parent, sizeof (search_state));
+  search_state *res = xmalloc (sizeof *res);
+  *res = *parent;
  parse_state_retain (res->states[0]);
  parse_state_retain (res->states[1]);
  return res;
@@ -583,6 +599,8 @@ search_state_print (search_state *ss)
  putc ('\n', stderr);
 }

+typedef gl_list_t search_state_list;
+
 static inline bool
 search_state_list_next (gl_list_iterator_t *it, search_state **ss)
 {
@@ -614,18 +632,20 @@ ss_set_parse_state (search_state *ss, int idx, parse_state *ps)
 */
 static counterexample *
 complete_diverging_examples (search_state *ss,
-                             symbol_number next_sym)
+                             symbol_number next_sym,
+                             bool shift_reduce)
 {
  derivation *new_derivs[2];
  for (int i = 0; i < 2; ++i)
    {
-      gl_list_t sitems;
+      state_item_list sitems;
      derivation_list derivs;
      parse_state_lists (ss->states[i], &sitems, &derivs);
      new_derivs[i] = complete_diverging_example (next_sym, sitems, derivs);
      gl_list_free (sitems);
    }
-  return new_counterexample (new_derivs[0], new_derivs[1], false, true);
+  return new_counterexample (new_derivs[0], new_derivs[1],
+                             shift_reduce, false, true);
 }

 /*
@@ -635,7 +655,7 @@ complete_diverging_examples (search_state *ss,
 */
 typedef struct
 {
-  gl_list_t states;
+  search_state_list states;
  int complexity;
 } search_state_bundle;

@@ -664,6 +684,8 @@ ssb_equals (const search_state_bundle *s1, const search_state_bundle *s2)
  return s1->complexity == s2->complexity;
 }

+typedef gl_list_t ssb_list;
+
 static size_t
 visited_hasher (const search_state *ss, size_t max)
 {
@@ -679,7 +701,7 @@ visited_comparator (const search_state *ss1, const search_state *ss2)
 }

 /* Priority queue for search states with minimal complexity. */
-static gl_list_t ssb_queue;
+static ssb_list ssb_queue;
 static Hash_table *visited;
 /* The set of parser states on the shortest lookahead-sensitive path. */
 static bitset scp_set = NULL;
@@ -702,7 +724,7 @@ ssb_append (search_state *ss)
  parse_state_free_contents_early (ss->states[1]);
  parse_state_retain (ss->states[0]);
  parse_state_retain (ss->states[1]);
-  search_state_bundle *ssb = xmalloc (sizeof (search_state_bundle));
+  search_state_bundle *ssb = xmalloc (sizeof *ssb);
  ssb->complexity = ss->complexity;
  gl_list_node_t n = gl_list_search (ssb_queue, ssb);
  if (!n)
@@ -756,12 +778,12 @@ reduction_cost (const parse_state *ps)
  return SHIFT_COST * shifts + PRODUCTION_COST * productions;
 }

-static gl_list_t
+static search_state_list
 reduction_step (search_state *ss, const item_number *conflict_item,
                int parser_state, int rule_len)
 {
  (void) conflict_item; // FIXME: Unused
-  gl_list_t result =
+  search_state_list result =
    gl_list_create_empty (GL_LINKED_LIST, NULL, NULL, NULL, 1);

  parse_state *ps = ss->states[parser_state];
@@ -897,7 +919,7 @@ search_state_prepend (search_state *ss, symbol_number sym, bitset guide)
 * the same prefix up to the dot.
 */
 static bool
-has_common_prefix (const item_number *itm1, const item_number *itm2)
+have_common_prefix (const item_number *itm1, const item_number *itm2)
 {
  int i = 0;
  for (; !item_number_is_rule_number (itm1[i]); ++i)
@@ -988,14 +1010,14 @@ generate_next_states (search_state *ss, state_item *conflict1,
      // prepended further, reduce.
      if (ready1 && ready2)
        {
-          gl_list_t reduced1 = reduction_step (ss, conflict1->item, 0, len1);
+          search_state_list reduced1 = reduction_step (ss, conflict1->item, 0, len1);
          gl_list_add_last (reduced1, ss);
          search_state *red1 = NULL;
          for (gl_list_iterator_t iter = gl_list_iterator (reduced1);
               search_state_list_next (&iter, &red1);
               )
            {
-              gl_list_t reduced2 =
+              search_state_list reduced2 =
                reduction_step (red1, conflict2->item, 1, len2);
              search_state *red2 = NULL;
              for (gl_list_iterator_t iter2 = gl_list_iterator (reduced2);
@@ -1011,7 +1033,7 @@ generate_next_states (search_state *ss, state_item *conflict1,
        }
      else if (ready1)
        {
-          gl_list_t reduced1 = reduction_step (ss, conflict1->item, 0, len1);
+          search_state_list reduced1 = reduction_step (ss, conflict1->item, 0, len1);
          search_state *red1 = NULL;
          for (gl_list_iterator_t iter = gl_list_iterator (reduced1);
               search_state_list_next (&iter, &red1);
@@ -1021,7 +1043,7 @@ generate_next_states (search_state *ss, state_item *conflict1,
        }
      else if (ready2)
        {
-          gl_list_t reduced2 = reduction_step (ss, conflict2->item, 1, len2);
+          search_state_list reduced2 = reduction_step (ss, conflict2->item, 1, len2);
          search_state *red2 = NULL;
          for (gl_list_iterator_t iter2 = gl_list_iterator (reduced2);
               search_state_list_next (&iter2, &red2);
@@ -1055,10 +1077,10 @@ static counterexample *
 unifying_example (state_item_number itm1,
                  state_item_number itm2,
                  bool shift_reduce,
-                  gl_list_t reduce_path, symbol_number next_sym)
+                  state_item_list reduce_path, symbol_number next_sym)
 {
-  state_item *conflict1 = state_items + itm1;
-  state_item *conflict2 = state_items + itm2;
+  state_item *conflict1 = &state_items[itm1];
+  state_item *conflict2 = &state_items[itm2];
  search_state *initial = initial_search_state (conflict1, conflict2);
  ssb_queue = gl_list_create_empty (GL_RBTREEHASH_LIST,
                                    (gl_listelement_equals_fn) ssb_equals,
@@ -1094,7 +1116,7 @@ unifying_example (state_item_number itm1,
              const state_item *si1src = parse_state_head (ps1);
              const state_item *si2src = parse_state_head (ps2);
              if (item_rule (si1src->item)->lhs == item_rule (si2src->item)->lhs
-                  && has_common_prefix (si1src->item, si2src->item))
+                  && have_common_prefix (si1src->item, si2src->item))
                {
                  // Stage 4: both paths share a prefix
                  derivation *d1 = parse_state_derivation (ps1);
@@ -1104,7 +1126,7 @@ unifying_example (state_item_number itm1,
                    {
                      // Once we have two derivations for the same symbol,
                      // we've found a unifying counterexample.
-                      cex = new_counterexample (d1, d2, true, false);
+                      cex = new_counterexample (d1, d2, shift_reduce, true, false);
                      derivation_retain (d1);
                      derivation_retain (d2);
                      goto cex_search_end;
@@ -1142,7 +1164,7 @@ cex_search_end:;
      // If a search state from Stage 3 is available, use it
      // to construct a more compact nonunifying counterexample.
      if (stage3result)
-        cex = complete_diverging_examples (stage3result, next_sym);
+        cex = complete_diverging_examples (stage3result, next_sym, shift_reduce);
      // Otherwise, construct a nonunifying counterexample that
      // begins from the start state using the shortest
      // lookahead-sensitive path to the reduce item.
@@ -1190,7 +1212,7 @@ counterexample_report (state_item_number itm1, state_item_number itm2,
 {
  // Compute the shortest lookahead-sensitive path and associated sets of
  // parser states.
-  gl_list_t shortest_path = shortest_path_from_start (itm1, next_sym);
+  state_item_list shortest_path = shortest_path_from_start (itm1, next_sym);
  bool reduce_prod_reached = false;
  const rule *reduce_rule = item_rule (state_items[itm1].item);

@@ -1219,6 +1241,8 @@ counterexample_report (state_item_number itm1, state_item_number itm2,
  free_counterexample (cex);
 }

+
+// ITM1 denotes a shift, ITM2 a reduce.
 static void
 counterexample_report_shift_reduce (state_item_number itm1, state_item_number itm2,
                                    symbol_number next_sym,
@@ -1226,7 +1250,8 @@ counterexample_report_shift_reduce (state_item_number itm1, state_item_number it
 {
  fputs (prefix, out);
  fprintf (out, _("Shift/reduce conflict on token %s:\n"), symbols[next_sym]->tag);
-  if (*prefix)
+  // In the report, print the items.
+  if (*prefix || trace_flag & trace_cex)
    {
      print_state_item (&state_items[itm1], out, prefix);
      print_state_item (&state_items[itm2], out, prefix);
@@ -1254,7 +1279,8 @@ counterexample_report_reduce_reduce (state_item_number itm1, state_item_number i
      }
    fputs (_(":\n"), out);
  }
-  if (*prefix)
+  // In the report, print the items.
+  if (*prefix || trace_flag & trace_cex)
    {
      print_state_item (&state_items[itm1], out, prefix);
      print_state_item (&state_items[itm2], out, prefix);
@@ -1265,7 +1291,7 @@ counterexample_report_reduce_reduce (state_item_number itm1, state_item_number i
 static state_item_number
 find_state_item_number (const rule *r, state_number sn)
 {
-  for (int i = state_item_map[sn]; i < state_item_map[sn + 1]; ++i)
+  for (state_item_number i = state_item_map[sn]; i < state_item_map[sn + 1]; ++i)
    if (!SI_DISABLED (i)
        && item_number_as_rule_number (*state_items[i].item) == r->number)
      return i;
@@ -1277,41 +1303,35 @@ counterexample_report_state (const state *s, FILE *out, const char *prefix)
 {
  const state_number sn = s->number;
  const reductions *reds = s->reductions;
+  bitset lookaheads = bitset_create (ntokens, BITSET_FIXED);
  for (int i = 0; i < reds->num; ++i)
    {
      const rule *r1 = reds->rules[i];
      const state_item_number c1 = find_state_item_number (r1, sn);
-      for (int j = state_item_map[sn]; j < state_item_map[sn + 1]; ++j)
-        if (!SI_DISABLED (j))
+      for (state_item_number c2 = state_item_map[sn]; c2 < state_item_map[sn + 1]; ++c2)
+        if (!SI_DISABLED (c2))
          {
-            state_item *si = state_items + j;
-            item_number conf = *si->item;
+            item_number conf = *state_items[c2].item;
            if (item_number_is_symbol_number (conf)
-              && bitset_test (reds->lookahead_tokens[i], conf))
-              counterexample_report_shift_reduce (c1, j, conf, out, prefix);
+                && bitset_test (reds->lookaheads[i], conf))
+              counterexample_report_shift_reduce (c1, c2, conf, out, prefix);
          }
      for (int j = i+1; j < reds->num; ++j)
        {
-          bitset conf = bitset_create (ntokens, BITSET_FIXED);
-          bitset_intersection (conf,
-                               reds->lookahead_tokens[i],
-                               reds->lookahead_tokens[j]);
-          if (!bitset_empty_p (conf))
-            {
-              const rule *r2 = reds->rules[j];
-              for (int k = state_item_map[sn]; k < state_item_map[sn + 1]; ++k)
-                if (!SI_DISABLED (k))
-                  {
-                    state_item *si = state_items + k;
-                    const rule *r = item_rule (si->item);
-                    if (r == r2)
-                      {
-                        counterexample_report_reduce_reduce (c1, k, conf, out, prefix);
-                        break;
-                      }
-                  }
-            }
-          bitset_free (conf);
+          const rule *r2 = reds->rules[j];
+          // Conflicts: common lookaheads.
+          bitset_intersection (lookaheads,
+                               reds->lookaheads[i],
+                               reds->lookaheads[j]);
+          if (!bitset_empty_p (lookaheads))
+            for (state_item_number c2 = state_item_map[sn]; c2 < state_item_map[sn + 1]; ++c2)
+              if (!SI_DISABLED (c2)
+                  && item_rule (state_items[c2].item) == r2)
+                {
+                  counterexample_report_reduce_reduce (c1, c2, lookaheads, out, prefix);
+                  break;
+                }
        }
    }
+  bitset_free (lookaheads);
 }
@@ -20,11 +20,17 @@
 #ifndef COUNTEREXAMPLE_H
 # define COUNTEREXAMPLE_H

-# include "state-item.h"
+# include "state.h"

+// Init/deinit this module.
 void counterexample_init (void);
 void counterexample_free (void);

-void counterexample_report_state (const state *s, FILE *out, const char *prefix);
+// Print the counterexamples for the conflicts of state S.
+//
+// Used both for the warnings on the terminal (OUT = stderr, PREFIX =
+// ""), and for the reports (OUT != stderr, PREFIX != "").
+void
+counterexample_report_state (const state *s, FILE *out, const char *prefix);

 #endif /* COUNTEREXAMPLE_H */
@@ -20,8 +20,11 @@
 #include <config.h>

 #include "derivation.h"
+#include "glyphs.h"

+#include <c-ctype.h>
 #include <gl_linked_list.h>
+#include <mbswidth.h>

 #include "system.h"
 #include "complain.h"
@@ -29,11 +32,15 @@
 struct derivation
 {
  symbol_number sym;
-  gl_list_t children;
+  derivation_list children;
  int reference_count;
+  // Color assigned for styling.  Guarantees that the derivation is
+  // always displayed with the same color, independently of the order
+  // in which the derivations are traversed.
+  int color;
 };

-static derivation d_dot = { -1, NULL, -1 };
+static derivation d_dot = { -1, NULL, -1, -1 };

 derivation *
 derivation_dot (void)
@@ -73,6 +80,7 @@ derivation_new (symbol_number sym, derivation_list children)
  deriv->sym = sym;
  deriv->children = children;
  deriv->reference_count = 0;
+  deriv->color = -1;
  return deriv;
 }

@@ -126,27 +134,243 @@ derivation_size (const derivation *deriv)
  return size;
 }

-/* Print DERIV, colored according to COUNTER.
-   Return false if nothing is printed.  */
+
+static int
+max (int a, int b)
+{
+  return a < b ? b : a;
+}
+
+// Longest distance from root to leaf.
+static int
+derivation_depth (const derivation *deriv)
+{
+  if (deriv->children)
+    {
+      // Children's depth cannot be 0, even if there are no children
+      // (the case of a derivation with an empty RHS).
+      int res = 1;
+      derivation *child;
+      for (gl_list_iterator_t it = gl_list_iterator (deriv->children);
+           derivation_list_next (&it, &child);
+           )
+        res = max (res, derivation_depth (child));
+      return res + 1;
+    }
+  else
+    return 1;
+}
+
 static bool
-derivation_print_impl (const derivation *deriv, FILE *f,
-                       bool leaves_only,
-                       int *counter, const char *prefix)
+all_spaces (const char *s)
+{
+  while (c_isspace (*s))
+    s++;
+  return *s == '\0';
+}
+
+// Printing the derivation as trees without trailing spaces is
+// painful: we cannot simply pad one "column" before moving to the
+// next:
+//
+// exp
+// ↳ x1          e1  foo1                      x1
+//   ↳ x2        ↳ ε ↳ foo2                    ↳ x2
+//     ↳ x3            ↳ foo3                    ↳ x3
+//       ↳ "X" •         ↳ x1        foo4          ↳ "X"
+//                         ↳ x2      ↳ "quuux"
+//                           ↳ x3
+//                             ↳ "X"
+//
+// It's hard for a column to know that it's "last" to decide whether
+// to output the right-padding or not.  So when we need to pad on the
+// right to complete a column, we don't output the spaces, we
+// accumulate the width of padding in *PADDING.
+//
+// Each time we actually print something (non space), we flush that
+// padding.  When we _don't_ print something, its width is added to
+// the current padding.
+//
+// This function implements this.
+//
+// When COND is true, put S on OUT, preceeded by *PADDING white
+// spaces.  Otherwise add the width to *PADDING.  Return the width of
+// S.
+static int
+fputs_if (bool cond, FILE *out, int *padding, const char *s)
+{
+  int res = mbswidth (s, 0);
+  if (cond && !all_spaces (s))
+    {
+      fprintf (out, "%*s%s", *padding, "", s);
+      *padding = 0;
+    }
+  else
+    {
+      *padding += res;
+    }
+  return res;
+}
+
+// The width taken to report this derivation recursively down to its
+// leaves.
+static int
+derivation_width (const derivation *deriv)
 {
+  if (deriv->children)
+    {
+      const symbol *sym = symbols[deriv->sym];
+      int self_width = mbswidth (sym->tag, 0);
+
+      // Arrow and space.
+      int children_width = down_arrow_width;
+      if (gl_list_size (deriv->children) == 0)
+        // Empty rhs.
+        children_width += empty_width;
+      else
+        {
+          derivation *child;
+          for (gl_list_iterator_t it = gl_list_iterator (deriv->children);
+               derivation_list_next (&it, &child);
+               )
+            children_width
+              += derivation_separator_width + derivation_width (child);
+          // No separator at the beginning.
+          children_width -= derivation_separator_width;
+        }
+      return max (self_width, children_width);
+    }
+  else if (deriv == &d_dot)
+    {
+      return dot_width;
+    }
+  else // leaf.
+    {
+      const symbol *sym = symbols[deriv->sym];
+      return mbswidth (sym->tag, 0);
+    }
+}
+
+
+// Print DERIV for DEPTH.
+//
+// The tree is printed from top to bottom with DEPTH ranging from 0 to
+// the total depth of the tree.  DERIV should only printed when we
+// reach its depth, i.e., then DEPTH is 0.
+//
+// When DEPTH is 1 and we're on a subderivation, then we print the RHS
+// of the derivation (in DEPTH 0 we printed its LHS).
+//
+// Return the "logical printed" width.  We might have not have reached
+// that width, in which case the missing spaces are in *PADDING.
+static int
+derivation_print_tree_impl (const derivation *deriv, FILE *out,
+                            int depth, int *padding)
+{
+  const int width = derivation_width (deriv);
+
+  int res = 0;
  if (deriv->children)
    {
      const symbol *sym = symbols[deriv->sym];
      char style[20];
-      snprintf (style, 20, "cex-%d", *counter);
+      snprintf (style, 20, "cex-%d", deriv->color);
+
+      if (depth == 0 || depth == 1)
+        {
+          begin_use_class (style, out);
+          begin_use_class ("cex-step", out);
+        }
+      if (depth == 0)
+        {
+          res += fputs_if (true, out, padding, sym->tag);
+        }
+      else
+        {
+          res += fputs_if (depth == 1, out, padding, down_arrow);
+          if (gl_list_size (deriv->children) == 0)
+            // Empty rhs.
+            res += fputs_if (depth == 1, out, padding, empty);
+          else
+            {
+              bool first = true;
+              derivation *child;
+              for (gl_list_iterator_t it = gl_list_iterator (deriv->children);
+                   derivation_list_next (&it, &child);
+                   )
+                {
+                  if (!first)
+                    res += fputs_if (depth == 1, out, padding, derivation_separator);
+                  res += derivation_print_tree_impl (child, out, depth - 1, padding);
+                  first = false;
+                }
+            }
+        }
+      if (depth == 0 || depth == 1)
+        {
+          end_use_class ("cex-step", out);
+          end_use_class (style, out);
+        }
+      *padding += width - res;
+      res = width;
+    }
+  else if (deriv == &d_dot)
+    {
+      if (depth == 0)
+        begin_use_class ("cex-dot", out);
+      res += fputs_if (depth == 0, out, padding, dot);
+      if (depth == 0)
+        end_use_class ("cex-dot", out);
+    }
+  else // leaf.
+    {
+      const symbol *sym = symbols[deriv->sym];
+      if (depth == 0)
+        begin_use_class ("cex-leaf", out);
+      res += fputs_if (depth == 0, out, padding, sym->tag);
+      if (depth == 0)
+        end_use_class ("cex-leaf", out);
+    }
+  return res;
+}
+
+static void
+derivation_print_tree (const derivation *deriv, FILE *out, const char *prefix)
+{
+  fputc ('\n', out);
+  for (int depth = 0, max_depth = derivation_depth (deriv);
+       depth < max_depth; ++depth)
+    {
+      int padding = 0;
+      fprintf (out, "    %s", prefix);
+      derivation_print_tree_impl (deriv, out, depth, &padding);
+      fputc ('\n', out);
+    }
+}
+
+
+/* Print DERIV, colored according to COUNTER.
+   Return false if nothing is printed.  */
+static bool
+derivation_print_flat_impl (derivation *deriv, FILE *out,
+                            bool leaves_only,
+                            int *counter, const char *prefix)
+{
+  if (deriv->children)
+    {
+      const symbol *sym = symbols[deriv->sym];
+      deriv->color = *counter;
      ++*counter;
-      begin_use_class (style, f);
+      char style[20];
+      snprintf (style, 20, "cex-%d", deriv->color);
+      begin_use_class (style, out);

      if (!leaves_only)
        {
-          fputs (prefix, f);
-          begin_use_class ("cex-step", f);
-          fprintf (f, "%s ::=[ ", sym->tag);
-          end_use_class ("cex-step", f);
+          fputs (prefix, out);
+          begin_use_class ("cex-step", out);
+          fprintf (out, "%s %s [ ", sym->tag, arrow);
+          end_use_class ("cex-step", out);
          prefix = "";
        }
      bool res = false;
@@ -155,7 +379,8 @@ derivation_print_impl (const derivation *deriv, FILE *f,
           derivation_list_next (&it, &child);
           )
        {
-          if (derivation_print_impl (child, f, leaves_only, counter, prefix))
+          if (derivation_print_flat_impl (child, out,
+                                          leaves_only, counter, prefix))
            {
              prefix = " ";
              res = true;
@@ -165,49 +390,57 @@ derivation_print_impl (const derivation *deriv, FILE *f,
        }
      if (!leaves_only)
        {
-          begin_use_class ("cex-step", f);
+          begin_use_class ("cex-step", out);
          if (res)
-            fputs (" ]", f);
+            fputs (" ]", out);
          else
-            fputs ("]", f);
-          end_use_class ("cex-step", f);
+            fputs ("]", out);
+          end_use_class ("cex-step", out);
        }
-      end_use_class (style, f);
+      end_use_class (style, out);
      return res;
    }
  else if (deriv == &d_dot)
    {
-      fputs (prefix, f);
-      begin_use_class ("cex-dot", f);
-      print_dot (f);
-      end_use_class ("cex-dot", f);
+      fputs (prefix, out);
+      begin_use_class ("cex-dot", out);
+      fputs (dot, out);
+      end_use_class ("cex-dot", out);
    }
  else // leaf.
    {
-      fputs (prefix, f);
+      fputs (prefix, out);
      const symbol *sym = symbols[deriv->sym];
-      begin_use_class ("cex-leaf", f);
-      fprintf (f, "%s", sym->tag);
-      end_use_class ("cex-leaf", f);
+      begin_use_class ("cex-leaf", out);
+      fprintf (out, "%s", sym->tag);
+      end_use_class ("cex-leaf", out);
    }
  return true;
 }

-void
-derivation_print (const derivation *deriv, FILE *out, const char *prefix)
+static void
+derivation_print_flat (const derivation *deriv, FILE *out, const char *prefix)
 {
  int counter = 0;
  fputs (prefix, out);
-  derivation_print_impl (deriv, out, false, &counter, "");
+  derivation_print_flat_impl ((derivation *)deriv, out, false, &counter, "");
  fputc ('\n', out);
 }

-
 void
 derivation_print_leaves (const derivation *deriv, FILE *out, const char *prefix)
 {
  int counter = 0;
  fputs (prefix, out);
-  derivation_print_impl (deriv, out, true, &counter, "");
+  derivation_print_flat_impl ((derivation *)deriv, out, true, &counter, "");
  fputc ('\n', out);
 }
+
+void
+derivation_print (const derivation *deriv, FILE *out, const char *prefix)
+{
+  if (getenv ("YYFLAT"))
+    derivation_print_flat (deriv, out, prefix);
+  else
+    derivation_print_tree (deriv, out, prefix);
+}
@@ -60,12 +60,15 @@ static inline derivation *derivation_new_leaf (symbol_number sym)
 {
  return derivation_new (sym, NULL);
 }
+
+// Number of symbols.
 size_t derivation_size (const derivation *deriv);
 void derivation_print (const derivation *deriv, FILE *out, const char *prefix);
 void derivation_print_leaves (const derivation *deriv, FILE *out, const char *prefix);
 void derivation_free (derivation *deriv);
 void derivation_retain (derivation *deriv);

+// A derivation denoting the position of the dot.
 derivation *derivation_dot (void);

 #endif /* DERIVATION_H */
@@ -221,7 +221,7 @@ static const argmatch_report_arg argmatch_report_args[] =
  { "none",            report_none },
  { "states",          report_states },
  { "itemsets",        report_states | report_itemsets },
-  { "lookaheads",      report_states | report_lookahead_tokens },
+  { "lookaheads",      report_states | report_lookaheads },
  { "solved",          report_states | report_solved_conflicts },
  { "counterexamples", report_cex },
  { "cex",             report_cex },
@@ -77,7 +77,7 @@ enum report
    report_none             = 0,
    report_states           = 1 << 0,
    report_itemsets         = 1 << 1,
-    report_lookahead_tokens = 1 << 2,
+    report_lookaheads       = 1 << 2,
    report_solved_conflicts = 1 << 3,
    report_cex              = 1 << 4,
    report_all              = ~0
@@ -0,0 +1,103 @@
+/* Graphical symbols.
+
+   Copyright (C) 2020 Free Software Foundation, Inc.
+
+   This file is part of Bison, the GNU Compiler Compiler.
+
+   This program is free software: you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation, either version 3 of the License, or
+   (at your option) any later version.
+
+   This program is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with this program.  If not, see <http://www.gnu.org/licenses/>.  */
+
+#include <config.h>
+
+#include "glyphs.h"
+
+#include <assert.h>
+#include <attribute.h>
+#include <stdbool.h>
+#include <string.h>
+#include <mbswidth.h>
+#include <unicodeio.h>
+
+// In gnulib/lib/unicodeio.h unicode_to_mb uses a buffer of 25 bytes.
+typedef char glyph_buffer_t[26];
+
+
+static glyph_buffer_t arrow_buf;
+const char *arrow;
+int arrow_width;
+
+static glyph_buffer_t down_arrow_buf;
+const char *down_arrow;
+int down_arrow_width;
+
+static glyph_buffer_t dot_buf;
+const char *dot;
+int dot_width;
+
+static glyph_buffer_t empty_buf;
+const char *empty;
+int empty_width;
+
+const char *derivation_separator = " ";
+int derivation_separator_width = 1;
+
+typedef struct
+{
+  const char **glyph;
+  char *buf;
+  const char *fallback;
+} callback_arg_t;
+
+
+static long
+on_success (const char *buf, size_t buflen, void *callback_arg)
+{
+  callback_arg_t *arg = (callback_arg_t *) callback_arg;
+  assert (buflen + 1 < sizeof arg->buf);
+  *stpncpy (arg->buf, buf, buflen) = '\0';
+  return 1;
+}
+
+static long
+on_failure (unsigned code MAYBE_UNUSED, const char *msg MAYBE_UNUSED,
+            void *callback_arg)
+{
+  callback_arg_t *arg = (callback_arg_t *) callback_arg;
+  assert (strlen (arg->fallback) + 1 < sizeof arg->buf);
+  strcpy (arg->buf, arg->fallback);
+  return 0;
+}
+
+static bool
+glyph_set (const char **glyph,
+           char buf[26], int *width,
+           unsigned code, const char *fallback)
+{
+  callback_arg_t arg = { glyph, buf, fallback };
+  int res = unicode_to_mb (code, on_success, on_failure, &arg);
+  *glyph = buf;
+  *width = mbswidth (*glyph, 0);
+  return res;
+}
+
+void
+glyphs_init (void)
+{
+  glyph_set (&arrow,      arrow_buf,      &arrow_width,      0x2192, "->");
+  glyph_set (&dot,        dot_buf,        &dot_width,        0x2022, ".");
+  glyph_set (&down_arrow, down_arrow_buf, &down_arrow_width, 0x21b3, "`->");
+  glyph_set (&empty,      empty_buf,      &empty_width,      0x03b5, "%empty");
+
+  strncat (down_arrow_buf, " ", sizeof down_arrow_buf - strlen (down_arrow_buf) - 1);
+  down_arrow_width += 1;
+}
@@ -0,0 +1,46 @@
+/* Graphical symbols.
+
+   Copyright (C) 2020 Free Software Foundation, Inc.
+
+   This file is part of Bison, the GNU Compiler Compiler.
+
+   This program is free software: you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation, either version 3 of the License, or
+   (at your option) any later version.
+
+   This program is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with this program.  If not, see <http://www.gnu.org/licenses/>.  */
+
+#ifndef GLYPHS_H
+# define GLYPHS_H
+
+/* Initialize the following variables.  */
+void glyphs_init (void);
+
+/* "→", separates the lhs of a rule from its rhs.  */
+extern const char *arrow;
+extern int arrow_width;
+
+/* "•", a point in an item (aka, a dotted rule).  */
+extern const char *dot;
+extern int dot_width;
+
+/* "↳ ", below an lhs to announce the rhs.  */
+extern const char *down_arrow;
+extern int down_arrow_width;
+
+/* "ε", an empty rhs.  */
+extern const char *empty;
+extern int empty_width;
+
+/* " ", separate symbols in the rhs of a derivation.  */
+extern const char *derivation_separator;
+extern int derivation_separator_width;
+
+#endif /* GLYPHS_H */
@@ -23,6 +23,7 @@

 #include "complain.h"
 #include "getargs.h"
+#include "glyphs.h"
 #include "gram.h"
 #include "print-xml.h"
 #include "reader.h"
@@ -56,8 +57,7 @@ item_print (item_number *item, rule const *previous_rule, FILE *out)

  for (item_number *sp = r->rhs; sp < item; sp++)
    fprintf (out, " %s", symbols[*sp]->tag);
-  putc (' ', out);
-  print_dot (out);
+  fprintf (out, " %s", dot);
  if (0 <= *r->rhs)
    for (item_number *sp = item; 0 <= *sp; ++sp)
      fprintf (out, " %s", symbols[*sp]->tag);
@@ -103,8 +103,6 @@

 # include "system.h"

-# include <unicodeio.h>
-
 # include "location.h"
 # include "symtab.h"

@@ -217,25 +215,6 @@ typedef struct
 extern rule *rules;
 extern rule_number nrules;

-/* Fallback in case we can't print "•".  */
-static inline long
-print_dot_fallback (unsigned int code _GL_UNUSED,
-                    const char *msg _GL_UNUSED,
-                    void *callback_arg)
-{
-  FILE *out = (FILE *) callback_arg;
-  putc ('.', out);
-  return -1;
-}
-
-/* Print "•", the symbol used to represent a point in an item (aka, a
-   dotted rule).  */
-static inline void
-print_dot (FILE *out)
-{
-  unicode_to_mb (0x2022, fwrite_success_callback, print_dot_fallback, out);
-}
-
 /* Get the rule associated to this item.  ITEM points inside RITEM.  */
 static inline rule const *
 item_rule (item_number const *item)
@@ -184,9 +184,9 @@ output_red (state const *s, reductions const *reds, FILE *fout)
      bool firste = true;
      rule_number ruleno = reds->rules[j]->number;

-      if (reds->lookahead_tokens)
+      if (reds->lookaheads)
        for (int i = 0; i < ntokens; i++)
-          if (bitset_test (reds->lookahead_tokens[j], i))
+          if (bitset_test (reds->lookaheads[j], i))
            {
              if (bitset_test (no_reduce_set, i))
                firstd = print_token (&dout, firstd, symbols[i]->tag);
@@ -1025,7 +1025,7 @@ ielr_split_states (bitsetv follow_kernel_items, bitsetv always_follows,
              {
                rule *this_rule = node->state->reductions->rules[r];
                bitset lookahead_set =
-                  node->state->reductions->lookahead_tokens[r];
+                  node->state->reductions->lookaheads[r];
                if (item_number_is_rule_number (*this_rule->rhs))
                  ielr_compute_goto_follow_set (follow_kernel_items,
                                                always_follows, node,
@@ -256,9 +256,9 @@ lookback_find_state (int lookback_index)
  state *res = NULL;
  for (int j = 0; j < nstates; ++j)
    if (states[j]->reductions
-        && states[j]->reductions->lookahead_tokens)
+        && states[j]->reductions->lookaheads)
      {
-        if (states[j]->reductions->lookahead_tokens - LA > lookback_index)
+        if (states[j]->reductions->lookaheads - LA > lookback_index)
          /* Went too far. */
          break;
        else
@@ -280,7 +280,7 @@ lookback_print (FILE *out)
    {
      fprintf (out, "   %3d = ", i);
      const state *s = lookback_find_state (i);
-      int rnum = i - (s->reductions->lookahead_tokens - LA);
+      int rnum = i - (s->reductions->lookaheads - LA);
      const rule *r = s->reductions->rules[rnum];
      fprintf (out, "(%3d, ", s->number);
      rule_print (r, NULL, out);
@@ -305,7 +305,7 @@ static void
 add_lookback_edge (state *s, rule const *r, goto_number gotono)
 {
  int ri = state_reduction_find (s, r);
-  int idx = (s->reductions->lookahead_tokens - LA) + ri;
+  int idx = (s->reductions->lookaheads - LA) + ri;
  lookback[idx] = goto_list_new (gotono, lookback[idx]);
 }

@@ -421,7 +421,7 @@ compute_follows (void)


 static void
-compute_lookahead_tokens (void)
+compute_lookaheads (void)
 {
  if (trace_flag & trace_automaton)
      lookback_print (stderr);
@@ -437,13 +437,12 @@ compute_lookahead_tokens (void)
 }


-/*----------------------------------------------------.
-| Count the number of lookahead tokens required for S |
-| (N_LOOKAHEAD_TOKENS member).                        |
-`----------------------------------------------------*/
+/*------------------------------------------------------.
+| Count the number of lookahead tokens required for S.  |
+`------------------------------------------------------*/

 static int
-state_lookahead_tokens_count (state *s, bool default_reduction_only_for_accept)
+state_lookaheads_count (state *s, bool default_reduction_only_for_accept)
 {
  const reductions *reds = s->reductions;
  const transitions *trans = s->transitions;
@@ -473,9 +472,9 @@ state_lookahead_tokens_count (state *s, bool default_reduction_only_for_accept)
 }


-/*----------------------------------------------------.
-| Compute LA, NLA, and the lookahead_tokens members.  |
-`----------------------------------------------------*/
+/*----------------------------------------------.
+| Compute LA, NLA, and the lookaheads members.  |
+`----------------------------------------------*/

 void
 initialize_LA (void)
@@ -491,25 +490,23 @@ initialize_LA (void)
  /* Compute the total number of reductions requiring a lookahead.  */
  nLA = 0;
  for (state_number i = 0; i < nstates; ++i)
-    nLA +=
-      state_lookahead_tokens_count (states[i],
-                                    default_reduction_only_for_accept);
+    nLA += state_lookaheads_count (states[i],
+                                   default_reduction_only_for_accept);
  /* Avoid having to special case 0.  */
  if (!nLA)
    nLA = 1;

  bitsetv pLA = LA = bitsetv_create (nLA, ntokens, BITSET_FIXED);

-  /* Initialize the members LOOKAHEAD_TOKENS for each state whose reductions
+  /* Initialize the members LOOKAHEADS for each state whose reductions
     require lookahead tokens.  */
  for (state_number i = 0; i < nstates; ++i)
    {
-      int count =
-        state_lookahead_tokens_count (states[i],
-                                      default_reduction_only_for_accept);
+      int count = state_lookaheads_count (states[i],
+                                          default_reduction_only_for_accept);
      if (count)
        {
-          states[i]->reductions->lookahead_tokens = pLA;
+          states[i]->reductions->lookaheads = pLA;
          pLA += count;
        }
    }
@@ -521,7 +518,7 @@ initialize_LA (void)
 `---------------------------------------------*/

 static void
-lookahead_tokens_print (FILE *out)
+lookaheads_print (FILE *out)
 {
  fputs ("Lookaheads:\n", out);
  for (state_number i = 0; i < nstates; ++i)
@@ -533,11 +530,11 @@ lookahead_tokens_print (FILE *out)
          for (int j = 0; j < reds->num; ++j)
            {
              fprintf (out, "    rule %d:", reds->rules[j]->number);
-              if (reds->lookahead_tokens)
+              if (reds->lookaheads)
              {
                bitset_iterator iter;
                int k;
-                BITSET_FOR_EACH (iter, reds->lookahead_tokens[j], k, 0)
+                BITSET_FOR_EACH (iter, reds->lookaheads[j], k, 0)
                  fprintf (out, " %s", symbols[k]->tag);
              }
              fputc ('\n', out);
@@ -564,10 +561,10 @@ lalr (void)
  lookback = xcalloc (nLA, sizeof *lookback);
  build_relations ();
  compute_follows ();
-  compute_lookahead_tokens ();
+  compute_lookaheads ();

  if (trace_flag & trace_sets)
-    lookahead_tokens_print (stderr);
+    lookaheads_print (stderr);
  if (trace_flag & trace_automaton)
    {
      begin_use_class ("trace0", stderr);
@@ -614,6 +611,6 @@ void
 lalr_free (void)
 {
  for (state_number s = 0; s < nstates; ++s)
-    states[s]->reductions->lookahead_tokens = NULL;
+    states[s]->reductions->lookaheads = NULL;
  bitsetv_free (LA);
 }
@@ -53,6 +53,8 @@ src_bison_SOURCES =                             \
  src/flex-scanner.h                            \
  src/getargs.c                                 \
  src/getargs.h                                 \
+  src/glyphs.c                                  \
+  src/glyphs.h                                  \
  src/gram.c                                    \
  src/gram.h                                    \
  src/graphviz.c                                \
@@ -82,8 +82,10 @@ lssi_comparator (lssi *s1, lssi *s2)
  return false;
 }

+typedef gl_list_t lssi_list;
+
 static inline bool
-append_lssi (lssi *sn, Hash_table *visited, gl_list_t queue)
+append_lssi (lssi *sn, Hash_table *visited, lssi_list queue)
 {
  if (hash_lookup (visited, sn))
    {
@@ -100,7 +102,7 @@ append_lssi (lssi *sn, Hash_table *visited, gl_list_t queue)
 static void
 lssi_print (lssi *l)
 {
-  print_state_item (state_items + l->si, stdout);
+  print_state_item (&state_items[l->si], stdout);
  if (l->lookahead)
    {
      printf ("FOLLOWL = { ");
@@ -121,7 +123,7 @@ static bitset
 eligible_state_items (state_item *target)
 {
  bitset result = bitset_create (nstate_items, BITSET_FIXED);
-  gl_list_t queue =
+  state_item_list queue =
    gl_list_create (GL_LINKED_LIST, NULL, NULL, NULL, true, 1,
                    (const void **) &target);
  while (gl_list_size (queue) > 0)
@@ -147,7 +149,7 @@ eligible_state_items (state_item *target)
 * this conflict. If optimized is true, only consider parser states
 * that can reach the conflict state.
 */
-gl_list_t
+state_item_list
 shortest_path_from_start (state_item_number target, symbol_number next_sym)
 {
  bitset eligible = eligible_state_items (&state_items[target]);
@@ -159,7 +161,7 @@ shortest_path_from_start (state_item_number target, symbol_number next_sym)
  bitset il = bitset_create (nsyms, BITSET_FIXED);
  bitset_set (il, 0);
  lssi *init = new_lssi (0, NULL, il, true);
-  gl_list_t queue = gl_list_create_empty (GL_LINKED_LIST, NULL, NULL,
+  lssi_list queue = gl_list_create_empty (GL_LINKED_LIST, NULL, NULL,
                                          NULL, true);
  append_lssi (init, visited, queue);
  // breadth-first search
@@ -175,7 +177,7 @@ shortest_path_from_start (state_item_number target, symbol_number next_sym)
          finished = true;
          break;
        }
-      state_item *si = state_items + last;
+      state_item *si = &state_items[last];
      // Transitions don't change follow_L
      if (si->trans >= 0)
        {
@@ -240,10 +242,10 @@ shortest_path_from_start (state_item_number target, symbol_number next_sym)
      fputs ("Cannot find shortest path to conflict state.", stderr);
      abort ();
    }
-  gl_list_t res =
+  state_item_list res =
    gl_list_create_empty (GL_LINKED_LIST, NULL, NULL, NULL, true);
  for (lssi *sn = n; sn != NULL; sn = sn->parent)
-    gl_list_add_first (res, state_items + sn->si);
+    gl_list_add_first (res, &state_items[sn->si]);

  hash_free (visited);
  gl_list_free (queue);
@@ -306,10 +308,10 @@ intersect (bitset ts, bitset syms)
 * Compute a list of state_items that have a production to n with respect
 * to its lookahead
 */
-gl_list_t
+state_item_list
 lssi_reverse_production (const state_item *si, bitset lookahead)
 {
-  gl_list_t result =
+  state_item_list result =
    gl_list_create_empty (GL_LINKED_LIST, NULL, NULL, NULL, true);
  if (SI_TRANSITION (si))
    return result;
@@ -320,7 +322,7 @@ lssi_reverse_production (const state_item *si, bitset lookahead)
  state_item_number sin;
  BITSET_FOR_EACH (biter, si->revs, sin, 0)
  {
-    state_item *prevsi = state_items + sin;
+    state_item *prevsi = &state_items[sin];
    if (!production_allowed (prevsi, si))
      continue;
    bitset prev_lookahead = prevsi->lookahead;
@@ -32,8 +32,8 @@
 * find shortest lookahead-sensitive path of state-items to target such that
 * next_sym is in the follow_L set of target in that position.
 */
-gl_list_t shortest_path_from_start (state_item_number target,
-                                    symbol_number next_sym);
+state_item_list shortest_path_from_start (state_item_number target,
+                                          symbol_number next_sym);

 /**
 * Determine if the given terminal is in the given symbol set or can begin
@@ -52,6 +52,6 @@ bool intersect (bitset ts, bitset syms);
 * to this state-item such that the resulting possible lookahead symbols are
 * as given.
 */
-gl_list_t lssi_reverse_production (const state_item *si, bitset lookahead);
+state_item_list lssi_reverse_production (const state_item *si, bitset lookahead);

 #endif /* LSSI_H */
@@ -38,6 +38,7 @@
 #include "files.h"
 #include "fixits.h"
 #include "getargs.h"
+#include "glyphs.h"
 #include "gram.h"
 #include "ielr.h"
 #include "lalr.h"
@@ -85,6 +86,7 @@ main (int argc, char *argv[])

  atexit (close_stdout);

+  glyphs_init ();
  uniqstrs_new ();
  muscle_init ();
  complain_init ();
@@ -34,7 +34,7 @@ typedef struct parse_state
  struct si_chunk
  {
    // elements newly added in this chunk
-    gl_list_t contents;
+    state_item_list contents;
    // properties of the linked list this chunk represents
    const state_item *head_elt;
    const state_item *tail_elt;
@@ -135,7 +135,7 @@ static parse_state *
 copy_parse_state (bool prepend, parse_state *parent)
 {
  parse_state *res = xmalloc (sizeof *res);
-  memcpy (res, parent, sizeof *res);
+  *res = *parent;
  res->state_items.contents
    = gl_list_create_empty (GL_LINKED_LIST, NULL, NULL, NULL, true);
  res->derivs.contents = derivation_list_new ();
@@ -252,7 +252,7 @@ parse_state_completed_steps (const parse_state *ps, int *shifts, int *production
  while (root_ps->parent)
    root_ps = root_ps->parent;

-  gl_list_t sis = root_ps->state_items.contents;
+  state_item_list sis = root_ps->state_items.contents;
  int count = 0;

  state_item *last = NULL;
@@ -337,19 +337,17 @@ parser_pop (parse_state *ps, int deriv_index,
  for (int i = 0; i < 4; ++i)
    chunks[i] = gl_list_create_empty (GL_LINKED_LIST, NULL, NULL, NULL, true);
  for (parse_state *pn = ps; pn != NULL; pn = pn->parent)
-    {
-      if (pn->prepend)
-        {
-          gl_list_add_last (chunks[0], pn->state_items.contents);
-          gl_list_add_last (chunks[2], pn->derivs.contents);
-        }
-      else
-        {
-          gl_list_add_first (chunks[1], pn->state_items.contents);
-          gl_list_add_first (chunks[3], pn->derivs.contents);
-        }
-    }
-  gl_list_t popped_derivs = derivation_list_new ();
+    if (pn->prepend)
+      {
+        gl_list_add_last (chunks[0], pn->state_items.contents);
+        gl_list_add_last (chunks[2], pn->derivs.contents);
+      }
+    else
+      {
+        gl_list_add_first (chunks[1], pn->state_items.contents);
+        gl_list_add_first (chunks[3], pn->derivs.contents);
+      }
+  derivation_list popped_derivs = derivation_list_new ();
  gl_list_t ret_chunks[4] = { ret->state_items.contents, NULL,
    ret->derivs.contents, popped_derivs
  };
@@ -390,7 +388,7 @@ parser_pop (parse_state *ps, int deriv_index,
 }

 void
-parse_state_lists (parse_state *ps, gl_list_t *sitems,
+parse_state_lists (parse_state *ps, state_item_list *sitems,
                   derivation_list *derivs)
 {
  parse_state *temp = empty_parse_state ();
@@ -418,12 +416,12 @@ nullable_closure (parse_state *ps, state_item *si, parse_state_list state_list)
  for (state_item_number sin = si->trans; sin != -1;
       prev_sin = sin, sin = state_items[sin].trans)
    {
-      state_item *psi = state_items + prev_sin;
+      state_item *psi = &state_items[prev_sin];
      symbol_number sp = item_number_as_symbol_number (*psi->item);
      if (ISTOKEN (sp) || !nullable[sp - ntokens])
        break;

-      state_item *nsi = state_items + sin;
+      state_item *nsi = &state_items[sin];
      current_ps = copy_parse_state (false, current_ps);
      ps_si_append (current_ps, nsi);
      ps_derivs_append (current_ps, derivation_new (sp, derivation_list_new ()));
@@ -431,14 +429,14 @@ nullable_closure (parse_state *ps, state_item *si, parse_state_list state_list)
    }
 }

-gl_list_t
+parse_state_list
 simulate_transition (parse_state *ps)
 {
  const state_item *si = ps->state_items.tail_elt;
  symbol_number sym = item_number_as_symbol_number (*si->item);
  // Transition on the same next symbol, taking nullable
  // symbols into account.
-  gl_list_t result = parse_state_list_new ();
+  parse_state_list result = parse_state_list_new ();
  state_item_number si_next = si->trans;
  // check for disabled transition, shouldn't happen
  // as any state_items that lead to these should be
@@ -446,11 +444,11 @@ simulate_transition (parse_state *ps)
  if (si_next < 0)
    return result;
  parse_state *next_ps = copy_parse_state (false, ps);
-  ps_si_append (next_ps, state_items + si_next);
+  ps_si_append (next_ps, &state_items[si_next]);
  ps_derivs_append (next_ps, derivation_new_leaf (sym));
  parse_state_list_append (result, next_ps);

-  nullable_closure (next_ps, state_items + si_next, result);
+  nullable_closure (next_ps, &state_items[si_next], result);
  return result;
 }

@@ -473,10 +471,10 @@ compatible (symbol_number sym1, symbol_number sym2)
    return false;
 }

-gl_list_t
+parse_state_list
 simulate_production (parse_state *ps, symbol_number compat_sym)
 {
-  gl_list_t result = parse_state_list_new ();
+  parse_state_list result = parse_state_list_new ();
  const state_item *si = parse_state_tail (ps);
  if (si->prods)
    {
@@ -486,7 +484,7 @@ simulate_production (parse_state *ps, symbol_number compat_sym)
        {
          // Take production step only if lhs is not nullable and
          // if first rhs symbol is compatible with compat_sym
-          state_item *next = state_items + sin;
+          state_item *next = &state_items[sin];
          item_number *itm1 = next->item;
          if (!compatible (*itm1, compat_sym) || !production_allowed (si, next))
            continue;
@@ -504,10 +502,10 @@ simulate_production (parse_state *ps, symbol_number compat_sym)
 // simulates a reduction on the given parse state, conflict_item is the
 // item associated with ps's conflict. symbol_set is a lookahead set this
 // reduction must be compatible with
-gl_list_t
+parse_state_list
 simulate_reduction (parse_state *ps, int rule_len, bitset symbol_set)
 {
-  gl_list_t result = parse_state_list_new ();
+  parse_state_list result = parse_state_list_new ();

  int s_size = ps->state_items.total_size;
  int d_size = ps->derivs.total_size;
@@ -529,7 +527,7 @@ simulate_reduction (parse_state *ps, int rule_len, bitset symbol_set)
  if (s_size != rule_len + 1)
    {
      state_item *tail = (state_item *) new_root->state_items.tail_elt;
-      ps_si_append (new_root, state_items + tail->trans);
+      ps_si_append (new_root, &state_items[tail->trans]);
      parse_state_list_append (result, new_root);
    }
  else
@@ -537,7 +535,7 @@ simulate_reduction (parse_state *ps, int rule_len, bitset symbol_set)
      // The head state_item is a production item, so we need to prepend
      // with possible source state-items.
      const state_item *head = ps->state_items.head_elt;
-      gl_list_t prev = lssi_reverse_production (head, symbol_set);
+      state_item_list prev = lssi_reverse_production (head, symbol_set);
      // TODO: better understand what causes this case.
      if (gl_list_size (prev) == 0)
        {
@@ -560,7 +558,7 @@ simulate_reduction (parse_state *ps, int rule_len, bitset symbol_set)
              copy = copy_parse_state (false, copy);
              struct si_chunk *sis = &copy->state_items;
              const state_item *tail = sis->tail_elt;
-              ps_si_append (copy, state_items + tail->trans);
+              ps_si_append (copy, &state_items[tail->trans]);
              parse_state_list_append (result, copy);
              nullable_closure (copy, (state_item *) sis->tail_elt, result);
            }
@@ -570,10 +568,10 @@ simulate_reduction (parse_state *ps, int rule_len, bitset symbol_set)
  return result;
 }

-gl_list_t
+parse_state_list
 parser_prepend (parse_state *ps)
 {
-  gl_list_t res = parse_state_list_new ();
+  parse_state_list res = parse_state_list_new ();
  const state_item *head = ps->state_items.head_elt;
  symbol_number prepend_sym =
    item_number_as_symbol_number (*(head->item - 1));
@@ -582,7 +580,7 @@ parser_prepend (parse_state *ps)
  BITSET_FOR_EACH (biter, head->revs, sin, 0)
  {
    parse_state *copy = copy_parse_state (true, ps);
-    ps_si_prepend (copy, state_items + sin);
+    ps_si_prepend (copy, &state_items[sin]);
    if (SI_TRANSITION (head))
      ps_derivs_prepend (copy, derivation_new_leaf (prepend_sym));
    parse_state_list_append (res, copy);
@@ -113,7 +113,7 @@ int parse_state_length (const parse_state *ps);
 int parse_state_depth (const parse_state *ps);

 /* returns the linked lists that the parse state is supposed to represent */
-void parse_state_lists (parse_state *ps, gl_list_t *state_items,
+void parse_state_lists (parse_state *ps, state_item_list *state_items,
                        derivation_list *derivs);

 /* various functions that return a list of states based off of
@@ -96,7 +96,7 @@ print_core (struct obstack *oout, state *s)
        obstack_sgrow (oout, " %empty");

      /* Experimental feature: display the lookahead tokens. */
-      if (report_flag & report_lookahead_tokens
+      if (report_flag & report_lookaheads
          && item_number_is_rule_number (*sp1))
        {
          /* Find the reduction we are handling.  */
@@ -104,13 +104,13 @@ print_core (struct obstack *oout, state *s)
          int redno = state_reduction_find (s, r);

          /* Print them if there are.  */
-          if (reds->lookahead_tokens && redno != -1)
+          if (reds->lookaheads && redno != -1)
            {
              bitset_iterator biter;
              int k;
              char const *sep = "";
              obstack_sgrow (oout, "  [");
-              BITSET_FOR_EACH (biter, reds->lookahead_tokens[redno], k, 0)
+              BITSET_FOR_EACH (biter, reds->lookaheads[redno], k, 0)
                {
                  obstack_sgrow (oout, sep);
                  obstack_backslash (oout, symbols[k]->tag);
@@ -86,12 +86,12 @@ print_core (FILE *out, int level, state *s)
          reductions *reds = s->reductions;
          int red = state_reduction_find (s, r);
          /* Print item with lookaheads if there are. */
-          if (reds->lookahead_tokens && red != -1)
+          if (reds->lookaheads && red != -1)
            {
              xml_printf (out, level + 1,
                          "<item rule-number=\"%d\" dot=\"%d\">",
                          r->number, sp1 - sp);
-              state_rule_lookahead_tokens_print_xml (s, r,
+              state_rule_lookaheads_print_xml (s, r,
                                                     out, level + 2);
              xml_puts (out, level + 1, "</item>");
              printed = true;
@@ -202,26 +202,26 @@ print_errs (FILE *out, int level, state *s)
 }


-/*-------------------------------------------------------------------------.
-| Report a reduction of RULE on LOOKAHEAD_TOKEN (which can be 'default').  |
-| If not ENABLED, the rule is masked by a shift or a reduce (S/R and       |
-| R/R conflicts).                                                          |
-`-------------------------------------------------------------------------*/
+/*-------------------------------------------------------------------.
+| Report a reduction of RULE on LOOKAHEAD (which can be 'default').  |
+| If not ENABLED, the rule is masked by a shift or a reduce (S/R and |
+| R/R conflicts).                                                    |
+`-------------------------------------------------------------------*/

 static void
-print_reduction (FILE *out, int level, char const *lookahead_token,
+print_reduction (FILE *out, int level, char const *lookahead,
                 rule *r, bool enabled)
 {
  if (r->number)
    xml_printf (out, level,
                "<reduction symbol=\"%s\" rule=\"%d\" enabled=\"%s\"/>",
-                xml_escape (lookahead_token),
+                xml_escape (lookahead),
                r->number,
                enabled ? "true" : "false");
  else
    xml_printf (out, level,
                "<reduction symbol=\"%s\" rule=\"accept\" enabled=\"%s\"/>",
-                xml_escape (lookahead_token),
+                xml_escape (lookahead),
                enabled ? "true" : "false");
 }

@@ -258,13 +258,13 @@ print_reductions (FILE *out, int level, state *s)
  if (default_reduction)
    report = true;

-  if (reds->lookahead_tokens)
+  if (reds->lookaheads)
    for (i = 0; i < ntokens; i++)
      {
        bool count = bitset_test (no_reduce_set, i);

        for (j = 0; j < reds->num; ++j)
-          if (bitset_test (reds->lookahead_tokens[j], i))
+          if (bitset_test (reds->lookaheads[j], i))
            {
              if (! count)
                {
@@ -289,14 +289,14 @@ print_reductions (FILE *out, int level, state *s)
  xml_puts (out, level, "<reductions>");

  /* Report lookahead tokens (or $default) and reductions.  */
-  if (reds->lookahead_tokens)
+  if (reds->lookaheads)
    for (i = 0; i < ntokens; i++)
      {
        bool defaulted = false;
        bool count = bitset_test (no_reduce_set, i);

        for (j = 0; j < reds->num; ++j)
-          if (bitset_test (reds->lookahead_tokens[j], i))
+          if (bitset_test (reds->lookaheads[j], i))
            {
              if (! count)
                {
@@ -382,14 +382,17 @@ print_grammar (FILE *out, int level)
  for (int i = 0; i < max_code + 1; i++)
    if (token_translations[i] != undeftoken->content->number)
      {
-        char const *tag = symbols[token_translations[i]]->tag;
-        int precedence = symbols[token_translations[i]]->content->prec;
-        assoc associativity = symbols[token_translations[i]]->content->assoc;
+        symbol const *sym = symbols[token_translations[i]];
+        char const *tag = sym->tag;
+        char const *type = sym->content->type_name;
+        int precedence = sym->content->prec;
+        assoc associativity = sym->content->assoc;
        xml_indent (out, level + 2);
        fprintf (out,
                 "<terminal symbol-number=\"%d\" token-number=\"%d\""
-                 " name=\"%s\" usefulness=\"%s\"",
-                 token_translations[i], i, xml_escape (tag),
+                 " name=\"%s\" type=\"%s\" usefulness=\"%s\"",
+                 token_translations[i], i, xml_escape_n (0, tag),
+                 type ? xml_escape_n (1, type) : "",
                 reduce_token_unused_in_grammar (token_translations[i])
                   ? "unused-in-grammar" : "useful");
        if (precedence)
@@ -404,12 +407,16 @@ print_grammar (FILE *out, int level)
  xml_puts (out, level + 1, "<nonterminals>");
  for (symbol_number i = ntokens; i < nsyms + nuseless_nonterminals; i++)
    {
-      char const *tag = symbols[i]->tag;
+      symbol const *sym = symbols[i];
+      char const *tag = sym->tag;
+      char const *type = sym->content->type_name;
      xml_printf (out, level + 2,
                  "<nonterminal symbol-number=\"%d\" name=\"%s\""
+                  " type=\"%s\""
                  " usefulness=\"%s\"/>",
-                  i, xml_escape (tag),
-                  reduce_nonterminal_useless_in_grammar (symbols[i]->content)
+                  i, xml_escape_n (0, tag),
+                  type ? xml_escape_n (1, type) : "",
+                  reduce_nonterminal_useless_in_grammar (sym->content)
                    ? "useless-in-grammar" : "useful");
    }
  xml_puts (out, level + 1, "</nonterminals>");
@@ -90,9 +90,9 @@ print_core (FILE *out, const state *s)
      previous_rule = r;

      /* Display the lookahead tokens?  */
-      if (report_flag & report_lookahead_tokens
+      if (report_flag & report_lookaheads
          && item_number_is_rule_number (*sp1))
-        state_rule_lookahead_tokens_print (s, r, out);
+        state_rule_lookaheads_print (s, r, out);
      fputc ('\n', out);
    }
 }
@@ -180,19 +180,19 @@ print_errs (FILE *out, const state *s)
 }


-/*-------------------------------------------------------------------------.
-| Report a reduction of RULE on LOOKAHEAD_TOKEN (which can be 'default').  |
-| If not ENABLED, the rule is masked by a shift or a reduce (S/R and       |
-| R/R conflicts).                                                          |
-`-------------------------------------------------------------------------*/
+/*-------------------------------------------------------------------.
+| Report a reduction of RULE on LOOKAHEAD (which can be 'default').  |
+| If not ENABLED, the rule is masked by a shift or a reduce (S/R and |
+| R/R conflicts).                                                    |
+`-------------------------------------------------------------------*/

 static void
 print_reduction (FILE *out, size_t width,
-                 const char *lookahead_token,
+                 const char *lookahead,
                 rule *r, bool enabled)
 {
-  fprintf (out, "    %s", lookahead_token);
-  for (int j = width - mbswidth (lookahead_token, 0); j > 0; --j)
+  fprintf (out, "    %s", lookahead);
+  for (int j = width - mbswidth (lookahead, 0); j > 0; --j)
    fputc (' ', out);
  if (!enabled)
    fputc ('[', out);
@@ -239,13 +239,13 @@ print_reductions (FILE *out, const state *s)
  if (default_reduction)
    width = mbswidth (_("$default"), 0);

-  if (reds->lookahead_tokens)
+  if (reds->lookaheads)
    for (int i = 0; i < ntokens; i++)
      {
        bool count = bitset_test (no_reduce_set, i);

        for (int j = 0; j < reds->num; ++j)
-          if (bitset_test (reds->lookahead_tokens[j], i))
+          if (bitset_test (reds->lookaheads[j], i))
            {
              if (! count)
                {
@@ -268,7 +268,7 @@ print_reductions (FILE *out, const state *s)
  bool default_reduction_only = true;

  /* Report lookahead tokens (or $default) and reductions.  */
-  if (reds->lookahead_tokens)
+  if (reds->lookaheads)
    for (int i = 0; i < ntokens; i++)
      {
        bool defaulted = false;
@@ -277,7 +277,7 @@ print_reductions (FILE *out, const state *s)
          default_reduction_only = false;

        for (int j = 0; j < reds->num; ++j)
-          if (bitset_test (reds->lookahead_tokens[j], i))
+          if (bitset_test (reds->lookaheads[j], i))
            {
              if (! count)
                {
@@ -377,11 +377,11 @@ print_terminal_symbols (FILE *out)
  for (int i = 0; i < max_code + 1; ++i)
    if (token_translations[i] != undeftoken->content->number)
      {
-        const char *tag = symbols[token_translations[i]]->tag;
+        const symbol *sym = symbols[token_translations[i]];
+        const char *tag = sym->tag;
        fprintf (out, "%4s%s", "", tag);
-        if (symbols[token_translations[i]]->content->type_name)
-          fprintf (out, " <%s>",
-                   symbols[token_translations[i]]->content->type_name);
+        if (sym->content->type_name)
+          fprintf (out, " <%s>", sym->content->type_name);
        fprintf (out, " (%d)", i);

        for (rule_number r = 0; r < nrules; r++)
@@ -403,7 +403,8 @@ print_nonterminal_symbols (FILE *out)
  fprintf (out, "%s\n\n", _("Nonterminals, with rules where they appear"));
  for (symbol_number i = ntokens; i < nsyms; i++)
    {
-      const char *tag = symbols[i]->tag;
+      const symbol *sym = symbols[i];
+      const char *tag = sym->tag;
      bool on_left = false;
      bool on_right = false;

@@ -418,9 +419,9 @@ print_nonterminal_symbols (FILE *out)

      int column = 4 + mbswidth (tag, 0);
      fprintf (out, "%4s%s", "", tag);
-      if (symbols[i]->content->type_name)
+      if (sym->content->type_name)
        column += fprintf (out, " <%s>",
-                           symbols[i]->content->type_name);
+                           sym->content->type_name);
      fprintf (out, " (%d)\n", i);

      if (on_left)
@@ -144,13 +144,13 @@ init_state_items (void)
      for (int j = 0; j < s->nitems; ++j)
        {
          state_item_set (sidx, s, s->items[j]);
-          state_item *si = state_items + sidx;
+          state_item *si = &state_items[sidx];
          const rule *r = item_rule (si->item);
          if (rule_search_idx < red->num && red->rules[rule_search_idx] < r)
            ++rule_search_idx;
          if (rule_search_idx < red->num && r == red->rules[rule_search_idx])
            {
-              bitsetv lookahead = red->lookahead_tokens;
+              bitsetv lookahead = red->lookaheads;
              if (lookahead)
                si->lookahead = lookahead[rule_search_idx];
            }
@@ -163,7 +163,7 @@ init_state_items (void)
          state_item_set (sidx, s, off);
          if (item_number_is_rule_number (ritem[off]))
            {
-              bitsetv lookahead = red->lookahead_tokens;
+              bitsetv lookahead = red->lookaheads;
              if (lookahead)
                state_items[sidx].lookahead = lookahead[rule_search_idx];
              ++rule_search_idx;
@@ -211,7 +211,7 @@ init_trans (void)
      for (int j = 0; j < t->num; ++j)
        if (!TRANSITION_IS_DISABLED (t, j))
          hash_xinsert (transition_set, t->states[j]);
-      for (int j = state_item_map[i]; j < state_item_map[i + 1]; ++j)
+      for (state_item_number j = state_item_map[i]; j < state_item_map[i + 1]; ++j)
        {
          item_number *item = state_items[j].item;
          if (item_number_is_rule_number (*item))
@@ -222,16 +222,14 @@ init_trans (void)
          // find the item in the destination state that corresponds
          // to the transition of item
          for (int k = 0; k < dst->nitems; ++k)
-            {
-              if (item + 1 == ritem + dst->items[k])
-                {
-                  state_item_number dstSI =
-                    state_item_index_lookup (dst->number, k);
+            if (item + 1 == ritem + dst->items[k])
+              {
+                state_item_number dstSI =
+                  state_item_index_lookup (dst->number, k);

-                  state_items[j].trans = dstSI;
-                  bitset_set (state_items[dstSI].revs, j);
-                  break;
-                }
+                state_items[j].trans = dstSI;
+                bitset_set (state_items[dstSI].revs, j);
+                break;
            }
        }
      hash_free (transition_set);
@@ -250,10 +248,10 @@ init_prods (void)

      // Add the nitems of state to skip to the production portion
      // of that state's state_items
-      for (int j = state_item_map[i] + s->nitems;
+      for (state_item_number j = state_item_map[i] + s->nitems;
           j < state_item_map[i + 1]; ++j)
        {
-          state_item *src = state_items + j;
+          state_item *src = &state_items[j];
          item_number *item = src->item;
          symbol_number lhs = item_rule (item)->lhs->number;
          bitset itms = hash_pair_lookup (closure_map, lhs);
@@ -266,9 +264,9 @@ init_prods (void)
        }
      // For each item with a dot followed by a nonterminal,
      // try to create a production edge.
-      for (int j = state_item_map[i]; j < state_item_map[i + 1]; ++j)
+      for (state_item_number j = state_item_map[i]; j < state_item_map[i + 1]; ++j)
        {
-          state_item *src = state_items + j;
+          state_item *src = &state_items[j];
          item_number item = *(src->item);
          // Skip reduce items and items with terminals after the dot
          if (item_number_is_rule_number (item) || ISTOKEN (item))
@@ -301,12 +299,12 @@ gen_lookaheads (void)
 {
  for (state_item_number i = 0; i < nstate_items; ++i)
    {
-      state_item *si = state_items + i;
+      state_item *si = &state_items[i];
      if (item_number_is_symbol_number (*(si->item)) || !si->lookahead)
        continue;

      bitset lookahead = si->lookahead;
-      gl_list_t queue =
+      state_item_list queue =
        gl_list_create (GL_LINKED_LIST, NULL, NULL, NULL, true, 1,
                        (const void **) &si);

@@ -339,7 +337,7 @@ init_firsts (void)
  firsts = bitsetv_create (nnterms, nsyms, BITSET_FIXED);
  for (rule_number i = 0; i < nrules; ++i)
    {
-      rule *r = rules + i;
+      rule *r = &rules[i];
      item_number *n = r->rhs;
      // Iterate through nullable nonterminals to try to find a terminal.
      while (item_number_is_symbol_number (*n) && ISVAR (*n)
@@ -357,7 +355,7 @@ init_firsts (void)
      change = false;
      for (rule_number i = 0; i < nrules; ++i)
        {
-          rule *r = rules + i;
+          rule *r = &rules[i];
          symbol_number lhs = r->lhs->number;
          bitset f_lhs = FIRSTS (lhs);
          for (item_number *n = r->rhs;
@@ -392,7 +390,7 @@ disable_state_item (state_item *si)
 static void
 prune_forward (const state_item *si)
 {
-  gl_list_t queue =
+  state_item_list queue =
    gl_list_create (GL_LINKED_LIST, NULL, NULL, NULL, true, 1,
                    (const void **) &si);

@@ -401,7 +399,7 @@ prune_forward (const state_item *si)
      state_item *dsi = (state_item *) gl_list_get_at (queue, 0);
      gl_list_remove_at (queue, 0);
      if (dsi->trans >= 0)
-        gl_list_add_last (queue, state_items + dsi->trans);
+        gl_list_add_last (queue, &state_items[dsi->trans]);

      if (dsi->prods)
        {
@@ -409,7 +407,7 @@ prune_forward (const state_item *si)
          state_item_number sin;
          BITSET_FOR_EACH (biter, dsi->prods, sin, 0)
            {
-              const state_item *prod = state_items + sin;
+              const state_item *prod = &state_items[sin];
              bitset_reset (prod->revs, dsi - state_items);
              if (bitset_empty_p (prod->revs))
                gl_list_add_last (queue, prod);
@@ -427,7 +425,7 @@ prune_forward (const state_item *si)
 static void
 prune_backward (const state_item *si)
 {
-  gl_list_t queue =
+  state_item_list queue =
    gl_list_create (GL_LINKED_LIST, NULL, NULL, NULL, true, 1,
                    (const void **) &si);

@@ -441,7 +439,7 @@ prune_backward (const state_item *si)
        {
          if (SI_DISABLED (sin))
            continue;
-          state_item *rev = state_items + sin;
+          state_item *rev = &state_items[sin];
          if (rev->prods)
            {
              bitset_reset (rev->prods, dsi - state_items);
@@ -466,7 +464,7 @@ prune_disabled_paths (void)
 {
  for (int i = nstate_items - 1; i >= 0; --i)
    {
-      state_item *si = state_items + i;
+      state_item *si = &state_items[i];
      if (si->trans == -1 && item_number_is_symbol_number (*si->item))
        {
          prune_forward (si);
@@ -494,9 +492,9 @@ state_items_report (void)
  for (state_number i = 0; i < nstates; ++i)
    {
      printf ("State %d:\n", i);
-      for (int j = state_item_map[i]; j < state_item_map[i + 1]; ++j)
+      for (state_item_number j = state_item_map[i]; j < state_item_map[i + 1]; ++j)
        {
-          state_item *si = state_items + j;
+          state_item *si = &state_items[j];
          item_print (si->item, NULL, stdout);
          if (SI_DISABLED (j))
            {
@@ -508,7 +506,7 @@ state_items_report (void)
          if (si->trans >= 0)
            {
              fputs ("    -> ", stdout);
-              print_state_item (state_items + si->trans, stdout, "");
+              print_state_item (&state_items[si->trans], stdout, "");
            }

          bitset sets[2] = { si->prods, si->revs };
@@ -523,7 +521,7 @@ state_items_report (void)
                  BITSET_FOR_EACH (biter, b, sin, 0)
                    {
                      fputs (txt[seti], stdout);
-                      print_state_item (state_items + sin, stdout, "");
+                      print_state_item (&state_items[sin], stdout, "");
                    }
                }
            }
@@ -562,10 +560,10 @@ state_items_init (void)
 void
 state_items_free (void)
 {
-  for (int i = 0; i < nstate_items; ++i)
+  for (state_item_number i = 0; i < nstate_items; ++i)
    if (!SI_DISABLED (i))
      {
-        state_item *si = state_items + i;
+        state_item *si = &state_items[i];
        if (si->prods)
          bitset_free (si->prods);
        bitset_free (si->revs);
@@ -594,5 +592,5 @@ production_allowed (const state_item *si, const state_item *next)
      if (prec1 == prec2 && s1->assoc == left_assoc)
        return false;
    }
-    return true;
+  return true;
 }
@@ -28,16 +28,16 @@
 # include "state.h"

 /* Initializes a graph connecting (state, production item) pairs to
-   pairs they can make a transition or production step to. This graph
+   pairs they can make a transition or production step to.  This graph
   is used to search for paths that represent counterexamples of some
   conflict.

   state_items is an array of state state-item pairs ordered by state.
   state_item_map maps state numbers to the first item which
-   corresponds to it in the array. A state's portion in state_items
-   begins with its items in the same order as it was in the
-   state. This is then followed by productions from the closure of the
-   state in order by rule.
+   corresponds to it in the array.  A state's portion in state_items
+   begins with its items in the same order as it was in the state.
+   This is then followed by productions from the closure of the state
+   in order by rule.

   There are two type of edges in this graph transitions and
   productions.  Transitions are the same as transitions from the
@@ -53,9 +53,9 @@
   production edges, and all others will have reverse transition
   edges. */

-# define SI_DISABLED(sin) (state_items[sin].trans == -2)
-# define SI_PRODUCTION(si) ((si) == state_items || *((si)->item - 1) < 0)
-# define SI_TRANSITION(si) ((si) != state_items && *((si)->item - 1) >= 0)
+# define SI_DISABLED(Sin) (state_items[Sin].trans == -2)
+# define SI_PRODUCTION(Si) ((Si) == state_items || *((Si)->item - 1) < 0)
+# define SI_TRANSITION(Si) ((Si) != state_items && *((Si)->item - 1) >= 0)

 typedef int state_item_number;

@@ -69,6 +69,9 @@ typedef struct
  bitset lookahead;
 } state_item;

+// A path of state-items.
+typedef gl_list_t state_item_list;
+
 extern bitsetv firsts;
 # define FIRSTS(sym) firsts[(sym) - ntokens]

@@ -92,6 +95,7 @@ void state_items_free (void);

 bool production_allowed (const state_item *si, const state_item *next);

+// Iterating on a state_item_list.
 static inline bool
 state_item_list_next (gl_list_iterator_t *it, state_item **si)
 {
@@ -101,7 +101,7 @@ reductions_new (int num, rule **reds)
  size_t rules_size = num * sizeof *reds;
  reductions *res = xmalloc (offsetof (reductions, rules) + rules_size);
  res->num = num;
-  res->lookahead_tokens = NULL;
+  res->lookaheads = NULL;
  memcpy (res->rules, reds, rules_size);
  return res;
 }
@@ -260,20 +260,20 @@ state_errs_set (state *s, int num, symbol **tokens)
 `--------------------------------------------------*/

 void
-state_rule_lookahead_tokens_print (state const *s, rule const *r, FILE *out)
+state_rule_lookaheads_print (state const *s, rule const *r, FILE *out)
 {
  /* Find the reduction we are handling.  */
  reductions *reds = s->reductions;
  int red = state_reduction_find (s, r);

  /* Print them if there are.  */
-  if (reds->lookahead_tokens && red != -1)
+  if (reds->lookaheads && red != -1)
    {
      bitset_iterator biter;
      int k;
      char const *sep = "";
      fprintf (out, "  [");
-      BITSET_FOR_EACH (biter, reds->lookahead_tokens[red], k, 0)
+      BITSET_FOR_EACH (biter, reds->lookaheads[red], k, 0)
        {
          fprintf (out, "%s%s", sep, symbols[k]->tag);
          sep = ", ";
@@ -283,7 +283,7 @@ state_rule_lookahead_tokens_print (state const *s, rule const *r, FILE *out)
 }

 void
-state_rule_lookahead_tokens_print_xml (state const *s, rule const *r,
+state_rule_lookaheads_print_xml (state const *s, rule const *r,
                                       FILE *out, int level)
 {
  /* Find the reduction we are handling.  */
@@ -291,12 +291,12 @@ state_rule_lookahead_tokens_print_xml (state const *s, rule const *r,
  int red = state_reduction_find (s, r);

  /* Print them if there are.  */
-  if (reds->lookahead_tokens && red != -1)
+  if (reds->lookaheads && red != -1)
    {
      bitset_iterator biter;
      int k;
      xml_puts (out, level, "<lookaheads>");
-      BITSET_FOR_EACH (biter, reds->lookahead_tokens[red], k, 0)
+      BITSET_FOR_EACH (biter, reds->lookaheads[red], k, 0)
        {
          xml_printf (out, level + 1, "<symbol>%s</symbol>",
                      xml_escape (symbols[k]->tag));
@@ -62,7 +62,7 @@

   Each reductions structure describes the possible reductions at the
   state whose number is in the number field.  rules is an array of
-   num rules.  lookahead_tokens is an array of bitsets, one per rule.
+   num rules.  lookaheads is an array of bitsets, one per rule.

   Conflict resolution can decide that certain tokens in certain
   states should explicitly be errors (for implementing %nonassoc).
@@ -187,7 +187,7 @@ errs *errs_new (int num, symbol **tokens);
 typedef struct
 {
  int num;
-  bitset *lookahead_tokens;
+  bitset *lookaheads;
  /* Sorted ascendingly on rule number.  */
  rule *rules[1];
 } reductions;
@@ -254,9 +254,9 @@ void state_errs_set (state *s, int num, symbol **errors);

 /* Print on OUT all the lookahead tokens such that this STATE wants to
   reduce R.  */
-void state_rule_lookahead_tokens_print (state const *s, rule const *r, FILE *out);
-void state_rule_lookahead_tokens_print_xml (state const *s, rule const *r,
-                                            FILE *out, int level);
+void state_rule_lookaheads_print (state const *s, rule const *r, FILE *out);
+void state_rule_lookaheads_print_xml (state const *s, rule const *r,
+                                      FILE *out, int level);

 /* Create/destroy the states hash table.  */
 void state_hash_new (void);
@@ -200,7 +200,7 @@ conflict_row (state *s)
        /* Find all reductions for token J, and record all that do not
           match ACTROW[J].  */
        for (int i = 0; i < reds->num; i += 1)
-          if (bitset_test (reds->lookahead_tokens[i], j)
+          if (bitset_test (reds->lookaheads[i], j)
              && (actrow[j]
                  != rule_number_as_item_number (reds->rules[i]->number)))
            {
@@ -247,7 +247,7 @@ action_row (state *s)

  reductions *reds = s->reductions;
  bool conflicted = false;
-  if (reds->lookahead_tokens)
+  if (reds->lookaheads)
    /* loop over all the rules available here which require
       lookahead (in reverse order to give precedence to the first
       rule) */
@@ -257,7 +257,7 @@ action_row (state *s)
      {
        bitset_iterator biter;
        int j;
-        BITSET_FOR_EACH (biter, reds->lookahead_tokens[i], j, 0)
+        BITSET_FOR_EACH (biter, reds->lookaheads[i], j, 0)
          {
            /* and record this rule as the rule to use if that
               token follows.  */
@@ -308,7 +308,7 @@ action_row (state *s)
    }

  /* Turn off default reductions where requested by the user.  See
-     state_lookahead_tokens_count in lalr.c to understand when states are
+     state_lookaheads_count in lalr.c to understand when states are
     labeled as consistent.  */
  {
    char *default_reductions =
@@ -865,8 +865,14 @@ State 5
        1 exp: exp OP exp .
        1 exp: exp . OP exp
      Example                  exp OP exp . OP exp
-      First derivation         exp ::=[ exp ::=[ exp OP exp . ] OP exp ]
-      Second derivation        exp ::=[ exp OP exp ::=[ exp . OP exp ] ]
+      Shift derivation
+        exp
+        `-> exp OP exp
+                   `-> exp . OP exp
+      Reduce derivation
+        exp
+        `-> exp              OP exp
+            `-> exp OP exp .

 ]])

@@ -1119,7 +1125,7 @@ m4_popdef([AT_TEST])
 # else.

 AT_SETUP([Defaulted Conflicted Reduction])
-AT_KEYWORDS([report])
+AT_KEYWORDS([cex report])

 AT_DATA([input.y],
 [[%%
@@ -1207,8 +1213,14 @@ State 1
        3 num: '0' .
        4 id: '0' .
      Example                  '0' .
-      First derivation         exp ::=[ num ::=[ '0' . ] ]
-      Second derivation        exp ::=[ id ::=[ '0' . ] ]
+      First derivation
+        exp
+        `-> num
+            `-> '0' .
+      Second derivation
+        exp
+        `-> id
+            `-> '0' .



@@ -1579,6 +1591,8 @@ AT_CLEANUP

 AT_SETUP([[Unreachable States After Conflict Resolution]])

+AT_KEYWORDS([cex report])
+
 # If conflict resolution makes states unreachable, remove those states, report
 # rules that are then unused, and don't report conflicts in those states.  Test
 # what happens when a nonterminal becomes useless as a result of state removal
@@ -1753,18 +1767,30 @@ State 4
    Shift/reduce conflict on token 'a':
       10 reported_conflicts: . %empty
        8 reported_conflicts: . 'a'
-      First example            resolved_conflict . 'a'
-      First derivation         start ::=[ resolved_conflict reported_conflicts ::=[ . ] 'a' ]
-      Second example           resolved_conflict . 'a' 'a'
-      Second derivation        start ::=[ resolved_conflict reported_conflicts ::=[ . 'a' ] 'a' ]
+      First example            resolved_conflict . 'a' 'a'
+      Shift derivation
+        start
+        `-> resolved_conflict reported_conflicts 'a'
+                              `-> . 'a'
+      Second example           resolved_conflict . 'a'
+      Reduce derivation
+        start
+        `-> resolved_conflict reported_conflicts 'a'
+                              `-> .

    Shift/reduce conflict on token 'a':
       10 reported_conflicts: . %empty
        9 reported_conflicts: . 'a'
-      First example            resolved_conflict . 'a'
-      First derivation         start ::=[ resolved_conflict reported_conflicts ::=[ . ] 'a' ]
-      Second example           resolved_conflict . 'a' 'a'
-      Second derivation        start ::=[ resolved_conflict reported_conflicts ::=[ . 'a' ] 'a' ]
+      First example            resolved_conflict . 'a' 'a'
+      Shift derivation
+        start
+        `-> resolved_conflict reported_conflicts 'a'
+                              `-> . 'a'
+      Second example           resolved_conflict . 'a'
+      Reduce derivation
+        start
+        `-> resolved_conflict reported_conflicts 'a'
+                              `-> .



@@ -1781,8 +1807,12 @@ State 5
        8 reported_conflicts: 'a' .
        9 reported_conflicts: 'a' .
      Example                  'a' .
-      First derivation         reported_conflicts ::=[ 'a' . ]
-      Second derivation        reported_conflicts ::=[ 'a' . ]
+      First derivation
+        reported_conflicts
+        `-> 'a' .
+      Second derivation
+        reported_conflicts
+        `-> 'a' .



@@ -1904,6 +1934,8 @@ AT_CLEANUP

 AT_SETUP([[%nonassoc error actions for multiple reductions in a state]])

+AT_KEYWORDS([cex report])
+
 AT_DATA([[input.y]],
 [[%nonassoc 'a' 'b' 'c'
 %%
@@ -1965,8 +1997,14 @@ AT_CHECK([[cat input.output | sed -n '/^State 0$/,/^State 1$/p']], 0,
       12 empty_c2: . %empty
       13 empty_c3: . %empty
      Example                  . 'c'
-      First derivation         start ::=[ empty_c2 ::=[ . ] 'c' ]
-      Second derivation        start ::=[ empty_c3 ::=[ . ] 'c' ]
+      First derivation
+        start
+        `-> empty_c2 'c'
+            `-> .
+      Second derivation
+        start
+        `-> empty_c3 'c'
+            `-> .



@@ -17,13 +17,23 @@

 AT_BANNER([[Counterexamples.]])

+# AT_BISON_CHECK_CEX(TREE, FLAT)
+# ------------------------------
 m4_define([AT_BISON_CHECK_CEX],
-[AT_DATA([experr], [$4])
-sed -e ['s/time limit exceeded: [0-9][.0-9]*/time limit exceeded: XXX/g'] \
-  experr >expout
-AT_BISON_CHECK([-Wcounterexamples $1], [$2], [$3], [stderr])
+[AT_KEYWORDS([cex])
+
+AT_BISON_CHECK([-Wcounterexamples input.y], [0], [], [stderr])
+# FIXME: Avoid trailing white spaces.
+AT_CHECK([[sed -e 's/time limit exceeded: [0-9][.0-9]*/time limit exceeded: XXX/g;s/ *$//;' stderr]],
+         [], [$1])
+
+m4_pushdef([AT_SET_ENV_IF],
+           [[YYFLAT=1; export YYFLAT;]]m4_defn([AT_SET_ENV_IF]))
+AT_BISON_CHECK([-Wcounterexamples input.y], [0], [], [stderr])
 AT_CHECK([[sed -e 's/time limit exceeded: [0-9][.0-9]*/time limit exceeded: XXX/g' stderr]],
-         [], [expout])
+         [], [$2])
+m4_popdef([AT_SET_ENV_IF])
+
 ])

 ## --------------------- ##
@@ -31,7 +41,6 @@ AT_CHECK([[sed -e 's/time limit exceeded: [0-9][.0-9]*/time limit exceeded: XXX/
 ## --------------------- ##

 AT_SETUP([Unifying S/R])
-AT_KEYWORDS([cex])

 AT_DATA([[input.y]],
 [[%token A B C
@@ -43,12 +52,26 @@ x: B | B C;
 y: A | A B;
 ]])

-AT_BISON_CHECK_CEX([input.y], [], [],
+AT_BISON_CHECK_CEX(
 [[input.y: warning: 1 shift/reduce conflict [-Wconflicts-sr]
 Shift/reduce conflict on token B:
  Example              A . B C
-  First derivation     s ::=[ a ::=[ A . ] x ::=[ B C ] ]
-  Second derivation    s ::=[ y ::=[ A . B ] c ::=[ C ] ]
+  Shift derivation
+    s
+    `-> y         c
+        `-> A . B `-> C
+  Reduce derivation
+    s
+    `-> a       x
+        `-> A . `-> B C
+
+input.y:4.4: warning: rule useless in parser due to conflicts [-Wother]
+]],
+[[input.y: warning: 1 shift/reduce conflict [-Wconflicts-sr]
+Shift/reduce conflict on token B:
+  Example              A . B C
+  Shift derivation     s -> [ y -> [ A . B ] c -> [ C ] ]
+  Reduce derivation    s -> [ a -> [ A . ] x -> [ B C ] ]

 input.y:4.4: warning: rule useless in parser due to conflicts [-Wother]
 ]])
@@ -60,7 +83,6 @@ AT_CLEANUP
 ## ------------------- ##

 AT_SETUP([Deep Unifying S/R])
-AT_KEYWORDS([cex])

 AT_DATA([[input.y]],
 [[%token A B C
@@ -72,17 +94,49 @@ a: A | A a;
 bc: B bc C | B C;
 ]])

-AT_BISON_CHECK_CEX([input.y], [], [],
+AT_BISON_CHECK_CEX(
 [[input.y: warning: 1 shift/reduce conflict [-Wconflicts-sr]
 Shift/reduce conflict on token B:
  Example              A . B C
-  First derivation     s ::=[ a ::=[ A . ] bc ::=[ B C ] ]
-  Second derivation    s ::=[ ac ::=[ A ac ::=[ b ::=[ . B ] ] C ] ]
+  Shift derivation
+    s
+    `-> ac
+        `-> A ac          C
+              `-> b
+                  `-> . B
+  Reduce derivation
+    s
+    `-> a       bc
+        `-> A . `-> B C

 Shift/reduce conflict on token B:
  Example              A A . B B C C
-  First derivation     s ::=[ a ::=[ A a ::=[ A . ] ] bc ::=[ B bc ::=[ B C ] C ] ]
-  Second derivation    s ::=[ ac ::=[ A ac ::=[ A ac ::=[ b ::=[ . b ::=[ B B ] ] ] C ] C ] ]
+  Shift derivation
+    s
+    `-> ac
+        `-> A ac                        C
+              `-> A ac                C
+                    `-> b
+                        `-> . b
+                              `-> B B
+  Reduce derivation
+    s
+    `-> a             bc
+        `-> A a       `-> B bc      C
+              `-> A .       `-> B C
+
+input.y:6.4: warning: rule useless in parser due to conflicts [-Wother]
+]],
+[[input.y: warning: 1 shift/reduce conflict [-Wconflicts-sr]
+Shift/reduce conflict on token B:
+  Example              A . B C
+  Shift derivation     s -> [ ac -> [ A ac -> [ b -> [ . B ] ] C ] ]
+  Reduce derivation    s -> [ a -> [ A . ] bc -> [ B C ] ]
+
+Shift/reduce conflict on token B:
+  Example              A A . B B C C
+  Shift derivation     s -> [ ac -> [ A ac -> [ A ac -> [ b -> [ . b -> [ B B ] ] ] C ] C ] ]
+  Reduce derivation    s -> [ a -> [ A a -> [ A . ] ] bc -> [ B bc -> [ B C ] C ] ]

 input.y:6.4: warning: rule useless in parser due to conflicts [-Wother]
 ]])
@@ -94,7 +148,6 @@ AT_CLEANUP
 ## ------------------------------------ ##

 AT_SETUP([S/R Conflict with Nullable Symbols])
-AT_KEYWORDS([cex])

 AT_DATA([[input.y]],
 [[%token A B X Y
@@ -107,18 +160,50 @@ y: %empty | Y y;
 xby: B | X xby Y;
 ]])

-AT_BISON_CHECK_CEX([input.y], [], [],
+AT_BISON_CHECK_CEX(
 [[input.y: warning: 2 shift/reduce conflicts [-Wconflicts-sr]
 Shift/reduce conflict on token B:
  Example              A . B
-  First derivation     s ::=[ ax ::=[ A x ::=[ . ] ] by ::=[ B y ::=[ ] ] ]
-  Second derivation    s ::=[ A xby ::=[ . B ] ]
+  Shift derivation
+    s
+    `-> A xby
+          `-> . B
+  Reduce derivation
+    s
+    `-> ax          by
+        `-> A x     `-> B y
+              `-> .       `-> %empty

 Shift/reduce conflict on token B:
-  First example        A X . B y $end
-  First derivation     $accept ::=[ s ::=[ ax ::=[ A x ::=[ X x ::=[ . ] ] ] by ::=[ B y ] ] $end ]
-  Second example       A X . B Y $end
-  Second derivation    $accept ::=[ s ::=[ A xby ::=[ X xby ::=[ . B ] Y ] ] $end ]
+  First example        A X . B Y $end
+  Shift derivation
+    $accept
+    `-> s                     $end
+        `-> A xby
+              `-> X xby     Y
+                    `-> . B
+  Second example       A X . B y $end
+  Reduce derivation
+    $accept
+    `-> s                             $end
+        `-> ax                by
+            `-> A x           `-> B y
+                  `-> X x
+                        `-> .
+
+input.y:5.4-9: warning: rule useless in parser due to conflicts [-Wother]
+]],
+[[input.y: warning: 2 shift/reduce conflicts [-Wconflicts-sr]
+Shift/reduce conflict on token B:
+  Example              A . B
+  Shift derivation     s -> [ A xby -> [ . B ] ]
+  Reduce derivation    s -> [ ax -> [ A x -> [ . ] ] by -> [ B y -> [ ] ] ]
+
+Shift/reduce conflict on token B:
+  First example        A X . B Y $end
+  Shift derivation     $accept -> [ s -> [ A xby -> [ X xby -> [ . B ] Y ] ] $end ]
+  Second example       A X . B y $end
+  Reduce derivation    $accept -> [ s -> [ ax -> [ A x -> [ X x -> [ . ] ] ] by -> [ B y ] ] $end ]

 input.y:5.4-9: warning: rule useless in parser due to conflicts [-Wother]
 ]])
@@ -130,7 +215,6 @@ AT_CLEANUP
 ## ---------------------------- ##

 AT_SETUP([Non-unifying Ambiguous S/R])
-AT_KEYWORDS([cex])

 AT_DATA([[input.y]],
 [[%token A B C D E
@@ -143,13 +227,32 @@ cd: C D;
 bc: B C;
 ]])

-AT_BISON_CHECK_CEX([input.y], [], [],
+AT_BISON_CHECK_CEX(
 [[input.y: warning: 1 shift/reduce conflict [-Wconflicts-sr]
 Shift/reduce conflict on token C:
-  First example        B . C D $end
-  First derivation     $accept ::=[ g ::=[ x ::=[ b ::=[ B . ] cd ::=[ C D ] ] ] $end ]
-  Second example       B . C $end
-  Second derivation    $accept ::=[ g ::=[ x ::=[ bc ::=[ B . C ] ] ] $end ]
+  First example        B . C $end
+  Shift derivation
+    $accept
+    `-> g                 $end
+        `-> x
+            `-> bc
+                `-> B . C
+  Second example       B . C D $end
+  Reduce derivation
+    $accept
+    `-> g                       $end
+        `-> x
+            `-> b       cd
+                `-> B . `-> C D
+
+input.y:6.4: warning: rule useless in parser due to conflicts [-Wother]
+]],
+[[input.y: warning: 1 shift/reduce conflict [-Wconflicts-sr]
+Shift/reduce conflict on token C:
+  First example        B . C $end
+  Shift derivation     $accept -> [ g -> [ x -> [ bc -> [ B . C ] ] ] $end ]
+  Second example       B . C D $end
+  Reduce derivation    $accept -> [ g -> [ x -> [ b -> [ B . ] cd -> [ C D ] ] ] $end ]

 input.y:6.4: warning: rule useless in parser due to conflicts [-Wother]
 ]])
@@ -161,7 +264,6 @@ AT_CLEANUP
 ## ------------------------------ ##

 AT_SETUP([Non-unifying Unambiguous S/R])
-AT_KEYWORDS([cex])

 AT_DATA([[input.y]],
 [[%token A B
@@ -172,13 +274,32 @@ x: A;
 y: A A B;
 ]])

-AT_BISON_CHECK_CEX([input.y], [], [],
+AT_BISON_CHECK_CEX(
 [[input.y: warning: 1 shift/reduce conflict [-Wconflicts-sr]
 Shift/reduce conflict on token A:
-  First example        A . A $end
-  First derivation     $accept ::=[ s ::=[ s ::=[ t ::=[ x ::=[ A . ] ] ] t ::=[ x ::=[ A ] ] ] $end ]
-  Second example       A . A B $end
-  Second derivation    $accept ::=[ s ::=[ t ::=[ y ::=[ A . A B ] ] ] $end ]
+  First example        A . A B $end
+  Shift derivation
+    $accept
+    `-> s                   $end
+        `-> t
+            `-> y
+                `-> A . A B
+  Second example       A . A $end
+  Reduce derivation
+    $accept
+    `-> s                             $end
+        `-> s               t
+            `-> t           `-> x
+                `-> x           `-> A
+                    `-> A .
+
+]],
+[[input.y: warning: 1 shift/reduce conflict [-Wconflicts-sr]
+Shift/reduce conflict on token A:
+  First example        A . A B $end
+  Shift derivation     $accept -> [ s -> [ t -> [ y -> [ A . A B ] ] ] $end ]
+  Second example       A . A $end
+  Reduce derivation    $accept -> [ s -> [ s -> [ t -> [ x -> [ A . ] ] ] t -> [ x -> [ A ] ] ] $end ]

 ]])

@@ -189,7 +310,6 @@ AT_CLEANUP
 ## ----------------------- ##

 AT_SETUP([S/R after first token])
-AT_KEYWORDS([cex])

 AT_DATA([[input.y]],
 [[%token A B X Y
@@ -205,18 +325,48 @@ xy: X Y;
 y: Y;
 ]])

-AT_BISON_CHECK_CEX([input.y], [], [],
+AT_BISON_CHECK_CEX(
 [[input.y: warning: 2 shift/reduce conflicts [-Wconflicts-sr]
 Shift/reduce conflict on token A:
  Example              b . A X X Y
-  First derivation     a ::=[ r ::=[ b . ] t ::=[ A x ::=[ X ] xy ::=[ X Y ] ] ]
-  Second derivation    a ::=[ s ::=[ b . xx ::=[ A X X ] y ::=[ Y ] ] ]
+  Shift derivation
+    a
+    `-> s
+        `-> b . xx        y
+                `-> A X X `-> Y
+  Reduce derivation
+    a
+    `-> r       t
+        `-> b . `-> A x     xy
+                      `-> X `-> X Y

 Shift/reduce conflict on token X:
-  First example        X . X xy
-  First derivation     a ::=[ x ::=[ X . ] t ::=[ X xy ] ]
-  Second example       A X . X
-  Second derivation    a ::=[ t ::=[ A xx ::=[ X . X ] ] ]
+  First example        A X . X
+  Shift derivation
+    a
+    `-> t
+        `-> A xx
+              `-> X . X
+  Second example       X . X xy
+  Reduce derivation
+    a
+    `-> x       t
+        `-> X . `-> X xy
+
+input.y:4.4: warning: rule useless in parser due to conflicts [-Wother]
+input.y:8.4: warning: rule useless in parser due to conflicts [-Wother]
+]],
+[[input.y: warning: 2 shift/reduce conflicts [-Wconflicts-sr]
+Shift/reduce conflict on token A:
+  Example              b . A X X Y
+  Shift derivation     a -> [ s -> [ b . xx -> [ A X X ] y -> [ Y ] ] ]
+  Reduce derivation    a -> [ r -> [ b . ] t -> [ A x -> [ X ] xy -> [ X Y ] ] ]
+
+Shift/reduce conflict on token X:
+  First example        A X . X
+  Shift derivation     a -> [ t -> [ A xx -> [ X . X ] ] ]
+  Second example       X . X xy
+  Reduce derivation    a -> [ x -> [ X . ] t -> [ X xy ] ]

 input.y:4.4: warning: rule useless in parser due to conflicts [-Wother]
 input.y:8.4: warning: rule useless in parser due to conflicts [-Wother]
@@ -229,7 +379,6 @@ AT_CLEANUP
 ## ----------------------------- ##

 AT_SETUP([Unifying R/R counterexample])
-AT_KEYWORDS([cex])

 AT_DATA([[input.y]],
 [[%token A
@@ -238,12 +387,25 @@ a : A b ;
 b : A | b;
 ]])

-AT_BISON_CHECK_CEX([input.y], [], [],
+AT_BISON_CHECK_CEX(
 [[input.y: warning: 1 reduce/reduce conflict [-Wconflicts-rr]
 Reduce/reduce conflict on token $end:
  Example              A b .
-  First derivation     a ::=[ A b . ]
-  Second derivation    a ::=[ A b ::=[ b . ] ]
+  First derivation
+    a
+    `-> A b .
+  Second derivation
+    a
+    `-> A b
+          `-> b .
+
+input.y:4.9: warning: rule useless in parser due to conflicts [-Wother]
+]],
+[[input.y: warning: 1 reduce/reduce conflict [-Wconflicts-rr]
+Reduce/reduce conflict on token $end:
+  Example              A b .
+  First derivation     a -> [ A b . ]
+  Second derivation    a -> [ A b -> [ b . ] ]

 input.y:4.9: warning: rule useless in parser due to conflicts [-Wother]
 ]])
@@ -255,7 +417,6 @@ AT_CLEANUP
 ## --------------------------------- ##

 AT_SETUP([Non-unifying R/R LR(1) conflict])
-AT_KEYWORDS([cex])

 AT_DATA([[input.y]],
 [[%token A B C D
@@ -265,13 +426,30 @@ a: D;
 b: D;
 ]])

-AT_BISON_CHECK_CEX([input.y], [], [],
+AT_BISON_CHECK_CEX(
 [[input.y: warning: 2 reduce/reduce conflicts [-Wconflicts-rr]
 Reduce/reduce conflict on tokens A, C:
  First example        D . A $end
-  First derivation     $accept ::=[ s ::=[ a ::=[ D . ] A ] $end ]
+  First derivation
+    $accept
+    `-> s             $end
+        `-> a       A
+            `-> D .
  Second example       B D . A $end
-  Second derivation    $accept ::=[ s ::=[ B b ::=[ D . ] A ] $end ]
+  Second derivation
+    $accept
+    `-> s               $end
+        `-> B b       A
+              `-> D .
+
+input.y:5.4: warning: rule useless in parser due to conflicts [-Wother]
+]],
+[[input.y: warning: 2 reduce/reduce conflicts [-Wconflicts-rr]
+Reduce/reduce conflict on tokens A, C:
+  First example        D . A $end
+  First derivation     $accept -> [ s -> [ a -> [ D . ] A ] $end ]
+  Second example       B D . A $end
+  Second derivation    $accept -> [ s -> [ B b -> [ D . ] A ] $end ]

 input.y:5.4: warning: rule useless in parser due to conflicts [-Wother]
 ]])
@@ -283,7 +461,6 @@ AT_CLEANUP
 ## --------------------------------- ##

 AT_SETUP([Non-unifying R/R LR(2) conflict])
-AT_KEYWORDS([cex])

 AT_DATA([[input.y]],
 [[%token H J K X
@@ -293,14 +470,32 @@ a: H i;
 i: X | i J K;
 ]])

-AT_BISON_CHECK_CEX([input.y], [], [],
+AT_BISON_CHECK_CEX(
 [[input.y: warning: 1 shift/reduce conflict [-Wconflicts-sr]
 Shift/reduce conflict on token J:
 time limit exceeded: XXX
-  First example        H i . J $end
-  First derivation     $accept ::=[ s ::=[ a ::=[ H i . ] J ] $end ]
-  Second example       H i . J K $end
-  Second derivation    $accept ::=[ a ::=[ H i ::=[ i . J K ] ] $end ]
+  First example        H i . J K $end
+  Shift derivation
+    $accept
+    `-> a                 $end
+        `-> H i
+              `-> i . J K
+  Second example       H i . J $end
+  Reduce derivation
+    $accept
+    `-> s               $end
+        `-> a         J
+            `-> H i .
+
+input.y:4.4-6: warning: rule useless in parser due to conflicts [-Wother]
+]],
+[[input.y: warning: 1 shift/reduce conflict [-Wconflicts-sr]
+Shift/reduce conflict on token J:
+time limit exceeded: XXX
+  First example        H i . J K $end
+  Shift derivation     $accept -> [ a -> [ H i -> [ i . J K ] ] $end ]
+  Second example       H i . J $end
+  Reduce derivation    $accept -> [ s -> [ a -> [ H i . ] J ] $end ]

 input.y:4.4-6: warning: rule useless in parser due to conflicts [-Wother]
 ]])
@@ -315,7 +510,6 @@ AT_CLEANUP
 # graph search

 AT_SETUP([Cex Search Prepend])
-AT_KEYWORDS([cex])

 AT_DATA([[input.y]],
 [[%token N A B C D
@@ -326,17 +520,48 @@ a: A;
 b: A B C | A B D;
 ]])

-AT_BISON_CHECK_CEX([input.y], [], [],
+AT_BISON_CHECK_CEX(
 [[input.y: warning: 1 shift/reduce conflict [-Wconflicts-sr]
 Shift/reduce conflict on token B:
  Example              N A . B C
-  First derivation     s ::=[ n ::=[ N a ::=[ A . ] B ] C ]
-  Second derivation    s ::=[ n ::=[ N b ::=[ A . B C ] ] ]
+  Shift derivation
+    s
+    `-> n
+        `-> N b
+              `-> A . B C
+  Reduce derivation
+    s
+    `-> n               C
+        `-> N a       B
+              `-> A .

 Shift/reduce conflict on token B:
  Example              N N A . B D C
-  First derivation     s ::=[ n ::=[ N n ::=[ N a ::=[ A . ] B ] D ] C ]
-  Second derivation    s ::=[ n ::=[ N n ::=[ N b ::=[ A . B D ] ] C ] ]
+  Shift derivation
+    s
+    `-> n
+        `-> N n                 C
+              `-> N b
+                    `-> A . B D
+  Reduce derivation
+    s
+    `-> n                       C
+        `-> N n               D
+              `-> N a       B
+                    `-> A .
+
+input.y:5.4: warning: rule useless in parser due to conflicts [-Wother]
+]],
+[[input.y: warning: 1 shift/reduce conflict [-Wconflicts-sr]
+Shift/reduce conflict on token B:
+  Example              N A . B C
+  Shift derivation     s -> [ n -> [ N b -> [ A . B C ] ] ]
+  Reduce derivation    s -> [ n -> [ N a -> [ A . ] B ] C ]
+
+Shift/reduce conflict on token B:
+  Example              N N A . B D C
+  Shift derivation     s -> [ n -> [ N n -> [ N b -> [ A . B D ] ] C ] ]
+  Reduce derivation    s -> [ n -> [ N n -> [ N a -> [ A . ] B ] D ] C ]

 input.y:5.4: warning: rule useless in parser due to conflicts [-Wother]
 ]])
@@ -351,7 +576,6 @@ AT_CLEANUP
 # precedence/associativity directives work.

 AT_SETUP([R/R cex with prec])
-AT_KEYWORDS([cex])

 AT_DATA([[input.y]],
 [[%left b
@@ -363,17 +587,49 @@ B : A b A;
 C : A c A;
 ]])

-AT_BISON_CHECK_CEX([input.y], [], [],
+AT_BISON_CHECK_CEX(
 [[input.y: warning: 4 reduce/reduce conflicts [-Wconflicts-rr]
 Reduce/reduce conflict on tokens b, c:
  Example              B . b c
-  First derivation     S ::=[ B ::=[ A ::=[ B . ] b A ::=[ ] ] C ::=[ A ::=[ ] c A ::=[ ] ] ]
-  Second derivation    S ::=[ B C ::=[ A ::=[ B ::=[ A ::=[ . ] b A ::=[ ] ] ] c A ::=[ ] ] ]
+  First derivation
+    S
+    `-> B                        C
+        `-> A       b A          `-> A          c A
+            `-> B .   `-> %empty     `-> %empty   `-> %empty
+  Second derivation
+    S
+    `-> B C
+          `-> A                          c A
+              `-> B                        `-> %empty
+                  `-> A     b A
+                      `-> .   `-> %empty

 Reduce/reduce conflict on tokens b, c:
  Example              C . c b
-  First derivation     S ::=[ C ::=[ A ::=[ C . ] c A ::=[ ] ] B ::=[ A ::=[ ] b A ::=[ ] ] ]
-  Second derivation    S ::=[ C B ::=[ A ::=[ C ::=[ A ::=[ . ] c A ::=[ ] ] ] b A ::=[ ] ] ]
+  First derivation
+    S
+    `-> C                        B
+        `-> A       c A          `-> A          b A
+            `-> C .   `-> %empty     `-> %empty   `-> %empty
+  Second derivation
+    S
+    `-> C B
+          `-> A                          b A
+              `-> C                        `-> %empty
+                  `-> A     c A
+                      `-> .   `-> %empty
+
+]],
+[[input.y: warning: 4 reduce/reduce conflicts [-Wconflicts-rr]
+Reduce/reduce conflict on tokens b, c:
+  Example              B . b c
+  First derivation     S -> [ B -> [ A -> [ B . ] b A -> [ ] ] C -> [ A -> [ ] c A -> [ ] ] ]
+  Second derivation    S -> [ B C -> [ A -> [ B -> [ A -> [ . ] b A -> [ ] ] ] c A -> [ ] ] ]
+
+Reduce/reduce conflict on tokens b, c:
+  Example              C . c b
+  First derivation     S -> [ C -> [ A -> [ C . ] c A -> [ ] ] B -> [ A -> [ ] b A -> [ ] ] ]
+  Second derivation    S -> [ C B -> [ A -> [ C -> [ A -> [ . ] c A -> [ ] ] ] b A -> [ ] ] ]

 ]])

@@ -384,7 +640,6 @@ AT_CLEANUP
 ## ------------------- ##

 AT_SETUP([Null nonterminals])
-AT_KEYWORDS([cex])

 AT_DATA([[input.y]],
 [[%token A
@@ -395,57 +650,187 @@ c : ;
 d : a | c A | d;
 ]])

-AT_BISON_CHECK_CEX([input.y], [], [],
+AT_BISON_CHECK_CEX(
 [[input.y: warning: 1 shift/reduce conflict [-Wconflicts-sr]
 input.y: warning: 6 reduce/reduce conflicts [-Wconflicts-rr]
 Reduce/reduce conflict on token A:
  First example        . c A A $end
-  First derivation     $accept ::=[ a ::=[ b ::=[ . ] d ::=[ c A A ] ] $end ]
+  First derivation
+    $accept
+    `-> a                   $end
+        `-> b     d
+            `-> . `-> c A A
  Second example       . c A A $end
-  Second derivation    $accept ::=[ a ::=[ c ::=[ . ] d ::=[ c A A ] ] $end ]
+  Second derivation
+    $accept
+    `-> a                   $end
+        `-> c     d
+            `-> . `-> c A A

 Reduce/reduce conflict on token A:
 time limit exceeded: XXX
  First example        b . c A A $end
-  First derivation     $accept ::=[ a ::=[ b d ::=[ a ::=[ b ::=[ . ] d ::=[ c A A ] ] ] ] $end ]
+  First derivation
+    $accept
+    `-> a                             $end
+        `-> b d
+              `-> a
+                  `-> b     d
+                      `-> . `-> c A A
  Second example       b . A $end
-  Second derivation    $accept ::=[ a ::=[ b d ::=[ c ::=[ . ] A ] ] $end ]
+  Second derivation
+    $accept
+    `-> a                 $end
+        `-> b d
+              `-> c     A
+                  `-> .

 Reduce/reduce conflict on token A:
 time limit exceeded: XXX
  First example        c . c A A $end
-  First derivation     $accept ::=[ a ::=[ c d ::=[ a ::=[ b ::=[ . ] d ::=[ c A A ] ] ] ] $end ]
+  First derivation
+    $accept
+    `-> a                             $end
+        `-> c d
+              `-> a
+                  `-> b     d
+                      `-> . `-> c A A
  Second example       c . A $end
-  Second derivation    $accept ::=[ a ::=[ c d ::=[ c ::=[ . ] A ] ] $end ]
+  Second derivation
+    $accept
+    `-> a                 $end
+        `-> c d
+              `-> c     A
+                  `-> .

 Shift/reduce conflict on token A:
 time limit exceeded: XXX
-  First example        b c . c A A $end
-  First derivation     $accept ::=[ a ::=[ b d ::=[ a ::=[ c d ::=[ a ::=[ b ::=[ . ] d ::=[ c A A ] ] ] ] ] ] $end ]
-  Second example       b c . A
-  Second derivation    a ::=[ b d ::=[ c . A ] ]
+  First example        b c . A
+  Shift derivation
+    a
+    `-> b d
+          `-> c . A
+  Second example       b c . c A A $end
+  Reduce derivation
+    $accept
+    `-> a                                       $end
+        `-> b d
+              `-> a
+                  `-> c d
+                        `-> a
+                            `-> b     d
+                                `-> . `-> c A A

 Reduce/reduce conflict on token A:
  First example        b c . c A A $end
-  First derivation     $accept ::=[ a ::=[ b d ::=[ a ::=[ c d ::=[ a ::=[ b ::=[ . ] d ::=[ c A A ] ] ] ] ] ] $end ]
+  First derivation
+    $accept
+    `-> a                                       $end
+        `-> b d
+              `-> a
+                  `-> c d
+                        `-> a
+                            `-> b     d
+                                `-> . `-> c A A
  Second example       b c . A $end
-  Second derivation    $accept ::=[ a ::=[ b d ::=[ a ::=[ c d ::=[ c ::=[ . ] A ] ] ] ] $end ]
+  Second derivation
+    $accept
+    `-> a                           $end
+        `-> b d
+              `-> a
+                  `-> c d
+                        `-> c     A
+                            `-> .

 Shift/reduce conflict on token A:
-  First example        b c . A $end
-  First derivation     $accept ::=[ a ::=[ b d ::=[ a ::=[ c d ::=[ c ::=[ . ] A ] ] ] ] $end ]
-  Second example       b c . A
-  Second derivation    a ::=[ b d ::=[ c . A ] ]
+  First example        b c . A
+  Shift derivation
+    a
+    `-> b d
+          `-> c . A
+  Second example       b c . A $end
+  Reduce derivation
+    $accept
+    `-> a                           $end
+        `-> b d
+              `-> a
+                  `-> c d
+                        `-> c     A
+                            `-> .

 Reduce/reduce conflict on token $end:
  Example              b d .
-  First derivation     a ::=[ b d . ]
-  Second derivation    a ::=[ b d ::=[ d . ] ]
+  First derivation
+    a
+    `-> b d .
+  Second derivation
+    a
+    `-> b d
+          `-> d .

 Reduce/reduce conflict on token $end:
  Example              c d .
-  First derivation     a ::=[ c d . ]
-  Second derivation    a ::=[ c d ::=[ d . ] ]
+  First derivation
+    a
+    `-> c d .
+  Second derivation
+    a
+    `-> c d
+          `-> d .
+
+input.y:5.4: warning: rule useless in parser due to conflicts [-Wother]
+input.y:6.15: warning: rule useless in parser due to conflicts [-Wother]
+]],
+[[input.y: warning: 1 shift/reduce conflict [-Wconflicts-sr]
+input.y: warning: 6 reduce/reduce conflicts [-Wconflicts-rr]
+Reduce/reduce conflict on token A:
+  First example        . c A A $end
+  First derivation     $accept -> [ a -> [ b -> [ . ] d -> [ c A A ] ] $end ]
+  Second example       . c A A $end
+  Second derivation    $accept -> [ a -> [ c -> [ . ] d -> [ c A A ] ] $end ]
+
+Reduce/reduce conflict on token A:
+time limit exceeded: XXX
+  First example        b . c A A $end
+  First derivation     $accept -> [ a -> [ b d -> [ a -> [ b -> [ . ] d -> [ c A A ] ] ] ] $end ]
+  Second example       b . A $end
+  Second derivation    $accept -> [ a -> [ b d -> [ c -> [ . ] A ] ] $end ]
+
+Reduce/reduce conflict on token A:
+time limit exceeded: XXX
+  First example        c . c A A $end
+  First derivation     $accept -> [ a -> [ c d -> [ a -> [ b -> [ . ] d -> [ c A A ] ] ] ] $end ]
+  Second example       c . A $end
+  Second derivation    $accept -> [ a -> [ c d -> [ c -> [ . ] A ] ] $end ]
+
+Shift/reduce conflict on token A:
+time limit exceeded: XXX
+  First example        b c . A
+  Shift derivation     a -> [ b d -> [ c . A ] ]
+  Second example       b c . c A A $end
+  Reduce derivation    $accept -> [ a -> [ b d -> [ a -> [ c d -> [ a -> [ b -> [ . ] d -> [ c A A ] ] ] ] ] ] $end ]
+
+Reduce/reduce conflict on token A:
+  First example        b c . c A A $end
+  First derivation     $accept -> [ a -> [ b d -> [ a -> [ c d -> [ a -> [ b -> [ . ] d -> [ c A A ] ] ] ] ] ] $end ]
+  Second example       b c . A $end
+  Second derivation    $accept -> [ a -> [ b d -> [ a -> [ c d -> [ c -> [ . ] A ] ] ] ] $end ]
+
+Shift/reduce conflict on token A:
+  First example        b c . A
+  Shift derivation     a -> [ b d -> [ c . A ] ]
+  Second example       b c . A $end
+  Reduce derivation    $accept -> [ a -> [ b d -> [ a -> [ c d -> [ c -> [ . ] A ] ] ] ] $end ]
+
+Reduce/reduce conflict on token $end:
+  Example              b d .
+  First derivation     a -> [ b d . ]
+  Second derivation    a -> [ b d -> [ d . ] ]
+
+Reduce/reduce conflict on token $end:
+  Example              c d .
+  First derivation     a -> [ c d . ]
+  Second derivation    a -> [ c d -> [ d . ] ]

 input.y:5.4: warning: rule useless in parser due to conflicts [-Wother]
 input.y:6.15: warning: rule useless in parser due to conflicts [-Wother]
@@ -458,7 +843,6 @@ AT_CLEANUP
 ## --------------------------- ##

 AT_SETUP([Non-unifying Prefix Share])
-AT_KEYWORDS([cex])

 # Tests for a counterexample which should start its derivation
 # at a shared symbol rather than the start symbol.
@@ -471,12 +855,27 @@ a: H i J J
 i: %empty | i J;
 ]])

-AT_BISON_CHECK_CEX([input.y], [], [],
+AT_BISON_CHECK_CEX(
 [[input.y: warning: 1 shift/reduce conflict [-Wconflicts-sr]
 Shift/reduce conflict on token J:
  Example              H i J . J J
-  First derivation     s ::=[ a ::=[ H i ::=[ i J . ] J J ] ]
-  Second derivation    s ::=[ a ::=[ H i J . J ] J ]
+  Shift derivation
+    s
+    `-> a             J
+        `-> H i J . J
+  Reduce derivation
+    s
+    `-> a
+        `-> H i         J J
+              `-> i J .
+
+input.y:5.13-15: warning: rule useless in parser due to conflicts [-Wother]
+]],
+[[input.y: warning: 1 shift/reduce conflict [-Wconflicts-sr]
+Shift/reduce conflict on token J:
+  Example              H i J . J J
+  Shift derivation     s -> [ a -> [ H i J . J ] J ]
+  Reduce derivation    s -> [ a -> [ H i -> [ i J . ] J J ] ]

 input.y:5.13-15: warning: rule useless in parser due to conflicts [-Wother]
 ]])
@@ -491,7 +890,6 @@ AT_CLEANUP
 # are derived correctly.

 AT_SETUP([Deep Null Unifying])
-AT_KEYWORDS([cex])

 AT_DATA([[input.y]],
 [[%token A D
@@ -503,12 +901,27 @@ c: %empty
 d: D;
 ]])

-AT_BISON_CHECK_CEX([input.y], [], [],
+AT_BISON_CHECK_CEX(
 [[input.y: warning: 1 shift/reduce conflict [-Wconflicts-sr]
 Shift/reduce conflict on token D:
  Example              A a . D
-  First derivation     s ::=[ A a a ::=[ b ::=[ c ::=[ . ] ] ] d ::=[ D ] ]
-  Second derivation    s ::=[ A a d ::=[ . D ] ]
+  Shift derivation
+    s
+    `-> A a d
+            `-> . D
+  Reduce derivation
+    s
+    `-> A a a             d
+            `-> b         `-> D
+                `-> c
+                    `-> .
+
+]],
+[[input.y: warning: 1 shift/reduce conflict [-Wconflicts-sr]
+Shift/reduce conflict on token D:
+  Example              A a . D
+  Shift derivation     s -> [ A a d -> [ . D ] ]
+  Reduce derivation    s -> [ A a a -> [ b -> [ c -> [ . ] ] ] d -> [ D ] ]

 ]])

@@ -521,7 +934,6 @@ AT_CLEANUP
 # Tests that expand_to_conflict works with nullable sybols

 AT_SETUP([Deep Null Non-unifying])
-AT_KEYWORDS([cex])

 AT_DATA([[input.y]],
 [[%token A D E
@@ -533,13 +945,31 @@ c: %empty
 d: D;
 ]])

-AT_BISON_CHECK_CEX([input.y], [], [],
+AT_BISON_CHECK_CEX(
 [[input.y: warning: 1 shift/reduce conflict [-Wconflicts-sr]
 Shift/reduce conflict on token D:
-  First example        A a . D E $end
-  First derivation     $accept ::=[ s ::=[ A a a ::=[ b ::=[ c ::=[ . ] ] ] d ::=[ D ] E ] $end ]
-  Second example       A a . D $end
-  Second derivation    $accept ::=[ s ::=[ A a d ::=[ . D ] ] $end ]
+  First example        A a . D $end
+  Shift derivation
+    $accept
+    `-> s               $end
+        `-> A a d
+                `-> . D
+  Second example       A a . D E $end
+  Reduce derivation
+    $accept
+    `-> s                             $end
+        `-> A a a             d     E
+                `-> b         `-> D
+                    `-> c
+                        `-> .
+
+]],
+[[input.y: warning: 1 shift/reduce conflict [-Wconflicts-sr]
+Shift/reduce conflict on token D:
+  First example        A a . D $end
+  Shift derivation     $accept -> [ s -> [ A a d -> [ . D ] ] $end ]
+  Second example       A a . D E $end
+  Reduce derivation    $accept -> [ s -> [ A a a -> [ b -> [ c -> [ . ] ] ] d -> [ D ] E ] $end ]

 ]])

@@ -28,6 +28,8 @@ m4_pushdef([AT_TEST],
 AT_SETUP([$1])
 AT_KEYWORDS([diagnostics])

+m4_if(m4_index([$1], [Counterexample]), [-1], [], [AT_KEYWORDS([cex])])
+
 # We need UTF-8 support for correct screen-width computation of UTF-8
 # characters.  Skip the test if not available.
 locale=`locale -a | $EGREP '^en_US\.(UTF-8|utf8)$' | sed 1q`
@@ -534,33 +536,115 @@ exp
 [1],
 [[input.y: <error>error:</error> shift/reduce conflicts: 4 found, 0 expected
 Shift/reduce conflict on token "+":
-  Example              <cex-0><cex-1><cex-leaf>exp</cex-leaf> <cex-leaf>"+"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-dot>•</cex-dot></cex-1> <cex-leaf>"+"</cex-leaf> <cex-leaf>exp</cex-leaf></cex-0>
-  First derivation     <cex-0><cex-step>exp ::=[ </cex-step><cex-1><cex-step>exp ::=[ </cex-step><cex-leaf>exp</cex-leaf> <cex-leaf>"+"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-dot>•</cex-dot><cex-step> ]</cex-step></cex-1> <cex-leaf>"+"</cex-leaf> <cex-leaf>exp</cex-leaf><cex-step> ]</cex-step></cex-0>
  Example              <cex-0><cex-leaf>exp</cex-leaf> <cex-leaf>"+"</cex-leaf><cex-1> <cex-leaf>exp</cex-leaf> <cex-dot>•</cex-dot> <cex-leaf>"+"</cex-leaf> <cex-leaf>exp</cex-leaf></cex-1></cex-0>
-  Second derivation    <cex-0><cex-step>exp ::=[ </cex-step><cex-leaf>exp</cex-leaf> <cex-leaf>"+"</cex-leaf><cex-1> <cex-step>exp ::=[ </cex-step><cex-leaf>exp</cex-leaf> <cex-dot>•</cex-dot> <cex-leaf>"+"</cex-leaf> <cex-leaf>exp</cex-leaf><cex-step> ]</cex-step></cex-1><cex-step> ]</cex-step></cex-0>
+  Shift derivation
+    <cex-0><cex-step>exp</cex-step></cex-0>
+    <cex-0><cex-step>↳ <cex-leaf>exp</cex-leaf><cex-leaf> "+"</cex-leaf><cex-1><cex-step> exp</cex-step></cex-1></cex-step></cex-0>
+    <cex-1><cex-step>          ↳ <cex-leaf>exp</cex-leaf><cex-dot> •</cex-dot><cex-leaf> "+"</cex-leaf><cex-leaf> exp</cex-leaf></cex-step></cex-1>
+  Example              <cex-0><cex-1><cex-leaf>exp</cex-leaf> <cex-leaf>"+"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-dot>•</cex-dot></cex-1> <cex-leaf>"+"</cex-leaf> <cex-leaf>exp</cex-leaf></cex-0>
+  Reduce derivation
+    <cex-0><cex-step>exp</cex-step></cex-0>
+    <cex-0><cex-step>↳ <cex-1><cex-step>exp</cex-step></cex-1><cex-leaf>             "+"</cex-leaf><cex-leaf> exp</cex-leaf></cex-step></cex-0>
+    <cex-1><cex-step>  ↳ <cex-leaf>exp</cex-leaf><cex-leaf> "+"</cex-leaf><cex-leaf> exp</cex-leaf><cex-dot> •</cex-dot></cex-step></cex-1>

 Shift/reduce conflict on token "else":
-  Example              <cex-0><cex-leaf>"if"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-leaf>"then"</cex-leaf><cex-1> <cex-leaf>"if"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-leaf>"then"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-dot>•</cex-dot></cex-1> <cex-leaf>"else"</cex-leaf> <cex-leaf>exp</cex-leaf></cex-0>
-  First derivation     <cex-0><cex-step>exp ::=[ </cex-step><cex-leaf>"if"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-leaf>"then"</cex-leaf><cex-1> <cex-step>exp ::=[ </cex-step><cex-leaf>"if"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-leaf>"then"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-dot>•</cex-dot><cex-step> ]</cex-step></cex-1> <cex-leaf>"else"</cex-leaf> <cex-leaf>exp</cex-leaf><cex-step> ]</cex-step></cex-0>
  Example              <cex-0><cex-leaf>"if"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-leaf>"then"</cex-leaf><cex-1> <cex-leaf>"if"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-leaf>"then"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-dot>•</cex-dot> <cex-leaf>"else"</cex-leaf> <cex-leaf>exp</cex-leaf></cex-1></cex-0>
-  Second derivation    <cex-0><cex-step>exp ::=[ </cex-step><cex-leaf>"if"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-leaf>"then"</cex-leaf><cex-1> <cex-step>exp ::=[ </cex-step><cex-leaf>"if"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-leaf>"then"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-dot>•</cex-dot> <cex-leaf>"else"</cex-leaf> <cex-leaf>exp</cex-leaf><cex-step> ]</cex-step></cex-1><cex-step> ]</cex-step></cex-0>
+  Shift derivation
+    <cex-0><cex-step>exp</cex-step></cex-0>
+    <cex-0><cex-step>↳ <cex-leaf>"if"</cex-leaf><cex-leaf> exp</cex-leaf><cex-leaf> "then"</cex-leaf><cex-1><cex-step> exp</cex-step></cex-1></cex-step></cex-0>
+    <cex-1><cex-step>                  ↳ <cex-leaf>"if"</cex-leaf><cex-leaf> exp</cex-leaf><cex-leaf> "then"</cex-leaf><cex-leaf> exp</cex-leaf><cex-dot> •</cex-dot><cex-leaf> "else"</cex-leaf><cex-leaf> exp</cex-leaf></cex-step></cex-1>
+  Example              <cex-0><cex-leaf>"if"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-leaf>"then"</cex-leaf><cex-1> <cex-leaf>"if"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-leaf>"then"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-dot>•</cex-dot></cex-1> <cex-leaf>"else"</cex-leaf> <cex-leaf>exp</cex-leaf></cex-0>
+  Reduce derivation
+    <cex-0><cex-step>exp</cex-step></cex-0>
+    <cex-0><cex-step>↳ <cex-leaf>"if"</cex-leaf><cex-leaf> exp</cex-leaf><cex-leaf> "then"</cex-leaf><cex-1><cex-step> exp</cex-step></cex-1><cex-leaf>                     "else"</cex-leaf><cex-leaf> exp</cex-leaf></cex-step></cex-0>
+    <cex-1><cex-step>                  ↳ <cex-leaf>"if"</cex-leaf><cex-leaf> exp</cex-leaf><cex-leaf> "then"</cex-leaf><cex-leaf> exp</cex-leaf><cex-dot> •</cex-dot></cex-step></cex-1>

 Shift/reduce conflict on token "+":
-  Example              <cex-0><cex-1><cex-leaf>"if"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-leaf>"then"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-dot>•</cex-dot></cex-1> <cex-leaf>"+"</cex-leaf> <cex-leaf>exp</cex-leaf></cex-0>
-  First derivation     <cex-0><cex-step>exp ::=[ </cex-step><cex-1><cex-step>exp ::=[ </cex-step><cex-leaf>"if"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-leaf>"then"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-dot>•</cex-dot><cex-step> ]</cex-step></cex-1> <cex-leaf>"+"</cex-leaf> <cex-leaf>exp</cex-leaf><cex-step> ]</cex-step></cex-0>
  Example              <cex-0><cex-leaf>"if"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-leaf>"then"</cex-leaf><cex-1> <cex-leaf>exp</cex-leaf> <cex-dot>•</cex-dot> <cex-leaf>"+"</cex-leaf> <cex-leaf>exp</cex-leaf></cex-1></cex-0>
-  Second derivation    <cex-0><cex-step>exp ::=[ </cex-step><cex-leaf>"if"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-leaf>"then"</cex-leaf><cex-1> <cex-step>exp ::=[ </cex-step><cex-leaf>exp</cex-leaf> <cex-dot>•</cex-dot> <cex-leaf>"+"</cex-leaf> <cex-leaf>exp</cex-leaf><cex-step> ]</cex-step></cex-1><cex-step> ]</cex-step></cex-0>
+  Shift derivation
+    <cex-0><cex-step>exp</cex-step></cex-0>
+    <cex-0><cex-step>↳ <cex-leaf>"if"</cex-leaf><cex-leaf> exp</cex-leaf><cex-leaf> "then"</cex-leaf><cex-1><cex-step> exp</cex-step></cex-1></cex-step></cex-0>
+    <cex-1><cex-step>                  ↳ <cex-leaf>exp</cex-leaf><cex-dot> •</cex-dot><cex-leaf> "+"</cex-leaf><cex-leaf> exp</cex-leaf></cex-step></cex-1>
+  Example              <cex-0><cex-1><cex-leaf>"if"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-leaf>"then"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-dot>•</cex-dot></cex-1> <cex-leaf>"+"</cex-leaf> <cex-leaf>exp</cex-leaf></cex-0>
+  Reduce derivation
+    <cex-0><cex-step>exp</cex-step></cex-0>
+    <cex-0><cex-step>↳ <cex-1><cex-step>exp</cex-step></cex-1><cex-leaf>                     "+"</cex-leaf><cex-leaf> exp</cex-leaf></cex-step></cex-0>
+    <cex-1><cex-step>  ↳ <cex-leaf>"if"</cex-leaf><cex-leaf> exp</cex-leaf><cex-leaf> "then"</cex-leaf><cex-leaf> exp</cex-leaf><cex-dot> •</cex-dot></cex-step></cex-1>

 Shift/reduce conflict on token "+":
-  Example              <cex-0><cex-1><cex-leaf>"if"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-leaf>"then"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-leaf>"else"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-dot>•</cex-dot></cex-1> <cex-leaf>"+"</cex-leaf> <cex-leaf>exp</cex-leaf></cex-0>
-  First derivation     <cex-0><cex-step>exp ::=[ </cex-step><cex-1><cex-step>exp ::=[ </cex-step><cex-leaf>"if"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-leaf>"then"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-leaf>"else"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-dot>•</cex-dot><cex-step> ]</cex-step></cex-1> <cex-leaf>"+"</cex-leaf> <cex-leaf>exp</cex-leaf><cex-step> ]</cex-step></cex-0>
  Example              <cex-0><cex-leaf>"if"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-leaf>"then"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-leaf>"else"</cex-leaf><cex-1> <cex-leaf>exp</cex-leaf> <cex-dot>•</cex-dot> <cex-leaf>"+"</cex-leaf> <cex-leaf>exp</cex-leaf></cex-1></cex-0>
-  Second derivation    <cex-0><cex-step>exp ::=[ </cex-step><cex-leaf>"if"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-leaf>"then"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-leaf>"else"</cex-leaf><cex-1> <cex-step>exp ::=[ </cex-step><cex-leaf>exp</cex-leaf> <cex-dot>•</cex-dot> <cex-leaf>"+"</cex-leaf> <cex-leaf>exp</cex-leaf><cex-step> ]</cex-step></cex-1><cex-step> ]</cex-step></cex-0>
+  Shift derivation
+    <cex-0><cex-step>exp</cex-step></cex-0>
+    <cex-0><cex-step>↳ <cex-leaf>"if"</cex-leaf><cex-leaf> exp</cex-leaf><cex-leaf> "then"</cex-leaf><cex-leaf> exp</cex-leaf><cex-leaf> "else"</cex-leaf><cex-1><cex-step> exp</cex-step></cex-1></cex-step></cex-0>
+    <cex-1><cex-step>                             ↳ <cex-leaf>exp</cex-leaf><cex-dot> •</cex-dot><cex-leaf> "+"</cex-leaf><cex-leaf> exp</cex-leaf></cex-step></cex-1>
+  Example              <cex-0><cex-1><cex-leaf>"if"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-leaf>"then"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-leaf>"else"</cex-leaf> <cex-leaf>exp</cex-leaf> <cex-dot>•</cex-dot></cex-1> <cex-leaf>"+"</cex-leaf> <cex-leaf>exp</cex-leaf></cex-0>
+  Reduce derivation
+    <cex-0><cex-step>exp</cex-step></cex-0>
+    <cex-0><cex-step>↳ <cex-1><cex-step>exp</cex-step></cex-1><cex-leaf>                                "+"</cex-leaf><cex-leaf> exp</cex-leaf></cex-step></cex-0>
+    <cex-1><cex-step>  ↳ <cex-leaf>"if"</cex-leaf><cex-leaf> exp</cex-leaf><cex-leaf> "then"</cex-leaf><cex-leaf> exp</cex-leaf><cex-leaf> "else"</cex-leaf><cex-leaf> exp</cex-leaf><cex-dot> •</cex-dot></cex-step></cex-1>

 ]])


+AT_TEST([[Deep Counterexamples]],
+[[%expect 0
+%%
+exp: x1 e1 foo1 x1 | y1 e2 bar1 y1
+foo1: foo2
+foo2: foo3
+foo3: x1 foo4
+foo4: "quuux"

+bar1: bar2
+bar2: bar3
+bar3: y1 bar4
+bar4: "quuux"
+
+x1: x2
+x2: x3
+x3: "X"
+
+y1: y2
+y2: y3
+y3: "X"
+
+e1:
+e2:
+]],
+[1],
+[[input.y:30.4: <warning>warning:</warning> empty rule without %empty [<warning>-Wempty-rule</warning>]
+   30 | e1:
+      |    <warning>^</warning>
+      |    <fixit-insert>%empty</fixit-insert>
+input.y:31.4: <warning>warning:</warning> empty rule without %empty [<warning>-Wempty-rule</warning>]
+   31 | e2:
+      |    <warning>^</warning>
+      |    <fixit-insert>%empty</fixit-insert>
+input.y: <error>error:</error> reduce/reduce conflicts: 1 found, 0 expected
+Reduce/reduce conflict on token "X":
+  Example              <cex-0><cex-1><cex-2><cex-3><cex-leaf>"X"</cex-leaf> <cex-dot>•</cex-dot></cex-3></cex-2></cex-1><cex-4></cex-4><cex-5><cex-6><cex-7><cex-8><cex-9><cex-10> <cex-leaf>"X"</cex-leaf></cex-10></cex-9></cex-8><cex-11> <cex-leaf>"quuux"</cex-leaf></cex-11></cex-7></cex-6></cex-5><cex-12><cex-13><cex-14> <cex-leaf>"X"</cex-leaf></cex-14></cex-13></cex-12></cex-0>
+  First derivation
+    <cex-0><cex-step>exp</cex-step></cex-0>
+    <cex-0><cex-step>↳ <cex-1><cex-step>x1</cex-step></cex-1><cex-4><cex-step>          e1</cex-step></cex-4><cex-5><cex-step>  foo1</cex-step></cex-5><cex-12><cex-step>                      x1</cex-step></cex-12></cex-step></cex-0>
+    <cex-1><cex-step>  ↳ <cex-2><cex-step>x2</cex-step></cex-2></cex-step></cex-1><cex-4><cex-step>        ↳ ε</cex-step></cex-4><cex-5><cex-step> ↳ <cex-6><cex-step>foo2</cex-step></cex-6></cex-step></cex-5><cex-12><cex-step>                    ↳ <cex-13><cex-step>x2</cex-step></cex-13></cex-step></cex-12>
+    <cex-2><cex-step>    ↳ <cex-3><cex-step>x3</cex-step></cex-3></cex-step></cex-2><cex-6><cex-step>            ↳ <cex-7><cex-step>foo3</cex-step></cex-7></cex-step></cex-6><cex-13><cex-step>                    ↳ <cex-14><cex-step>x3</cex-step></cex-14></cex-step></cex-13>
+    <cex-3><cex-step>      ↳ <cex-leaf>"X"</cex-leaf><cex-dot> •</cex-dot></cex-step></cex-3><cex-7><cex-step>         ↳ <cex-8><cex-step>x1</cex-step></cex-8><cex-11><cex-step>        foo4</cex-step></cex-11></cex-step></cex-7><cex-14><cex-step>          ↳ <cex-leaf>"X"</cex-leaf></cex-step></cex-14>
+    <cex-8><cex-step>                        ↳ <cex-9><cex-step>x2</cex-step></cex-9></cex-step></cex-8><cex-11><cex-step>      ↳ <cex-leaf>"quuux"</cex-leaf></cex-step></cex-11>
+    <cex-9><cex-step>                          ↳ <cex-10><cex-step>x3</cex-step></cex-10></cex-step></cex-9>
+    <cex-10><cex-step>                            ↳ <cex-leaf>"X"</cex-leaf></cex-step></cex-10>
+  Example              <cex-0><cex-1><cex-2><cex-3><cex-leaf>"X"</cex-leaf> <cex-dot>•</cex-dot></cex-3></cex-2></cex-1><cex-4></cex-4><cex-5><cex-6><cex-7><cex-8><cex-9><cex-10> <cex-leaf>"X"</cex-leaf></cex-10></cex-9></cex-8><cex-11> <cex-leaf>"quuux"</cex-leaf></cex-11></cex-7></cex-6></cex-5><cex-12><cex-13><cex-14> <cex-leaf>"X"</cex-leaf></cex-14></cex-13></cex-12></cex-0>
+  Second derivation
+    <cex-0><cex-step>exp</cex-step></cex-0>
+    <cex-0><cex-step>↳ <cex-1><cex-step>y1</cex-step></cex-1><cex-4><cex-step>          e2</cex-step></cex-4><cex-5><cex-step>  bar1</cex-step></cex-5><cex-12><cex-step>                      y1</cex-step></cex-12></cex-step></cex-0>
+    <cex-1><cex-step>  ↳ <cex-2><cex-step>y2</cex-step></cex-2></cex-step></cex-1><cex-4><cex-step>        ↳ ε</cex-step></cex-4><cex-5><cex-step> ↳ <cex-6><cex-step>bar2</cex-step></cex-6></cex-step></cex-5><cex-12><cex-step>                    ↳ <cex-13><cex-step>y2</cex-step></cex-13></cex-step></cex-12>
+    <cex-2><cex-step>    ↳ <cex-3><cex-step>y3</cex-step></cex-3></cex-step></cex-2><cex-6><cex-step>            ↳ <cex-7><cex-step>bar3</cex-step></cex-7></cex-step></cex-6><cex-13><cex-step>                    ↳ <cex-14><cex-step>y3</cex-step></cex-14></cex-step></cex-13>
+    <cex-3><cex-step>      ↳ <cex-leaf>"X"</cex-leaf><cex-dot> •</cex-dot></cex-step></cex-3><cex-7><cex-step>         ↳ <cex-8><cex-step>y1</cex-step></cex-8><cex-11><cex-step>        bar4</cex-step></cex-11></cex-step></cex-7><cex-14><cex-step>          ↳ <cex-leaf>"X"</cex-leaf></cex-step></cex-14>
+    <cex-8><cex-step>                        ↳ <cex-9><cex-step>y2</cex-step></cex-9></cex-step></cex-8><cex-11><cex-step>      ↳ <cex-leaf>"quuux"</cex-leaf></cex-step></cex-11>
+    <cex-9><cex-step>                          ↳ <cex-10><cex-step>y3</cex-step></cex-10></cex-step></cex-9>
+    <cex-10><cex-step>                            ↳ <cex-leaf>"X"</cex-leaf></cex-step></cex-10>
+
+input.y: <warning>warning:</warning> fix-its can be applied.  Rerun with option '--update'. [<warning>-Wother</warning>]
+]])

 m4_popdef([AT_TEST])

@@ -1169,11 +1169,16 @@ m4_define([AT_BISON_CHECK_XML],
  [cp xml-tests/test.output expout]
  AT_CHECK([[$XSLTPROC \
             `]]AT_SET_ENV[[ bison --print-datadir`/xslt/xml2text.xsl \
-             xml-tests/test.xml]], [[0]], [expout])
+             xml-tests/test.xml]], [[0]], [stdout])
+  # xml2text and xml2dot always use '•', while --report uses '•' or '.'
+  # depending on the locale, and the test suite is run with the plain
+  # C locale.
+  AT_CHECK([[sed -e 's/•/./g' stdout]], [], [expout])
  [sort xml-tests/test.gv > expout]
  AT_CHECK([[$XSLTPROC \
             `]]AT_SET_ENV[[ bison --print-datadir`/xslt/xml2dot.xsl \
-             xml-tests/test.xml | sort]], [[0]], [expout])
+             xml-tests/test.xml | sort | sed -e 's/•/./g']],
+           [[0]], [stdout])
  [rm -rf xml-tests expout]
  AT_RESTORE_SPECIAL_FILES
 [fi]])
@@ -36,6 +36,14 @@ $(top_srcdir)/%D%/package.m4: $(top_srcdir)/configure
 	} >$@.tmp
 	$(AM_V_at)mv $@.tmp $@

+
+# Update the test cases.  Consider the latest test results to be the
+# correct expectations, and change the test cases to match them.
+.PHONY: update-tests
+update-tests:
+	$(AM_V_GEN)cd $(top_srcdir) \
+	  && build-aux/update-test $(abs_builddir)/%D%/testsuite.dir/*/testsuite.log
+
 ## ------------------------- ##
 ## Generate the test suite.  ##
 ## ------------------------- ##
@@ -764,24 +764,24 @@ AT_CHECK([[sed -e 's/bison-xml-report version="[^"]*"/bison-xml-report version="
      </rule>
    </rules>
    <terminals>
-      <terminal symbol-number="0" token-number="0" name="&quot;end of file&quot;" usefulness="useful"/>
-      <terminal symbol-number="1" token-number="256" name="error" usefulness="useful"/>
-      <terminal symbol-number="3" token-number="258" name="&quot;:=&quot;" usefulness="useful"/>
-      <terminal symbol-number="4" token-number="259" name="&quot;incr&quot;" usefulness="useful"/>
-      <terminal symbol-number="5" token-number="260" name="&quot;identifier&quot;" usefulness="useful"/>
-      <terminal symbol-number="6" token-number="261" name="&quot;number&quot;" usefulness="useful"/>
-      <terminal symbol-number="7" token-number="262" name="&quot;(&quot;" usefulness="useful"/>
-      <terminal symbol-number="8" token-number="263" name="&quot;)&quot;" usefulness="useful"/>
+      <terminal symbol-number="0" token-number="0" name="&quot;end of file&quot;" type="" usefulness="useful"/>
+      <terminal symbol-number="1" token-number="256" name="error" type="" usefulness="useful"/>
+      <terminal symbol-number="3" token-number="258" name="&quot;:=&quot;" type="" usefulness="useful"/>
+      <terminal symbol-number="4" token-number="259" name="&quot;incr&quot;" type="" usefulness="useful"/>
+      <terminal symbol-number="5" token-number="260" name="&quot;identifier&quot;" type="std::string" usefulness="useful"/>
+      <terminal symbol-number="6" token-number="261" name="&quot;number&quot;" type="int" usefulness="useful"/>
+      <terminal symbol-number="7" token-number="262" name="&quot;(&quot;" type="" usefulness="useful"/>
+      <terminal symbol-number="8" token-number="263" name="&quot;)&quot;" type="" usefulness="useful"/>
    </terminals>
    <nonterminals>
-      <nonterminal symbol-number="9" name="$accept" usefulness="useful"/>
-      <nonterminal symbol-number="10" name="unit" usefulness="useful"/>
-      <nonterminal symbol-number="11" name="assignments" usefulness="useful"/>
-      <nonterminal symbol-number="12" name="assignment" usefulness="useful"/>
-      <nonterminal symbol-number="13" name="id" usefulness="useful"/>
-      <nonterminal symbol-number="14" name="exp" usefulness="useful"/>
-      <nonterminal symbol-number="15" name="@1" usefulness="useful"/>
-      <nonterminal symbol-number="16" name="@2" usefulness="useful"/>
+      <nonterminal symbol-number="9" name="$accept" type="" usefulness="useful"/>
+      <nonterminal symbol-number="10" name="unit" type="" usefulness="useful"/>
+      <nonterminal symbol-number="11" name="assignments" type="" usefulness="useful"/>
+      <nonterminal symbol-number="12" name="assignment" type="" usefulness="useful"/>
+      <nonterminal symbol-number="13" name="id" type="std::string" usefulness="useful"/>
+      <nonterminal symbol-number="14" name="exp" type="int" usefulness="useful"/>
+      <nonterminal symbol-number="15" name="@1" type="int" usefulness="useful"/>
+      <nonterminal symbol-number="16" name="@2" type="int" usefulness="useful"/>
    </nonterminals>
  </grammar>

@@ -1137,6 +1137,373 @@ AT_CHECK([[sed -e 's/bison-xml-report version="[^"]*"/bison-xml-report version="
 </bison-xml-report>
 ]])

+# Check HTML output.
+if test x"$XSLTPROC" != x""; then
+  AT_CHECK([[$XSLTPROC \
+           `]]AT_SET_ENV[[ bison --print-datadir`/xslt/xml2xhtml.xsl \
+           input.xml | \
+           sed -e 's/GNU Bison [0-9][-.0-9a-z]*/GNU Bison VERSION/' \
+           >input.html]])
+
+  AT_CHECK([cat input.html], [],
+[[<?xml version="1.0" encoding="UTF-8"?>
+<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
+<html xmlns="http://www.w3.org/1999/xhtml" xmlns:bison="http://www.gnu.org/software/bison/">
+  <head>
+    <meta http-equiv="Content-Type" content="text/html;charset=UTF-8" />
+    <title>input.yy - GNU Bison XML Automaton Report</title>
+    <style type="text/css">
+      body {
+        font-family: "Nimbus Sans L", Arial, sans-serif;
+        font-size: 9pt;
+      }
+      a:link {
+        color: #1f00ff;
+        text-decoration: none;
+      }
+      a:visited {
+        color: #1f00ff;
+        text-decoration: none;
+      }
+      a:hover {
+        color: red;
+      }
+      #menu a {
+        text-decoration: underline;
+      }
+      .i {
+        font-style: italic;
+      }
+      .pre {
+        font-family: monospace;
+        white-space: pre;
+      }
+      ol.decimal {
+        list-style-type: decimal;
+      }
+      ol.lower-alpha {
+        list-style-type: lower-alpha;
+      }
+      .dot {
+        color: #cc0000;
+      }
+      #footer {
+        margin-top: 3.5em;
+        font-size: 7pt;
+      }
+      </style>
+  </head>
+  <body><h1>GNU Bison XML Automaton Report</h1><p>
+    input grammar: <span class="i">input.yy</span></p>
+
+<h3>Table of Contents</h3><ul id="menu"><li><a href="#reductions">Reductions</a><ul class="lower-alpha"><li><a href="#nonterminals_useless_in_grammar">Nonterminals useless in grammar</a></li><li><a href="#terminals_unused_in_grammar">Terminals unused in grammar</a></li><li><a href="#rules_useless_in_grammar">Rules useless in grammar</a></li></ul></li><li><a href="#conflicts">Conflicts</a></li><li><a href="#grammar">Grammar</a><ul class="lower-alpha"><li><a href="#grammar">Itemset</a></li><li><a href="#terminals">Terminal symbols</a></li><li><a href="#nonterminals">Nonterminal symbols</a></li></ul></li><li><a href="#automaton">Automaton</a></li></ul><h2><a name="reductions" id="reductions"></a> Reductions</h2><h3><a name="nonterminals_useless_in_grammar" id="nonterminals_useless_in_grammar"></a> Nonterminals useless in grammar</h3>
+
+<h3><a name="terminals_unused_in_grammar" id="terminals_unused_in_grammar"></a> Terminals unused in grammar</h3>
+
+<h3><a name="rules_useless_in_grammar" id="rules_useless_in_grammar"></a> Rules useless in grammar</h3>
+<h2><a name="conflicts" id="conflicts"></a> Conflicts</h2>
+
+<h2><a name="grammar" id="grammar"></a> Grammar</h2>
+<p class="pre">
+  <a name="rule_0" id="rule_0">  0</a> <span class="i">$accept</span> → <span class="i">unit</span> <b>"end of file"</b>
+
+  <a name="rule_1" id="rule_1">  1</a> <span class="i">unit</span> → <span class="i">assignments</span> <span class="i">exp</span>
+
+  <a name="rule_2" id="rule_2">  2</a> <span class="i">assignments</span> → %empty
+  <a name="rule_3" id="rule_3">  3</a>             | <span class="i">assignments</span> <span class="i">assignment</span>
+
+  <a name="rule_4" id="rule_4">  4</a> <span class="i">assignment</span> → <span class="i">id</span> <b>":="</b> <span class="i">exp</span>
+
+  <a name="rule_5" id="rule_5">  5</a> <span class="i">id</span> → <b>"identifier"</b>
+
+  <a name="rule_6" id="rule_6">  6</a> <span class="i">@1</span> → %empty
+
+  <a name="rule_7" id="rule_7">  7</a> <span class="i">@2</span> → %empty
+
+  <a name="rule_8" id="rule_8">  8</a> <span class="i">exp</span> → <b>"incr"</b> <span class="i">exp</span> <span class="i">@1</span> <span class="i">@2</span> <span class="i">exp</span>
+  <a name="rule_9" id="rule_9">  9</a>     | <b>"("</b> <span class="i">exp</span> <b>")"</b>
+  <a name="rule_10" id="rule_10"> 10</a>     | <b>"identifier"</b>
+  <a name="rule_11" id="rule_11"> 11</a>     | <b>"number"</b>
+</p>
+
+<h3><a name="terminals" id="terminals"></a> Terminals, with rules where they appear</h3>
+
+<ul>
+  <li><b>"end of file"</b> (0) <a href="#rule_0">0</a></li>
+  <li><b>error</b> (256)</li>
+  <li><b>":="</b> (258) <a href="#rule_4">4</a></li>
+  <li><b>"incr"</b> (259) <a href="#rule_8">8</a></li>
+  <li><b>"identifier"</b> &lt;std::string&gt; (260) <a href="#rule_5">5</a> <a href="#rule_10">10</a></li>
+  <li><b>"number"</b> &lt;int&gt; (261) <a href="#rule_11">11</a></li>
+  <li><b>"("</b> (262) <a href="#rule_9">9</a></li>
+  <li><b>")"</b> (263) <a href="#rule_9">9</a></li>
+</ul>
+
+<h3><a name="nonterminals" id="nonterminals"></a> Nonterminals, with rules where they appear</h3>
+
+<ul>
+  <li><b>$accept</b> (9)
+    <ul>
+      <li>on left: <a href="#rule_0">0</a></li>
+    </ul>
+  </li>
+  <li><b>unit</b> (10)
+    <ul>
+      <li>on left: <a href="#rule_1">1</a></li>
+      <li>on right: <a href="#rule_0">0</a></li>
+    </ul>
+  </li>
+  <li><b>assignments</b> (11)
+    <ul>
+      <li>on left: <a href="#rule_2">2</a> <a href="#rule_3">3</a></li>
+      <li>on right: <a href="#rule_1">1</a> <a href="#rule_3">3</a></li>
+    </ul>
+  </li>
+  <li><b>assignment</b> (12)
+    <ul>
+      <li>on left: <a href="#rule_4">4</a></li>
+      <li>on right: <a href="#rule_3">3</a></li>
+    </ul>
+  </li>
+  <li><b>id</b> &lt;std::string&gt; (13)
+    <ul>
+      <li>on left: <a href="#rule_5">5</a></li>
+      <li>on right: <a href="#rule_4">4</a></li>
+    </ul>
+  </li>
+  <li><b>exp</b> &lt;int&gt; (14)
+    <ul>
+      <li>on left: <a href="#rule_8">8</a> <a href="#rule_9">9</a> <a href="#rule_10">10</a> <a href="#rule_11">11</a></li>
+      <li>on right: <a href="#rule_1">1</a> <a href="#rule_4">4</a> <a href="#rule_8">8</a> <a href="#rule_9">9</a></li>
+    </ul>
+  </li>
+  <li><b>@1</b> &lt;int&gt; (15)
+    <ul>
+      <li>on left: <a href="#rule_6">6</a></li>
+      <li>on right: <a href="#rule_8">8</a></li>
+    </ul>
+  </li>
+  <li><b>@2</b> &lt;int&gt; (16)
+    <ul>
+      <li>on left: <a href="#rule_7">7</a></li>
+      <li>on right: <a href="#rule_8">8</a></li>
+    </ul>
+  </li>
+</ul><h2><a name="automaton" id="automaton"></a> Automaton</h2>
+
+<h3><a name="state_0" id="state_0"></a>State 0</h3>
+
+<p class="pre">  <a href="#rule_0">  0</a> <span class="i">$accept</span> → <span class="dot">•</span> <span class="i">unit</span> <b>"end of file"</b>
+  <a href="#rule_1">  1</a> <span class="i">unit</span> → <span class="dot">•</span> <span class="i">assignments</span> <span class="i">exp</span>
+  <a href="#rule_2">  2</a> <span class="i">assignments</span> → <span class="dot">•</span> %empty
+  <a href="#rule_3">  3</a>             | <span class="dot">•</span> <span class="i">assignments</span> <span class="i">assignment</span>
+
+    $default  <a href="#rule_2">reduce using rule 2</a> (assignments)
+
+    unit         <a href="#state_1">go to state 1</a>
+    assignments  <a href="#state_2">go to state 2</a>
+</p>
+
+<h3><a name="state_1" id="state_1"></a>State 1</h3>
+
+<p class="pre">  <a href="#rule_0">  0</a> <span class="i">$accept</span> → <span class="i">unit</span> <span class="dot">•</span> <b>"end of file"</b>
+
+    "end of file"  <a href="#state_3">shift, and go to state 3</a>
+</p>
+
+<h3><a name="state_2" id="state_2"></a>State 2</h3>
+
+<p class="pre">  <a href="#rule_1">  1</a> <span class="i">unit</span> → <span class="i">assignments</span> <span class="dot">•</span> <span class="i">exp</span>
+  <a href="#rule_3">  3</a> <span class="i">assignments</span> → <span class="i">assignments</span> <span class="dot">•</span> <span class="i">assignment</span>
+  <a href="#rule_4">  4</a> <span class="i">assignment</span> → <span class="dot">•</span> <span class="i">id</span> <b>":="</b> <span class="i">exp</span>
+  <a href="#rule_5">  5</a> <span class="i">id</span> → <span class="dot">•</span> <b>"identifier"</b>
+  <a href="#rule_8">  8</a> <span class="i">exp</span> → <span class="dot">•</span> <b>"incr"</b> <span class="i">exp</span> <span class="i">@1</span> <span class="i">@2</span> <span class="i">exp</span>
+  <a href="#rule_9">  9</a>     | <span class="dot">•</span> <b>"("</b> <span class="i">exp</span> <b>")"</b>
+  <a href="#rule_10"> 10</a>     | <span class="dot">•</span> <b>"identifier"</b>
+  <a href="#rule_11"> 11</a>     | <span class="dot">•</span> <b>"number"</b>
+
+    "incr"        <a href="#state_4">shift, and go to state 4</a>
+    "identifier"  <a href="#state_5">shift, and go to state 5</a>
+    "number"      <a href="#state_6">shift, and go to state 6</a>
+    "("           <a href="#state_7">shift, and go to state 7</a>
+
+    assignment  <a href="#state_8">go to state 8</a>
+    id          <a href="#state_9">go to state 9</a>
+    exp         <a href="#state_10">go to state 10</a>
+</p>
+
+<h3><a name="state_3" id="state_3"></a>State 3</h3>
+
+<p class="pre">  <a href="#rule_0">  0</a> <span class="i">$accept</span> → <span class="i">unit</span> <b>"end of file"</b> <span class="dot">•</span>
+
+    $default  accept
+</p>
+
+<h3><a name="state_4" id="state_4"></a>State 4</h3>
+
+<p class="pre">  <a href="#rule_8">  8</a> <span class="i">exp</span> → <span class="dot">•</span> <b>"incr"</b> <span class="i">exp</span> <span class="i">@1</span> <span class="i">@2</span> <span class="i">exp</span>
+  <a href="#rule_8">  8</a>     | <b>"incr"</b> <span class="dot">•</span> <span class="i">exp</span> <span class="i">@1</span> <span class="i">@2</span> <span class="i">exp</span>
+  <a href="#rule_9">  9</a>     | <span class="dot">•</span> <b>"("</b> <span class="i">exp</span> <b>")"</b>
+  <a href="#rule_10"> 10</a>     | <span class="dot">•</span> <b>"identifier"</b>
+  <a href="#rule_11"> 11</a>     | <span class="dot">•</span> <b>"number"</b>
+
+    "incr"        <a href="#state_4">shift, and go to state 4</a>
+    "identifier"  <a href="#state_11">shift, and go to state 11</a>
+    "number"      <a href="#state_6">shift, and go to state 6</a>
+    "("           <a href="#state_7">shift, and go to state 7</a>
+
+    exp  <a href="#state_12">go to state 12</a>
+</p>
+
+<h3><a name="state_5" id="state_5"></a>State 5</h3>
+
+<p class="pre">  <a href="#rule_5">  5</a> <span class="i">id</span> → <b>"identifier"</b> <span class="dot">•</span>  [":="]
+  <a href="#rule_10"> 10</a> <span class="i">exp</span> → <b>"identifier"</b> <span class="dot">•</span>  ["end of file"]
+
+    "end of file"  <a href="#rule_10">reduce using rule 10</a> (exp)
+    $default       <a href="#rule_5">reduce using rule 5</a> (id)
+</p>
+
+<h3><a name="state_6" id="state_6"></a>State 6</h3>
+
+<p class="pre">  <a href="#rule_11"> 11</a> <span class="i">exp</span> → <b>"number"</b> <span class="dot">•</span>
+
+    $default  <a href="#rule_11">reduce using rule 11</a> (exp)
+</p>
+
+<h3><a name="state_7" id="state_7"></a>State 7</h3>
+
+<p class="pre">  <a href="#rule_8">  8</a> <span class="i">exp</span> → <span class="dot">•</span> <b>"incr"</b> <span class="i">exp</span> <span class="i">@1</span> <span class="i">@2</span> <span class="i">exp</span>
+  <a href="#rule_9">  9</a>     | <span class="dot">•</span> <b>"("</b> <span class="i">exp</span> <b>")"</b>
+  <a href="#rule_9">  9</a>     | <b>"("</b> <span class="dot">•</span> <span class="i">exp</span> <b>")"</b>
+  <a href="#rule_10"> 10</a>     | <span class="dot">•</span> <b>"identifier"</b>
+  <a href="#rule_11"> 11</a>     | <span class="dot">•</span> <b>"number"</b>
+
+    "incr"        <a href="#state_4">shift, and go to state 4</a>
+    "identifier"  <a href="#state_11">shift, and go to state 11</a>
+    "number"      <a href="#state_6">shift, and go to state 6</a>
+    "("           <a href="#state_7">shift, and go to state 7</a>
+
+    exp  <a href="#state_13">go to state 13</a>
+</p>
+
+<h3><a name="state_8" id="state_8"></a>State 8</h3>
+
+<p class="pre">  <a href="#rule_3">  3</a> <span class="i">assignments</span> → <span class="i">assignments</span> <span class="i">assignment</span> <span class="dot">•</span>
+
+    $default  <a href="#rule_3">reduce using rule 3</a> (assignments)
+</p>
+
+<h3><a name="state_9" id="state_9"></a>State 9</h3>
+
+<p class="pre">  <a href="#rule_4">  4</a> <span class="i">assignment</span> → <span class="i">id</span> <span class="dot">•</span> <b>":="</b> <span class="i">exp</span>
+
+    ":="  <a href="#state_14">shift, and go to state 14</a>
+</p>
+
+<h3><a name="state_10" id="state_10"></a>State 10</h3>
+
+<p class="pre">  <a href="#rule_1">  1</a> <span class="i">unit</span> → <span class="i">assignments</span> <span class="i">exp</span> <span class="dot">•</span>
+
+    $default  <a href="#rule_1">reduce using rule 1</a> (unit)
+</p>
+
+<h3><a name="state_11" id="state_11"></a>State 11</h3>
+
+<p class="pre">  <a href="#rule_10"> 10</a> <span class="i">exp</span> → <b>"identifier"</b> <span class="dot">•</span>
+
+    $default  <a href="#rule_10">reduce using rule 10</a> (exp)
+</p>
+
+<h3><a name="state_12" id="state_12"></a>State 12</h3>
+
+<p class="pre">  <a href="#rule_6">  6</a> <span class="i">@1</span> → <span class="dot">•</span> %empty
+  <a href="#rule_8">  8</a> <span class="i">exp</span> → <b>"incr"</b> <span class="i">exp</span> <span class="dot">•</span> <span class="i">@1</span> <span class="i">@2</span> <span class="i">exp</span>
+
+    $default  <a href="#rule_6">reduce using rule 6</a> (@1)
+
+    @1  <a href="#state_15">go to state 15</a>
+</p>
+
+<h3><a name="state_13" id="state_13"></a>State 13</h3>
+
+<p class="pre">  <a href="#rule_9">  9</a> <span class="i">exp</span> → <b>"("</b> <span class="i">exp</span> <span class="dot">•</span> <b>")"</b>
+
+    ")"  <a href="#state_16">shift, and go to state 16</a>
+</p>
+
+<h3><a name="state_14" id="state_14"></a>State 14</h3>
+
+<p class="pre">  <a href="#rule_4">  4</a> <span class="i">assignment</span> → <span class="i">id</span> <b>":="</b> <span class="dot">•</span> <span class="i">exp</span>
+  <a href="#rule_8">  8</a> <span class="i">exp</span> → <span class="dot">•</span> <b>"incr"</b> <span class="i">exp</span> <span class="i">@1</span> <span class="i">@2</span> <span class="i">exp</span>
+  <a href="#rule_9">  9</a>     | <span class="dot">•</span> <b>"("</b> <span class="i">exp</span> <b>")"</b>
+  <a href="#rule_10"> 10</a>     | <span class="dot">•</span> <b>"identifier"</b>
+  <a href="#rule_11"> 11</a>     | <span class="dot">•</span> <b>"number"</b>
+
+    "incr"        <a href="#state_4">shift, and go to state 4</a>
+    "identifier"  <a href="#state_11">shift, and go to state 11</a>
+    "number"      <a href="#state_6">shift, and go to state 6</a>
+    "("           <a href="#state_7">shift, and go to state 7</a>
+
+    exp  <a href="#state_17">go to state 17</a>
+</p>
+
+<h3><a name="state_15" id="state_15"></a>State 15</h3>
+
+<p class="pre">  <a href="#rule_7">  7</a> <span class="i">@2</span> → <span class="dot">•</span> %empty
+  <a href="#rule_8">  8</a> <span class="i">exp</span> → <b>"incr"</b> <span class="i">exp</span> <span class="i">@1</span> <span class="dot">•</span> <span class="i">@2</span> <span class="i">exp</span>
+
+    $default  <a href="#rule_7">reduce using rule 7</a> (@2)
+
+    @2  <a href="#state_18">go to state 18</a>
+</p>
+
+<h3><a name="state_16" id="state_16"></a>State 16</h3>
+
+<p class="pre">  <a href="#rule_9">  9</a> <span class="i">exp</span> → <b>"("</b> <span class="i">exp</span> <b>")"</b> <span class="dot">•</span>
+
+    $default  <a href="#rule_9">reduce using rule 9</a> (exp)
+</p>
+
+<h3><a name="state_17" id="state_17"></a>State 17</h3>
+
+<p class="pre">  <a href="#rule_4">  4</a> <span class="i">assignment</span> → <span class="i">id</span> <b>":="</b> <span class="i">exp</span> <span class="dot">•</span>
+
+    $default  <a href="#rule_4">reduce using rule 4</a> (assignment)
+</p>
+
+<h3><a name="state_18" id="state_18"></a>State 18</h3>
+
+<p class="pre">  <a href="#rule_8">  8</a> <span class="i">exp</span> → <span class="dot">•</span> <b>"incr"</b> <span class="i">exp</span> <span class="i">@1</span> <span class="i">@2</span> <span class="i">exp</span>
+  <a href="#rule_8">  8</a>     | <b>"incr"</b> <span class="i">exp</span> <span class="i">@1</span> <span class="i">@2</span> <span class="dot">•</span> <span class="i">exp</span>
+  <a href="#rule_9">  9</a>     | <span class="dot">•</span> <b>"("</b> <span class="i">exp</span> <b>")"</b>
+  <a href="#rule_10"> 10</a>     | <span class="dot">•</span> <b>"identifier"</b>
+  <a href="#rule_11"> 11</a>     | <span class="dot">•</span> <b>"number"</b>
+
+    "incr"        <a href="#state_4">shift, and go to state 4</a>
+    "identifier"  <a href="#state_11">shift, and go to state 11</a>
+    "number"      <a href="#state_6">shift, and go to state 6</a>
+    "("           <a href="#state_7">shift, and go to state 7</a>
+
+    exp  <a href="#state_19">go to state 19</a>
+</p>
+
+<h3><a name="state_19" id="state_19"></a>State 19</h3>
+
+<p class="pre">  <a href="#rule_8">  8</a> <span class="i">exp</span> → <b>"incr"</b> <span class="i">exp</span> <span class="i">@1</span> <span class="i">@2</span> <span class="i">exp</span> <span class="dot">•</span>
+
+    $default  <a href="#rule_8">reduce using rule 8</a> (exp)
+</p>
+
+<div id="footer"><hr />This document was generated using
+      <a href="http://www.gnu.org/software/bison/" title="GNU Bison">
+      GNU Bison VERSION</a>
+      XML Automaton Report.<br />
+      Verbatim copying and distribution of this entire page is
+      permitted in any medium, provided this notice is preserved.</div></body>
+</html>
+]])
+fi

 AT_BISON_OPTION_POPDEFS
 AT_CLEANUP
@@ -1172,39 +1539,74 @@ AT_CHECK([LC_ALL="$locale" bison -fno-caret -o input.cc -rall -Wcex --graph=inpu
 input.y: warning: 3 reduce/reduce conflicts [-Wconflicts-rr]
 Shift/reduce conflict on token "⊕":
  Example              exp "+" exp • "⊕" exp
-  First derivation     exp ::=[ exp ::=[ exp "+" exp • ] "⊕" exp ]
-  Second derivation    exp ::=[ exp "+" exp ::=[ exp • "⊕" exp ] ]
+  Shift derivation
+    exp
+    ↳ exp "+" exp
+              ↳ exp • "⊕" exp
+  Reduce derivation
+    exp
+    ↳ exp             "⊕" exp
+      ↳ exp "+" exp •

 Reduce/reduce conflict on tokens $end, "+", "⊕":
  Example              exp "+" exp •
-  First derivation     exp ::=[ exp "+" exp • ]
-  Second derivation    exp ::=[ exp "+" exp • ]
+  First derivation
+    exp
+    ↳ exp "+" exp •
+  Second derivation
+    exp
+    ↳ exp "+" exp •

 Shift/reduce conflict on token "⊕":
  Example              exp "+" exp • "⊕" exp
-  First derivation     exp ::=[ exp ::=[ exp "+" exp • ] "⊕" exp ]
-  Second derivation    exp ::=[ exp "+" exp ::=[ exp • "⊕" exp ] ]
+  Shift derivation
+    exp
+    ↳ exp "+" exp
+              ↳ exp • "⊕" exp
+  Reduce derivation
+    exp
+    ↳ exp             "⊕" exp
+      ↳ exp "+" exp •

 Shift/reduce conflict on token "⊕":
  Example              exp "⊕" exp • "⊕" exp
-  First derivation     exp ::=[ exp ::=[ exp "⊕" exp • ] "⊕" exp ]
-  Second derivation    exp ::=[ exp "⊕" exp ::=[ exp • "⊕" exp ] ]
+  Shift derivation
+    exp
+    ↳ exp "⊕" exp
+              ↳ exp • "⊕" exp
+  Reduce derivation
+    exp
+    ↳ exp             "⊕" exp
+      ↳ exp "⊕" exp •

 Shift/reduce conflict on token "+":
  Example              exp "⊕" exp • "+" exp
-  First derivation     exp ::=[ exp ::=[ exp "⊕" exp • ] "+" exp ]
-  Second derivation    exp ::=[ exp "⊕" exp ::=[ exp • "+" exp ] ]
+  Shift derivation
+    exp
+    ↳ exp "⊕" exp
+              ↳ exp • "+" exp
+  Reduce derivation
+    exp
+    ↳ exp             "+" exp
+      ↳ exp "⊕" exp •

 Shift/reduce conflict on token "+":
  Example              exp "⊕" exp • "+" exp
-  First derivation     exp ::=[ exp ::=[ exp "⊕" exp • ] "+" exp ]
-  Second derivation    exp ::=[ exp "⊕" exp ::=[ exp • "+" exp ] ]
+  Shift derivation
+    exp
+    ↳ exp "⊕" exp
+              ↳ exp • "+" exp
+  Reduce derivation
+    exp
+    ↳ exp             "+" exp
+      ↳ exp "⊕" exp •

 input.y:6.3-13: warning: rule useless in parser due to conflicts [-Wother]
 ]])

 # Check the contents of the report.
-AT_CHECK([cat input.output], [],
+# FIXME: Avoid trailing white spaces.
+AT_CHECK([sed -e 's/ *$//' input.output], [],
 [[Rules useless in parser due to conflicts

    3 exp: exp "+" exp
@@ -1347,22 +1749,38 @@ State 7
        2 exp: exp "+" exp •
        1 exp: exp • "⊕" exp
      Example                  exp "+" exp • "⊕" exp
-      First derivation         exp ::=[ exp ::=[ exp "+" exp • ] "⊕" exp ]
-      Second derivation        exp ::=[ exp "+" exp ::=[ exp • "⊕" exp ] ]
+      Shift derivation
+        exp
+        ↳ exp "+" exp
+                  ↳ exp • "⊕" exp
+      Reduce derivation
+        exp
+        ↳ exp             "⊕" exp
+          ↳ exp "+" exp •

    Reduce/reduce conflict on tokens $end, "+", "⊕":
        2 exp: exp "+" exp •
        3 exp: exp "+" exp •
      Example                  exp "+" exp •
-      First derivation         exp ::=[ exp "+" exp • ]
-      Second derivation        exp ::=[ exp "+" exp • ]
+      First derivation
+        exp
+        ↳ exp "+" exp •
+      Second derivation
+        exp
+        ↳ exp "+" exp •

    Shift/reduce conflict on token "⊕":
        3 exp: exp "+" exp •
        1 exp: exp • "⊕" exp
      Example                  exp "+" exp • "⊕" exp
-      First derivation         exp ::=[ exp ::=[ exp "+" exp • ] "⊕" exp ]
-      Second derivation        exp ::=[ exp "+" exp ::=[ exp • "⊕" exp ] ]
+      Shift derivation
+        exp
+        ↳ exp "+" exp
+                  ↳ exp • "⊕" exp
+      Reduce derivation
+        exp
+        ↳ exp             "⊕" exp
+          ↳ exp "+" exp •



@@ -1384,22 +1802,40 @@ State 8
        1 exp: exp "⊕" exp •
        1 exp: exp • "⊕" exp
      Example                  exp "⊕" exp • "⊕" exp
-      First derivation         exp ::=[ exp ::=[ exp "⊕" exp • ] "⊕" exp ]
-      Second derivation        exp ::=[ exp "⊕" exp ::=[ exp • "⊕" exp ] ]
+      Shift derivation
+        exp
+        ↳ exp "⊕" exp
+                  ↳ exp • "⊕" exp
+      Reduce derivation
+        exp
+        ↳ exp             "⊕" exp
+          ↳ exp "⊕" exp •

    Shift/reduce conflict on token "+":
        1 exp: exp "⊕" exp •
        2 exp: exp • "+" exp
      Example                  exp "⊕" exp • "+" exp
-      First derivation         exp ::=[ exp ::=[ exp "⊕" exp • ] "+" exp ]
-      Second derivation        exp ::=[ exp "⊕" exp ::=[ exp • "+" exp ] ]
+      Shift derivation
+        exp
+        ↳ exp "⊕" exp
+                  ↳ exp • "+" exp
+      Reduce derivation
+        exp
+        ↳ exp             "+" exp
+          ↳ exp "⊕" exp •

    Shift/reduce conflict on token "+":
        1 exp: exp "⊕" exp •
        3 exp: exp • "+" exp
      Example                  exp "⊕" exp • "+" exp
-      First derivation         exp ::=[ exp ::=[ exp "⊕" exp • ] "+" exp ]
-      Second derivation        exp ::=[ exp "⊕" exp ::=[ exp • "+" exp ] ]
+      Shift derivation
+        exp
+        ↳ exp "⊕" exp
+                  ↳ exp • "+" exp
+      Reduce derivation
+        exp
+        ↳ exp             "+" exp
+          ↳ exp "⊕" exp •

 ]])

@@ -1514,16 +1950,16 @@ AT_CHECK([[sed -e 's/bison-xml-report version="[^"]*"/bison-xml-report version="
      </rule>
    </rules>
    <terminals>
-      <terminal symbol-number="0" token-number="0" name="$end" usefulness="useful"/>
-      <terminal symbol-number="1" token-number="256" name="error" usefulness="useful"/>
-      <terminal symbol-number="3" token-number="258" name="&quot;+&quot;" usefulness="useful" prec="1" assoc="left"/>
-      <terminal symbol-number="4" token-number="259" name="&quot;⊕&quot;" usefulness="useful"/>
-      <terminal symbol-number="5" token-number="260" name="&quot;number&quot;" usefulness="useful"/>
-      <terminal symbol-number="6" token-number="261" name="&quot;Ñùṃéℝô&quot;" usefulness="useful"/>
+      <terminal symbol-number="0" token-number="0" name="$end" type="" usefulness="useful"/>
+      <terminal symbol-number="1" token-number="256" name="error" type="" usefulness="useful"/>
+      <terminal symbol-number="3" token-number="258" name="&quot;+&quot;" type="" usefulness="useful" prec="1" assoc="left"/>
+      <terminal symbol-number="4" token-number="259" name="&quot;⊕&quot;" type="" usefulness="useful"/>
+      <terminal symbol-number="5" token-number="260" name="&quot;number&quot;" type="" usefulness="useful"/>
+      <terminal symbol-number="6" token-number="261" name="&quot;Ñùṃéℝô&quot;" type="" usefulness="useful"/>
    </terminals>
    <nonterminals>
-      <nonterminal symbol-number="7" name="$accept" usefulness="useful"/>
-      <nonterminal symbol-number="8" name="exp" usefulness="useful"/>
+      <nonterminal symbol-number="7" name="$accept" type="" usefulness="useful"/>
+      <nonterminal symbol-number="8" name="exp" type="" usefulness="useful"/>
    </nonterminals>
  </grammar>

@@ -1725,6 +2161,243 @@ AT_CHECK([[sed -e 's/bison-xml-report version="[^"]*"/bison-xml-report version="
 </bison-xml-report>
 ]])

+# Check HTML output.
+if test x"$XSLTPROC" != x""; then
+  AT_CHECK([[$XSLTPROC \
+           `]]AT_SET_ENV[[ bison --print-datadir`/xslt/xml2xhtml.xsl \
+           input.xml | \
+           sed -e 's/GNU Bison [0-9][-.0-9a-z]*/GNU Bison VERSION/' \
+           >input.html]])
+
+  AT_CHECK([cat input.html], [],
+[[<?xml version="1.0" encoding="UTF-8"?>
+<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
+<html xmlns="http://www.w3.org/1999/xhtml" xmlns:bison="http://www.gnu.org/software/bison/">
+  <head>
+    <meta http-equiv="Content-Type" content="text/html;charset=UTF-8" />
+    <title>input.y - GNU Bison XML Automaton Report</title>
+    <style type="text/css">
+      body {
+        font-family: "Nimbus Sans L", Arial, sans-serif;
+        font-size: 9pt;
+      }
+      a:link {
+        color: #1f00ff;
+        text-decoration: none;
+      }
+      a:visited {
+        color: #1f00ff;
+        text-decoration: none;
+      }
+      a:hover {
+        color: red;
+      }
+      #menu a {
+        text-decoration: underline;
+      }
+      .i {
+        font-style: italic;
+      }
+      .pre {
+        font-family: monospace;
+        white-space: pre;
+      }
+      ol.decimal {
+        list-style-type: decimal;
+      }
+      ol.lower-alpha {
+        list-style-type: lower-alpha;
+      }
+      .dot {
+        color: #cc0000;
+      }
+      #footer {
+        margin-top: 3.5em;
+        font-size: 7pt;
+      }
+      </style>
+  </head>
+  <body><h1>GNU Bison XML Automaton Report</h1><p>
+    input grammar: <span class="i">input.y</span></p>
+
+<h3>Table of Contents</h3><ul id="menu"><li><a href="#reductions">Reductions</a><ul class="lower-alpha"><li><a href="#nonterminals_useless_in_grammar">Nonterminals useless in grammar</a></li><li><a href="#terminals_unused_in_grammar">Terminals unused in grammar</a></li><li><a href="#rules_useless_in_grammar">Rules useless in grammar</a></li><li><a href="#rules_useless_in_parser">Rules useless in parser due to conflicts</a></li></ul></li><li><a href="#conflicts">Conflicts</a></li><li><a href="#grammar">Grammar</a><ul class="lower-alpha"><li><a href="#grammar">Itemset</a></li><li><a href="#terminals">Terminal symbols</a></li><li><a href="#nonterminals">Nonterminal symbols</a></li></ul></li><li><a href="#automaton">Automaton</a></li></ul><h2><a name="reductions" id="reductions"></a> Reductions</h2><h3><a name="nonterminals_useless_in_grammar" id="nonterminals_useless_in_grammar"></a> Nonterminals useless in grammar</h3>
+
+<h3><a name="terminals_unused_in_grammar" id="terminals_unused_in_grammar"></a> Terminals unused in grammar</h3>
+
+<h3><a name="rules_useless_in_grammar" id="rules_useless_in_grammar"></a> Rules useless in grammar</h3>
+<h2><a name="rules_useless_in_parser" id="rules_useless_in_parser"></a> Rules useless in parser due to conflicts</h2>
+<p class="pre">
+  <a href="#rule_3">  3</a> <span class="i">exp</span> → <span class="i">exp</span> <b>"+"</b> <span class="i">exp</span>
+</p>
+
+<h2><a name="conflicts" id="conflicts"></a> Conflicts</h2>
+
+<p class="pre"><a href="#state_7">State 7</a> conflicts: 1 shift/reduce, 3 reduce/reduce
+<a href="#state_8">State 8</a> conflicts: 2 shift/reduce
+
+
+</p><h2><a name="grammar" id="grammar"></a> Grammar</h2>
+<p class="pre">
+  <a name="rule_0" id="rule_0">  0</a> <span class="i">$accept</span> → <span class="i">exp</span> <b>$end</b>
+
+  <a name="rule_1" id="rule_1">  1</a> <span class="i">exp</span> → <span class="i">exp</span> <b>"⊕"</b> <span class="i">exp</span>
+  <a name="rule_2" id="rule_2">  2</a>     | <span class="i">exp</span> <b>"+"</b> <span class="i">exp</span>
+  <a name="rule_3" id="rule_3">  3</a>     | <span class="i">exp</span> <b>"+"</b> <span class="i">exp</span>
+  <a name="rule_4" id="rule_4">  4</a>     | <b>"number"</b>
+  <a name="rule_5" id="rule_5">  5</a>     | <b>"Ñùṃéℝô"</b>
+</p>
+
+<h3><a name="terminals" id="terminals"></a> Terminals, with rules where they appear</h3>
+
+<ul>
+  <li><b>$end</b> (0) <a href="#rule_0">0</a></li>
+  <li><b>error</b> (256)</li>
+  <li><b>"+"</b> (258) <a href="#rule_2">2</a> <a href="#rule_3">3</a></li>
+  <li><b>"⊕"</b> (259) <a href="#rule_1">1</a></li>
+  <li><b>"number"</b> (260) <a href="#rule_4">4</a></li>
+  <li><b>"Ñùṃéℝô"</b> (261) <a href="#rule_5">5</a></li>
+</ul>
+
+<h3><a name="nonterminals" id="nonterminals"></a> Nonterminals, with rules where they appear</h3>
+
+<ul>
+  <li><b>$accept</b> (7)
+    <ul>
+      <li>on left: <a href="#rule_0">0</a></li>
+    </ul>
+  </li>
+  <li><b>exp</b> (8)
+    <ul>
+      <li>on left: <a href="#rule_1">1</a> <a href="#rule_2">2</a> <a href="#rule_3">3</a> <a href="#rule_4">4</a> <a href="#rule_5">5</a></li>
+      <li>on right: <a href="#rule_0">0</a> <a href="#rule_1">1</a> <a href="#rule_2">2</a> <a href="#rule_3">3</a></li>
+    </ul>
+  </li>
+</ul><h2><a name="automaton" id="automaton"></a> Automaton</h2>
+
+<h3><a name="state_0" id="state_0"></a>State 0</h3>
+
+<p class="pre">  <a href="#rule_0">  0</a> <span class="i">$accept</span> → <span class="dot">•</span> <span class="i">exp</span> <b>$end</b>
+  <a href="#rule_1">  1</a> <span class="i">exp</span> → <span class="dot">•</span> <span class="i">exp</span> <b>"⊕"</b> <span class="i">exp</span>
+  <a href="#rule_2">  2</a>     | <span class="dot">•</span> <span class="i">exp</span> <b>"+"</b> <span class="i">exp</span>
+  <a href="#rule_3">  3</a>     | <span class="dot">•</span> <span class="i">exp</span> <b>"+"</b> <span class="i">exp</span>
+  <a href="#rule_4">  4</a>     | <span class="dot">•</span> <b>"number"</b>
+  <a href="#rule_5">  5</a>     | <span class="dot">•</span> <b>"Ñùṃéℝô"</b>
+
+    "number"  <a href="#state_1">shift, and go to state 1</a>
+    "Ñùṃéℝô"  <a href="#state_2">shift, and go to state 2</a>
+
+    exp  <a href="#state_3">go to state 3</a>
+</p>
+
+<h3><a name="state_1" id="state_1"></a>State 1</h3>
+
+<p class="pre">  <a href="#rule_4">  4</a> <span class="i">exp</span> → <b>"number"</b> <span class="dot">•</span>
+
+    $default  <a href="#rule_4">reduce using rule 4</a> (exp)
+</p>
+
+<h3><a name="state_2" id="state_2"></a>State 2</h3>
+
+<p class="pre">  <a href="#rule_5">  5</a> <span class="i">exp</span> → <b>"Ñùṃéℝô"</b> <span class="dot">•</span>
+
+    $default  <a href="#rule_5">reduce using rule 5</a> (exp)
+</p>
+
+<h3><a name="state_3" id="state_3"></a>State 3</h3>
+
+<p class="pre">  <a href="#rule_0">  0</a> <span class="i">$accept</span> → <span class="i">exp</span> <span class="dot">•</span> <b>$end</b>
+  <a href="#rule_1">  1</a> <span class="i">exp</span> → <span class="i">exp</span> <span class="dot">•</span> <b>"⊕"</b> <span class="i">exp</span>
+  <a href="#rule_2">  2</a>     | <span class="i">exp</span> <span class="dot">•</span> <b>"+"</b> <span class="i">exp</span>
+  <a href="#rule_3">  3</a>     | <span class="i">exp</span> <span class="dot">•</span> <b>"+"</b> <span class="i">exp</span>
+
+    $end  <a href="#state_4">shift, and go to state 4</a>
+    "+"   <a href="#state_5">shift, and go to state 5</a>
+    "⊕"   <a href="#state_6">shift, and go to state 6</a>
+</p>
+
+<h3><a name="state_4" id="state_4"></a>State 4</h3>
+
+<p class="pre">  <a href="#rule_0">  0</a> <span class="i">$accept</span> → <span class="i">exp</span> <b>$end</b> <span class="dot">•</span>
+
+    $default  accept
+</p>
+
+<h3><a name="state_5" id="state_5"></a>State 5</h3>
+
+<p class="pre">  <a href="#rule_1">  1</a> <span class="i">exp</span> → <span class="dot">•</span> <span class="i">exp</span> <b>"⊕"</b> <span class="i">exp</span>
+  <a href="#rule_2">  2</a>     | <span class="dot">•</span> <span class="i">exp</span> <b>"+"</b> <span class="i">exp</span>
+  <a href="#rule_2">  2</a>     | <span class="i">exp</span> <b>"+"</b> <span class="dot">•</span> <span class="i">exp</span>
+  <a href="#rule_3">  3</a>     | <span class="dot">•</span> <span class="i">exp</span> <b>"+"</b> <span class="i">exp</span>
+  <a href="#rule_3">  3</a>     | <span class="i">exp</span> <b>"+"</b> <span class="dot">•</span> <span class="i">exp</span>
+  <a href="#rule_4">  4</a>     | <span class="dot">•</span> <b>"number"</b>
+  <a href="#rule_5">  5</a>     | <span class="dot">•</span> <b>"Ñùṃéℝô"</b>
+
+    "number"  <a href="#state_1">shift, and go to state 1</a>
+    "Ñùṃéℝô"  <a href="#state_2">shift, and go to state 2</a>
+
+    exp  <a href="#state_7">go to state 7</a>
+</p>
+
+<h3><a name="state_6" id="state_6"></a>State 6</h3>
+
+<p class="pre">  <a href="#rule_1">  1</a> <span class="i">exp</span> → <span class="dot">•</span> <span class="i">exp</span> <b>"⊕"</b> <span class="i">exp</span>
+  <a href="#rule_1">  1</a>     | <span class="i">exp</span> <b>"⊕"</b> <span class="dot">•</span> <span class="i">exp</span>
+  <a href="#rule_2">  2</a>     | <span class="dot">•</span> <span class="i">exp</span> <b>"+"</b> <span class="i">exp</span>
+  <a href="#rule_3">  3</a>     | <span class="dot">•</span> <span class="i">exp</span> <b>"+"</b> <span class="i">exp</span>
+  <a href="#rule_4">  4</a>     | <span class="dot">•</span> <b>"number"</b>
+  <a href="#rule_5">  5</a>     | <span class="dot">•</span> <b>"Ñùṃéℝô"</b>
+
+    "number"  <a href="#state_1">shift, and go to state 1</a>
+    "Ñùṃéℝô"  <a href="#state_2">shift, and go to state 2</a>
+
+    exp  <a href="#state_8">go to state 8</a>
+</p>
+
+<h3><a name="state_7" id="state_7"></a>State 7</h3>
+
+<p class="pre">  <a href="#rule_1">  1</a> <span class="i">exp</span> → <span class="i">exp</span> <span class="dot">•</span> <b>"⊕"</b> <span class="i">exp</span>
+  <a href="#rule_2">  2</a>     | <span class="i">exp</span> <span class="dot">•</span> <b>"+"</b> <span class="i">exp</span>
+  <a href="#rule_2">  2</a>     | <span class="i">exp</span> <b>"+"</b> <span class="i">exp</span> <span class="dot">•</span>  [$end, "+", "⊕"]
+  <a href="#rule_3">  3</a>     | <span class="i">exp</span> <span class="dot">•</span> <b>"+"</b> <span class="i">exp</span>
+  <a href="#rule_3">  3</a>     | <span class="i">exp</span> <b>"+"</b> <span class="i">exp</span> <span class="dot">•</span>  [$end, "+", "⊕"]
+
+    "⊕"  <a href="#state_6">shift, and go to state 6</a>
+
+    $end      <a href="#rule_2">reduce using rule 2</a> (exp)
+    $end      [<a href="#rule_3">reduce using rule 3</a> (exp)]
+    "+"       <a href="#rule_2">reduce using rule 2</a> (exp)
+    "+"       [<a href="#rule_3">reduce using rule 3</a> (exp)]
+    "⊕"       [<a href="#rule_2">reduce using rule 2</a> (exp)]
+    "⊕"       [<a href="#rule_3">reduce using rule 3</a> (exp)]
+    $default  <a href="#rule_2">reduce using rule 2</a> (exp)
+
+    Conflict between <a href="#rule_2">rule 2</a> and token "+" resolved as reduce (%left "+").
+</p>
+
+<h3><a name="state_8" id="state_8"></a>State 8</h3>
+
+<p class="pre">  <a href="#rule_1">  1</a> <span class="i">exp</span> → <span class="i">exp</span> <span class="dot">•</span> <b>"⊕"</b> <span class="i">exp</span>
+  <a href="#rule_1">  1</a>     | <span class="i">exp</span> <b>"⊕"</b> <span class="i">exp</span> <span class="dot">•</span>  [$end, "+", "⊕"]
+  <a href="#rule_2">  2</a>     | <span class="i">exp</span> <span class="dot">•</span> <b>"+"</b> <span class="i">exp</span>
+  <a href="#rule_3">  3</a>     | <span class="i">exp</span> <span class="dot">•</span> <b>"+"</b> <span class="i">exp</span>
+
+    "+"  <a href="#state_5">shift, and go to state 5</a>
+    "⊕"  <a href="#state_6">shift, and go to state 6</a>
+
+    "+"       [<a href="#rule_1">reduce using rule 1</a> (exp)]
+    "⊕"       [<a href="#rule_1">reduce using rule 1</a> (exp)]
+    $default  <a href="#rule_1">reduce using rule 1</a> (exp)
+</p>
+
+<div id="footer"><hr />This document was generated using
+      <a href="http://www.gnu.org/software/bison/" title="GNU Bison">
+      GNU Bison VERSION</a>
+      XML Automaton Report.<br />
+      Verbatim copying and distribution of this entire page is
+      permitted in any medium, provided this notice is preserved.</div></body>
+</html>
+]])
+fi

 AT_BISON_OPTION_POPDEFS
 AT_CLEANUP
@@ -283,11 +283,11 @@ m4_popdef([AT_TEST])
 ## Many lookahead tokens.   ##
 ## ------------------------ ##

-# AT_DATA_LOOKAHEAD_TOKENS_GRAMMAR(FILE-NAME, SIZE)
-# --------------------------------------------------
+# AT_DATA_LOOKAHEADS_GRAMMAR(FILE-NAME, SIZE)
+# -------------------------------------------
 # Create FILE-NAME, containing a self checking parser for a grammar
 # requiring SIZE lookahead tokens.
-m4_define([AT_DATA_LOOKAHEAD_TOKENS_GRAMMAR],
+m4_define([AT_DATA_LOOKAHEADS_GRAMMAR],
 [AT_BISON_OPTION_PUSHDEFS
 AT_DATA([[gengram.pl]],
 [[#! /usr/bin/perl -w
@@ -384,7 +384,7 @@ AT_BISON_OPTION_POPDEFS

 AT_SETUP([Many lookahead tokens])

-AT_DATA_LOOKAHEAD_TOKENS_GRAMMAR([input.y], [1000])
+AT_DATA_LOOKAHEADS_GRAMMAR([input.y], [1000])

 # GNU m4 requires about 70 MiB for this test on a 32-bit host.
 # Ask for 200 MiB, which should be plenty even on a 64-bit host.
Author	SHA1	Message	Date
Akim Demaille	cf890692f1	version 3.6.92 * NEWS: Record release date.	2020-07-19 09:24:57 +02:00
Akim Demaille	6932023f4d	style: avoid strncpy syntax-check seems to dislike strncpy. The GNU Coreutils replaced their uses of strncpy with stpncpy. strlcpy is not an option. http://sources.redhat.com/ml/libc-alpha/2002-01/msg00159.html http://sources.redhat.com/ml/libc-alpha/2002-01/msg00011.html http://lists.gnu.org/archive/html/bug-gnulib/2004-09/msg00181.html * src/glyphs.c: Use stpncpy.	2020-07-19 09:23:42 +02:00
Akim Demaille	fff17fe8fe	cex: display derivations as trees Sometimes, understanding the derivations is difficult, because they are serialized to fit in one line. For instance, the example taken from the NEWS file: %token ID %% s: a ID a: expr expr: expr ID ',' \| "expr" gave First example expr • ID ',' ID $end Shift derivation $accept → [ s → [ a → [ expr → [ expr • ID ',' ] ] ID ] $end ] Second example expr • ID $end Reduce derivation $accept → [ s → [ a → [ expr • ] ID ] $end ] Printing as trees, it gives: First example expr • ID ',' ID $end Shift derivation $accept ↳ s $end ↳ a ID ↳ expr ↳ expr • ID ',' Second example expr • ID $end Reduce derivation $accept ↳ s $end ↳ a ID ↳ expr • * src/glyphs.h, src/glyphs.c (down_arrow, empty, derivation_separator): New. * src/derivation.c (derivation_print, derivation_print_impl): Rename as... (derivation_print_flat, derivation_print_flat_impl): These. (fputs_if, derivation_depth, derivation_width, derivation_print_tree) (derivation_print_tree_impl, derivation_print): New. * src/counterexample.c (print_counterexample): Adjust. * tests/conflicts.at, tests/counterexample.at, tests/diagnostics.at, * tests/report.at: Adjust.	2020-07-18 07:54:02 +02:00
Akim Demaille	5544615a59	cex: use the glyphs * src/derivation.c: here. * src/gram.h, src/gram.c (print_arrow, print_dot, print_fallback): Remove.	2020-07-16 07:31:25 +02:00
Akim Demaille	346ba14f15	cex: factor the handling of graphical symbols * src/glyphs.h, src/glyphs.c: New.	2020-07-16 07:31:24 +02:00
Akim Demaille	4d18195ebc	cex: style changes * src/counterexample.c: here.	2020-07-15 06:41:07 +02:00
Akim Demaille	dd3e7b3895	cex: simplify tests * tests/counterexample.at (AT_BISON_CHECK_CEX): Handle the keyword. Simplify the signature.	2020-07-15 06:38:36 +02:00
Akim Demaille	64a3b6546a	cex: more colors Provided by Daniela Becker. * data/bison-default.css: More colors.	2020-07-15 06:38:36 +02:00
Akim Demaille	bad07a7f66	style: comments changes * src/print.c: here.	2020-07-14 14:26:02 +02:00
Akim Demaille	88bd814bf1	doc: update GLR sections Reported by Christian Schoenebeck. * doc/bison.texi (GLR Parsers): Minor fixes. (Compiler Requirements for GLR): Remove, quite useless today.	2020-07-14 06:56:15 +02:00
Akim Demaille	4f9ae5de07	cex: display shifts before reductions When reporting counterexamples for s/r conflicts, put the shift first. This is more natural, and displays the default resolution first, which is also what happens for r/r conflicts where the smallest rule number is displayed first, and "wins". * src/counterexample.c (counterexample): Add a shift_reduce member. (new_counterexample): Adjust. Swap the derivations when this is a s/r conflict. (print_counterexample): For s/r conflicts, prefer "Shift derivation" and "Reduce derivation" rather than "First/Second derivation". * tests/conflicts.at, tests/counterexample.at, tests/report.at: Adjust. * NEWS, doc/bison.texi: Ditto.	2020-07-14 06:48:48 +02:00
Akim Demaille	78f72a4516	style: s/lookahead_tokens/lookaheads/g Currently we use both names. Let's stick to the short one. * src/AnnotationList.c, src/conflicts.c, src/counterexample.c, * src/getargs.c, src/getargs.h, src/graphviz.c, src/ielr.c, * src/lalr.c, src/print-graph.c, src/print-xml.c, src/print.c, * src/state-item.c, src/state.c, src/state.h, src/tables.c: s/lookahead_token/lookahead/gi.	2020-07-14 06:48:48 +02:00
Akim Demaille	c04693d651	cex: factor memory allocation * src/counterexample.c (counterexample_report_state): Allocate once per conflicted state, instead of once per r/r conflict.	2020-07-14 06:48:48 +02:00
Akim Demaille	12191911ba	cex: use state_item_number consistently * src/counterexample.c, src/state-item.c: here. (counterexample_report_state): While at it, prefer c2 to j/k, to match c1.	2020-07-14 06:48:48 +02:00
Akim Demaille	d7f27477f4	cex: more consistent memory allocation/copy * src/counterexample.c, src/parse-simulation.c: It is more usual in Bison to use sizeof on expressions than on types, especially for allocation. Let the compiler do it's job instead of calling memcpy ourselves.	2020-07-14 06:48:48 +02:00
Akim Demaille	5bad15d7ea	cex: minor renaming * src/counterexample.c (has_common_prefix): Rename as... (have_common_prefix): this.	2020-07-14 06:48:48 +02:00
Akim Demaille	cd099edf2d	cex: use better type names There are too many gl_list_t in there, it's hard to understand what is going on. Introduce and use more precise types. I sure can be wrong in some places, it's hard to tell without proper tool support. * src/counterexample.c, src/lssi.c, src/lssi.h, src/parse-simulation.c, * src/parse-simulation.h, src/state-item.c, src/state-item.h (si_bfs_node_list, search_state_list, ssb_list, lssi_list) (state_item_list): New.	2020-07-14 06:48:48 +02:00
Akim Demaille	1e12219775	cex: minor style changes * src/counterexample.h, src/derivation.h, src/derivation.c: More comments. Use `out` for FILE*, as elsewhere.	2020-07-14 06:48:48 +02:00
Akim Demaille	d8c2af56c1	tests: beware of version numbers from git describe * tests/report.at: Be robust to version numbers such as 3.6.4.133-fbac-dirty.	2020-07-14 06:48:48 +02:00
Akim Demaille	cc11bb037c	tests: fix expectations Broken in `ee86ea8839`. * tests/diagnostics.at: here.	2020-07-14 06:48:48 +02:00
Akim Demaille	121dd98508	doc: makeinfo wants @arrow{}, not @arrow * doc/bison.texi: here.	2020-07-12 08:15:44 +02:00
Akim Demaille	2eddbd0ac4	gnulib: update	2020-07-11 19:11:19 +02:00
Akim Demaille	ee86ea8839	cex: prefer → to ::= It does not make a lot of sense to use ::= in our counterexamples, that's not something that belongs to the Bison "vocabulary". Using the colon makes sense, but it's too discreet. Let's use the arrow, which we already use in some reports (HTML and Dot). * src/gram.h (print_dot_fallback): Generalize into... (print_fallback): this. (print_arrow): New. * src/derivation.c: Use it. * NEWS, tests/conflicts.at, tests/counterexample.at, * tests/diagnostics.at, tests/report.at: Adjust. * doc/bison.texi: Ditto. Unfortunately the literal `→` is output as `↦`. So we need to use @arrow.	2020-07-11 18:43:46 +02:00
Akim Demaille	a2ad33dca6	style: cex: prefer the array notation Prefer `&foos[i]` to `foos + i` when `foos` is an array. IMHO, it makes the semantics clearer. * src/counterexample.c, src/lssi.c, src/parse-simulation.c, * src/state-item.c: With arrays, prefer the array notation rather than the pointer one.	2020-07-11 18:07:09 +02:00
Akim Demaille	5b2b7b1ffb	style: cex: remove variables that don't make it simpler to read * src/counterexample.c: With arrays, prefer the array notation rather than the pointer one.	2020-07-11 18:07:09 +02:00
Akim Demaille	dc72b3566d	bistromathic: demonstrate caret-diagnostics * examples/c/bistromathic/parse.y (user_context): We need the current line. (yyreport_syntax_error): Quote the guilty line, with squiggles. * examples/c/bistromathic/bistromathic.test: Adjust.	2020-07-11 18:06:45 +02:00
Akim Demaille	c47e1174d4	bistromathic: do not display parse errors on completion Currently autocompletion on a line with errors leaks the error messages. It can be useful to let the user know, but GNU Readline does not provide us with an nice way to display the error. So we actually break into the current line of the user. So instead, do not show these errors. * examples/c/bistromathic/parse.y (user_context): New. Use %param to pass it to the parser and scanner. Keep quiet when in computing autocompletion.	2020-07-11 18:05:50 +02:00
Akim Demaille	093eeb27e9	bistromathic: don't stupidly reset the location for each token That quite defeats the whole point of locations... But anyway, we should not see these messages at all. * examples/c/bistromathic/parse.y (expected_tokens): Fix (useless) location tracking.	2020-07-11 18:05:41 +02:00
Akim Demaille	dab23c4a21	bistromathic: promote yytoken_kind_t * examples/c/bistromathic/parse.y: Use yytoken_kind_t rather than int.	2020-07-11 18:05:35 +02:00
Akim Demaille	38a169bec1	html: capitalize titles * data/xslt/xml2xhtml.xsl: Use "State 0", not "state 0". As we do in text reports.	2020-07-11 12:58:45 +02:00
Akim Demaille	dc77d6500f	html: don't define several times the same anchors Currently when we output useless rules, they appear before the grammar, but using the same invocation. As a result, the anchor is defined twice, and the wrong one, being first, is honored. * data/xslt/xml2xhtml.xsl (rule): Take a new 'anchor' parameter to decide whether being an anchor, or a target. Let it be true when output the grammar. * tests/report.at: Adjust.	2020-07-11 12:58:44 +02:00
Akim Demaille	8262c7dc22	html: simplify * data/xslt/xml2xhtml.xsl: Merge two identical when-clauses.	2020-07-11 12:58:44 +02:00
Akim Demaille	44f28d10ee	reports: let html reports catch up with --report and --graph * data/xslt/xml2xhtml.xsl: Show the symbol types. * tests/report.at: Adjust.	2020-07-11 12:58:44 +02:00
Akim Demaille	44ad466a32	reports: let xml reports catch up with --report and --graph The text and Dot reports are expected to be identical when generated directly (--report, --graph) or indirectly (via XML). The xml testsuite had not be run for ages, let it catch up a bit. * src/print-xml.c: Pass the type of the symbols. * data/xslt/xml2text.xsl Catch up with the new layout. Display the symbol types. Use '•', not '.' * tests/local.at: Smash '•' to '.' when matching against the direct text report. * tests/report.at: Adjust XML expectations.	2020-07-11 12:58:44 +02:00
Akim Demaille	a839f4c461	reports: update html ouput * data/xslt/xml2xhtml.xsl: Improve indentation. Use ul/li rather that pre.	2020-07-11 12:58:44 +02:00
Akim Demaille	9a51c6a128	tests: check html * tests/report.at: here.	2020-07-11 12:58:44 +02:00
Akim Demaille	2608b0cf12	style: factor complex expressions * src/print-xml.c, src/print.c: Introduce a variable pointing to the current symbol.	2020-07-11 12:58:44 +02:00
Akim Demaille	aa766d1560	maint: make it easier to update expectations * tests/local.mk (update-tests): New.	2020-07-11 12:58:44 +02:00
Akim Demaille	a7ed13b25f	maint: post-release administrivia * NEWS: Add header line for next release. * .prev-version: Record previous version. * cfg.mk (old_NEWS_hash): Auto-update.	2020-07-09 21:38:23 +02:00
@@ -1 +1 @@
 .6.90
 .6.91