Skip to content

Tools: Docker

Jump to bottom Edit New page
petermr edited this page Sep 17, 2020 · 16 revisions

Docker

what it is

An infrastructure that containerizes programs to make them easy to distribute and run.

public resources:

what we use it for

Wrapping complex programs and environment for distribution. Currently aimed at

  • getpapers to avoid installing Node
  • ami and amidict to distribute code, dictionaries, stylesheets etc.
  • KNIME to distribute precompiled workflows
  • bundling resources (e.g. articles, derived material, etc.)

installation of Docker

Platform problems

Windows Home

Problems have been reported on this, but not on Windows Pro

Docker Mentors

  • Anugrah?

installation of container

  • We can set up Docker Hub Automated Builds to make it easy to save, distribute and install the container images.
  • Any version or platform concerns?
  • Any resource constraints? (e.g. memory)
  • how to document and maintain?
  • how to distribute gigabyte files

projects

getpapers

Bundle:

  • Node / or installation script?
  • getpapers.js or installation script?

ami

AMI was Dockerised for the forest-plot subset. Cannot find code ATM

Also See https://github.com/petermr/ami3/blob/master/Dockerfile.

FROM maven:3-jdk-8 as builder

# Alternative: clone remote code > not used because Dockerfile lifes in code repo
# get and build ami3 (integrating https://github.com/petermr/cephis and https://github.com/petermr/normami)
# WORKDIR /app
# RUN git clone --depth 1 https://github.com/petermr/ami3.git && \

WORKDIR /app/ami3
COPY src src
COPY pom.xml pom.xml

RUN mvn -Dmaven.test.skip=true install

# unstaged build with just this PATH adjustment works!
ENV PATH /app/ami3/target/appassembler/bin/:${PATH}

# remove unused .bat files
WORKDIR /app/ami3/target/appassembler/bin/
RUN rm *.bat

FROM openjdk:8
# would like to copy more specifically the jar and binary files here, if possible, see #2
COPY --from=builder /app/ami3/target/ /bin/ami3/
ENV PATH /bin/ami3/appassembler/bin/:${PATH}

# Add additional tools needed to handle PDF workflows and support Python tools:
RUN apt-get update && \
    apt-get -y install tesseract-ocr gocr default-jre python3 python3-pip && \
    rm -rf /var/lib/apt/lists/*

## And install Tika and GROBID:
RUN curl -k -o /opt/tika.jar https://www.mirrorservice.org/sites/ftp.apache.org/tika/tika-app-1.24.jar && \
    curl -k -L -o /opt/grobid-src-0.5.3.zip https://github.com/kermitt2/grobid/archive/0.5.3.zip && \
    curl -k -L -o /opt/grobid-core-0.5.3-onejar.jar https://github.com/kermitt2/grobid/releases/download/0.5.3/grobid-core-0.5.3-onejar.jar && \
    cd /opt && unzip -o grobid-src-0.5.3.zip && mkdir -p /opt/grobid-0.5.3/grobid-core/build/libs/ && mv /opt/grobid-core-0.5.3-onejar.jar /opt/grobid-0.5.3/grobid-core/build/libs/ && \
    mkdir -p /opt/grobid-0.5.3/grobid-home/tmp && chmod a+rwx /opt/grobid-0.5.3/grobid-home/tmp

CMD [ "/bin/bash" ]
# docker build --tag ami3 .
# docker run --rm -it norami-docker:0.0.1 ami-test a b c

KNIME

to be added

Docker

getpapers

In order to run getpapers from a docker container the following Dockerfile may be used:

(PMR: Please upload a copy to the openVirus site)

FROM node:slim

WORKDIR /usr/src/app

RUN npm install --global getpapers

This may be run from terminal using the following shell script

#!/usr/bin/env bash

docker build -t paper_getter .

docker run -it --rm --name get-papers \
 -v $(pwd)/results:/results \
 paper_getter \
 getpapers -o /results --query 'c4 photosynthesis flaveria'

Here the query is set by docker run. This also sets the folder the results will be stored in using the mounted volume. This allows for queries to be run automatically if desired.

Running ami using Docker

First step is to build the clone the ami repository

git clone https://github.com/petermr/openVirus.git

Then you will need to move into the freshly cloned repo and build the docker image using the command

docker build -t ami3 .

This might take a bit of time since its quite a complicated image. Now we can start using the image.

In order to run the image we will run docker using the following format

docker run -it --rm --name test-ami3 \
 -v $(pwd)/<tree-folder>:<tree-folder> \
ami3 \
ami <ami_command>

for example if we have written our downloaded files into $(pwd)/xml_results

docker run -it --rm --name test-ami3 \
 -v $(pwd)/xml_results:/xml_results \
 ami3 \
ami -p ./xml_results pdfbox

Running a ami-search using Docker

docker run -it --rm --name test-ami3 \
 -v $(pwd)/../scraper/xml_results:/xml_results \
 -v $(pwd)/output:/output \
 -v $(pwd)/logs:/logs \
 ami3 \
ami -p xml_results/ search --dictionary country

Creating a library using Docker. This will put the dictionary into the folder, target.

docker run -it --rm --name test-ami3 \
 -v $(pwd)/../scraper/xml_results:/xml_results \
 -v $(pwd)/output:/output \
 -v $(pwd)/logs:/logs \
 -v $(pwd)/target:/target \
 ami3 \
 amidict  --dictionary myterpenes  --directory=target/dictionary/create  --inputname miniterpenes \
create --informat list --terms thymol menthol borneol junkolol --wikilinks wikidata wikipedia --outformats xml
Add a custom sidebar
Clone this wiki locally