use launcher script

Original version use python module to call R code. This would lock all python threads resulting in missing heartbeats.

Author: robkooper

sample-extractors/simple-r-extractor/Dockerfile

@@ -1,15 +1,16 @@
-FROM clowder/pyclowder
+FROM clowder/pyclowder:latest
 
 ENV R_SCRIPT="" \
     R_FUNCTION=""
 
-RUN apt-get -q -q update \
-    && apt-get -y install --no-install-recommends r-base-core r-base-dev python-dev \
+RUN echo "deb http://cran.rstudio.com/bin/linux/ubuntu xenial/" > /etc/apt/sources.list.d/R.list \
+    && apt-key adv --keyserver keyserver.ubuntu.com --recv E084DAB9 \
+    && apt-get -q -q update \
+    && apt-get -y install --no-install-recommends r-base-core r-base-dev \
     && rm -rf /var/lib/apt/lists/* \
-    && pip install rpy2==2.8.6
-RUN Rscript --vanilla -e "install.packages('jsonlite', repos='http://cran.rstudio.com/'); print(system.file(package = 'jsonlite')); q(status=as.integer(system.file(package = 'jsonlite') == ''))"
+    && Rscript --vanilla -e "install.packages('jsonlite', repos='http://cran.rstudio.com/'); print(system.file(package = 'jsonlite')); q(status=as.integer(system.file(package = 'jsonlite') == ''))"
 
-COPY r_extractor.py /home/clowder/
+COPY r_extractor.py launcher.R /home/clowder/
 
 # install any packages
 ONBUILD COPY packages.* Dockerfile /home/clowder/

sample-extractors/simple-r-extractor/launcher.R

@@ -0,0 +1,27 @@
+#!/usr/bin/env Rscript
+
+# get environment variables
+r_script <- Sys.getenv('R_SCRIPT')
+r_function <- Sys.getenv('R_FUNCTION')
+if (r_function == '') {
+  stop("Need a function to call.")
+}
+
+# command line arguments
+args <- commandArgs(trailingOnly=TRUE)
+if (length(args) != 2) {
+  stop("Need 2 arguments (input_file, json_file)")
+}
+input_file <- args[1]
+json_file <- args[2]
+
+# source script file
+if (r_script != '') {
+  source(r_script)
+}
+
+# call function
+result <- do.call(r_function, list(input_file))
+
+# write result as json
+write(jsonlite::toJSON(result, auto_unbox=TRUE), json_file)

sample-extractors/simple-r-extractor/r_extractor.py

@@ -1,25 +1,17 @@
 #!/usr/bin/env python
 
 import json
-import os
+import subprocess
+import tempfile
 
 from pyclowder.extractors import SimpleExtractor
-import rpy2.robjects as robjects
-
-r_script = os.getenv("R_SCRIPT")
-r_function = os.getenv("R_FUNCTION")
 
 
 class RExtractor(SimpleExtractor):
     def process_file(self, input_file):
-        r_result = robjects.r('''
-                    if ("%s" != "") {
-                        source("%s")
-                    }
-                    result <- do.call("%s", list("%s"))
-                    jsonlite::toJSON(result, auto_unbox=TRUE)
-                ''' % (r_script, r_script, r_function, input_file))
-        return json.loads(str(r_result))
+        with tempfile.NamedTemporaryFile(suffix=".json") as json_file:
+            subprocess.check_call(['/home/clowder/launcher.R', input_file, json_file.name])
+            return json.load(json_file.file)
 
 
 RExtractor().start()