Re: extra lookahead sets in report

[Joel E. Denny]

On Thu, 18 Oct 2007, Joel E. Denny wrote:

> On Thu, 18 Oct 2007, Akim Demaille wrote:
> 
> > You don't need to, it's obvious.  But what's not obvious is
> > what exactly is incorrect.  Your comments seem to imply that
> > Bison is correct in its implementation of the "incorrect"
> > algorithm it uses, therefore, albeit incorrect, the result
> > is faithful.  So I would just keep it, unless it's only
> > misleading.
> 
> Is the following any clearer?

I committed the following, which I hope is even clearer.  Let me know if 
it still needs improvement.

Index: ChangeLog
===================================================================
RCS file: /sources/bison/bison/ChangeLog,v
retrieving revision 1.1739
diff -p -u -r1.1739 ChangeLog
--- ChangeLog   17 Oct 2007 18:13:53 -0000      1.1739
+++ ChangeLog   19 Oct 2007 02:42:25 -0000
@@ -1,3 +1,11 @@
+2007-10-19  Joel E. Denny  <address@hidden>
+
+       * NEWS: Add entry for recent .output file lookahead set fix.
+       * doc/bison.texinfo (Understanding): Remove a bogus lookahead set in
+       the example .output text.
+       * tests/regression.at (Extra lookahead sets in report): Improve wording
+       of comments.
+
 2007-10-17  Wojciech Polak  <address@hidden>
 
        * src/print-xml.c (print_grammar): Renamed
Index: NEWS
===================================================================
RCS file: /sources/bison/bison/NEWS,v
retrieving revision 1.176
diff -p -u -r1.176 NEWS
--- NEWS        15 Aug 2007 20:21:19 -0000      1.176
+++ NEWS        19 Oct 2007 02:42:26 -0000
@@ -15,6 +15,15 @@ Changes in version 2.3a+ (????-??-??):
     3. For any rule used only in such states, Bison now reports the rule as
        "never reduced because of conflicts".
 
+* When instructed to generate a `.output' file including lookahead sets
+  (using `--report=lookahead', for example), Bison now prints each reduction's
+  lookahead set only next to the associated state's one item that (1) is
+  associated with the same rule as the reduction and (2) has its dot at the end
+  of its RHS.  Previously, Bison also erroneously printed the lookahead set
+  next to all of the state's other items associated with the same rule.  This
+  bug affected only the `.output' file and not the generated parser source
+  code.
+
 * The -g and --graph options now output graphs in Graphviz DOT format,
   not VCG format.
 
Index: doc/bison.texinfo
===================================================================
RCS file: /sources/bison/bison/doc/bison.texinfo,v
retrieving revision 1.239
diff -p -u -r1.239 bison.texinfo
--- doc/bison.texinfo   17 Oct 2007 04:35:34 -0000      1.239
+++ doc/bison.texinfo   19 Oct 2007 02:42:29 -0000
@@ -7387,7 +7387,7 @@ with some set of possible lookahead toke
 @example
 state 8
 
-    exp  ->  exp . '+' exp  [$, '+', '-', '/']   (rule 1)
+    exp  ->  exp . '+' exp   (rule 1)
     exp  ->  exp '+' exp .  [$, '+', '-', '/']   (rule 1)
     exp  ->  exp . '-' exp   (rule 2)
     exp  ->  exp . '*' exp   (rule 3)
Index: tests/regression.at
===================================================================
RCS file: /sources/bison/bison/tests/regression.at,v
retrieving revision 1.115
diff -p -u -r1.115 regression.at
--- tests/regression.at 17 Oct 2007 04:35:35 -0000      1.115
+++ tests/regression.at 19 Oct 2007 02:42:30 -0000
@@ -1149,11 +1149,12 @@ AT_CLEANUP
 
 AT_SETUP([[Extra lookahead sets in report]])
 
-# Within a single state, Bison used to print a reduction's lookahead set next
-# to all items of the associated rule.  Instead, it should only print it next
-# to the item that is of the associated rule and whose dot is at the end of the
-# RHS.  Printing it next to items whose dot is not at the end of the RHS is
-# sometimes redundant and, as in this test case, is sometimes incorrect.
+# Bison prints each reduction's lookahead set only next to the associated
+# state's one item that (1) is associated with the same rule as the reduction
+# and (2) has its dot at the end of its RHS.  Previously, Bison also
+# erroneously printed the lookahead set next to all of the state's other items
+# associated with the same rule.  This bug affected only the `.output' file and
+# not the generated parser source code.
 
 AT_DATA([[input.y]],
 [[%%