small fix

This commit is contained in:
xunbu
2025-05-21 12:52:49 +08:00
parent 5e62fc1107
commit ba2b4abba5
5 changed files with 20 additions and 71 deletions

6
.idea/workspace.xml generated
View File

@@ -6,7 +6,9 @@
<component name="ChangeListManager"> <component name="ChangeListManager">
<list default="true" id="6b18b44a-df57-4212-a857-9e291ebe5dd2" name="更改" comment=""> <list default="true" id="6b18b44a-df57-4212-a857-9e291ebe5dd2" name="更改" comment="">
<change beforePath="$PROJECT_DIR$/.idea/workspace.xml" beforeDir="false" afterPath="$PROJECT_DIR$/.idea/workspace.xml" afterDir="false" /> <change beforePath="$PROJECT_DIR$/.idea/workspace.xml" beforeDir="false" afterPath="$PROJECT_DIR$/.idea/workspace.xml" afterDir="false" />
<change beforePath="$PROJECT_DIR$/docutranslate/app.py" beforeDir="false" afterPath="$PROJECT_DIR$/docutranslate/app.py" afterDir="false" />
<change beforePath="$PROJECT_DIR$/docutranslate/static/index.html" beforeDir="false" afterPath="$PROJECT_DIR$/docutranslate/static/index.html" afterDir="false" /> <change beforePath="$PROJECT_DIR$/docutranslate/static/index.html" beforeDir="false" afterPath="$PROJECT_DIR$/docutranslate/static/index.html" afterDir="false" />
<change beforePath="$PROJECT_DIR$/docutranslate/utils/convert.py" beforeDir="false" />
<change beforePath="$PROJECT_DIR$/pyproject.toml" beforeDir="false" afterPath="$PROJECT_DIR$/pyproject.toml" afterDir="false" /> <change beforePath="$PROJECT_DIR$/pyproject.toml" beforeDir="false" afterPath="$PROJECT_DIR$/pyproject.toml" afterDir="false" />
</list> </list>
<option name="SHOW_DIALOG" value="false" /> <option name="SHOW_DIALOG" value="false" />
@@ -602,7 +604,7 @@
<workItem from="1747739438735" duration="826000" /> <workItem from="1747739438735" duration="826000" />
<workItem from="1747740341909" duration="145000" /> <workItem from="1747740341909" duration="145000" />
<workItem from="1747752718385" duration="81000" /> <workItem from="1747752718385" duration="81000" />
<workItem from="1747754618316" duration="14549000" /> <workItem from="1747754618316" duration="16566000" />
</task> </task>
<servers /> <servers />
</component> </component>
@@ -621,7 +623,7 @@
</option> </option>
</component> </component>
<component name="com.intellij.coverage.CoverageDataManagerImpl"> <component name="com.intellij.coverage.CoverageDataManagerImpl">
<SUITE FILE_PATH="coverage/filetranslate$app_test__1_.coverage" NAME="app_test (1) 覆盖结果" MODIFIED="1747798850500" SOURCE_PROVIDER="com.intellij.coverage.DefaultCoverageFileProvider" RUNNER="coverage.py" COVERAGE_BY_TEST_ENABLED="false" COVERAGE_TRACING_ENABLED="false" WORKING_DIRECTORY="$PROJECT_DIR$/tests" /> <SUITE FILE_PATH="coverage/filetranslate$app_test__1_.coverage" NAME="app_test (1) 覆盖结果" MODIFIED="1747802494815" SOURCE_PROVIDER="com.intellij.coverage.DefaultCoverageFileProvider" RUNNER="coverage.py" COVERAGE_BY_TEST_ENABLED="false" COVERAGE_TRACING_ENABLED="false" WORKING_DIRECTORY="$PROJECT_DIR$/tests" />
<SUITE FILE_PATH="coverage/filetranslate$test.coverage" NAME="test 覆盖结果" MODIFIED="1747472297913" SOURCE_PROVIDER="com.intellij.coverage.DefaultCoverageFileProvider" RUNNER="coverage.py" COVERAGE_BY_TEST_ENABLED="false" COVERAGE_TRACING_ENABLED="false" WORKING_DIRECTORY="$PROJECT_DIR$/tests" /> <SUITE FILE_PATH="coverage/filetranslate$test.coverage" NAME="test 覆盖结果" MODIFIED="1747472297913" SOURCE_PROVIDER="com.intellij.coverage.DefaultCoverageFileProvider" RUNNER="coverage.py" COVERAGE_BY_TEST_ENABLED="false" COVERAGE_TRACING_ENABLED="false" WORKING_DIRECTORY="$PROJECT_DIR$/tests" />
<SUITE FILE_PATH="coverage/filetranslate$convert.coverage" NAME="convert 覆盖结果" MODIFIED="1746963490689" SOURCE_PROVIDER="com.intellij.coverage.DefaultCoverageFileProvider" RUNNER="coverage.py" COVERAGE_BY_TEST_ENABLED="false" COVERAGE_TRACING_ENABLED="false" WORKING_DIRECTORY="$PROJECT_DIR$/docutranslate/utils" /> <SUITE FILE_PATH="coverage/filetranslate$convert.coverage" NAME="convert 覆盖结果" MODIFIED="1746963490689" SOURCE_PROVIDER="com.intellij.coverage.DefaultCoverageFileProvider" RUNNER="coverage.py" COVERAGE_BY_TEST_ENABLED="false" COVERAGE_TRACING_ENABLED="false" WORKING_DIRECTORY="$PROJECT_DIR$/docutranslate/utils" />
<SUITE FILE_PATH="coverage/PDFtranslate$PDFtranslater__1_.coverage" NAME="PDFtranslater (1) 覆盖结果" MODIFIED="1746633258205" SOURCE_PROVIDER="com.intellij.coverage.DefaultCoverageFileProvider" RUNNER="coverage.py" COVERAGE_BY_TEST_ENABLED="false" COVERAGE_TRACING_ENABLED="false" WORKING_DIRECTORY="$PROJECT_DIR$/pdftranslate_packages" /> <SUITE FILE_PATH="coverage/PDFtranslate$PDFtranslater__1_.coverage" NAME="PDFtranslater (1) 覆盖结果" MODIFIED="1746633258205" SOURCE_PROVIDER="com.intellij.coverage.DefaultCoverageFileProvider" RUNNER="coverage.py" COVERAGE_BY_TEST_ENABLED="false" COVERAGE_TRACING_ENABLED="false" WORKING_DIRECTORY="$PROJECT_DIR$/pdftranslate_packages" />

View File

@@ -177,20 +177,18 @@ async def _perform_translation(params: Dict[str, Any], file_contents: bytes, ori
# --- API Endpoints --- # --- API Endpoints ---
@app.get("/", response_class=HTMLResponse) @app.get("/", response_class=HTMLResponse)
async def main_page(request: Request): async def main_page(request: Request):
# Serve index.html from the static directory or root project directory
# Assuming index.html is at the same level as app.py or in STATIC_DIR
# For simplicity, if index.html is at root:
# return FileResponse(Path(__file__).parent / "index.html")
# If using Jinja2Templates and index.html is in "templates" folder:
# return templates.TemplateResponse("index.html", {"request": request})
# Using FileResponse for index.html directly:
index_path = Path("index.html") # Adjust if index.html is elsewhere index_path = Path("index.html") # Adjust if index.html is elsewhere
if not index_path.exists(): if not index_path.exists():
# Fallback to static dir if not in root # Fallback to static dir if not in root
index_path = STATIC_DIR / "index.html" index_path = STATIC_DIR / "index.html"
if not index_path.exists(): if not index_path.exists():
raise HTTPException(status_code=404, detail="index.html not found") raise HTTPException(status_code=404, detail="index.html not found")
return FileResponse(index_path) no_cache_headers = {
"Cache-Control": "no-store, no-cache, must-revalidate, max-age=0",
"Pragma": "no-cache", # 兼容 HTTP/1.0
"Expires": "0", # 兼容旧版代理/缓存
}
return FileResponse(index_path,headers=no_cache_headers)
@app.post("/translate") @app.post("/translate")
@@ -408,7 +406,7 @@ async def download_html(filename_with_ext: str):
def run_app(): def run_app():
print("正在启动 DocuTranslate WebUI") print("正在启动 DocuTranslate WebUI")
print("请访问 http://127.0.0.1:8010") print("请访问 http://127.0.0.1:8010 ctrl+点击链接即可打开)")
uvicorn.run(app, host="127.0.0.1", port=8010, workers=1) uvicorn.run(app, host="127.0.0.1", port=8010, workers=1)

View File

@@ -229,7 +229,7 @@
<div class="form-group"> <div class="form-group">
<label for="convert_engin">转换引擎</label> <label for="convert_engin">转换引擎</label>
<select id="convert_engin" name="convert_engin"> <select id="convert_engin" name="convert_engin">
<option value="mineru" selected>Mineru</option> <option value="mineru" selected>minerU</option>
<option value="docling" id="docling">Docling</option> <option value="docling" id="docling">Docling</option>
</select> </select>
</div> </div>
@@ -360,6 +360,7 @@
return defaultValue; return defaultValue;
} }
} }
//api访问地址到获取地址的映射 //api访问地址到获取地址的映射
const apiHrefMap = { const apiHrefMap = {
"https://openrouter.ai/api/v1": "https://openrouter.ai/settings/keys", "https://openrouter.ai/api/v1": "https://openrouter.ai/settings/keys",
@@ -513,12 +514,14 @@
console.warn(`get engine list failed: ${response.status}`); console.warn(`get engine list failed: ${response.status}`);
return; return;
} }
const status = await response.json(); const enginList = await response.json();
statusMsg.textContent = '正在初始化'; statusMsg.textContent = '正在初始化';
status.forEach((engin) => { let options = convertEnginSelect.querySelectors(`option[value="${engin}"]`);
let option = convertEnginSelect.querySelector(`option[value="${engin}"]`); options.forEach((option) => {
option.disabled = true; if (!enginList.includes(option.value)) {
option.textContent += "(不可用)" option.disabled = true;
option.textContent += "(不可用)"
}
}) })
if (status.includes(convertEnginSelect.value)) { if (status.includes(convertEnginSelect.value)) {
convertEnginSelect.value = "mineru"; convertEnginSelect.value = "mineru";

View File

@@ -1,53 +0,0 @@
import os
from pathlib import Path
from docling.datamodel.base_models import InputFormat
from docling.datamodel.document import DocumentStream
from docling.datamodel.pipeline_options import PdfPipelineOptions, TableFormerMode
from docling.datamodel.settings import settings
from docling.document_converter import DocumentConverter, PdfFormatOption
from docling_core.types.doc import ImageRefMode
from huggingface_hub.errors import LocalEntryNotFoundError
from docutranslate.logger import translater_logger
IMAGE_RESOLUTION_SCALE = 4
def file2markdown_embed_images(file_path: Path | str | DocumentStream, formula=False, code=False,
artifacts_path: Path | str | None = None) -> str:
translater_logger.info(f"正在将文档转换为markdown")
pipeline_options = PdfPipelineOptions(artifacts_path=artifacts_path)
pipeline_options.do_ocr = False
pipeline_options.images_scale = IMAGE_RESOLUTION_SCALE
pipeline_options.generate_picture_images = True
# pipeline_options.table_structure_options.mode = TableFormerMode.FAST
pipeline_options.table_structure_options.do_cell_matching = False
if formula:
pipeline_options.do_formula_enrichment = True
if code:
pipeline_options.do_code_enrichment = True
# pipeline_options.accelerator_options= AcceleratorOptions(
# num_threads=4, device=AcceleratorDevice.AUTO
# )
# 打印时间
settings.debug.profile_pipeline_timings = True
converter = DocumentConverter(format_options={
InputFormat.PDF: PdfFormatOption(pipeline_options=pipeline_options)
})
try:
conversion_result = converter.convert(file_path)
result = conversion_result.document.export_to_markdown(image_mode=ImageRefMode.EMBEDDED)
except LocalEntryNotFoundError:
translater_logger.info(f"无法连接huggingface正在尝试换源")
os.environ['HF_ENDPOINT'] = 'https://hf-mirror.com'
conversion_result = converter.convert(file_path)
result = conversion_result.document.export_to_markdown(image_mode=ImageRefMode.EMBEDDED)
translater_logger.info(f"已转换为markdown")
translater_logger.info(f"pdf转换耗时: {conversion_result.timings["pipeline_total"].times}")
return result
if __name__ == '__main__':
pass

View File

@@ -8,7 +8,6 @@ dependencies = [
"httpx", "httpx",
"markdown2", "markdown2",
"fastapi[standard]>=0.115.12", "fastapi[standard]>=0.115.12",
"docling>=2.33.0",
] ]
[project.scripts] [project.scripts]
docutranslate = "docutranslate.cli:main" docutranslate = "docutranslate.cli:main"