arpadboda commented on a change in pull request #784:
URL: https://github.com/apache/nifi-minifi-cpp/pull/784#discussion_r442041040
##########
File path: extensions/script/python/ExecutePythonProcessor.cpp
##########
@@ -35,155 +35,184 @@ namespace python {
namespace processors {
core::Property ExecutePythonProcessor::ScriptFile("Script File", // NOLINT
- R"(Path to script file to execute)", "");
+ R"(Path to script file to execute. Only one of Script File or Script Body
may be used)", "");
+core::Property ExecutePythonProcessor::ScriptBody("Script Body", // NOLINT
Review comment:
Using property builder (withdescription, blah blah) can help breaking
these lines without linter errors. And improve readability as well.
You can also specify some to be required.
##########
File path: extensions/script/python/ExecutePythonProcessor.cpp
##########
@@ -35,155 +35,184 @@ namespace python {
namespace processors {
core::Property ExecutePythonProcessor::ScriptFile("Script File", // NOLINT
- R"(Path to script file to execute)", "");
+ R"(Path to script file to execute. Only one of Script File or Script Body
may be used)", "");
+core::Property ExecutePythonProcessor::ScriptBody("Script Body", // NOLINT
+ R"(Script to execute. Only one of Script File or Script Body may be
used)", "");
core::Property ExecutePythonProcessor::ModuleDirectory("Module Directory", //
NOLINT
- R"(Comma-separated list of paths to files and/or directories which
- contain modules required by
the script)", "");
+ R"(Comma-separated list of paths to files and/or directories which contain
modules required by the script)", "");
core::Relationship ExecutePythonProcessor::Success("success", "Script
successes"); // NOLINT
core::Relationship ExecutePythonProcessor::Failure("failure", "Script
failures"); // NOLINT
void ExecutePythonProcessor::initialize() {
// initialization requires that we do a little leg work prior to onSchedule
// so that we can provide manifest our processor identity
- std::set<core::Property> properties;
-
- std::string prop;
- getProperty(ScriptFile.getName(), prop);
-
- properties.insert(ScriptFile);
- properties.insert(ModuleDirectory);
- setSupportedProperties(properties);
-
- std::set<core::Relationship> relationships;
- relationships.insert(Success);
- relationships.insert(Failure);
- setSupportedRelationships(std::move(relationships));
- setAcceptAllProperties();
- if (!prop.empty()) {
- setProperty(ScriptFile, prop);
- std::shared_ptr<script::ScriptEngine> engine;
- python_logger_ =
logging::LoggerFactory<ExecutePythonProcessor>::getAliasedLogger(getName());
+ if (getProperties().empty()) {
+ setSupportedProperties({
+ ScriptFile,
+ ScriptBody,
+ ModuleDirectory
+ });
+ setAcceptAllProperties();
+ setSupportedRelationships({
+ Success,
+ Failure
+ });
+ valid_init_ = false;
+ return;
+ }
- engine = createEngine<python::PythonScriptEngine>();
+ python_logger_ =
logging::LoggerFactory<ExecutePythonProcessor>::getAliasedLogger(getName());
- if (engine == nullptr) {
- throw std::runtime_error("No script engine available");
- }
+ getProperty(ModuleDirectory.getName(), module_directory_);
- try {
- engine->evalFile(prop);
- auto me = shared_from_this();
- triggerDescribe(engine, me);
- triggerInitialize(engine, me);
+ valid_init_ = false;
+ appendPathForImportModules();
+ loadScript();
+ try {
+ if (script_to_exec_.size()) {
+ std::shared_ptr<python::PythonScriptEngine> engine = getScriptEngine();
+ engine->eval(script_to_exec_);
+ auto shared_this = shared_from_this();
+ engine->describe(shared_this);
+ engine->onInitialize(shared_this);
+ handleEngineNoLongerInUse(std::move(engine));
valid_init_ = true;
- } catch (std::exception &exception) {
- logger_->log_error("Caught Exception %s", exception.what());
- engine = nullptr;
- std::rethrow_exception(std::current_exception());
- valid_init_ = false;
- } catch (...) {
- logger_->log_error("Caught Exception");
- engine = nullptr;
- std::rethrow_exception(std::current_exception());
- valid_init_ = false;
}
-
+ }
+ catch (const std::exception& exception) {
+ logger_->log_error("Caught Exception: %s", exception.what());
+ std::rethrow_exception(std::current_exception());
+ }
+ catch (...) {
+ logger_->log_error("Caught Exception");
+ std::rethrow_exception(std::current_exception());
}
}
void ExecutePythonProcessor::onSchedule(const
std::shared_ptr<core::ProcessContext> &context, const
std::shared_ptr<core::ProcessSessionFactory> &sessionFactory) {
if (!valid_init_) {
- throw std::runtime_error("Could not correctly in initialize " + getName());
- }
- context->getProperty(ScriptFile.getName(), script_file_);
- context->getProperty(ModuleDirectory.getName(), module_directory_);
- if (script_file_.empty() && script_engine_.empty()) {
- logger_->log_error("Script File must be defined");
- return;
+ throw std::runtime_error("Could not correctly initialize " + getName());
}
-
try {
- std::shared_ptr<script::ScriptEngine> engine;
-
- // Use an existing engine, if one is available
- if (script_engine_q_.try_dequeue(engine)) {
- logger_->log_debug("Using available %s script engine instance",
script_engine_);
- } else {
- logger_->log_info("Creating new %s script instance", script_engine_);
- logger_->log_info("Approximately %d %s script instances created for this
processor", script_engine_q_.size_approx(), script_engine_);
-
- engine = createEngine<python::PythonScriptEngine>();
-
- if (engine == nullptr) {
- throw std::runtime_error("No script engine available");
- }
-
- if (!script_file_.empty()) {
- engine->evalFile(script_file_);
- } else {
- throw std::runtime_error("No Script File is available to execute");
- }
+ reloadScriptIfUsingScriptFileProperty();
+ if (script_to_exec_.empty()) {
+ throw std::runtime_error("Neither Script Body nor Script File is
available to execute");
}
+ std::shared_ptr<python::PythonScriptEngine> engine = getScriptEngine();
- triggerSchedule(engine, context);
+ engine->eval(script_to_exec_);
+ engine->onSchedule(context);
- // Make engine available for use again
- if (script_engine_q_.size_approx() < getMaxConcurrentTasks()) {
- logger_->log_debug("Releasing %s script engine", script_engine_);
- script_engine_q_.enqueue(engine);
- } else {
- logger_->log_info("Destroying script engine because it is no longer
needed");
- }
- } catch (std::exception &exception) {
- logger_->log_error("Caught Exception %s", exception.what());
- } catch (...) {
+ handleEngineNoLongerInUse(std::move(engine));
+ }
+ catch (const std::exception& exception) {
+ logger_->log_error("Caught Exception: %s", exception.what());
+ }
+ catch (...) {
logger_->log_error("Caught Exception");
}
}
void ExecutePythonProcessor::onTrigger(const
std::shared_ptr<core::ProcessContext> &context, const
std::shared_ptr<core::ProcessSession> &session) {
+ if (!valid_init_) {
+ throw std::runtime_error("Could not correctly initialize " + getName());
+ }
try {
- std::shared_ptr<script::ScriptEngine> engine;
-
- // Use an existing engine, if one is available
- if (script_engine_q_.try_dequeue(engine)) {
- logger_->log_debug("Using available %s script engine instance",
script_engine_);
- } else {
- logger_->log_info("Creating new %s script instance", script_engine_);
- logger_->log_info("Approximately %d %s script instances created for this
processor", script_engine_q_.size_approx(), script_engine_);
-
- engine = createEngine<python::PythonScriptEngine>();
-
- if (engine == nullptr) {
- throw std::runtime_error("No script engine available");
- }
-
- if (!script_file_.empty()) {
- engine->evalFile(script_file_);
- } else {
- throw std::runtime_error("No Script File is available to execute");
- }
+ // TODO(hunyadi): When using "Script File" property, we currently re-read
the script file content every time the processor is triggered. This should
change to single-read when we release 1.0.0
+ // https://issues.apache.org/jira/browse/MINIFICPP-1223
+ reloadScriptIfUsingScriptFileProperty();
+ if (script_to_exec_.empty()) {
+ throw std::runtime_error("Neither Script Body nor Script File is
available to execute");
}
- triggerEngineProcessor(engine, context, session);
-
- // Make engine available for use again
- if (script_engine_q_.size_approx() < getMaxConcurrentTasks()) {
- logger_->log_debug("Releasing %s script engine", script_engine_);
- script_engine_q_.enqueue(engine);
- } else {
- logger_->log_info("Destroying script engine because it is no longer
needed");
- }
- } catch (std::exception &exception) {
- logger_->log_error("Caught Exception %s", exception.what());
+ std::shared_ptr<python::PythonScriptEngine> engine = getScriptEngine();
+ engine->onTrigger(context, session);
+ handleEngineNoLongerInUse(std::move(engine));
+ }
+ catch (const std::exception &exception) {
+ logger_->log_error("Caught Exception: %s", exception.what());
this->yield();
- } catch (...) {
+ }
+ catch (...) {
logger_->log_error("Caught Exception");
this->yield();
}
}
+// TODO(hunyadi): This is potentially not what we want. See
https://issues.apache.org/jira/browse/MINIFICPP-1222
+std::shared_ptr<python::PythonScriptEngine>
ExecutePythonProcessor::getScriptEngine() {
+ std::shared_ptr<python::PythonScriptEngine> engine;
+ // Use an existing engine, if one is available
+ if (script_engine_q_.try_dequeue(engine)) {
+ logger_->log_debug("Using available [%p] script engine instance",
engine.get());
+ return engine;
+ }
+ engine = createEngine<python::PythonScriptEngine>();
+ logger_->log_info("Created new [%p] script engine instance. Number of
instances: approx. %d / %d.", engine.get(), script_engine_q_.size_approx(),
getMaxConcurrentTasks());
+ if (engine == nullptr) {
+ throw std::runtime_error("No script engine available");
+ }
+ return engine;
+}
+
+void
ExecutePythonProcessor::handleEngineNoLongerInUse(std::shared_ptr<python::PythonScriptEngine>&&
engine) {
+ // Make engine available for use again
+ if (script_engine_q_.size_approx() < getMaxConcurrentTasks()) {
+ logger_->log_debug("Releasing [%p] script engine", engine.get());
+ script_engine_q_.enqueue(engine);
+ } else {
+ logger_->log_info("Destroying script engine because it is no longer
needed");
+ }
+}
+
+void ExecutePythonProcessor::appendPathForImportModules() {
+ // TODO(hunyadi): I have spent some time trying to figure out pybind11, but
+ // could not get this working yet. It is up to be implemented later
+ // https://issues.apache.org/jira/browse/MINIFICPP-1224
+ if (module_directory_.size()) {
+ logger_->log_error("Not supported property: Module Directory.");
+ }
+}
+
+void ExecutePythonProcessor::loadScriptFromFile(const std::string& file_path) {
+ std::ifstream file_handle(file_path);
+ if (!file_handle.is_open()) {
+ script_to_exec_ = "";
+ throw std::runtime_error("Failed to read Script File: " + file_path);
+ }
+ script_to_exec_ = std::string{
(std::istreambuf_iterator<char>(file_handle)),
(std::istreambuf_iterator<char>()) };
+}
+
+void ExecutePythonProcessor::loadScript() {
+ std::string script_file;
+ std::string script_body;
+ getProperty(ScriptFile.getName(), script_file);
+ getProperty(ScriptBody.getName(), script_body);
+ if (script_file.size()) {
+ if (script_body.size()) {
+ throw std::runtime_error("Only one of Script File or Script Body may be
used");
+ }
+ loadScriptFromFile(script_file);
+ return;
+ }
+ if (script_body.size()) {
+ script_to_exec_ = script_body;
+ return;
+ }
+ throw std::runtime_error("Neither Script Body nor Script File is available
to execute");
Review comment:
Not a big deal, but I prefer a structure of the function that throws in
error branches.
```
if (script_body.size() == 0) {
throw std::runtime_error("Neither Script Body nor Script File is available
to execute");
}
script_to_exec_ = script_body;
```
This way it seems to be shorter and better structured to me.
##########
File path: libminifi/test/script-tests/ExecutePythonProcessorTests.cpp
##########
@@ -0,0 +1,220 @@
+/**
+ *
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements. See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+#define CATCH_CONFIG_MAIN
+
+#include <memory>
+#include <string>
+#include <set>
+
+#include "../TestBase.h"
+
+#include "processors/GetFile.h"
+#include "python/ExecutePythonProcessor.h"
+#include "processors/LogAttribute.h"
+#include "processors/PutFile.h"
+#include "utils/file/FileUtils.h"
+#include "utils/TestUtils.h"
+
+namespace {
+using org::apache::nifi::minifi::utils::createTempDir;
+using org::apache::nifi::minifi::utils::putFileToDir;
+using org::apache::nifi::minifi::utils::createTempDirWithFile;
+using org::apache::nifi::minifi::utils::getFileContent;
+
+class ExecutePythonProcessorTestBase {
+ public:
+ ExecutePythonProcessorTestBase() :
+ logTestController_(LogTestController::getInstance()),
+
logger_(logging::LoggerFactory<org::apache::nifi::minifi::python::processors::ExecutePythonProcessor>::getLogger())
{
+ reInitialize();
+ }
+ virtual ~ExecutePythonProcessorTestBase() {
+ logTestController_.reset();
+ logTestController_.setDebug<TestPlan>();
+
logTestController_.setDebug<minifi::python::processors::ExecutePythonProcessor>();
+ logTestController_.setDebug<minifi::processors::PutFile>();
+ logTestController_.setDebug<minifi::processors::PutFile::ReadCallback>();
+ }
+
+ protected:
+ void reInitialize() {
+ testController_.reset(new TestController());
+ plan_ = testController_->createPlan();
+ }
+
+ std::string getScriptFullPath(const std::string& script_file_name) {
+ return SCRIPT_FILES_DIRECTORY + utils::file::FileUtils::get_separator() +
script_file_name;
+ }
+
+ static const std::string TEST_FILE_NAME;
+ static const std::string TEST_FILE_CONTENT;
+ static const std::string SCRIPT_FILES_DIRECTORY;
+
+ std::unique_ptr<TestController> testController_;
+ std::shared_ptr<TestPlan> plan_;
+ LogTestController& logTestController_;
+ std::shared_ptr<logging::Logger> logger_;
+};
+
+const std::string ExecutePythonProcessorTestBase::TEST_FILE_NAME{
"test_file.txt" };
+const std::string ExecutePythonProcessorTestBase::TEST_FILE_CONTENT{ "Test
text\n" };
+const std::string ExecutePythonProcessorTestBase::SCRIPT_FILES_DIRECTORY{
"test_scripts" };
+
+class SimplePythonFlowFileTransferTest : public ExecutePythonProcessorTestBase
{
+ public:
+ enum class Expectation {
+ OUTPUT_FILE_MATCHES_INPUT,
+ RUNTIME_RELATIONSHIP_EXCEPTION,
+ PROCESSOR_INITIALIZATION_EXCEPTION
+ };
+ SimplePythonFlowFileTransferTest() : ExecutePythonProcessorTestBase{} {}
+
+ protected:
+ void testSimpleFilePassthrough(const Expectation expectation, const
core::Relationship& execute_python_out_conn, const std::string&
used_as_script_file, const std::string& used_as_script_body) {
+ reInitialize();
+ const std::string input_dir = createTempDirWithFile(testController_.get(),
TEST_FILE_NAME, TEST_FILE_CONTENT);
+ const std::string output_dir = createTempDir(testController_.get());
+
+ addGetFileProcessorToPlan(input_dir);
+ if (Expectation::PROCESSOR_INITIALIZATION_EXCEPTION == expectation) {
+ REQUIRE_THROWS(addExecutePythonProcessorToPlan(used_as_script_file,
used_as_script_body));
+ return;
+ }
+ REQUIRE_NOTHROW(addExecutePythonProcessorToPlan(used_as_script_file,
used_as_script_body));
+ addPutFileProcessorToPlan(execute_python_out_conn, output_dir);
+
+ plan_->runNextProcessor(); // GetFile
+ if (Expectation::RUNTIME_RELATIONSHIP_EXCEPTION == expectation) {
+ REQUIRE_THROWS(plan_->runNextProcessor()); // ExecutePythonProcessor
+ return;
+ }
+ REQUIRE_NOTHROW(plan_->runNextProcessor()); // ExecutePythonProcessor
+ plan_->runNextProcessor(); // PutFile
+
+ const std::string output_file_path = output_dir +
utils::file::FileUtils::get_separator() + TEST_FILE_NAME;
+
+ if (Expectation::OUTPUT_FILE_MATCHES_INPUT == expectation) {
+ const std::string output_file_content{ getFileContent(output_file_path)
};
+ REQUIRE(TEST_FILE_CONTENT == output_file_content);
+ }
+ }
+ void testsStatefulProcessor() {
+ reInitialize();
+ const std::string output_dir = createTempDir(testController_.get());
+
+ auto executePythonProcessor =
plan_->addProcessor("ExecutePythonProcessor", "executePythonProcessor");
+ plan_->setProperty(executePythonProcessor,
org::apache::nifi::minifi::python::processors::ExecutePythonProcessor::ScriptFile.getName(),
getScriptFullPath("stateful_processor.py"));
+ executePythonProcessor->initialize();
+
+ addPutFileProcessorToPlan(core::Relationship("success", "description"),
output_dir);
+ plan_->runNextProcessor(); // ExecutePythonProcessor
+ for (std::size_t i = 0; i < 10; ++i) {
+ plan_->runCurrentProcessor(); // ExecutePythonProcessor
+ }
+ plan_->runNextProcessor(); // PutFile
+ for (std::size_t i = 0; i < 10; ++i) {
+ plan_->runCurrentProcessor(); // PutFile
+ const std::string state_name = std::to_string(i);
+ const std::string output_file_path = output_dir +
utils::file::FileUtils::get_separator() + state_name;
Review comment:
There is ```concat_path``` in FileUtils.h.
##########
File path: extensions/script/python/ExecutePythonProcessor.h
##########
@@ -18,12 +18,16 @@
* limitations under the License.
*/
-#ifndef NIFI_MINIFI_CPP_EXECUTEPYPROC_H
-#define NIFI_MINIFI_CPP_EXECUTEPYPROC_H
+#ifndef EXTENSIONS_SCRIPT_PYTHON_EXECUTEPYTHONPROCESSOR_H_
+#define EXTENSIONS_SCRIPT_PYTHON_EXECUTEPYTHONPROCESSOR_H_
-#include <concurrentqueue.h>
-#include <core/Resource.h>
-#include <core/Processor.h>
+#include <memory>
+#include <string>
+#include <vector>
+
+#include <concurrentqueue.h> // NOLINT
+#include <core/Resource.h> // NOLINT
Review comment:
Linter detects these as C headers because "<" is used instead quotes.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
