Merge lp:~zorba-coders/zorba/bug-analyze_string into lp:zorba

[Matthias Brantner]

proper return type specification for analyze-string

proper return type specification for analyze-string done manually instead of automatically done by the code generation.

[Zorba Build Bot (zorba-buildbot)]

Validation queue starting for merge proposal.
Log at: http://zorbatest.lambda.nu:8080/remotequeue/bug-analyze_string-2012-03-15T15-01-13.07Z/log.html

[Paul J. Lucas (paul-lucas)]

Why isn't code-generation improved so it *can* be done via code-generation instead?

[Paul J. Lucas (paul-lucas)]

> Why isn't code-generation improved so it *can* be done via code-generation
> instead?

Automatic code-generation is going to be even more important (and necessary) in light of the forthcoming LLVM stuff. There's going to be even more code that ought to be generated automatically, e.g., C thunks, and lots of boilerplate LLVM C++ API code.

[Zorba Build Bot (zorba-buildbot)]

Validation queue job bug-analyze_string-2012-03-15T15-01-13.07Z is finished. The final status was:

All tests succeeded!

[Zorba Build Bot (zorba-buildbot)]

Voting does not meet specified criteria. Required: Approve > 1, Disapprove < 1, Needs Fixing < 1, Pending < 1. Got: 3 Pending.

[Matthias Brantner (matthias-brantner)]

> Why isn't code-generation improved so it *can* be done via code-generation
> instead?
There are plenty of cases to specify the user-defined types. Handling each of them automatically (i.e. generate the correct create_* function call to create the type) is not worth it for the 2 functions that we have _in the core_ that require this. I don't think there will be a lot more such functions in the core with user-defined return types. Hence, doing it manually seems to be the right thing to me.

[Matthias Brantner (matthias-brantner)]

> > Why isn't code-generation improved so it *can* be done via code-generation
> > instead?
>
> Automatic code-generation is going to be even more important (and necessary)
> in light of the forthcoming LLVM stuff. There's going to be even more code
> that ought to be generated automatically, e.g., C thunks, and lots of
> boilerplate LLVM C++ API code.
I totally agree. However, as far as I understand the issue is orthogonal because this mostly concerns the runtime code and not the code for declaring the functions in the compiler/context.

[Paul J. Lucas (paul-lucas)]

review: Approve

[Markos Zaharioudakis (markos-za)]

review: Approve

[Zorba Build Bot (zorba-buildbot)]

Validation queue starting for merge proposal.
Log at: http://zorbatest.lambda.nu:8080/remotequeue/bug-analyze_string-2012-04-18T13-55-07.571Z/log.html

[Zorba Build Bot (zorba-buildbot)]

Validation queue job bug-analyze_string-2012-04-18T13-55-07.571Z is finished. The final status was:

All tests succeeded!

=== modified file 'src/functions/func_strings_impl.cpp'
--- src/functions/func_strings_impl.cpp	2012-01-11 17:30:25 +0000
+++ src/functions/func_strings_impl.cpp	2012-03-15 15:00:28 +0000
@@ -61,6 +61,44 @@
   return NULL;
 }
 
+
+/*******************************************************************************
+
+********************************************************************************/
+void populate_context_strings_impl(static_context* sctx)
+{
+  xqtref_t lAnalyzeStringResultType =
+    GENV_TYPESYSTEM.create_node_type(
+        store::StoreConsts::elementNode,
+        createQName("http://www.w3.org/2005/xpath-functions","","analyze-string-result"),
+        NULL,
+        TypeConstants::QUANT_ONE,
+        false,
+        false
+      );
+
+  {
+    DECL_WITH_KIND(sctx, fn_analyze_string_3_0,
+        (createQName("http://www.w3.org/2005/xpath-functions","","analyze-string"), 
+        GENV_TYPESYSTEM.STRING_TYPE_QUESTION, 
+        GENV_TYPESYSTEM.STRING_TYPE_ONE, 
+        lAnalyzeStringResultType),
+        FunctionConsts::FN_ANALYZE_STRING_2);
+
+  }
+
+
+  {
+    DECL_WITH_KIND(sctx, fn_analyze_string_3_0,
+        (createQName("http://www.w3.org/2005/xpath-functions","","analyze-string"), 
+        GENV_TYPESYSTEM.STRING_TYPE_QUESTION, 
+        GENV_TYPESYSTEM.STRING_TYPE_ONE, 
+        GENV_TYPESYSTEM.STRING_TYPE_ONE, 
+        lAnalyzeStringResultType),
+        FunctionConsts::FN_ANALYZE_STRING_3);
+  }
+}
+
 }
 
 /* vim:set et sw=2 ts=2: */

=== added file 'src/functions/func_strings_impl.h'
--- src/functions/func_strings_impl.h	1970-01-01 00:00:00 +0000
+++ src/functions/func_strings_impl.h	2012-03-15 15:00:28 +0000
@@ -0,0 +1,38 @@
+/*
+ * Copyright 2006-2012 The FLWOR Foundation.
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+#pragma once
+#ifndef ZORBA_FUNCTIONS_STRINGS_IMPL_H
+#define ZORBA_FUNCTIONS_STRINGS_IMPL_H
+
+
+#include "common/shared_types.h"
+#include "functions/function.h"
+#include "functions/function_impl.h"
+#include "functions/function_consts.h"
+
+
+namespace zorba {
+
+
+void populate_context_strings_impl(static_context* sctx);
+
+
+} //namespace zorba
+
+
+#endif 
+/* vim:set et sw=2 ts=2: */
+

=== modified file 'src/functions/library.cpp'
--- src/functions/library.cpp	2012-03-07 15:49:25 +0000
+++ src/functions/library.cpp	2012-03-15 15:00:28 +0000
@@ -57,6 +57,7 @@
 #include "functions/func_sequences.h"
 #include "functions/func_sequences_impl.h"
 #include "functions/func_strings.h"
+#include "functions/func_strings_impl.h"
 #include "functions/func_uris.h"
 #include "functions/func_json.h"
 #include "functions/func_var_decl.h"
@@ -122,6 +123,7 @@
   populate_context_schema(sctx);
   populate_context_sctx(sctx);
   populate_context_strings(sctx);
+  populate_context_strings_impl(sctx);
   populate_context_uris(sctx);
   populate_context_sequences(sctx);
   populate_context_sequences_impl(sctx);

=== modified file 'src/functions/pregenerated/func_strings.cpp'
--- src/functions/pregenerated/func_strings.cpp	2012-02-16 12:48:17 +0000
+++ src/functions/pregenerated/func_strings.cpp	2012-03-15 15:00:28 +0000
@@ -853,33 +853,6 @@
   {
     
 
-    DECL_WITH_KIND(sctx, fn_analyze_string_3_0,
-        (createQName("http://www.w3.org/2005/xpath-functions","","analyze-string"), 
-        GENV_TYPESYSTEM.STRING_TYPE_QUESTION, 
-        GENV_TYPESYSTEM.STRING_TYPE_ONE, 
-        GENV_TYPESYSTEM.ELEMENT_TYPE_ONE),
-        FunctionConsts::FN_ANALYZE_STRING_2);
-
-  }
-
-
-  {
-    
-
-    DECL_WITH_KIND(sctx, fn_analyze_string_3_0,
-        (createQName("http://www.w3.org/2005/xpath-functions","","analyze-string"), 
-        GENV_TYPESYSTEM.STRING_TYPE_QUESTION, 
-        GENV_TYPESYSTEM.STRING_TYPE_ONE, 
-        GENV_TYPESYSTEM.STRING_TYPE_ONE, 
-        GENV_TYPESYSTEM.ELEMENT_TYPE_ONE),
-        FunctionConsts::FN_ANALYZE_STRING_3);
-
-  }
-
-
-  {
-    
-
     DECL_WITH_KIND(sctx, fn_zorba_string_materialize,
         (createQName("http://www.zorba-xquery.com/modules/string","","materialize"), 
         GENV_TYPESYSTEM.STRING_TYPE_ONE, 

=== modified file 'src/runtime/spec/strings/strings.xml'
--- src/runtime/spec/strings/strings.xml	2012-01-11 17:30:25 +0000
+++ src/runtime/spec/strings/strings.xml	2012-03-15 15:00:28 +0000
@@ -668,7 +668,7 @@
     fn:analyze-string
   </zorba:description>
 
-  <zorba:function>
+  <zorba:function generateDECL="false">
     <zorba:signature localname="analyze-string" prefix="fn" version="3.0">
       <zorba:param>xs:string?</zorba:param>
       <zorba:param>xs:string</zorba:param>