Merge pull request #23 from crim-ca/fix-nlp-docker

.gitignore

@@ -23,3 +23,6 @@
 # disallow notebooks at root
 ./*.ipynb
 **/.ipynb_checkpoints/
+
+### Outputs
+*.log

Makefile

@@ -0,0 +1,18 @@
+# Included custom configs change the value of MAKEFILE_LIST
+# Extract the required reference beforehand so we can use it for help target
+MAKEFILE_NAME := $(word $(words $(MAKEFILE_LIST)),$(MAKEFILE_LIST))
+# Include custom config if it is available
+-include Makefile.config
+
+# Application
+APP_ROOT    := $(abspath $(lastword $(MAKEFILE_NAME))/..)
+APP_NAME    := $(shell basename $(APP_ROOT))
+APP_DOMAINS ?= eo nlp
+DOCKER_REPO ?= crim-ca/pavics-jupyter-images
+
+DOCKER_BUILDS := $(addprefix docker-build-, $(APP_DOMAINS))
+$(DOCKER_BUILDS): docker-build-%:
+	docker build -t $(DOCKER_REPO)/$*:latest "$(APP_ROOT)/$(*)" 2>&1 | tee "$(APP_ROOT)/make-$@.log"
+
+.PHONY: docker-build
+docker-build: $(DOCKER_BUILDS)

nlp/Dockerfile

@@ -2,14 +2,16 @@ FROM birdhouse/pavics-jupyter-base:0.5.0
 
 # must update conda env as root, because of a permission error when having pip dependencies in the input yml file
 USER root
+# install pre-requirement for compiling some dependencies
+RUN apt install -y pkg-config libpcre++-dev
 
 COPY environment.yml /environment.yml
-COPY notebook_config.yml /notebook_config.yml
 
 # update env "birdy"
 # use umask 0000 so that package files for the updated environment are usable by the user for the jupyter-conda-extension
-RUN umask 0000 && mamba env update -f /environment.yml \
-    && mamba clean -a
+RUN umask 0000 && \
+    mamba env update -f /environment.yml && \
+    mamba clean -y -a
 
 # Set the encoding to UTF-8, this is needed for heideltime to work properly
 ENV LANG=C.UTF-8
@@ -19,40 +21,49 @@ RUN python -m spacy download en_core_web_trf
 
 # Downloading the ner-large flair model
 RUN mkdir flair_models && \
-    curl -L -o flair_models/ner-large  https://huggingface.co/flair/ner-english-large/resolve/main/pytorch_model.bin > /dev/null
+    curl -L -o flair_models/ner-large \
+        https://huggingface.co/flair/ner-english-large/resolve/main/pytorch_model.bin > /dev/null
 
 # Heideltime Tree-tagger Installation
 RUN mkdir -p heideltime/tree-tagger-linux-3.2.3 && cd heideltime/tree-tagger-linux-3.2.3 && \
-    curl -o tree-tagger-linux-3.2.3.tar.gz  https://www.cis.lmu.de/~schmid/tools/TreeTagger/data/tree-tagger-linux-3.2.3.tar.gz && \
-    curl -o tagger-scripts.tar.gz https://www.cis.lmu.de/~schmid/tools/TreeTagger/data/tagger-scripts.tar.gz && \
-    curl -o english.par.gz https://www.cis.lmu.de/~schmid/tools/TreeTagger/data/english.par.gz && \
-    curl -o install-tagger.sh https://www.cis.lmu.de/~schmid/tools/TreeTagger/data/install-tagger.sh && \
+    curl -o tree-tagger-linux-3.2.3.tar.gz \
+        https://www.cis.lmu.de/~schmid/tools/TreeTagger/data/tree-tagger-linux-3.2.3.tar.gz && \
+    curl -o tagger-scripts.tar.gz \
+        https://www.cis.lmu.de/~schmid/tools/TreeTagger/data/tagger-scripts.tar.gz && \
+    curl -o english.par.gz \
+        https://www.cis.lmu.de/~schmid/tools/TreeTagger/data/english.par.gz && \
+    curl -o install-tagger.sh \
+        https://www.cis.lmu.de/~schmid/tools/TreeTagger/data/install-tagger.sh && \
     tar -xvzf tree-tagger-linux-3.2.3.tar.gz  && \
     sh install-tagger.sh && \
     rm tree-tagger-linux-3.2.3.tar.gz tagger-scripts.tar.gz english.par.gz install-tagger.sh
 
 # Download and extract heideltime.standalone.jar
-RUN cd heideltime && \   
-    curl -L -o heideltime-standalone-2.2.1.tar.gz https://github.com/HeidelTime/heideltime/releases/download/VERSION2.2.1/heideltime-standalone-2.2.1.tar.gz && \
+RUN cd heideltime && \
+    curl -L -o heideltime-standalone-2.2.1.tar.gz \
+        https://github.com/HeidelTime/heideltime/releases/download/VERSION2.2.1/heideltime-standalone-2.2.1.tar.gz && \
     tar -xzvf heideltime-standalone-2.2.1.tar.gz heideltime-standalone/de.unihd.dbs.heideltime.standalone.jar && \
     mv heideltime-standalone/de.unihd.dbs.heideltime.standalone.jar . && \ 
     rmdir heideltime-standalone && \
     rm heideltime-standalone-2.2.1.tar.gz
 
+# Give read&write permission to jenkins for config
+RUN chown -R jenkins heideltime
+
 # Setup Haskell for Duckling server
-# https://github.com/facebook/duckling
-RUN curl -sSL https://get.haskellstack.org/ | sh && \
+RUN curl -sSL https://get.haskellstack.org/ | bash && \
     git clone https://github.com/facebook/duckling && \
     cd duckling && \
     stack build && \
     stack install && \
+    cd .. && \
     rm -fr duckling
-
-# Give read&write permission to jenkins for config
-RUN chown -R jenkins heideltime
+ENV PATH="/root/.local/bin:$PATH"
 
 # Give ownership of the conda cache folder to jenkins, to enable installing packages by the user from JupyterLab
 RUN mkdir /opt/conda/pkgs/cache && chown -R 1000:1000 /opt/conda/pkgs/cache
 
+COPY notebook_config.yml /notebook_config.yml
+
 # specify user because of problem running start-notebook.sh when being root
 USER jenkins