Refactoring PoParser

bun.lock

@@ -220,6 +220,7 @@
         "open": "^10.1.0",
         "php-parser": "^3.1.5",
         "plist": "^3.1.0",
+        "pofile": "^1.1.4",
         "preferred-pm": "^4.1.1",
         "properties": "^1.2.1",
         "rambda": "^9.4.2",
@@ -2916,6 +2917,8 @@
 
     "pngjs": ["[email protected]", "", {}, "sha512-NCrCHhWmnQklfH4MtJMRjZ2a8c80qXeMlQMv2uVp9ISJMTt562SbGd6n2oq0PaPgKm7Z6pL9E2UlLIhC+SHL3w=="],
 
+    "pofile": ["[email protected]", "", {}, "sha512-r6Q21sKsY1AjTVVjOuU02VYKVNQGJNQHjTIvs4dEbeuuYfxgYk/DGD2mqqq4RDaVkwdSq0VEtmQUOPe/wH8X3g=="],
+
     "possible-typed-array-names": ["[email protected]", "", {}, "sha512-d7Uw+eZoloe0EHDIYoe+bQ5WXnGMOpmiZFTuMWCwpjzzkL2nTjcKiAk4hh8TjnGye2TwWOk3UXucZ+3rbmBa8Q=="],
 
     "postcss": ["[email protected]", "", { "dependencies": { "nanoid": "^3.3.8", "picocolors": "^1.1.1", "source-map-js": "^1.2.1" } }, "sha512-6oz2beyjc5VMn/KV1pPw8fliQkhBXrVn1Z3TVyqZxU8kZpzEKhBdmCFqI6ZbmGtamQvQGuU1sgPTk8ZrXDD7jQ=="],

packages/cli/package.json

@@ -42,6 +42,7 @@
     "open": "^10.1.0",
     "php-parser": "^3.1.5",
     "plist": "^3.1.0",
+    "pofile": "^1.1.4",
     "preferred-pm": "^4.1.1",
     "properties": "^1.2.1",
     "rambda": "^9.4.2",

packages/cli/src/parsers/__tests__/po.test.ts

@@ -58,10 +58,14 @@ msgstr ""
       const input = `
 msgid "with_quotes"
 msgstr "text with "quotes" inside"
+
+msgid "with_escaped_quotes"
+msgstr "text with escaped \\"quotes\\" inside"
 `;
       const result = await parser.parse(input);
       expect(result).toEqual({
         with_quotes: 'text with "quotes" inside',
+        with_escaped_quotes: 'text with escaped "quotes" inside',
       });
     });
 
@@ -70,6 +74,37 @@ msgstr "text with "quotes" inside"
       const result = await parser.parse(input);
       expect(result).toEqual({});
     });
+
+    test("handles headers in input", async () => {
+      const input = `
+"POT-Creation-Date: 2025-07-11 19:57+0900\n"
+"MIME-Version: 1.0\n"
+"Content-Type: text/plain; charset=utf-8\n"
+"Content-Transfer-Encoding: 8bit\n"
+"X-Generator: @lingui/cli\n"
+"Language: ko\n"
+"Project-Id-Version: \n"
+"Report-Msgid-Bugs-To: \n"
+"PO-Revision-Date: \n"
+"Last-Translator: \n"
+"Language-Team: \n"
+"Plural-Forms: \n"
+
+# This is a comment
+#: another comment
+msgid "key"
+msgstr "value"
+`;
+      const result = await parser.parse(input);
+      expect(result).toEqual({
+        key: "value",
+      });
+
+      const serialized = await parser.serialize("ko", result);
+      expect(serialized).toBe(
+        'msgid ""\nmsgstr ""\n"POT-Creation-Date: 2025-07-11 19:57+0900\\n"\n": \\n"\n"MIME-Version: 1.0\\n"\n": \\n"\n"Content-Type: text/plain; charset=utf-8\\n"\n": \\n"\n"Content-Transfer-Encoding: 8bit\\n"\n": \\n"\n"X-Generator: @lingui/cli\\n"\n": \\n"\n"Language: ko\\n"\n": \\n"\n"Project-Id-Version: \\n"\n": \\n"\n"Report-Msgid-Bugs-To: \\n"\n": \\n"\n"PO-Revision-Date: \\n"\n": \\n"\n"Last-Translator: \\n"\n": \\n"\n"Language-Team: \\n"\n": \\n"\n"Plural-Forms: \\n"\n": \\n"\n\n# This is a comment\n#: another comment\nmsgid "key"\nmsgstr "value"\n',
+      );
+    });
   });
 
   describe("serialize", () => {
@@ -80,7 +115,7 @@ msgstr "text with "quotes" inside"
       };
       const result = await parser.serialize("en", input);
       expect(result).toBe(
-        'msgid "hello"\nmsgstr "world"\n\nmsgid "test"\nmsgstr "value"\n',
+        'msgid ""\nmsgstr ""\n\nmsgid "hello"\nmsgstr "world"\n\nmsgid "test"\nmsgstr "value"\n',
       );
     });
 
@@ -89,7 +124,7 @@ msgstr "text with "quotes" inside"
         empty: "",
       };
       const result = await parser.serialize("en", input);
-      expect(result).toBe('msgid "empty"\nmsgstr ""\n');
+      expect(result).toBe('msgid ""\nmsgstr ""\n\nmsgid "empty"\nmsgstr ""\n');
     });
 
     test("serializes translations with quotes", async () => {
@@ -98,14 +133,14 @@ msgstr "text with "quotes" inside"
       };
       const result = await parser.serialize("en", input);
       expect(result).toBe(
-        'msgid "with_quotes"\nmsgstr "text with "quotes" inside"\n',
+        'msgid ""\nmsgstr ""\n\nmsgid "with_quotes"\nmsgstr "text with \\"quotes\\" inside"\n',
       );
     });
 
     test("handles empty object", async () => {
       const input = {};
       const result = await parser.serialize("en", input);
-      expect(result).toBe("");
+      expect(result).toBe('msgid ""\nmsgstr ""\n');
     });
 
     test("adds newline at end of file", async () => {
@@ -121,14 +156,14 @@ msgstr "text with "quotes" inside"
       };
       const result = await parser.serialize("en", translations);
       expect(result).toBe(
-        'msgid "hello"\nmsgstr "world"\n\nmsgid "keep"\nmsgstr "value"\n',
+        'msgid ""\nmsgstr ""\n\nmsgid "hello"\nmsgstr "world"\n\nmsgid "keep"\nmsgstr "value"\n',
       );
     });
 
     test("handles object with no translations", async () => {
       const translations = {};
       const result = await parser.serialize("en", translations);
-      expect(result).toBe("");
+      expect(result).toBe('msgid ""\nmsgstr ""\n');
     });
   });
 });

packages/cli/src/parsers/formats/po.ts

@@ -1,33 +1,22 @@
+import PO from "pofile";
 import { BaseParser } from "../core/base-parser.js";
 
 export class PoParser extends BaseParser {
+  #po: PO = null!;
+
   async parse(input: string) {
     try {
+      this.#po = PO.parse(input);
       const result: Record<string, string> = {};
-      const lines = input.split("\n");
-      let currentKey = "";
-      let currentValue = "";
-
-      for (const line of lines) {
-        const trimmed = line.trim();
 
-        if (isSkippableLine(trimmed)) {
+      for (const item of this.#po.items) {
+        if (!item.msgid) {
           continue;
         }
 
-        if (trimmed.startsWith("msgid")) {
-          if (currentKey) {
-            result[currentKey] = currentValue;
-          }
-          currentKey = parseMsgId(trimmed);
-          currentValue = "";
-        } else if (trimmed.startsWith("msgstr")) {
-          currentValue = parseMsgStr(trimmed);
-        }
-      }
+        const value = item.msgstr.at(0) || "";
 
-      if (currentKey) {
-        result[currentKey] = currentValue;
+        result[item.msgid] = value;
       }
 
       return result;
@@ -44,39 +33,38 @@ export class PoParser extends BaseParser {
     _originalData?: Record<string, string>,
   ): Promise<string> {
     try {
+      if (!this.#po) {
+        this.#po = new PO();
+      }
+
       if (Object.keys(data).length === 0) {
-        return "";
+        return this.#po.toString();
       }
 
-      const result = Object.entries(data)
-        .map(([key, value]) => {
-          return `msgid "${key}"\nmsgstr "${value}"`;
-        })
-        .join("\n\n");
+      const originalPoItems = Object.fromEntries(
+        this.#po.items.map((item) => [item.msgid, item]),
+      );
+
+      this.#po.items = Object.entries(data).map(([key, value]) => {
+        let item = originalPoItems[key];
 
-      return `${result}\n`;
+        if (!item) {
+          item = new PO.Item();
+          item.msgid = key;
+        }
+
+        if (value) {
+          item.msgstr = [value];
+        }
+
+        return item;
+      });
+
+      return this.#po.toString();
     } catch (error) {
       throw new Error(
         `Failed to serialize PO: ${error instanceof Error ? error.message : String(error)}`,
       );
     }
   }
 }
-
-function isSkippableLine(line: string): boolean {
-  return !line || line.startsWith("#");
-}
-
-function parseMsgId(line: string): string {
-  const match = line.match(/msgid "(.*)"/);
-  return match ? unescapeQuotes(match[1]) : "";
-}
-
-function parseMsgStr(line: string): string {
-  const match = line.match(/msgstr "(.*)"/);
-  return match ? unescapeQuotes(match[1]) : "";
-}
-
-function unescapeQuotes(str: string): string {
-  return str.replace(/\\"/g, '"');
-}