gpt-po 1.2.1 → 1.2.4

package/README.md

@@ -19,6 +19,11 @@ Set `OPENAI_API_KEY` before using this tool.
 
 **It is recommended to use the paid OpenAI API to improve translation speed, as the free OpenAI API is slower (only 3 translations per minute) and has usage restrictions.**
 
+### Environment Variables
+- `OPENAI_API_KEY`: OpenAI API key.
+- `OPENAI_API_HOST`: OpenAI API host (default: https://api.openai.com).
+- `OPENAI_MODEL_TMP`: OpenAI model temperature (default: 0.1).
+
 ## Usage Scenarios
 
 - `gpt-po sync --po <file> --pot <file>` Update the po file based on the pot file, while preserving the original translations.
@@ -60,7 +65,9 @@ Options:
   --verbose              show verbose log
   -l, --lang <lang>      target language (ISO 639-1 code)
   -o, --output <file>    output file path, overwirte po file by default
-  --context              context file path (provides additional context to the bot)
+  --context <file>       context file path (provides additional context to the bot)
+  --context-length <length>  Maximum accumulated length of source strings (msgid) to translate in each API request (default: 2000, env: API_CONTEXT_LENGTH)
+  --timeout <ms>         Timeout in milliseconds for API requests (default: 20000, env: API_TIMEOUT)
   -h, --help             display help for command
 ```
 
@@ -90,4 +97,4 @@ Options:
   --po <file>   po file path
   --pot <file>  pot file path
   -h, --help    display help for command
-```
+```

package/README_zh-CN.md

@@ -21,6 +21,11 @@ npm install gpt-po
 
 *国内用户要设置`HTTPS_PROXY`环境变量上梯子才能用*
 
+### 环境变量
+- `OPENAI_API_KEY`: OpenAI API密钥。
+- `OPENAI_API_HOST`: OpenAI API主机（默认：https://api.openai.com）。
+- `OPENAI_MODEL_TMP`: OpenAI模型温度（默认：0.1）。
+
 ## 常见用法
 
 - `gpt-po sync --po <file> --pot <file>` 根据 pot 文件更新 po 文件，同时保留原有翻译。
@@ -62,7 +67,9 @@ npm install gpt-po
   --verbose              显示详细日志
   -l, --lang <lang>      目标语言 (ISO 639-1 代码)
   -o, --output <file>    输出文件路径，默认覆盖 po 文件
-  --context              上下文文件路径（为机器人提供额外的上下文）
+  --context <file>       上下文文件路径（为机器人提供额外的上下文）
+  --context-length <length>  每次 API 请求中源字符串（msgid）的最大累计长度（默认：2000，环境变量：API_CONTEXT_LENGTH）
+  --timeout <ms>         API 请求超时时间（单位：毫秒，默认：20000，环境变量：API_TIMEOUT）
   -h, --help             显示命令帮助
 ```
 

package/lib/package.json

@@ -1,6 +1,6 @@
 {
     "name": "gpt-po",
-    "version": "1.2.1",
+    "version": "1.2.4",
     "description": "command tool for translate po files by gpt",
     "main": "lib/src/index.js",
     "type": "module",

package/lib/src/index.d.ts

@@ -1,2 +1,2 @@
-#!/usr/bin/env node --no-warnings=ExperimentalWarning
+#!/usr/bin/env -S node --no-warnings=ExperimentalWarning
 export {};

package/lib/src/index.js

@@ -1,4 +1,4 @@
-#!/usr/bin/env node --no-warnings=ExperimentalWarning
+#!/usr/bin/env -S node --no-warnings=ExperimentalWarning
 import { Command, Option } from "commander";
 import path from "path";
 import { fileURLToPath } from "url";
@@ -39,10 +39,12 @@ const translateCommand = new SharedOptionsCommand("translate")
     .option("-l, --lang <lang>", "target language (ISO 639-1)")
     .option("--verbose", "print verbose log")
     .option("--context <file>", "text file that provides the bot additional context")
+    .addOption(new Option("--context-length <length>", "maximum accumulated length of source strings (msgid) to translate in each API request").env("API_CONTEXT_LENGTH").default("2000"))
+    .addOption(new Option("--timeout <ms>", "timeout in milliseconds for API requests").env("API_TIMEOUT").default("20000"))
     .addOption(new Option("-o, --output <file>", "output file path, overwirte po file by default").conflicts("dir"))
     .addCompileOptions()
     .action(async (args) => {
-    const { key, host, model, po, dir, source, lang, verbose, output, context } = args;
+    const { key, host, model, po, dir, source, lang, verbose, output, context, contextLength, timeout } = args;
     if (host) {
         process.env.OPENAI_API_HOST = host;
     }
@@ -57,10 +59,10 @@ const translateCommand = new SharedOptionsCommand("translate")
     init();
     const compileOptions = getCompileOptions(args);
     if (po) {
-        await translatePo(model, po, source, lang, verbose, output, context, compileOptions);
+        await translatePo(model, po, source, lang, verbose, output, context, parseInt(contextLength), parseInt(timeout), compileOptions);
     }
     else if (dir) {
-        await translatePoDir(model, dir, source, lang, verbose, context, compileOptions);
+        await translatePoDir(model, dir, source, lang, verbose, context, parseInt(contextLength), parseInt(timeout), compileOptions);
     }
     else {
         console.error("po file or directory is required");

package/lib/src/translate.d.ts

@@ -2,6 +2,6 @@ import { GetTextTranslation } from "gettext-parser";
 import { OpenAI } from "openai";
 import { CompileOptions } from "./utils.js";
 export declare function init(force?: boolean): OpenAI;
-export declare function translate(src: string, lang: string, model: string, translations: GetTextTranslation[], contextFile: string): Promise<void>;
-export declare function translatePo(model: string, po: string, source: string, lang: string, verbose: boolean, output: string, contextFile: string, compileOptions?: CompileOptions): Promise<void>;
-export declare function translatePoDir(model: string | undefined, dir: string, source: string, lang: string, verbose: boolean, contextFile: string, compileOptions?: CompileOptions): Promise<void>;
+export declare function translate(src: string, lang: string, model: string, translations: GetTextTranslation[], contextFile: string, timeout: number): Promise<void>;
+export declare function translatePo(model: string, po: string, source: string, lang: string, verbose: boolean, output: string, contextFile: string, contextLength: number, timeout: number, compileOptions?: CompileOptions): Promise<void>;
+export declare function translatePoDir(model: string | undefined, dir: string, source: string, lang: string, verbose: boolean, contextFile: string, contextLength: number, timeout: number, compileOptions?: CompileOptions): Promise<void>;

package/lib/src/translate.js

@@ -36,7 +36,7 @@ export function init(force) {
     }
     return _openai;
 }
-export async function translate(src, lang, model, translations, contextFile) {
+export async function translate(src, lang, model, translations, contextFile, timeout) {
     const lang_code = lang
         .toLowerCase()
         .trim()
@@ -58,11 +58,14 @@ export async function translate(src, lang, model, translations, contextFile) {
         .join("\n");
     const context = contextFile ? "\n\nContext: " + fs.readFileSync(contextFile, "utf-8") : "";
     const translationsContent = translations
-        .map((tr, idx) => `<translate index="${idx + dicts.user.length + 1}">${tr.msgid}</translate>`)
+        .map((tr, idx) => {
+        const contextAttr = tr.msgctxt ? ` context="${tr.msgctxt}"` : "";
+        return `<translate index="${idx + dicts.user.length + 1}"${contextAttr}>${tr.msgid}</translate>`;
+    })
         .join("\n");
     const res = await _openai.chat.completions.create({
         model: model,
-        temperature: 0.1,
+        temperature: process.env.OPENAI_MODEL_TMP ? parseFloat(process.env.OPENAI_MODEL_TMP) : 0.1,
         messages: [
             {
                 role: "system",
@@ -70,12 +73,12 @@ export async function translate(src, lang, model, translations, contextFile) {
             },
             {
                 role: "user",
-                content: `${_userprompt}\n\nWait for my incoming message in "${src}" and translate it into "${lang}"(a language code and an optional region code). ` +
+                content: `${_userprompt}\n\nWait for my incoming message(s) in \`${src}\` and translate them into \`${lang}\` (\`${src}\` and \`${lang}\` are XPG/POSIX locale names, used in Unix-like systems and GNU Gettext). ` +
                     notes
             },
             {
                 role: "assistant",
-                content: `Understood, I will translate your incoming "${src}" message into "${lang}", carefully following guidelines. Please go ahead and send your message for translation.`
+                content: `Understood, I will translate your incoming \`${src}\` message(s) into \`${lang}\`, carefully following guidelines. Please go ahead and send your message(s) for translation.`
             },
             // add userdict
             ...(dicts.user.length > 0
@@ -91,7 +94,7 @@ export async function translate(src, lang, model, translations, contextFile) {
             }
         ]
     }, {
-        timeout: 20000,
+        timeout,
         stream: false
     });
     const content = res.choices[0].message.content ?? "";
@@ -107,7 +110,7 @@ export async function translate(src, lang, model, translations, contextFile) {
         }
     });
 }
-export async function translatePo(model, po, source, lang, verbose, output, contextFile, compileOptions) {
+export async function translatePo(model, po, source, lang, verbose, output, contextFile, contextLength, timeout, compileOptions) {
     const potrans = await parsePo(po);
     if (!lang)
         lang = potrans.headers["Language"];
@@ -132,11 +135,16 @@ export async function translatePo(model, po, source, lang, verbose, output, cont
     let trimed = false;
     for (const [ctx, entries] of Object.entries(potrans.translations)) {
         for (const [msgid, trans] of Object.entries(entries)) {
-            if (msgid == "")
+            if (msgid === "")
                 continue;
             if (!trans.msgstr[0]) {
-                list.push(trans);
-                continue;
+                list.push({
+                    msgctxt: trans.msgctxt || ctx,
+                    msgid,
+                    msgid_plural: trans.msgid_plural,
+                    msgstr: trans.msgstr,
+                    comments: trans.comments
+                });
             }
             else if (trimRegx.test(trans.msgstr[0])) {
                 trimed = true;
@@ -162,13 +170,13 @@ export async function translatePo(model, po, source, lang, verbose, output, cont
             await new Promise((resolve) => setTimeout(resolve, 20000));
         }
         const trans = list[i];
-        if (c < 2000) {
+        if (c < contextLength) {
             translations.push(trans);
             c += trans.msgid.length;
         }
-        if (c >= 2000 || i == list.length - 1) {
+        if (c >= contextLength || i == list.length - 1) {
             try {
-                await translate(source, lang, model, translations, contextFile);
+                await translate(source, lang, model, translations, contextFile, timeout);
                 if (verbose) {
                     translations.forEach((trans) => {
                         console.log(trans.msgid);
@@ -179,7 +187,7 @@ export async function translatePo(model, po, source, lang, verbose, output, cont
                 c = 0;
                 // update progress
                 printProgress(i + 1, list.length);
-                // save po file after each 2000 characters
+                // save po file after each 2000 characters by default
                 await compilePo(potrans, output || po, compileOptions);
             }
             catch (error) {
@@ -205,13 +213,13 @@ export async function translatePo(model, po, source, lang, verbose, output, cont
     }
     console.log("done.");
 }
-export async function translatePoDir(model = "gpt-3.5-turbo", dir, source, lang, verbose, contextFile, compileOptions) {
+export async function translatePoDir(model = "gpt-3.5-turbo", dir, source, lang, verbose, contextFile, contextLength, timeout, compileOptions) {
     const files = fs.readdirSync(dir);
     for (const file of files) {
         if (file.endsWith(".po")) {
             const po = path.join(dir, file);
             console.log(`translating ${po}`);
-            await translatePo(model, po, source, lang, verbose, po, contextFile, compileOptions);
+            await translatePo(model, po, source, lang, verbose, po, contextFile, contextLength, timeout, compileOptions);
         }
     }
 }

package/lib/src/userprompt.txt

@@ -12,11 +12,18 @@ Translation guidelines are as follows:
    - Input messages will be wrapped in `<translate>` XML tags with an index attribute, e.g., `<translate index="1">`.
    - Respond with the translated message wrapped in `<translated>` XML tags, including the same index attribute, e.g., `<translated index="1">`.
 
-4. **Multiple Translations**:
+4. **Context Handling**:
+   - Some messages will include a context attribute in the translate tag, e.g., `<translate index="1" context="Menu">`.
+   - Use this context to inform your translation but only return the translated text.
+   - Example (translating from `en_GB` to `fr_FR`):
+     Input: `<translate index="1" context="Menu">File</translate>`
+     Output: `<translated index="1">Fichier</translated>`
+
+5. **Multiple Translations**:
    - You may receive multiple translation requests in a single input, each with a unique index.
    - Ensuring the complete number of requests are translated, even if they are repeated, while maintaining the original order when possible.
 
-**Examples**:
+**Example (translating from `en_US` to `zh_CN`):**
 - Input:
   `<translate index="1">This is a message. </translate>`
   `<translate index="2"> Hello %s</translate>`

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "gpt-po",
-  "version": "1.2.1",
+  "version": "1.2.4",
   "description": "command tool for translate po files by gpt",
   "main": "lib/src/index.js",
   "type": "module",