mbeckerle commented on a change in pull request #259: Incremental progress on schema compilation space/speed issue. URL: https://github.com/apache/incubator-daffodil/pull/259#discussion_r303501595
########## File path: daffodil-test/src/test/resources/org/apache/daffodil/section23/dfdl_expressions/expressions3.tdml ########## @@ -0,0 +1,400 @@ +<?xml version="1.0" encoding="UTF-8"?> +<!-- + Licensed to the Apache Software Foundation (ASF) under one or more + contributor license agreements. See the NOTICE file distributed with + this work for additional information regarding copyright ownership. + The ASF licenses this file to You 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. +--> + +<tdml:testSuite + suiteName="expressions3" + xmlns:tdml="http://www.ibm.com/xmlns/dfdl/testData"; + xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"; + xmlns:xs="http://www.w3.org/2001/XMLSchema"; + xmlns:dfdl="http://www.ogf.org/dfdl/dfdl-1.0/"; + xmlns:dfdlx="http://www.ogf.org/dfdl/dfdl-1.0/extensions"; + xmlns:ex="http://example.com"; + xmlns:fn="http://www.w3.org/2005/xpath-functions"; + defaultRoundTrip="true"> + + <tdml:defineSchema name="s1" elementFormDefault="unqualified"> + <xs:include schemaLocation="org/apache/daffodil/xsd/DFDLGeneralFormat.dfdl.xsd" /> + <dfdl:format ref="ex:GeneralFormat" lengthKind="delimited"/> + + <xs:group name="g"> + <!-- + The path expression below is going to have different type + elements for the different instances of this group in the + final schema. + + Ultimately they all are converted to string. + --> + <xs:sequence> + <xs:choice dfdl:choiceDispatchKey="{ ../foo/bar }"> + <xs:element name="c1" type="xs:string" dfdl:choiceBranchKey="1" /> + <xs:element name="c2" type="xs:string" dfdl:choiceBranchKey="2" /> + </xs:choice> + </xs:sequence> + </xs:group> + + <xs:element name="r"> + <xs:complexType> + <xs:sequence dfdl:separator="|"> + <xs:element name="e1"> + <xs:complexType> + <xs:sequence> + <xs:element name="foo"> + <xs:complexType> + <xs:sequence dfdl:separator="|"> + <xs:element name="bar" type="xs:string" /> + <xs:group ref="ex:g" /> + </xs:sequence> + </xs:complexType> + </xs:element> + </xs:sequence> + </xs:complexType> + </xs:element> + <xs:element name="e2"> + <xs:complexType> + <xs:sequence> + <xs:element name="foo"> + <xs:complexType> + <xs:sequence dfdl:separator="|"> + <xs:element name="bar" type="xs:int" /> + <xs:group ref="ex:g" /> + </xs:sequence> + </xs:complexType> + </xs:element> + </xs:sequence> + </xs:complexType> + </xs:element> + </xs:sequence> + </xs:complexType> + </xs:element> + </tdml:defineSchema> + + <tdml:parserTestCase name="test_polymorphic_expr_1" model="s1" root="r" + description="Polymorphic path expression"> + <tdml:document>1|a|2|b</tdml:document> + <tdml:infoset> + <tdml:dfdlInfoset> + <ex:r> + <e1> + <foo> + <bar>1</bar> + <c1>a</c1> + </foo> + </e1> + <e2> + <foo> + <bar>2</bar> + <c2>b</c2> + </foo> + </e2> + </ex:r> + </tdml:dfdlInfoset> + </tdml:infoset> + <tdml:warnings> + <tdml:warning>should be manually cast</tdml:warning> + </tdml:warnings> + </tdml:parserTestCase> + + <tdml:defineSchema name="s2a" elementFormDefault="unqualified"> Review comment: DFDL does not have anyURI type. I do have example of a path that is shared and has type int/float, and float/complex-type. This is a point where the DFDL spec is unclear. It takes no position on whether expressions are shared or copied/inlined per context. What I've done is make it so that if properties are expressed on a reference term (element ref, sequence ref, or choice ref) then the properties are copied. This is a natural consequence of the fact that the properties might appear on the ref, not on the shared object being referenced, so they may in fact be distinct property values despite referencing a shared definition. However, if an expression appears on a local element, sequence or choice, then it is inside an element decl, or group definition, so then it is not being shared, and so no copying is being done. BUT, a relative path within it it that is upward-and-outward (e.g., ../../foo/bar) may reference outside the global decl/definition. This is the case in my test examples. In this case the path is (now, with this change set) being interpreted as a polymorphic expression ... one that must make sense in all contexts, and must have a common definition shared definition in all contexts. So if we have int and float interpretations we will get double behavior since the LUB of int and float is SignedNumeric, and the most general representation of that (the type all such can be converted into) is double. Same with float and decimal. But if it is int and decimal, you'll get decimal behavior since decimal is a supertype of int. I can add examples that show the int/decimal case, so as to show you don't always end up with double. The only way to test this in TDML is to have a test that requires an auto-coercion be added, and then detect the warning for this, or possibly set a flag that makes the lack of this auto-coercion to cause an error. The error can be scrutinized for "decimal". I will construct one of these examples. ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services
