[commit/Ada 2/3] processId: Do not modify already encoded IDs

[Joel Brobecker <brobecker@adacore.com>]

The processID function is supposed to take a symbol name, and process it
in a way that allows us to look that symbol up.  This patch is adding
a guard to make sure that we do not apply any transformation if we detect
that we are given an already-encoded symbol name.  For instance:

    gv___XR_pck__global_variable___XE

This happens in the case where we are trying to print the value of
a renaming. To do this, we simply parse and evaluate the XR symbol
name as an expression. Without this change, the expression parser
transforms gv___XR_pck__global_variable___XE into somethink like
gv___xr_pck__global_variable___xe, which then screws up the rest
of the renaming evaluation.

gdb/ChangeLog:

        * ada-lex.p (processId): Do not modify already encoded IDs.
        Update function documentation.

Tested on x86_64-linux, checked in.

---
 gdb/ChangeLog |    5 +++++
 gdb/ada-lex.l |   15 +++++++++++++--
 2 files changed, 18 insertions(+), 2 deletions(-)

diff --git a/gdb/ChangeLog b/gdb/ChangeLog
index c6a6d8f..621da5b 100644
--- a/gdb/ChangeLog
+++ b/gdb/ChangeLog
@@ -1,5 +1,10 @@
 2012-02-29  Joel Brobecker  <brobecker@adacore.com>
 
+	* ada-lex.p (processId): Do not modify already encoded IDs.
+	Update function documentation.
+
+2012-02-29  Joel Brobecker  <brobecker@adacore.com>
+
 	* ada-lang.h (ada_find_renaming_symbol): Replace parameter
 	"name" with "struct symbol *name_sym".
 	* ada-exp.y (write_var_or_type): Update call to
diff --git a/gdb/ada-lex.l b/gdb/ada-lex.l
index 48667d0..5102ff4 100644
--- a/gdb/ada-lex.l
+++ b/gdb/ada-lex.l
@@ -410,7 +410,9 @@ processReal (const char *num0)
 
 
 /* Store a canonicalized version of NAME0[0..LEN-1] in yylval.ssym.  The
-   resulting string is valid until the next call to ada_parse.  It differs
+   resulting string is valid until the next call to ada_parse.  If
+   NAME0 contains the substring "___", it is assumed to be already
+   encoded and the resulting name is equal to it.  Otherwise, it differs
    from NAME0 in that:
     + Characters between '...' or <...> are transfered verbatim to 
       yylval.ssym.
@@ -430,8 +432,18 @@ processId (const char *name0, int len)
   int i0, i;
   struct stoken result;
 
+  result.ptr = name;
   while (len > 0 && isspace (name0[len-1]))
     len -= 1;
+
+  if (strstr (name0, "___") != NULL)
+    {
+      strncpy (name, name0, len);
+      name[len] = '\000';
+      result.length = len;
+      return result;
+    }
+
   i = i0 = 0;
   while (i0 < len)
     {
@@ -471,7 +483,6 @@ processId (const char *name0, int len)
     }
   name[i] = '\000';
 
-  result.ptr = name;
   result.length = i;
   return result;
 }
-- 
1.7.1