fix

2025-08-05 00:10:46 +08:00
parent 262abe88c6
commit fb99ed78e3
5 changed files with 72 additions and 46 deletions
--- a/docutranslate/agents/segments_agent.py
+++ b/docutranslate/agents/segments_agent.py
@@ -1,4 +1,5 @@
 import json
+from json_repair import json_repair
 from dataclasses import dataclass
 from json import JSONDecodeError

@@ -28,12 +29,18 @@ class SegmentsTranslateAgent(Agent):
 翻译后的片段应该与源格式尽量相同
 如果待翻译片段已经是目标语言，则保持原样
 # 输出
-翻译后的片段序列，以json文本表示（文本而非代码块）。其中键是片段编号，值是翻译后的片段
+翻译后的片段序列，以json文本表示（注意不是代码块）。其中键是片段编号，值是翻译后的片段。
+返回的json文本必须能被json.loads转换为形如{{"片段编号":"译文"}}的字典。
 # 示例
 ## 输入
 {r'{"0":"hello","1":"apple","2":true,"3":"false"}'}
 ## 输出
 {r'{"0":"你好","1":"苹果","2":true,"3":"错误"}'}
+# 错误输出示例
+"{r'"{\"0\":\"你好\",\"1\":\"苹果\"}"'}"
+# 正确输出示例
+{r'{"0":"你好","1":"苹果"}'}
+警告：绝不要将整个JSON对象用引号包裹成一个字符串。
 """
        if config.custom_prompt:
            self.system_prompt += "\n# 重要规则或背景【非常重要】\n" + config.custom_prompt + '\n'
@@ -41,26 +48,40 @@ class SegmentsTranslateAgent(Agent):
    def send_segments(self, segments: list[str], chunk_size: int):
        indexed_originals = {str(i): text for i, text in enumerate(segments)}
        chunks = flat_json_split(indexed_originals, chunk_size)
-        prompts = [json.dumps(chunk) for chunk in chunks]
+        prompts = [json.dumps(chunk,ensure_ascii=False) for chunk in chunks]
        translated_chunks = super().send_prompts(prompts=prompts)
        indexed_translated = indexed_originals.copy()
        for chunk_str in translated_chunks:
-            translated_part = json.loads(chunk_str)
-            indexed_translated.update(translated_part)
+            try:
+                translated_part = json_repair.loads(chunk_str)
+                for key,val in translated_part:
+                    if key in indexed_translated:
+                        indexed_translated[key]=val
+            except JSONDecodeError as e:
+                self.logger.info(f"json解析错误，解析文本:{chunk_str}，错误:{e.__repr__()}")
+            except ValueError as e:
+                self.logger.info(f"value错误，更新对象:{indexed_translated}，错误:{e.__repr__()}")
+
        return list(indexed_translated.values())

    #todo:增加协程粒度
    async def send_segments_async(self, segments: list[str], chunk_size: int):
        indexed_originals = {str(i): text for i, text in enumerate(segments)}
        chunks = flat_json_split(indexed_originals, chunk_size)
-        prompts = [json.dumps(chunk) for chunk in chunks]
+        prompts = [json.dumps(chunk,ensure_ascii=False) for chunk in chunks]
        translated_chunks = await super().send_prompts_async(prompts=prompts)
        indexed_translated = indexed_originals.copy()
        for chunk_str in translated_chunks:
            try:
-                translated_part = json.loads(chunk_str)
-                indexed_translated.update(translated_part)
+                translated_part:dict = json_repair.loads(chunk_str)
+                for key,val in translated_part.items():
+                    if key in indexed_translated:
+                        indexed_translated[key]=val
            except JSONDecodeError as e:
-                self.logger.info(f"json解析错误，解析文本:{chunk_str}，错误:{e.__repr__()}")
+                self.logger.error(f"json解析错误，解析文本:{chunk_str}，错误:{e.__repr__()}")
+            except ValueError as e:
+                self.logger.error(f"value错误，更新对象:{indexed_translated}，错误:{e.__repr__()}")
+            except AttributeError as e:
+                self.logger.error(f"属性错误,chunk_str:{chunk_str}，错误:{e.__repr__()}")

        return list(indexed_translated.values())
--- a/docutranslate/exporter/xlsx/xlsx2html_exporter.py
+++ b/docutranslate/exporter/xlsx/xlsx2html_exporter.py
@@ -3,6 +3,7 @@ from io import BytesIO

 import jinja2
 import openpyxl
+from xlsx2html import xlsx2html

 from docutranslate.exporter.base import ExporterConfig
 from docutranslate.exporter.xlsx.base import XlsxExporter
@@ -22,40 +23,5 @@ class Xlsx2HTMLExporter(XlsxExporter):
        self.cdn = config.cdn

    def export(self, document: Document) -> Document:
-
-        # 1. 加载工作簿和工作表
-        workbook = openpyxl.load_workbook(BytesIO(document.content))
-        sheet = workbook.active
-
-        # 2. 手动构建HTML字符串
-        table = '<table border="1">\n'
-
-        # 处理表头
-        table += '  <thead>\n    <tr>\n'
-        for cell in sheet[1]:  # 假设第一行是表头
-            table += f'      <th>{cell.value}</th>\n'
-        table += '    </tr>\n  </thead>\n'
-
-        # 处理数据行
-        table += '  <tbody>\n'
-        # iter_rows(min_row=2) 从第二行开始遍历
-        for row in sheet.iter_rows(min_row=2):
-            table += '    <tr>\n'
-            for cell in row:
-                # 处理None值，防止在HTML中显示"None"
-                cell_value = cell.value if cell.value is not None else ""
-                table += f'      <td>{cell_value}</td>\n'
-            table += '    </tr>\n'
-        table += '  </tbody>\n'
-
-        table += '</table>'
-
-        html_template = resource_path("template/xlsx.html").read_text(encoding="utf-8")
-
-        pico = f'<style>{resource_path("static/pico.css").read_text(encoding="utf-8")}</style>' if not self.cdn else r'<link rel="stylesheet" href="https://s4.zstatic.net/ajax/libs/picocss/2.1.1/pico.min.css" integrity="sha512-+4kjFgVD0n6H3xt19Ox84B56MoS7srFn60tgdWFuO4hemtjhySKyW4LnftYZn46k3THUEiTTsbVjrHai+0MOFw==" crossorigin="anonymous" referrerpolicy="no-referrer" />'
-        render = jinja2.Template(html_template).render(
-            title=document.stem,
-            pico=pico,
-            body=table,
-        )
-        return Document.from_bytes(content=render.encode("utf-8"), suffix=".html", stem=document.stem)
+        html_content = xlsx2html(BytesIO(document.content), output=None).getvalue()
+        return Document.from_bytes(content=html_content.encode("utf-8"), suffix=".html", stem=document.stem)
--- a/docutranslate/static/index.html
+++ b/docutranslate/static/index.html
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -9,6 +9,8 @@ dependencies = [
    "fastapi[standard]>=0.115.12",
    "jsonpath-ng>=1.7.0",
    "openpyxl>=3.1.5",
+    "xlsx2html>=0.6.2",
+    "json-repair>=0.48.0",
 ]
 dynamic = ["version"]

--- a/uv.lock
+++ b/uv.lock
@@ -51,6 +51,15 @@ wheels = [
    { url = "https://files.pythonhosted.org/packages/77/06/bb80f5f86020c4551da315d78b3ab75e8228f89f0162f2c3a819e407941a/attrs-25.3.0-py3-none-any.whl", hash = "sha256:427318ce031701fea540783410126f03899a97ffc6f61596ad581ac2e40e3bc3", size = 63815 },
 ]

+[[package]]
+name = "babel"
+version = "2.17.0"
+source = { registry = "https://pypi.org/simple" }
+sdist = { url = "https://files.pythonhosted.org/packages/7d/6b/d52e42361e1aa00709585ecc30b3f9684b3ab62530771402248b1b1d6240/babel-2.17.0.tar.gz", hash = "sha256:0c54cffb19f690cdcc52a3b50bcbf71e07a808d1c80d549f2459b9d2cf0afb9d", size = 9951852 }
+wheels = [
+    { url = "https://files.pythonhosted.org/packages/b7/b8/3fe70c75fe32afc4bb507f75563d39bc5642255d1d94f1f23604725780bf/babel-2.17.0-py3-none-any.whl", hash = "sha256:4d0b53093fdfb4b21c92b5213dba5a1b23885afa8383709427046b21c366e5f2", size = 10182537 },
+]
+
 [[package]]
 name = "beautifulsoup4"
 version = "4.13.4"
@@ -283,9 +292,11 @@ source = { editable = "." }
 dependencies = [
    { name = "fastapi", extra = ["standard"] },
    { name = "httpx" },
+    { name = "json-repair" },
    { name = "jsonpath-ng" },
    { name = "markdown2" },
    { name = "openpyxl" },
+    { name = "xlsx2html" },
 ]

 [package.optional-dependencies]
@@ -306,10 +317,12 @@ requires-dist = [
    { name = "docling", marker = "extra == 'docling'", specifier = ">=2.40.0" },
    { name = "fastapi", extras = ["standard"], specifier = ">=0.115.12" },
    { name = "httpx", specifier = "==0.27.2" },
+    { name = "json-repair", specifier = ">=0.48.0" },
    { name = "jsonpath-ng", specifier = ">=1.7.0" },
    { name = "markdown2", specifier = ">=2.5.3" },
    { name = "opencv-python", marker = "extra == 'docling'", specifier = ">=4.11.0.86" },
    { name = "openpyxl", specifier = ">=3.1.5" },
+    { name = "xlsx2html", specifier = ">=0.6.2" },
 ]
 provides-extras = ["docling"]

@@ -553,6 +566,15 @@ wheels = [
    { url = "https://files.pythonhosted.org/packages/62/a1/3d680cbfd5f4b8f15abc1d571870c5fc3e594bb582bc3b64ea099db13e56/jinja2-3.1.6-py3-none-any.whl", hash = "sha256:85ece4451f492d0c13c5dd7c13a64681a86afae63a5f347908daf103ce6d2f67", size = 134899 },
 ]

+[[package]]
+name = "json-repair"
+version = "0.48.0"
+source = { registry = "https://pypi.org/simple" }
+sdist = { url = "https://files.pythonhosted.org/packages/fc/b5/dd0e703abd6f69507c7ec0494c4d0bf5ecefaabaa454801bebcc8b80ff73/json_repair-0.48.0.tar.gz", hash = "sha256:030f826e6867dbc465be7163dfc23458c0776002c0878d239b29136cd2ae8f39", size = 34736 }
+wheels = [
+    { url = "https://files.pythonhosted.org/packages/ba/e9/22315fd481ed9dc007646181b8c2ebf8b63cd512e59c750d03d50a9ed838/json_repair-0.48.0-py3-none-any.whl", hash = "sha256:c3eb34518c39a7a58d963dbbbda8cdb44e16d819169a85d6d1882f7d2fa24774", size = 26354 },
+]
+
 [[package]]
 name = "jsonlines"
 version = "3.1.0"
@@ -2406,6 +2428,21 @@ wheels = [
    { url = "https://files.pythonhosted.org/packages/fa/a8/5b41e0da817d64113292ab1f8247140aac61cbf6cfd085d6a0fa77f4984f/websockets-15.0.1-py3-none-any.whl", hash = "sha256:f7a866fbc1e97b5c617ee4116daaa09b722101d4a3c170c787450ba409f9736f", size = 169743 },
 ]

+[[package]]
+name = "xlsx2html"
+version = "0.6.2"
+source = { registry = "https://pypi.org/simple" }
+dependencies = [
+    { name = "babel" },
+    { name = "openpyxl" },
+    { name = "packaging" },
+    { name = "six" },
+]
+sdist = { url = "https://files.pythonhosted.org/packages/c6/1b/770c7c3d73b7dd0309f8aa640be19782f6762f03caa871edcb8c3fbd0ae6/xlsx2html-0.6.2.tar.gz", hash = "sha256:e3be926dca7c3217eabe6b9e4e50447b1c65cd2a16a711e395f5e9a1ac18ce9e", size = 19638 }
+wheels = [
+    { url = "https://files.pythonhosted.org/packages/0c/26/a43d76e5ffddb1c2d958915fec5d99552a80559e6403ee3f0dd3db28fdea/xlsx2html-0.6.2-py2.py3-none-any.whl", hash = "sha256:d10d6c18be2e563a9fe8aee5dcadb2140be38e1090db717ede1b8ff09648452a", size = 17547 },
+]
+
 [[package]]
 name = "xlsxwriter"
 version = "3.2.3"