Fix #102: Quote configurable delimiter character, not just commas (#107)

swiftcsv · Aug 28, 2022 · ada8ce6 · ada8ce6
1 parent 50a62e4
commit ada8ce6
Show file tree

Hide file tree

Showing 6 changed files with 72 additions and 36 deletions.
diff --git a/SwiftCSV.xcodeproj/project.pbxproj b/SwiftCSV.xcodeproj/project.pbxproj
@@ -82,6 +82,7 @@
 		DF94FE462898F3A3008FD3F9 /* utf8_with_bom.csv in Resources */ = {isa = PBXBuildFile; fileRef = DF94FE452898F3A3008FD3F9 /* utf8_with_bom.csv */; };
 		DF94FE472898F3A3008FD3F9 /* utf8_with_bom.csv in Resources */ = {isa = PBXBuildFile; fileRef = DF94FE452898F3A3008FD3F9 /* utf8_with_bom.csv */; };
 		DF94FE482898F3A3008FD3F9 /* utf8_with_bom.csv in Resources */ = {isa = PBXBuildFile; fileRef = DF94FE452898F3A3008FD3F9 /* utf8_with_bom.csv */; };
+		DFAD8B7B28B601EB0042BB56 /* Serializer.swift in Sources */ = {isa = PBXBuildFile; fileRef = DFAD8B7A28B601EB0042BB56 /* Serializer.swift */; };
 		E46085921CCB1E8F00385286 /* large.csv in Resources */ = {isa = PBXBuildFile; fileRef = E46085911CCB1E8F00385286 /* large.csv */; };
 		E46085941CCB1F5C00385286 /* PerformanceTest.swift in Sources */ = {isa = PBXBuildFile; fileRef = E46085931CCB1F5C00385286 /* PerformanceTest.swift */; };
 		F5C19F502283243C00920B06 /* ResourceHelper.swift in Sources */ = {isa = PBXBuildFile; fileRef = F5C19F4F2283243C00920B06 /* ResourceHelper.swift */; };
@@ -153,6 +154,7 @@
 		BE6C86061CB5CE44009A351D /* QuotedTests.swift */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = sourcecode.swift; path = QuotedTests.swift; sourceTree = "<group>"; };
 		BE9B02D71CBE57B8009FE424 /* Parser.swift */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = sourcecode.swift; path = Parser.swift; sourceTree = "<group>"; };
 		DF94FE452898F3A3008FD3F9 /* utf8_with_bom.csv */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = text; path = utf8_with_bom.csv; sourceTree = "<group>"; };
+		DFAD8B7A28B601EB0042BB56 /* Serializer.swift */ = {isa = PBXFileReference; lastKnownFileType = sourcecode.swift; path = Serializer.swift; sourceTree = "<group>"; };
 		E46085911CCB1E8F00385286 /* large.csv */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = text; path = large.csv; sourceTree = "<group>"; };
 		E46085931CCB1F5C00385286 /* PerformanceTest.swift */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = sourcecode.swift; path = PerformanceTest.swift; sourceTree = "<group>"; };
 		F5C19F4F2283243C00920B06 /* ResourceHelper.swift */ = {isa = PBXFileReference; fileEncoding = 4; lastKnownFileType = sourcecode.swift; path = ResourceHelper.swift; sourceTree = "<group>"; };
@@ -255,6 +257,7 @@
 				3D444BCC1C7D88290001C60C /* String+Lines.swift */,
 				BE9B02D71CBE57B8009FE424 /* Parser.swift */,
 				508975D61DBF34CF006F3DBE /* ParsingState.swift */,
+				DFAD8B7A28B601EB0042BB56 /* Serializer.swift */,
 			);
 			path = SwiftCSV;
 			sourceTree = "<group>";
@@ -588,6 +591,7 @@
 			buildActionMask = 2147483647;
 			files = (
 				50B3EEA4286F8A84007B3956 /* CSVDelimiter.swift in Sources */,
+				DFAD8B7B28B601EB0042BB56 /* Serializer.swift in Sources */,
 				508975D21DBB897A006F3DBE /* NamedCSVView.swift in Sources */,
 				508CA0FB2771F2E70084C8E8 /* CSV+DelimiterGuessing.swift in Sources */,
 				3DAAEE9C1C74C7EC00A933DB /* CSV.swift in Sources */,

diff --git a/SwiftCSV/CSV.swift b/SwiftCSV/CSV.swift
@@ -128,14 +128,6 @@ extension CSV: CustomStringConvertible {
     }
 }
 
-func enquoteContentsIfNeeded(cell: String) -> String {
-    // Add quotes if value contains a comma
-    if cell.contains(",") {
-        return "\"\(cell)\""
-    }
-    return cell
-}
-
 extension CSV {
     /// Load a CSV file from `url`.
     ///

diff --git a/SwiftCSV/EnumeratedCSVView.swift b/SwiftCSV/EnumeratedCSVView.swift
@@ -44,19 +44,9 @@ public struct Enumerated: CSVView {
     }
 
     public func serialize(header: [String], delimiter: CSVDelimiter) -> String {
-        let separator = String(delimiter.rawValue)
-
-        let head = header
-            .map(enquoteContentsIfNeeded(cell:))
-            .joined(separator: separator) + "\n"
-
-        let content = rows.map { row in
-            row.map(enquoteContentsIfNeeded(cell:))
-                .joined(separator: separator)
-        }.joined(separator: "\n")
-
-        return head + content
+        return Serializer.serialize(header: header, rows: rows, delimiter: delimiter)
     }
+
 }
 
 extension Collection {

diff --git a/SwiftCSV/NamedCSVView.swift b/SwiftCSV/NamedCSVView.swift
@@ -35,19 +35,11 @@ public struct Named: CSVView {
     }
 
     public func serialize(header: [String], delimiter: CSVDelimiter) -> String {
-        let separator = String(delimiter.rawValue)
+        let rowsOrderingCellsByHeader = rows.map { row in
+            header.map { cellID in row[cellID]! }
+        }
 
-        let head = header
-            .map(enquoteContentsIfNeeded(cell:))
-            .joined(separator: separator) + "\n"
-
-        let content = rows.map { row in
-            header
-                .map { cellID in row[cellID]! }
-                .map(enquoteContentsIfNeeded(cell:))
-                .joined(separator: separator)
-        }.joined(separator: "\n")
-
-        return head + content
+        return Serializer.serialize(header: header, rows: rowsOrderingCellsByHeader, delimiter: delimiter)
     }
+
 }
diff --git a/SwiftCSV/Serializer.swift b/SwiftCSV/Serializer.swift
@@ -0,0 +1,46 @@
+//
+//  Serializer.swift
+//  SwiftCSV
+//
+
+import Foundation
+
+enum Serializer {
+
+    static let newline = "\n"
+
+    static func serialize(header: [String], rows: [[String]], delimiter: CSVDelimiter) -> String {
+        let head = serializeRow(row: header, delimiter: delimiter) + newline
+
+        let content = rows.map { row in
+            serializeRow(row: row, delimiter: delimiter)
+        }.joined(separator: newline)
+
+        return head + content
+    }
+
+
+    static func serializeRow(row: [String], delimiter: CSVDelimiter) -> String {
+        let separator = String(delimiter.rawValue)
+
+        let content = row.map { cell in
+            cell.enquoted(whenContaining: separator)
+        }.joined(separator: separator)
+
+        return content
+    }
+
+}
+
+fileprivate extension String {
+
+    func enquoted(whenContaining separator: String) -> String {
+        // Add quotes if value contains a delimiter
+        if self.contains(separator) {
+            return "\"\(self)\""
+        }
+
+        return self
+    }
+
+}
diff --git a/SwiftCSVTests/EnumeratedCSVViewTests.swift b/SwiftCSVTests/EnumeratedCSVViewTests.swift
@@ -76,12 +76,24 @@ class EnumeratedViewTests: XCTestCase {
         XCTAssertEqual(csv.columns, expectedColumns)
     }
 
-    func testSerialization() {
+    func testSerialization() throws {
+        // Comma-separated values.
         XCTAssertEqual(csv.serialized, "id,name,age\n1,Alice,18\n2,Bob,19\n3,Charlie,20")
-    }
 
-    func testSerializationWithDoubleQuotes() throws {
+        // Comma-separated values with double quotes and embedded delimiters in cells.
         csv = try CSV<Enumerated>(string: "id,\"the, name\",age\n1,\"Alice, In, Wonderland\",18\n2,Bob,19\n3,Charlie,20")
         XCTAssertEqual(csv.serialized, "id,\"the, name\",age\n1,\"Alice, In, Wonderland\",18\n2,Bob,19\n3,Charlie,20")
+
+        // Tab-separated values with implicit delimiter (delimiter guessing).
+        csv = try CSV<Enumerated>(string: "id\tname\tage\n1\tAlice\t18\n2\tBob\t19\n3\tCharlie\t20")
+        XCTAssertEqual(csv.serialized, "id\tname\tage\n1\tAlice\t18\n2\tBob\t19\n3\tCharlie\t20")
+
+        // Tab-separated values with double quotes and embedded delimiters in cells.
+        csv = try CSV<Enumerated>(string: "id\t\"the\t name\"\tage\n1\t\"Alice\t In\t Wonderland\"\t18\n2\tBob\t19\n3\tCharlie\t20")
+        XCTAssertEqual(csv.serialized, "id\t\"the\t name\"\tage\n1\t\"Alice\t In\t Wonderland\"\t18\n2\tBob\t19\n3\tCharlie\t20")
+
+        // Tab-separated values with explicit alternate delimiter (tab) and embedded default delimiters (commas) in cells.
+        csv = try CSV<Enumerated>(string: "id\tthe, name,age\n1\tAlice, In, Wonderland\t18\n2\tBob\t19\n3\tCharlie\t20", delimiter: .tab)
+        XCTAssertEqual(csv.serialized, "id\tthe, name,age\n1\tAlice, In, Wonderland\t18\n2\tBob\t19\n3\tCharlie\t20")
     }
 }