feat(#7): python venv

.dockerignore

@@ -0,0 +1 @@
+renovate.json

Dockerfile

@@ -29,11 +29,19 @@ RUN apt-get update -y --fix-missing
 RUN apt-get install -y --no-install-recommends make=4.3-4.1build2
 RUN apt-get install -y --no-install-recommends git
 RUN apt-get install -y --no-install-recommends npm
+RUN apt-get install -y --no-install-recommends python3
+RUN apt-get install -y --no-install-recommends python3-pip
+RUN apt-get install -y --no-install-recommends python3-venv
 
+# Setup Python venv.
 WORKDIR /srdataset
+RUN python3 -m venv .venv && /srdataset/.venv/bin/python -m pip install pandas
+ENV PYTHON=/srdataset/.venv/bin/python
+
 COPY Makefile /srdataset/
 ENV LOCAL /srdataset
 COPY steps/install.sh /srdataset/steps/
+COPY steps/requirements.txt /srdataset/steps/
 RUN sh steps/install.sh
 COPY . /srdataset
-ENTRYPOINT make install collect metrics
+ENTRYPOINT make collect metrics

Makefile

@@ -38,12 +38,12 @@ test:
 
 # Install.
 install:
-	echo "Installing..."
+	chmod +x steps/install.sh &&./steps/install.sh
 
 # Collect repositories from GitHub API.
 collect:
 	echo "Collecting..."
 
 # Measure repositories with SR metrics.
 metrics:
-	echo "Measuring SRs..."
+	chmod +x steps/metrics.sh &&./steps/metrics.sh

steps/collect.sh

@@ -24,6 +24,14 @@
 set -e
 set -o pipefail
 
+# create pats.txt > $PATS
+# stars:2..100000 size:>=20 mirror:false template:false topic:(c | python | java | javascript | ruby)
+
+# java=32,258
+# c=10,434
+# python=86,627
+# javascript=72,496
+# ruby=4,997
 declare -a args=( \
         "--query=${SEARCH_QUERY}" \
         "--start=${START_DATE}" \
@@ -32,3 +40,8 @@ declare -a args=( \
         "--tokens=${PATS}"
 )
 ghminer "${args[@]}"
+if [ ! -f result.csv ]; then
+  echo "File result.csv was not created!"
+  echo "Failing..."
+  exit 0
+fi

steps/install.sh

@@ -23,6 +23,6 @@
 
 set -e
 
+# Install ghminer.
 npm install -g ghminer@0.0.3
-pip3 install -r requirements.txt
 echo "All dependencies are installed and up to date! Now you can run 'make' and build the dataset."

steps/metrics.sh

@@ -25,4 +25,6 @@ set -e
 set -o pipefail
 
 # Compute CPD, RC, IC metrics.
-python3 metrics.py
+cd steps
+echo "Measuring SRs..."
+$PYTHON metrics.py