Merge pull request #16 from se4bio/dockerfile

Dockerise

.github/workflows/validate-python.yml

@@ -14,11 +14,7 @@ jobs:
       - name: 'Validate tests'
         working-directory: ./cbioportal-core
         run: |
-          docker run -v ${PWD}:/cbioportal-core python:3.6 /bin/sh -c '
+          docker run -v ${PWD}:/cbioportal-core python:3.6 /bin/bash -c '
             cd cbioportal-core &&
             pip install -r requirements.txt &&
-            export PYTHONPATH=/cbioportal-core/src/main/resources/scripts &&
-            cd src/test/scripts/ &&
-            python unit_tests_validate_data.py &&
-            python system_tests_validate_data.py &&
-            python system_tests_validate_studies.py'
+            source test_scripts.sh'

.gitignore

@@ -31,3 +31,14 @@ replay_pid*
 
 .idea
 /dependency-reduced-pom.xml
+
+# python
+__pycache__/
+*.pyc
+.venv/
+
+# test html reports
+*.html~
+
+# app properties that contains credentials
+/application.properties

Dockerfile

@@ -0,0 +1,45 @@
+FROM maven:3-eclipse-temurin-21 as jar_builder
+
+# Set the working directory in the Maven image
+WORKDIR /app
+
+# Copy the java source files and the pom.xml file into the image
+COPY src ./src
+COPY pom.xml .
+
+# Build the application
+RUN mvn clean package -DskipTests
+
+FROM maven:3-eclipse-temurin-21
+
+# download system dependencies first to take advantage of docker caching
+RUN apt-get update; apt-get install -y --no-install-recommends \
+        build-essential \
+        default-mysql-client \
+        default-libmysqlclient-dev \
+        python3 \
+        python3-setuptools \
+        python3-dev \
+        python3-pip \
+        unzip \
+        perl \
+    && rm -rf /var/lib/apt/lists/* \
+    && pip3 install wheel
+
+# Install any needed packages specified in requirements.txt
+COPY requirements.txt ./
+RUN pip install --no-cache-dir -r requirements.txt
+
+RUN ln -s $(which python3) /usr/local/bin/python || true
+
+COPY --from=jar_builder /app/core-*.jar /
+COPY scripts/ scripts/
+RUN chmod -R a+x /scripts/
+
+# Set the working directory in the container
+WORKDIR /scripts/
+
+ENV PORTAL_HOME=/
+
+# This file is empty. It has to be overriden by bind mounting the actual application.properties
+RUN touch /application.properties

README.md

@@ -44,3 +44,90 @@ Use Maven to run the integration tests. Ensure you are in the root directory of
 ```
 mvn integration-test
 ```
+
+## Development
+
+### Prerequisites
+To contribute to `cbioportal-core`, ensure you have the following tools installed:
+
+- Python 3: Required for study validation and orchestration scripts. These scripts utilize the underlying loader jar.
+- Perl: Specify the version required based on script compatibility. Necessary for data loading scripts interfacing with lookup tables.
+- JDK 21: Essential for developing the data loader component.
+- Maven 3.8.3: Used to compile and test the loader jar. Review this [issue](https://github.com/cBioPortal/cbioportal-core/issues/15) before starting.
+
+### Setup
+
+1. Create a Python virtual environment (first-time setup):
+```bash
+python -m venv .venv
+```
+
+2. Activate the virtual environment:
+```bash
+source .venv/bin/activate
+```
+
+3. Install required Python dependencies (first-time setup or when dependencies have changed):
+```bash
+pip install -r requirements.txt
+```
+
+### Building and Testing
+
+After you are done with the setup, you can build and test the project.
+
+1. Execute tests through the provided script:
+```bash
+source test_scripts.sh
+```
+
+2. Build the loader jar using Maven (includes testing):
+```bash
+mvn clean package
+```
+*Note:* The Maven configuration is set to place the jar in the project's root directory to ensure consistent paths in both development and production.
+
+### Configuring Application Properties
+
+The loader requires specific properties set to establish a connection to the database. These properties should be defined in the application.properties file within your project.
+
+#### Creating the Properties File
+
+1. Begin by creating your application.properties file. This can be done by copying from an example or template provided in the project:
+```bash
+cp application.properties.example application.properties
+```
+
+2. Open application.properties in your preferred text editor and modify the properties to match your database configuration and other environment-specific settings.
+
+#### Setting the PORTAL_HOME Environment Variable
+
+The PORTAL_HOME environment variable should be set to the directory containing your application.properties file, typically the root of your project:
+```
+export PORTAL_HOME=$(pwd)
+```
+Ensure this command is run in the root directory of your project, where the application.properties file is located. This setup is crucial for the loader to correctly access the required properties.
+
+#### maven.properties
+TODO: Document role of `maven.properties` file.
+
+### Script Execution with Loader Jar
+
+To run scripts that require the loader jar, ensure the jar file is in the project root.
+The script will search for `core-*.jar` in the root of the project:
+```bash
+python scripts/importer/metaImport.py -s tests/test_data/study_es_0 -p tests/test_data/api_json_unit_tests -o
+```
+
+## Running in docker
+
+Build docker image with:
+```bash
+docker build -t cbioportal-core .
+```
+
+Example of how to start the loading:
+```bash
+docker run -it -v $(pwd)/data/:/data/ -v $(pwd)/application.properties:/application.properties cbioportal-core python importer/metaImport.py -s /data/study_es_0 -p /data/api_json -o
+```
+

application.properties.example

@@ -0,0 +1,5 @@
+spring.datasource.url=jdbc:mysql://localhost:3306/cbioportal?useSSL=false
+spring.datasource.username=cbio
+spring.datasource.password=P@ssword1
+spring.datasource.driver-class-name=com.mysql.jdbc.Driver
+spring.jpa.database-platform=org.hibernate.dialect.MySQL5InnoDBDialect

pom.xml

@@ -236,7 +236,11 @@
 								<resource>
 									<directory>${project.basedir}</directory>
 									<includes>
+										<!-- FIXME Remove these includes when running loader from the cbioportal docker image is deprecated. Loading should be done from cbioportal-core docker image directly instead. -->
+										<!-- Includes python/perl/shell scripts with the jar to transfer to cbioportal project -->
+										<!-- See here https://github.com/cBioPortal/cbioportal/blob/14e00a6c5580700b96152c45f34baf150e652c89/docker/web-and-data/Dockerfile#L48 -->
 										<include>requirements.txt</include>
+										<include>scripts/</include>
 									</includes>
 								</resource>
 							</resources>
@@ -323,6 +327,13 @@
 					</execution>
 				</executions>
 			</plugin>
+			<plugin>
+				<groupId>org.apache.maven.plugins</groupId>
+				<artifactId>maven-jar-plugin</artifactId>
+				<configuration>
+					<outputDirectory>${project.basedir}</outputDirectory>
+				</configuration>
+			</plugin>
 		</plugins>
 	</build>
 

src/main/resources/scripts/env.pl → scripts/env.pl

@@ -39,18 +39,18 @@
 }
 
 # Set up Classpath to use the scripts jar
-sub locate_src_root {
+sub locate_root {
     # isolate the directory this code file is in
     my ($volume, $script_dir, undef) = File::Spec->splitpath(__FILE__);
-    # go up from cbioportal/core/src/main/scripts/ to cbioportal/
-    my $src_root_dir = File::Spec->catdir($script_dir, (File::Spec->updir()) x 4);
+    # go up one level
+    my $root_dir = File::Spec->catdir($script_dir, File::Spec->updir());
     # reassamble the path and resolve updirs (/../)
-    return abs_path(File::Spec->catpath($volume, $src_root_dir));
+    return abs_path(File::Spec->catpath($volume, $root_dir));
 }
-$src_root = locate_src_root();
-@jar_files = glob("$src_root/scripts/target/scripts-*.jar");
+$root_dir = locate_root();
+@jar_files = glob("$root_dir/core-*.jar");
 if (scalar @jar_files != 1) {
-    die "Expected to find 1 scripts-*.jar, but found: " . scalar @jar_files;
+    die "Expected to find 1 core-*.jar, but found: " . scalar @jar_files;
 }
 $cp = pop @jar_files;
 

src/main/resources/scripts/envSimple.pl → scripts/envSimple.pl

@@ -34,18 +34,18 @@
 }
 
 # Set up Classpath to use the scripts jar
-sub locate_src_root {
+sub locate_root {
     # isolate the directory this code file is in
     my ($volume, $script_dir, undef) = File::Spec->splitpath(__FILE__);
-    # go up from cbioportal/core/src/main/scripts/ to cbioportal/
-    my $src_root_dir = File::Spec->catdir($script_dir, (File::Spec->updir()) x 1);
+    # go up one level
+    my $root_dir = File::Spec->catdir($script_dir, File::Spec->updir());
     # reassamble the path and resolve updirs (/../)
-    return abs_path(File::Spec->catpath($volume, $src_root_dir));
+    return abs_path(File::Spec->catpath($volume, $root_dir));
 }
-$src_root = locate_src_root();
-@jar_files = glob("$src_root/core-*.jar");
+$root_dir = locate_root();
+@jar_files = glob("$root_dir/core-*.jar");
 if (scalar @jar_files != 1) {
-    die "Expected to find 1 scripts-*.jar, but found: " . scalar @jar_files;
+    die "Expected to find 1 core-*.jar, but found: " . scalar @jar_files;
 }
 $cp = pop @jar_files;
 

src/main/resources/scripts/importer/cbioportalImporter.py → scripts/importer/cbioportalImporter.py

@@ -501,10 +501,11 @@ def locate_jar():
     """
     # get the directory name of the currently running script,
     # resolving any symlinks
-    script_dir = Path(__file__).resolve().parent
-    # go up from core/scripts/importer/ to core/
-    src_root = script_dir.parent.parent
-    jars = list((src_root ).glob('core-*.jar'))
+    this_file = Path(__file__).resolve()
+    importer_dir = this_file.parent
+    scripts_dir = importer_dir.parent
+    root_dir = scripts_dir.parent
+    jars = list((root_dir).glob('core-*.jar'))
     if len(jars) != 1:
         raise FileNotFoundError(
             'Expected to find 1 scripts-*.jar, but found ' + str(len(jars)))

src/main/resources/scripts/importer/cbioportal_common.py → scripts/importer/cbioportal_common.py

@@ -997,7 +997,9 @@ def run_java(*args):
         java_command = os.path.join(java_home, 'bin', 'java')
     else:
         java_command = 'java'
-    process = Popen([java_command] + list(args), stdout=PIPE, stderr=STDOUT,
+    full_cmd = [java_command] + list(args)
+    print(">", " ".join(full_cmd))
+    process = Popen(full_cmd, stdout=PIPE, stderr=STDOUT,
                     universal_newlines=True)
     ret = []
     while process.poll() is None:

src/main/resources/scripts/importer/validateData.py → scripts/importer/validateData.py

@@ -957,7 +957,7 @@ def load_chromosome_lengths(reference_genome, logger):
                                     'downloadChromosomeSizes.py.')
 
         logger.debug("Retrieving chromosome lengths from '%s'",
-                     chrom_size_file)
+                     os.path.basename(chrom_size_file))
 
         try:
             chrom_size_dict = chrom_sizes[reference_genome]

test_scripts.sh

@@ -0,0 +1 @@
+pushd tests/; PYTHONPATH=../scripts:$PYTHONPATH python -m unittest *.py; popd