Merge pull request #264 from codelynx/master

bugfix: supporting namespace for querying the same attribute name with different namespace
tid-kijyun · Jan 19, 2024 · 86bca4f · 86bca4f
2 parents bf19cf5 + c050bda
commit 86bca4f
Show file tree

Hide file tree

Showing 7 changed files with 225 additions and 13 deletions.
diff --git a/Kanna.xcodeproj/project.pbxproj b/Kanna.xcodeproj/project.pbxproj
@@ -26,6 +26,8 @@
 		1EB4A01F204C1F240003D7A2 /* KannaCSSTests.swift in Sources */ = {isa = PBXBuildFile; fileRef = 1EB4A01E204C1F240003D7A2 /* KannaCSSTests.swift */; };
 		1EB4A021204C20760003D7A2 /* KannaXMLTests.swift in Sources */ = {isa = PBXBuildFile; fileRef = 1EB4A020204C20760003D7A2 /* KannaXMLTests.swift */; };
 		1EC805FA1FA2FB2F0067D3DA /* Deprecated.swift in Sources */ = {isa = PBXBuildFile; fileRef = 1EC805F91FA2FB2F0067D3DA /* Deprecated.swift */; };
+		CC02AE172B5A1A210075B74A /* pptx-presentation.xml.rels in Resources */ = {isa = PBXBuildFile; fileRef = CC02AE162B5A1A210075B74A /* pptx-presentation.xml.rels */; };
+		CC02AE192B5A1A480075B74A /* pptx-presentation.xml in Resources */ = {isa = PBXBuildFile; fileRef = CC02AE182B5A1A480075B74A /* pptx-presentation.xml */; };
 /* End PBXBuildFile section */
 
 /* Begin PBXContainerItemProxy section */
@@ -66,6 +68,8 @@
 		1EB4A01E204C1F240003D7A2 /* KannaCSSTests.swift */ = {isa = PBXFileReference; lastKnownFileType = sourcecode.swift; path = KannaCSSTests.swift; sourceTree = "<group>"; };
 		1EB4A020204C20760003D7A2 /* KannaXMLTests.swift */ = {isa = PBXFileReference; lastKnownFileType = sourcecode.swift; path = KannaXMLTests.swift; sourceTree = "<group>"; };
 		1EC805F91FA2FB2F0067D3DA /* Deprecated.swift */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = sourcecode.swift; path = Deprecated.swift; sourceTree = "<group>"; };
+		CC02AE162B5A1A210075B74A /* pptx-presentation.xml.rels */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = text.xml; path = "pptx-presentation.xml.rels"; sourceTree = "<group>"; };
+		CC02AE182B5A1A480075B74A /* pptx-presentation.xml */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = text.xml; path = "pptx-presentation.xml"; sourceTree = "<group>"; };
 /* End PBXFileReference section */
 
 /* Begin PBXFrameworksBuildPhase section */
@@ -171,6 +175,8 @@
 				1E7ADC9C1DF5907F006E1815 /* sample.html */,
 				1E7ADC9D1DF5907F006E1815 /* test_HTML4.html */,
 				1E7ADC9E1DF5907F006E1815 /* test_XML_ExcelWorkbook.xml */,
+				CC02AE162B5A1A210075B74A /* pptx-presentation.xml.rels */,
+				CC02AE182B5A1A480075B74A /* pptx-presentation.xml */,
 				1E7ADC9F1DF5907F006E1815 /* versions.xml */,
 			);
 			name = Data;
@@ -282,9 +288,11 @@
 			isa = PBXResourcesBuildPhase;
 			buildActionMask = 2147483647;
 			files = (
+				CC02AE172B5A1A210075B74A /* pptx-presentation.xml.rels in Resources */,
 				1E7ADCA21DF5907F006E1815 /* test_HTML4.html in Resources */,
 				1E7ADCA11DF5907F006E1815 /* sample.html in Resources */,
 				1E7ADCA01DF5907F006E1815 /* libraries.xml in Resources */,
+				CC02AE192B5A1A480075B74A /* pptx-presentation.xml in Resources */,
 				1E7ADCA41DF5907F006E1815 /* versions.xml in Resources */,
 				1E7ADCA31DF5907F006E1815 /* test_XML_ExcelWorkbook.xml in Resources */,
 			);
@@ -403,16 +411,19 @@
 				GCC_WARN_UNINITIALIZED_AUTOS = YES_AGGRESSIVE;
 				GCC_WARN_UNUSED_FUNCTION = YES;
 				GCC_WARN_UNUSED_VARIABLE = YES;
-				MACOSX_DEPLOYMENT_TARGET = 10.9;
+				IPHONEOS_DEPLOYMENT_TARGET = 12.0;
+				MACOSX_DEPLOYMENT_TARGET = 10.13;
 				MTL_ENABLE_DEBUG_INFO = YES;
 				ONLY_ACTIVE_ARCH = YES;
 				SDKROOT = macosx;
 				SWIFT_ACTIVE_COMPILATION_CONDITIONS = DEBUG;
 				SWIFT_INCLUDE_PATHS = "";
 				SWIFT_OPTIMIZATION_LEVEL = "-Onone";
 				SWIFT_VERSION = 5.0;
+				TVOS_DEPLOYMENT_TARGET = 12.0;
 				VERSIONING_SYSTEM = "apple-generic";
 				VERSION_INFO_PREFIX = "";
+				WATCHOS_DEPLOYMENT_TARGET = 4.0;
 			};
 			name = Debug;
 		};
@@ -462,14 +473,17 @@
 				GCC_WARN_UNINITIALIZED_AUTOS = YES_AGGRESSIVE;
 				GCC_WARN_UNUSED_FUNCTION = YES;
 				GCC_WARN_UNUSED_VARIABLE = YES;
-				MACOSX_DEPLOYMENT_TARGET = 10.9;
+				IPHONEOS_DEPLOYMENT_TARGET = 12.0;
+				MACOSX_DEPLOYMENT_TARGET = 10.13;
 				MTL_ENABLE_DEBUG_INFO = NO;
 				SDKROOT = macosx;
 				SWIFT_INCLUDE_PATHS = "";
 				SWIFT_OPTIMIZATION_LEVEL = "-Owholemodule";
 				SWIFT_VERSION = 5.0;
+				TVOS_DEPLOYMENT_TARGET = 12.0;
 				VERSIONING_SYSTEM = "apple-generic";
 				VERSION_INFO_PREFIX = "";
+				WATCHOS_DEPLOYMENT_TARGET = 4.0;
 			};
 			name = Release;
 		};
@@ -487,12 +501,16 @@
 				FRAMEWORK_VERSION = A;
 				INFOPLIST_FILE = Sources/Kanna/Info.plist;
 				INSTALL_PATH = "$(LOCAL_LIBRARY_DIR)/Frameworks";
+				IPHONEOS_DEPLOYMENT_TARGET = 12.0;
 				LD_RUNPATH_SEARCH_PATHS = "$(inherited) @executable_path/../Frameworks @loader_path/Frameworks";
+				MACOSX_DEPLOYMENT_TARGET = 10.13;
 				OTHER_LDFLAGS = "-lxml2";
 				PRODUCT_BUNDLE_IDENTIFIER = com.tid.Kanna;
 				PRODUCT_NAME = "$(TARGET_NAME)";
 				SKIP_INSTALL = YES;
 				SWIFT_VERSION = 5.0;
+				TVOS_DEPLOYMENT_TARGET = 12.0;
+				WATCHOS_DEPLOYMENT_TARGET = 4.0;
 			};
 			name = Debug;
 		};
@@ -510,12 +528,16 @@
 				FRAMEWORK_VERSION = A;
 				INFOPLIST_FILE = Sources/Kanna/Info.plist;
 				INSTALL_PATH = "$(LOCAL_LIBRARY_DIR)/Frameworks";
+				IPHONEOS_DEPLOYMENT_TARGET = 12.0;
 				LD_RUNPATH_SEARCH_PATHS = "$(inherited) @executable_path/../Frameworks @loader_path/Frameworks";
+				MACOSX_DEPLOYMENT_TARGET = 10.13;
 				OTHER_LDFLAGS = "-lxml2";
 				PRODUCT_BUNDLE_IDENTIFIER = com.tid.Kanna;
 				PRODUCT_NAME = "$(TARGET_NAME)";
 				SKIP_INSTALL = YES;
 				SWIFT_VERSION = 5.0;
+				TVOS_DEPLOYMENT_TARGET = 12.0;
+				WATCHOS_DEPLOYMENT_TARGET = 4.0;
 			};
 			name = Release;
 		};
@@ -528,10 +550,12 @@
 				COMBINE_HIDPI_IMAGES = YES;
 				DEVELOPMENT_TEAM = DP9Q5R8635;
 				LD_RUNPATH_SEARCH_PATHS = "$(inherited) @executable_path/../Frameworks @loader_path/../Frameworks";
+				MACOSX_DEPLOYMENT_TARGET = 10.13;
 				PRODUCT_BUNDLE_IDENTIFIER = com.tid.KannaTests;
 				PRODUCT_NAME = "$(TARGET_NAME)";
 				SWIFT_OPTIMIZATION_LEVEL = "-Onone";
 				SWIFT_VERSION = 5.0;
+				TVOS_DEPLOYMENT_TARGET = 12.0;
 			};
 			name = Debug;
 		};
@@ -544,9 +568,11 @@
 				COMBINE_HIDPI_IMAGES = YES;
 				DEVELOPMENT_TEAM = DP9Q5R8635;
 				LD_RUNPATH_SEARCH_PATHS = "$(inherited) @executable_path/../Frameworks @loader_path/../Frameworks";
+				MACOSX_DEPLOYMENT_TARGET = 10.13;
 				PRODUCT_BUNDLE_IDENTIFIER = com.tid.KannaTests;
 				PRODUCT_NAME = "$(TARGET_NAME)";
 				SWIFT_VERSION = 5.0;
+				TVOS_DEPLOYMENT_TARGET = 12.0;
 			};
 			name = Release;
 		};

diff --git a/Sources/Kanna/Kanna.swift b/Sources/Kanna/Kanna.swift
@@ -180,6 +180,15 @@ public protocol XMLDocument: AnyObject, SearchableNode {
     var namespaces: [Namespace] { get }
 }
 
+public extension XMLDocument {
+	var namespaceDictionary: [String: String]? {
+		let dictionary = self.namespaces.reduce(into: [:]) {
+			// when prefix is blank, treat prefix "" as "xmlns", or xpath cannot specify "" as prefix
+			$0[$1.prefix == "" ? "xmlns": $1.prefix] = $1.name
+		}
+		return dictionary.count > 0 ? dictionary : nil
+	}
+}
 /**
 HTMLDocument
 */

diff --git a/Sources/Kanna/libxmlHTMLDocument.swift b/Sources/Kanna/libxmlHTMLDocument.swift
@@ -307,6 +307,7 @@ final class libxmlXMLDocument: XMLDocument {
     }
 
     func xpath(_ xpath: String, namespaces: [String: String]? = nil) -> XPathObject {
+    	let namespaces = namespaces ?? self.namespaceDictionary
         guard let docPtr = docPtr else { return .none }
         return XPath(doc: self, docPtr: docPtr).xpath(xpath, namespaces: namespaces)
     }
@@ -336,6 +337,7 @@ struct XPath {
         guard let ctxt = xmlXPathNewContext(docPtr) else { return .none }
         defer { xmlXPathFreeContext(ctxt) }
 
+    	let namespaces = namespaces ?? self.doc.namespaceDictionary
         if let nsDictionary = namespaces {
             for (ns, name) in nsDictionary {
                 xmlXPathRegisterNs(ctxt, ns, name)

diff --git a/Sources/Kanna/libxmlHTMLNode.swift b/Sources/Kanna/libxmlHTMLNode.swift
@@ -115,17 +115,17 @@ final class libxmlHTMLNode: XMLElement {
         get {
             var attr = nodePtr.pointee.properties
             while attr != nil {
-                let mem = attr?.pointee
-                if let tagName = String(validatingUTF8: UnsafeRawPointer((mem?.name)!).assumingMemoryBound(to: CChar.self)) {
-                    if attributeName == tagName {
-                        if let children = mem?.children {
-                            return libxmlGetNodeContent(children)
-                        } else {
-                            return ""
-                        }
-                    }
-                }
-                attr = attr?.pointee.next
+                let mem = attr!.pointee
+                let prefix = mem.ns.flatMap { $0.pointee.prefix.string }
+                let tagName = [prefix, mem.name.string].compactMap { $0 }.joined(separator: ":")
+				if attributeName == tagName {
+					if let children = mem.children {
+						return libxmlGetNodeContent(children)
+					} else {
+						return ""
+					}
+				}
+                attr = attr!.pointee.next
             }
             return nil
         }
@@ -233,3 +233,10 @@ private func escape(_ str: String) -> String {
     }
     return newStr
 }
+
+fileprivate extension UnsafePointer<UInt8> {
+	var string: String? {
+		let string = String(validatingUTF8: UnsafePointer<CChar>(OpaquePointer(self)))
+		return string
+	}
+}
diff --git a/Tests/KannaTests/Data/pptx-presentation.xml b/Tests/KannaTests/Data/pptx-presentation.xml
@@ -0,0 +1,118 @@
+<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
+<p:presentation xmlns:a="http://schemas.openxmlformats.org/drawingml/2006/main"
+    xmlns:r="http://schemas.openxmlformats.org/officeDocument/2006/relationships"
+    xmlns:p="http://schemas.openxmlformats.org/presentationml/2006/main" saveSubsetFonts="1"
+    autoCompressPictures="0">
+    <p:sldMasterIdLst>
+        <p:sldMasterId id="2147483648" r:id="rId1" />
+    </p:sldMasterIdLst>
+    <p:sldIdLst>
+        <p:sldId id="256" r:id="rId2" />
+    </p:sldIdLst>
+    <p:sldSz cx="12192000" cy="6858000" />
+    <p:notesSz cx="6858000" cy="9144000" />
+    <p:defaultTextStyle>
+        <a:defPPr>
+            <a:defRPr lang="en-US" />
+        </a:defPPr>
+        <a:lvl1pPr marL="0" algn="l" defTabSz="914400" rtl="0" eaLnBrk="1" latinLnBrk="0"
+            hangingPunct="1">
+            <a:defRPr sz="1800" kern="1200">
+                <a:solidFill>
+                    <a:schemeClr val="tx1" />
+                </a:solidFill>
+                <a:latin typeface="+mn-lt" />
+                <a:ea typeface="+mn-ea" />
+                <a:cs typeface="+mn-cs" />
+            </a:defRPr>
+        </a:lvl1pPr>
+        <a:lvl2pPr marL="457200" algn="l" defTabSz="914400" rtl="0" eaLnBrk="1" latinLnBrk="0"
+            hangingPunct="1">
+            <a:defRPr sz="1800" kern="1200">
+                <a:solidFill>
+                    <a:schemeClr val="tx1" />
+                </a:solidFill>
+                <a:latin typeface="+mn-lt" />
+                <a:ea typeface="+mn-ea" />
+                <a:cs typeface="+mn-cs" />
+            </a:defRPr>
+        </a:lvl2pPr>
+        <a:lvl3pPr marL="914400" algn="l" defTabSz="914400" rtl="0" eaLnBrk="1" latinLnBrk="0"
+            hangingPunct="1">
+            <a:defRPr sz="1800" kern="1200">
+                <a:solidFill>
+                    <a:schemeClr val="tx1" />
+                </a:solidFill>
+                <a:latin typeface="+mn-lt" />
+                <a:ea typeface="+mn-ea" />
+                <a:cs typeface="+mn-cs" />
+            </a:defRPr>
+        </a:lvl3pPr>
+        <a:lvl4pPr marL="1371600" algn="l" defTabSz="914400" rtl="0" eaLnBrk="1" latinLnBrk="0"
+            hangingPunct="1">
+            <a:defRPr sz="1800" kern="1200">
+                <a:solidFill>
+                    <a:schemeClr val="tx1" />
+                </a:solidFill>
+                <a:latin typeface="+mn-lt" />
+                <a:ea typeface="+mn-ea" />
+                <a:cs typeface="+mn-cs" />
+            </a:defRPr>
+        </a:lvl4pPr>
+        <a:lvl5pPr marL="1828800" algn="l" defTabSz="914400" rtl="0" eaLnBrk="1" latinLnBrk="0"
+            hangingPunct="1">
+            <a:defRPr sz="1800" kern="1200">
+                <a:solidFill>
+                    <a:schemeClr val="tx1" />
+                </a:solidFill>
+                <a:latin typeface="+mn-lt" />
+                <a:ea typeface="+mn-ea" />
+                <a:cs typeface="+mn-cs" />
+            </a:defRPr>
+        </a:lvl5pPr>
+        <a:lvl6pPr marL="2286000" algn="l" defTabSz="914400" rtl="0" eaLnBrk="1" latinLnBrk="0"
+            hangingPunct="1">
+            <a:defRPr sz="1800" kern="1200">
+                <a:solidFill>
+                    <a:schemeClr val="tx1" />
+                </a:solidFill>
+                <a:latin typeface="+mn-lt" />
+                <a:ea typeface="+mn-ea" />
+                <a:cs typeface="+mn-cs" />
+            </a:defRPr>
+        </a:lvl6pPr>
+        <a:lvl7pPr marL="2743200" algn="l" defTabSz="914400" rtl="0" eaLnBrk="1" latinLnBrk="0"
+            hangingPunct="1">
+            <a:defRPr sz="1800" kern="1200">
+                <a:solidFill>
+                    <a:schemeClr val="tx1" />
+                </a:solidFill>
+                <a:latin typeface="+mn-lt" />
+                <a:ea typeface="+mn-ea" />
+                <a:cs typeface="+mn-cs" />
+            </a:defRPr>
+        </a:lvl7pPr>
+        <a:lvl8pPr marL="3200400" algn="l" defTabSz="914400" rtl="0" eaLnBrk="1" latinLnBrk="0"
+            hangingPunct="1">
+            <a:defRPr sz="1800" kern="1200">
+                <a:solidFill>
+                    <a:schemeClr val="tx1" />
+                </a:solidFill>
+                <a:latin typeface="+mn-lt" />
+                <a:ea typeface="+mn-ea" />
+                <a:cs typeface="+mn-cs" />
+            </a:defRPr>
+        </a:lvl8pPr>
+        <a:lvl9pPr marL="3657600" algn="l" defTabSz="914400" rtl="0" eaLnBrk="1" latinLnBrk="0"
+            hangingPunct="1">
+            <a:defRPr sz="1800" kern="1200">
+                <a:solidFill>
+                    <a:schemeClr val="tx1" />
+                </a:solidFill>
+                <a:latin typeface="+mn-lt" />
+                <a:ea typeface="+mn-ea" />
+                <a:cs typeface="+mn-cs" />
+            </a:defRPr>
+        </a:lvl9pPr>
+    </p:defaultTextStyle>
+</p:presentation>
diff --git a/Tests/KannaTests/Data/pptx-presentation.xml.rels b/Tests/KannaTests/Data/pptx-presentation.xml.rels
@@ -0,0 +1,21 @@
+<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
+<Relationships xmlns="http://schemas.openxmlformats.org/package/2006/relationships">
+    <Relationship Id="rId3"
+        Type="http://schemas.openxmlformats.org/officeDocument/2006/relationships/presProps"
+        Target="presProps.xml" />
+    <Relationship Id="rId2"
+        Type="http://schemas.openxmlformats.org/officeDocument/2006/relationships/slide"
+        Target="slides/slide1.xml" />
+    <Relationship Id="rId1"
+        Type="http://schemas.openxmlformats.org/officeDocument/2006/relationships/slideMaster"
+        Target="slideMasters/slideMaster1.xml" />
+    <Relationship Id="rId6"
+        Type="http://schemas.openxmlformats.org/officeDocument/2006/relationships/tableStyles"
+        Target="tableStyles.xml" />
+    <Relationship Id="rId5"
+        Type="http://schemas.openxmlformats.org/officeDocument/2006/relationships/theme"
+        Target="theme/theme1.xml" />
+    <Relationship Id="rId4"
+        Type="http://schemas.openxmlformats.org/officeDocument/2006/relationships/viewProps"
+        Target="viewProps.xml" />
+</Relationships>
diff --git a/Tests/KannaTests/KannaXMLTests.swift b/Tests/KannaTests/KannaXMLTests.swift
@@ -88,6 +88,35 @@ class KannaXMLTests: XCTestCase {
             XCTAssertEqual(namespaces.sorted(), arry.sorted())
         }
     }
+
+    func testNamespaces_multipleNamespaces() {
+    	// namespaces: "xmlns:a", "xmlns:r", "xmlns:p"
+    	let url = Bundle(for: KannaXMLTests.self).url(forResource: "pptx-presentation", withExtension: "xml")
+    	XCTAssertNotNil(url)
+    	let doc = try? XML(url: url!, encoding: .utf8)
+    	XCTAssertNotNil(doc)
+    	let nodes = Array(doc!.xpath("//p:sldId"))
+    	XCTAssert(nodes.count == 1)
+    	let sldId = nodes[0]
+    	XCTAssert(sldId.tagName == "sldId")
+    	XCTAssert(sldId["id"] == "256")
+    	XCTAssert(sldId["r:id"] == "rId2")
+    }
+
+    func testNamespaces_singleNamespace() {
+    	// namespaces: "xmlns"
+    	let url = Bundle(for: KannaXMLTests.self).url(forResource: "pptx-presentation", withExtension: "xml.rels")
+    	XCTAssertNotNil(url)
+    	let doc = try? XML(url: url!, encoding: .utf8)
+    	XCTAssertNotNil(doc)
+    	let nodes1 = Array(doc!.xpath("//Relationship"))
+    	XCTAssert(nodes1.count == 0)
+    	let nodes2 = Array(doc!.xpath("//xmlns:Relationship"))
+    	XCTAssert(nodes2.count == 6)
+    	let (relationship0, relationship1) = (nodes2[0], nodes2[1])
+    	XCTAssert(relationship0["Id"] == "rId3")
+    	XCTAssert(relationship1["Id"] == "rId2")
+    }
 }
 
 extension KannaXMLTests {