From:	byterock@cvs.perl.org	Date:	Fri Oct  3 12:51:50 2008
Subject:	[svn:dbd-oracle] r11920 - dbd-oracle/branches/utf8_ea

Author: byterock
Date: Fri Oct  3 11:51:48 2008
New Revision: 11920

Modified:
   dbd-oracle/branches/utf8_ea/Oracle.pm
   dbd-oracle/branches/utf8_ea/dbdimp.c
   dbd-oracle/branches/utf8_ea/dbdimp.h
   dbd-oracle/branches/utf8_ea/oci8.c

Log:
hold this for now

Modified: dbd-oracle/branches/utf8_ea/Oracle.pm
==============================================================================
--- dbd-oracle/branches/utf8_ea/Oracle.pm	(original)
+++ dbd-oracle/branches/utf8_ea/Oracle.pm	Fri Oct  3 11:51:48 2008
@@ -935,7 +935,8 @@
                  ora_parse_lang		=> undef,
                  ora_placeholders	=> undef,
                  ora_auto_lob		=> undef,
-                 ora_check_sql		=> undef
+                 ora_check_sql		=> undef,
+                 ora_ary_is_utf8        => undef
                 };
     }
    
@@ -2054,6 +2055,49 @@
 
    $dbh->{ora_ph_csform} = SQLCS_NCHAR; # default for all future placeholders
 
+B<Sending Array Data using Execute_array>
+
+The Array data bound to a placeholder, which is to be executed with execute_array,
+by default does not handle Unicode strings containing non-ASCII characters, unless
+the default client character set is AL32UTF8.
+
+If your data contains Perl Unicode strings (UTF-8) then you will have to explictly
+tell DBD::Oracle how to handle this data at the prepare stage or at the bind levle.
+
+Use ora_
+
+
+
+
+
+If your array data If one of the tuble As the nature of Array Data  except for a few special
+cases. These are listed here with the highest precedence first:
+
+If the C<ora_csid> attribute is given to bind_param() then that
+is passed to Oracle and takes precedence.
+
+If the value is a Perl Unicode string (UTF-8) then DBD::Oracle
+ensures that Oracle uses the Unicode character set, regardless of
+the NLS_LANG and NLS_NCHAR settings.
+
+If the placeholder is for inserting an NCLOB then the client NLS_NCHAR
+character set is used. (That's useful but inconsistent with the other behaviour
+so may change. Best to be explicit by using the C<ora_csform>
+attribute.)
+
+If the C<ora_csform> attribute is given to bind_param() then that
+determines if the value should be assumed to be in the default
+(NLS_LANG) or NCHAR (NLS_NCHAR) client character set. 
+
+
+   use DBD::Oracle qw( SQLCS_IMPLICIT SQLCS_NCHAR );
+   ...
+   $sth->bind_param(1, $value, { ora_csform => SQLCS_NCHAR }); 
+
+or
+
+   $dbh->{ora_ph_csform} = SQLCS_NCHAR; # default for all future placeholders
+
 B<Sending Data using SQL>
 
 Oracle assumes the SQL statement is in the default client character

Modified: dbd-oracle/branches/utf8_ea/dbdimp.c
==============================================================================
--- dbd-oracle/branches/utf8_ea/dbdimp.c	(original)
+++ dbd-oracle/branches/utf8_ea/dbdimp.c	Fri Oct  3 11:51:48 2008
@@ -2759,11 +2759,15 @@
 		    if ( (svp=hv_fetch((HV*)SvRV(attribs), "ora_field",9, 0)) != NULL) {
 				phs->ora_field = SvREFCNT_inc(*svp);
 		    }
+			if ( (svp=hv_fetch((HV*)SvRV(attribs), "ora_ary_is_utf8",15, 0)) != NULL) {
+				phs->utf8_array = 1;
+		    }
 		    if ( (svp=hv_fetch((HV*)SvRV(attribs), "ora_csform", 10, 0)) != NULL) {
 				if (SvIV(*svp) == SQLCS_IMPLICIT || SvIV(*svp) == SQLCS_NCHAR)
 				    phs->csform = (ub1)SvIV(*svp);
 				else warn("ora_csform must be 1 (SQLCS_IMPLICIT) or 2 (SQLCS_NCHAR), not %d", SvIV(*svp));
 		    }
+
 		    if ( (svp=hv_fetch((HV*)SvRV(attribs), "ora_maxdata_size", 16, 0)) != NULL) {
 				phs->maxdata_size = SvUV(*svp);
     		}
@@ -3131,8 +3135,8 @@
 
     OCIAttrGet_log_stat(phs->bndhp, OCI_HTYPE_BIND, &phs->csid_orig, (ub4)0 ,
 		OCI_ATTR_CHARSET_ID, imp_sth->errhp, status);
- 
-PerlIO_printf(DBILOGFP, "OCI_ATTR_CHARSET_ID=%d\n\n",phs->csid_orig); 
+
+PerlIO_printf(DBILOGFP, "OCI_ATTR_CHARSET_ID=%d\n\n",phs->csid_orig);
     OCIBindDynamic_log(phs->bndhp, imp_sth->errhp,
                        (dvoid *)phs, dbd_phs_in,
                        (dvoid *)phs, dbd_phs_out, status);
@@ -3191,45 +3195,15 @@
    PerlIO_printf(DBILOGFP, " ARRAY_BIND_UTF8=%d CS_IS_NOT_UTF8=%d \n",ARRAY_BIND_UTF8,CS_IS_NOT_UTF8(csid));
 
  PerlIO_printf(DBILOGFP, " ARRAY_BIND_UTF8=%d CS_IS_NOT_UTF8=%d \n",ARRAY_BIND_UTF8,CS_IS_UTF8(csid));
-	if ((utf8 & ARRAY_BIND_UTF8) && CS_IS_NOT_UTF8(csid)) {
-		/* if the specified or default csid is not utf8 _compatible_ AND we have */
-		/* mixed utf8 and native (non-utf8) data, then it's a fatal problem */
-	    /* utf8 _compatible_ means, can be upgraded to utf8, ie. utf8 or ascii */
-
-   PerlIO_printf(DBILOGFP, " phs->csid=%d  phs->csform=%d \n", phs->csid , phs->csform);
-   PerlIO_printf(DBILOGFP, " csid=%d \n", csid);
-
-          csid = (utf8 && !CS_IS_UTF8(csid)) ? utf8_csid : CSFORM_IMPLIED_CSID(csform);
-          phs->csid = csid;
-            phs->csform = csform;  
-
-
-   PerlIO_printf(DBILOGFP, " out phs->csid=%d  phs->csform=%d \n", phs->csid , phs->csform);
- PerlIO_printf(DBILOGFP, " out csid=%d \n", csid);
-
-
-
-/*#	    if ((utf8 & ARRAY_BIND_NATIVE) && !CS_IS_UTF8_COMPATIBLE(csid)) {*/
-  PerlIO_printf(DBILOGFP, "Can't mix utf8 and non-utf8 in array bind\n");
-	/*	croak("Can't mix utf8 and non-utf8 in array bind");*/
-	/*    } else {
-		csid = utf8_csid;  not al32utf8_csid here on purpose 
-            }*/
-	}
-
-
-#ifdef OCI_ATTR_CHARSET_ID
-
-  PerlIO_printf(DBILOGFP, "OCI_ATTR_CHARSET_ID is set \n");
-
-
-#else
-	        /* Effectively only used so AL32UTF8 works properly */
-     PerlIO_printf(DBILOGFP, "OCI_ATTR_CHARSET_ID is Not set \n");
-
-
-#endif /* OCI_ATTR_CHARSET_ID */
-		/* if data is utf8 but charset isn't then switch to utf8 csid */
+	 if ((utf8 & ARRAY_BIND_UTF8) && !CS_IS_UTF8(csid)) {
+	      /* if the specified or default csid is not utf8 _compatible_ AND we have
+	       * mixed utf8 and native (non-utf8) data, then it's a fatal problem
+	       * utf8 _compatible_ means, can be upgraded to utf8, ie. utf8 or ascii */
+	      if ((utf8 & ARRAY_BIND_NATIVE) && !CS_IS_UTF8_COMPATIBLE(csid)) {
+	        warn("Can't mix utf8 and non-utf8 in array bind");
+	      }
+	      csid = utf8_csid; /* not al32utf8_csid here on purpose */
+	    }
 
 	if (trace_level >= 3 || dbd_verbose <= 3 )
 		PerlIO_printf(DBILOGFP, "do_bind_array_exec() (3): bind %s <== [array of values] "
@@ -3385,6 +3359,7 @@
         }
         av = (AV*)SvRV(sv);
         for(i = 0; i < param_count; i++) {
+
             if(!phs[i]) {
                SV **phs_svp;
                sprintf(namebuf, ":p%d", i+1);
@@ -3433,35 +3408,13 @@
 
 			/* update the utf8_flgs for this value */
 
-			PerlIO_printf(DBILOGFP, "\n SvUTF8(sv)=%d. and sv=%s..\n",SvUTF8(sv),neatsvpv(sv,1));
-
-
-
-
 			if (SvUTF8(sv)) {
-				/*STRLEN	len2;
-								char*	xx;
-								xx=SvPVutf8( sv, len2);*/
 				utf8_flgs[i] |= ARRAY_BIND_UTF8;
 			}
 			else {
-				/*STRLEN	len2;
-				char*	xx;
-				bool    is_utf8;
-				xx=SvPVutf8( sv, len2);
-				PerlIO_printf(DBILOGFP, " xx=%s,len2=%d...\n",xx,len2);
-
-				is_utf8=is_utf8_string((U8*)sv,1);
-
-				PerlIO_printf(DBILOGFP, "is_utf8=%d...\n",is_utf8);
-*/
-
 				utf8_flgs[i] |= ARRAY_BIND_NATIVE;
-
 			}
 
-
-
         	/* Do OCI bind calls on last iteration. */
         	if( ((unsigned int) j ) == exe_count - 1 ) {
         	    if(!do_bind_array_exec(sth, imp_sth, phs[i], utf8_flgs[i])) {

Modified: dbd-oracle/branches/utf8_ea/dbdimp.h
==============================================================================
--- dbd-oracle/branches/utf8_ea/dbdimp.h	(original)
+++ dbd-oracle/branches/utf8_ea/dbdimp.h	Fri Oct  3 11:51:48 2008
@@ -213,6 +213,7 @@
     ub2 csid_orig;	/* original oracle default csid 	*/
     ub2 csid;		/* 0 for automatic			*/
     ub1 csform;		/* 0 for automatic			*/
+    ub1 utf8_array;  /* Array has non ascii characters */
     ub4 maxdata_size;	/* set OCI_ATTR_MAXDATA_SIZE if >0	*/
     bool is_inout;
 
@@ -256,7 +257,6 @@
 extern ub2 ncharsetid;
 extern ub2 us7ascii_csid;
 extern ub2 utf8_csid;
-extern ub2 we8mswin1252; /*default for windows */
 extern ub2 al32utf8_csid;
 extern ub2 al16utf16_csid;
 
@@ -264,11 +264,11 @@
   (  ( cs == us7ascii_csid ) )
 
 #define CS_IS_UTF8( cs ) \
-   (  ( cs == utf8_csid ) || ( cs == al32utf8_csid )|| ( cs == we8mswin1252 ) )
+   (  ( cs == utf8_csid ) || ( cs == al32utf8_csid ) )
+
+#define CS_IS_UTF8_COMPATIBLE( cs ) \
+  ( CS_IS_UTF8(cs) || ( (cs) == us7ascii_csid ) )
 
-/*#define CS_IS_UTF8_COMPATIBLE( cs ) \
-  ( CS_IS_UTF8(cs) || ( (cs) == us7ascii_csid ) || ( (cs) == we8mswin1252 ) )
-*/
 
 #define CS_IS_UTF16( cs ) ( cs == al16utf16_csid )
 

Modified: dbd-oracle/branches/utf8_ea/oci8.c
==============================================================================
--- dbd-oracle/branches/utf8_ea/oci8.c	(original)
+++ dbd-oracle/branches/utf8_ea/oci8.c	Fri Oct  3 11:51:48 2008
@@ -518,6 +518,7 @@
     ub4	oparse_lng   	 = 1;  /* auto v6 or v7 as suits db connected to	*/
     int ora_check_sql 	 = 1;	/* to force a describe to check SQL	*/
     IV  ora_placeholders = 1;	/* find and handle placeholders */
+    IV  ora_ary_is_utf8  = 0;
 	/* XXX we set ora_check_sql on for now to force setup of the	*/
 	/* row cache. Change later to set up row cache using just a	*/
 	/* a memory size, perhaps also default $RowCacheSize to a	*/
@@ -556,7 +557,7 @@
 		DBD_ATTRIB_GET_IV(  attribs, "ora_clbk_lob", 12, svp, ora_clbk_lob);
 		DBD_ATTRIB_GET_IV(  attribs, "ora_piece_lob", 13, svp, ora_piece_lob);
 	    DBD_ATTRIB_GET_IV(  attribs, "ora_piece_size", 14, svp, ora_piece_size);
-
+        DBD_ATTRIB_GET_IV(  attribs, "ora_ary_is_utf8", 15, svp, ora_ary_is_utf8);
 		imp_sth->auto_lob = (ora_auto_lob) ? 1 : 0;
 		imp_sth->pers_lob = (ora_pers_lob) ? 1 : 0;
 		imp_sth->clbk_lob = (ora_clbk_lob) ? 1 : 0;
@@ -2082,7 +2083,7 @@
     sv_setpvn(dest_sv, (char*)fb_ary->cb_abuf,(STRLEN)actual_bufl);
 
   	if (fbh->ftype != SQLT_BIN){
-		
+
 		if (CSFORM_IMPLIES_UTF8(fbh->csform) ){ /* do the UTF 8 magic*/
 			SvUTF8_on(dest_sv);
 		}
@@ -3396,7 +3397,7 @@
     	char new_tablename[100];
     	ub4 syn_schema_len = 0, syn_name_len = 0,tn_len;
       	OCIAttrGet_log_stat(imp_sth->dschp,  OCI_HTYPE_DESCRIBE,
-				  &parmhp, 0, OCI_ATTR_PARAM, errhp, status);				  
+				  &parmhp, 0, OCI_ATTR_PARAM, errhp, status);
       	OCIAttrGet_log_stat(parmhp, OCI_DTYPE_PARAM,
       		      &syn_schema, &syn_schema_len, OCI_ATTR_SCHEMA_NAME, errhp, status);
 		OCIAttrGet_log_stat(parmhp, OCI_DTYPE_PARAM,