Fix the encoding for some missing chinese fonts

It fixes #20489.

Author: calixteman

src/core/evaluator.js

@@ -3588,7 +3588,7 @@ class PartialEvaluator {
     if (properties.composite) {
       // CIDSystemInfo helps to match CID to glyphs
       const cidSystemInfo = dict.get("CIDSystemInfo");
-      if (cidSystemInfo instanceof Dict) {
+      if (cidSystemInfo instanceof Dict && !properties.cidSystemInfo) {
         properties.cidSystemInfo = {
           registry: stringToPDFString(cidSystemInfo.get("Registry")),
           ordering: stringToPDFString(cidSystemInfo.get("Ordering")),
@@ -3670,6 +3670,51 @@ class PartialEvaluator {
       baseEncodingName = null;
     }
 
+    // Ignore incorrectly specified WinAnsiEncoding for non-embedded CJK fonts
+    // (fixes issue20489). Some chinese fonts often have WinAnsiEncoding in the
+    // PDF even though they should use Identity-H or GB-EUC-H encoding.
+    if (
+      baseEncodingName === "WinAnsiEncoding" &&
+      nonEmbeddedFont &&
+      properties.name?.charCodeAt(0) >= 0xb7
+    ) {
+      const fontName = properties.name;
+      // This list is built from some names from Pdfium and mupdf:
+      //  - https://pdfium.googlesource.com/pdfium/+/master/core/fpdfapi/font/cpdf_font.cpp#41
+      //  - https://fossies.org/linux/mupdf/source/pdf/pdf-font.c#l_820
+      const chineseFontNames = [
+        "\xCB\xCE\xCC\xE5", // SimSun
+        "\xBA\xDA\xCC\xE5", // SimHei
+        "\xBF\xAC\xCC\xE5", // SimKai
+        "\xB7\xC2\xCB\xCE", // SimFang
+        "\xBF\xAC\xCC\xE5_GB2312", // SimKai
+        "\xB7\xC2\xCB\xCE_GB2312", // SimFang
+        "\xC1\xA5\xCA\xE9", // SimLi
+        "\xD0\xC2\xCB\xCE", // SimSun
+      ];
+
+      // Check for common Chinese font names and their GBK-encoded equivalents
+      // (which may appear as Latin-1 when incorrectly decoded).
+      if (chineseFontNames.includes(fontName)) {
+        baseEncodingName = null;
+        properties.defaultEncoding = "Adobe-GB1-UCS2";
+        properties.composite = true;
+        properties.cidEncoding = Name.get("GBK-EUC-H");
+        const cMap = await CMapFactory.create({
+          encoding: properties.cidEncoding,
+          fetchBuiltInCMap: this._fetchBuiltInCMapBound,
+          useCMap: null,
+        });
+        properties.cMap = cMap;
+        properties.vertical = properties.cMap.vertical;
+        properties.cidSystemInfo = {
+          registry: "Adobe",
+          ordering: "GB1",
+          supplement: 0,
+        };
+      }
+    }
+
     if (baseEncodingName) {
       properties.defaultEncoding = getEncoding(baseEncodingName);
     } else {

src/core/font_substitutions.js

@@ -306,6 +306,64 @@ const substitutionMap = new Map([
       alias: "Wingdings",
     },
   ],
+  [
+    "\xCB\xCE\xCC\xE5",
+    {
+      local: ["SimSun", "SimSun Regular", "NSimSun"],
+      style: NORMAL,
+      ultimate: "serif",
+    },
+  ],
+  [
+    "\xBA\xDA\xCC\xE5",
+    {
+      local: ["SimHei", "SimHei Regular"],
+      style: NORMAL,
+      ultimate: "sans-serif",
+    },
+  ],
+  [
+    "\xBF\xAC\xCC\xE5",
+    {
+      local: ["KaiTi", "SimKai", "SimKai Regular"],
+      style: NORMAL,
+      ultimate: "sans-serif",
+    },
+  ],
+  [
+    "\xB7\xC2\xCB\xCE",
+    {
+      local: ["FangSong", "SimFang", "SimFang Regular"],
+      style: NORMAL,
+      ultimate: "serif",
+    },
+  ],
+  [
+    "\xBF\xAC\xCC\xE5_GB2312",
+    {
+      alias: "\xBF\xAC\xCC\xE5",
+    },
+  ],
+  [
+    "\xB7\xC2\xCB\xCE_GB2312",
+    {
+      alias: "\xB7\xC2\xCB\xCE",
+    },
+  ],
+  [
+    "\xC1\xA5\xCA\xE9",
+    {
+      local: ["SimLi", "SimLi Regular"],
+      style: NORMAL,
+      ultimate: "serif",
+    },
+  ],
+  [
+    "\xD0\xC2\xCB\xCE",
+    {
+      alias: "\xCB\xCE\xCC\xE5",
+    },
+  ],
 ]);
 
 const fontAliases = new Map([["Arial-Black", "ArialBlack"]]);

test/pdfs/.gitignore

@@ -767,3 +767,4 @@
 !issue20225.pdf
 !issue20513.pdf
 !issue20516.pdf
+!issue20489.pdf

test/pdfs/issue20489.pdf

@@ -13106,5 +13106,12 @@
     "md5": "19a3a347773518242fa3cf1c04a9a1e4",
     "rounds": 1,
     "type": "eq"
+  },
+  {
+    "id": "issue20489",
+    "file": "pdfs/issue20489.pdf",
+    "md5": "b85c798b9a4cc2cd4337d335321cc612",
+    "rounds": 1,
+    "type": "eq"
   }
 ]