| # copied/adapted from oriingal docker file here: https://github.com/NeotomaDB/MetaExtractor/tree/main/docker/data-review-tool | |
| FROM python:3.10 | |
| RUN git clone https://github.com/NeotomaDB/MetaExtractor.git | |
| # Copy the requirements.txt file to the working directory | |
| # COPY ./requirements.txt . | |
| # Install the Python dependencies | |
| RUN pip install --no-cache-dir -r MetaExtractor/docker/data-review-tool/requirements.txt | |
| WORKDIR MetaExtractor/ | |
| ENV LOG_LEVEL=DEBUG | |
| # Expose the port your Dash app is running on | |
| # changed from default 8050 to match HF spaces requirements | |
| # edit the port in the app.py file | |
| # inspired from: https://stackoverflow.com/questions/27713362/editing-files-from-dockerfile | |
| RUN sed -i "s|port=8050|port=7860 |g" ./src/data_review_tool/app.py | |
| EXPOSE 7860 | |
| RUN pip install pyarrow | |
| RUN mkdir -p ./inputs | |
| # set defualt env variables to sample files in the repo | |
| ENV ARTICLE_RELEVANCE_BATCH=article-relevance-output.parquet | |
| ENV ENTITY_EXTRACTION_BATCH=entity-extraction-output.zip | |
| # instead of vlume mapping copy sample files to expected input folder in regular image | |
| RUN cp ./data/data-review-tool/article-relevance-output.parquet ./inputs \ | |
| && cp ./data/data-review-tool/entity-extraction-output.zip ./inputs | |
| ENTRYPOINT python src/data_review_tool/app.py |