diff --git a/runtime/ops/formatter/file_exporter/process.py b/runtime/ops/formatter/file_exporter/process.py index 5b5b48b..c4934c4 100644 --- a/runtime/ops/formatter/file_exporter/process.py +++ b/runtime/ops/formatter/file_exporter/process.py @@ -23,6 +23,7 @@ class FileExporter(Mapper): super(FileExporter, self).__init__(*args, **kwargs) self.last_ops = True self.text_support_ext = kwargs.get("text_support_ext", ['txt', 'html', 'md', 'markdown', + 'xlsx', 'xls', 'csv', 'pptx', 'ppt', 'xml', 'json', 'doc', 'docx', 'pdf']) self.data_support_ext = kwargs.get("data_support_ext", ['jpg', 'jpeg', 'png', 'bmp']) self.medical_support_ext = kwargs.get("medical_support_ext", ['svs', 'tif', 'tiff']) diff --git a/runtime/ops/requirements.txt b/runtime/ops/requirements.txt index 8c842b0..b214fb7 100644 --- a/runtime/ops/requirements.txt +++ b/runtime/ops/requirements.txt @@ -19,4 +19,4 @@ xmltodict==1.0.2 zhconv==1.4.3 sqlalchemy==2.0.40 pymysql==1.1.1 -unstructured[docx]==0.18.15 \ No newline at end of file +unstructured[docx,csv,xlsx,pptx]==0.18.15 \ No newline at end of file diff --git a/scripts/images/runtime/Dockerfile b/scripts/images/runtime/Dockerfile index 92900df..86d08da 100644 --- a/scripts/images/runtime/Dockerfile +++ b/scripts/images/runtime/Dockerfile @@ -8,7 +8,7 @@ COPY scripts/images/runtime/start.sh /opt/runtime/start.sh ENV PYTHONPATH=/opt/runtime/datamate/ RUN apt update \ - && apt install -y libgl1 libglib2.0-0 vim poppler-utils tesseract-ocr tesseract-ocr-chi-sim libmagic1t64 libreoffice\ + && apt install -y libgl1 libglib2.0-0 vim libmagic1t64 libreoffice\ && apt clean \ && rm -rf /var/lib/apt/lists/*