LMAssembly // Let LMInstantiator summarize user data.

Packages/vChewing_LangModelAssembly/Sources/LangModelAssembly/LMInstantiator_UserDictionarySummarized.swift

@@ -0,0 +1,33 @@
+// (c) 2021 and onwards The vChewing Project (MIT-NTL License).
+// ====================
+// This code is released under the MIT license (SPDX-License-Identifier: MIT)
+// ... with NTL restriction stating that:
+// No trademark license is granted to use the trade names, trademarks, service
+// marks, or product names of Contributor, except as required to fulfill notice
+// requirements defined in MIT License.
+
+import Foundation
+
+public extension LMAssembly {
+  struct UserDictionarySummarized: Codable {
+    let isCHS: Bool
+    let userPhrases: [String: [String]]
+    let filter: [String: [String]]
+    let userSymbols: [String: [String]]
+    let replacements: [String: String]
+    let associates: [String: [String]]
+  }
+}
+
+public extension LMAssembly.LMInstantiator {
+  func summarize(all: Bool) -> LMAssembly.UserDictionarySummarized {
+    LMAssembly.UserDictionarySummarized(
+      isCHS: isCHS,
+      userPhrases: lmUserPhrases.dictRepresented,
+      filter: lmFiltered.dictRepresented,
+      userSymbols: lmUserSymbols.dictRepresented,
+      replacements: lmReplacements.dictRepresented,
+      associates: all ? lmAssociates.dictRepresented : [:]
+    )
+  }
+}

Packages/vChewing_LangModelAssembly/Sources/LangModelAssembly/SubLMs/lmAssociates.swift

@@ -11,7 +11,7 @@ import Megrez
 extension LMAssembly {
   struct LMAssociates {
     public private(set) var filePath: String?
-    var rangeMap: [String: [(Range<String.Index>, Int)]] = [:]
+    var rangeMap: [String: [(Range<String.Index>, Int)]] = [:] // Range 只可能是一整行，所以必須得有 index。
     var strData: String = ""
 
     public var count: Int { rangeMap.count }
@@ -97,15 +97,9 @@ extension LMAssembly {
 
     public func valuesFor(pair: Megrez.KeyValuePaired) -> [String] {
       var pairs: [String] = []
-      if let arrRangeRecords: [(Range<String.Index>, Int)] = rangeMap[pair.toNGramKey] {
-        for (netaRange, index) in arrRangeRecords {
-          let neta = strData[netaRange].split(separator: " ")
-          let theValue: String = .init(neta[index])
-          pairs.append(theValue)
-        }
-      }
-      if let arrRangeRecords: [(Range<String.Index>, Int)] = rangeMap[pair.value] {
-        for (netaRange, index) in arrRangeRecords {
+      let availableResults = [rangeMap[pair.toNGramKey], rangeMap[pair.value]].compactMap { $0 }
+      availableResults.forEach { arrRangeRecords in
+        arrRangeRecords.forEach { netaRange, index in
           let neta = strData[netaRange].split(separator: " ")
           let theValue: String = .init(neta[index])
           pairs.append(theValue)
@@ -120,3 +114,17 @@ extension LMAssembly {
     }
   }
 }
+
+extension LMAssembly.LMAssociates {
+  var dictRepresented: [String: [String]] {
+    var result = [String: [String]]()
+    rangeMap.forEach { key, arrRangeRecords in
+      arrRangeRecords.forEach { netaRange, index in
+        let neta = strData[netaRange].split(separator: " ")
+        let theValue: String = .init(neta[index])
+        result[key, default: []].append(theValue)
+      }
+    }
+    return result
+  }
+}

Packages/vChewing_LangModelAssembly/Sources/LangModelAssembly/SubLMs/lmCoreEX.swift

@@ -184,3 +184,15 @@ extension LMAssembly {
     }
   }
 }
+
+extension LMAssembly.LMCoreEX {
+  var dictRepresented: [String: [String]] {
+    var result = [String: [String]]()
+    rangeMap.forEach { key, arrValueRanges in
+      result[key, default: []] = arrValueRanges.map { currentRange in
+        strData[currentRange].description
+      }
+    }
+    return result
+  }
+}

Packages/vChewing_LangModelAssembly/Sources/LangModelAssembly/SubLMs/lmReplacements.swift

@@ -98,3 +98,13 @@ extension LMAssembly {
     }
   }
 }
+
+extension LMAssembly.LMReplacements {
+  var dictRepresented: [String: String] {
+    var result = [String: String]()
+    rangeMap.forEach { key, valueRange in
+      result[key] = strData[valueRange].description
+    }
+    return result
+  }
+}