2 Commits

Author SHA1 Message Date
Aaron Po
366cb99e1d logging updates 2026-05-17 02:58:47 -04:00
Aaron Po
5d80b53351 Update main.cc 2026-05-17 01:29:32 -04:00
16 changed files with 368 additions and 250 deletions

View File

@@ -43,8 +43,7 @@ class LogProducer final : public ILogger {
* *
* Blocks while the channel applies backpressure. * Blocks while the channel applies backpressure.
*/ */
void Log(LogLevel level, PipelinePhase phase, void Log(LogEntry const& entry) override;
std::string_view message) override;
private: private:
BoundedChannel<LogEntry>& channel_; BoundedChannel<LogEntry>& channel_;

View File

@@ -34,12 +34,10 @@ class ILogger {
/** /**
* @brief Submit a log message to the logging subsystem. * @brief Submit a log message to the logging subsystem.
* *
* @param level Severity of the message. * @param entry Structured log record containing message and metadata.
* @param phase Pipeline execution phase associated with the message. *
* @param message Log message text.
*/ */
virtual void Log(LogLevel level, PipelinePhase phase, virtual void Log(const LogEntry& entry) = 0;
std::string_view message) = 0;
}; };
#endif // BIERGARTEN_PIPELINE_INCLUDES_SERVICES_LOGGING_LOGGER_H_ #endif // BIERGARTEN_PIPELINE_INCLUDES_SERVICES_LOGGING_LOGGER_H_

View File

@@ -1,14 +1,14 @@
#include "services/logging/logger.h" #include <chrono>
#include <iostream> #include <iostream>
#include <optional> #include <optional>
#include <sstream> #include <sstream>
#include <string> #include <string>
#include "data_model/models.h" #include "data_model/models.h"
#include "services/logging/logger.h"
std::optional<ApplicationOptions> ParseArguments(const int argc, char** argv, std::optional<ApplicationOptions> ParseArguments(
std::shared_ptr<ILogger> logger) { const int argc, char** argv, std::shared_ptr<ILogger> logger) {
prog_opts::options_description desc("Pipeline Options"); prog_opts::options_description desc("Pipeline Options");
auto opt = desc.add_options(); auto opt = desc.add_options();
@@ -70,8 +70,12 @@ std::optional<ApplicationOptions> ParseArguments(const int argc, char** argv,
return usage_stream.str(); return usage_stream.str();
})(); })();
if (logger) { if (logger) {
logger->Log(LogLevel::Info, PipelinePhase::Startup, title); logger->Log({.level = LogLevel::Info,
logger->Log(LogLevel::Info, PipelinePhase::Startup, usage); .phase = PipelinePhase::Startup,
.message = title});
logger->Log({.level = LogLevel::Info,
.phase = PipelinePhase::Startup,
.message = usage});
} else { } else {
std::cout << title << std::endl << usage << std::endl; std::cout << title << std::endl << usage << std::endl;
} }
@@ -87,7 +91,9 @@ std::optional<ApplicationOptions> ParseArguments(const int argc, char** argv,
std::stringstream help_stream; std::stringstream help_stream;
help_stream << "\n" << desc; help_stream << "\n" << desc;
if (logger) { if (logger) {
logger->Log(LogLevel::Info, PipelinePhase::Startup, help_stream.str()); logger->Log({.level = LogLevel::Info,
.phase = PipelinePhase::Startup,
.message = help_stream.str()});
} else { } else {
std::cout << help_stream.str() << std::endl; std::cout << help_stream.str() << std::endl;
} }
@@ -99,8 +105,7 @@ std::optional<ApplicationOptions> ParseArguments(const int argc, char** argv,
options.pipeline.output_path = var_map["output"].as<std::string>(); options.pipeline.output_path = var_map["output"].as<std::string>();
options.pipeline.log_path = var_map["log-path"].as<std::string>(); options.pipeline.log_path = var_map["log-path"].as<std::string>();
options.pipeline.prompt_dir = var_map["prompt-dir"].as<std::string>(); options.pipeline.prompt_dir = var_map["prompt-dir"].as<std::string>();
options.pipeline.location_count = options.pipeline.location_count = var_map["location-count"].as<uint32_t>();
var_map["location-count"].as<uint32_t>();
const bool use_mocked = var_map["mocked"].as<bool>(); const bool use_mocked = var_map["mocked"].as<bool>();
const std::string model_path = var_map["model"].as<std::string>(); const std::string model_path = var_map["model"].as<std::string>();
@@ -111,7 +116,9 @@ std::optional<ApplicationOptions> ParseArguments(const int argc, char** argv,
const std::string msg = const std::string msg =
"Invalid arguments: --mocked and --model are mutually exclusive"; "Invalid arguments: --mocked and --model are mutually exclusive";
if (logger) { if (logger) {
logger->Log(LogLevel::Error, PipelinePhase::Startup, msg); logger->Log({.level = LogLevel::Error,
.phase = PipelinePhase::Startup,
.message = msg});
} else { } else {
std::cerr << msg << std::endl; std::cerr << msg << std::endl;
} }
@@ -122,7 +129,9 @@ std::optional<ApplicationOptions> ParseArguments(const int argc, char** argv,
const std::string msg = const std::string msg =
"Invalid arguments: either --mocked or --model must be specified"; "Invalid arguments: either --mocked or --model must be specified";
if (logger) { if (logger) {
logger->Log(LogLevel::Error, PipelinePhase::Startup, msg); logger->Log({.level = LogLevel::Error,
.phase = PipelinePhase::Startup,
.message = msg});
} else { } else {
std::cerr << msg << std::endl; std::cerr << msg << std::endl;
} }
@@ -135,7 +144,9 @@ std::optional<ApplicationOptions> ParseArguments(const int argc, char** argv,
const std::string msg = const std::string msg =
"Invalid arguments: --prompt-dir is required when not using --mocked"; "Invalid arguments: --prompt-dir is required when not using --mocked";
if (logger) { if (logger) {
logger->Log(LogLevel::Error, PipelinePhase::Startup, msg); logger->Log({.level = LogLevel::Error,
.phase = PipelinePhase::Startup,
.message = msg});
} else { } else {
std::cerr << msg << std::endl; std::cerr << msg << std::endl;
} }
@@ -162,7 +173,9 @@ std::optional<ApplicationOptions> ParseArguments(const int argc, char** argv,
const std::string msg = const std::string msg =
"Sampling parameters are ignored when using --mocked"; "Sampling parameters are ignored when using --mocked";
if (logger) { if (logger) {
logger->Log(LogLevel::Warn, PipelinePhase::Startup, msg); logger->Log({.level = LogLevel::Warn,
.phase = PipelinePhase::Startup,
.message = msg});
} else { } else {
std::cerr << msg << std::endl; std::cerr << msg << std::endl;
} }
@@ -186,15 +199,20 @@ std::optional<ApplicationOptions> ParseArguments(const int argc, char** argv,
std::string("Failed to parse command-line arguments: ") + std::string("Failed to parse command-line arguments: ") +
exception.what(); exception.what();
if (logger) { if (logger) {
logger->Log(LogLevel::Error, PipelinePhase::Startup, msg); logger->Log({.level = LogLevel::Error,
.phase = PipelinePhase::Startup,
.message = msg});
} else { } else {
std::cerr << msg << std::endl; std::cerr << msg << std::endl;
} }
return std::nullopt; return std::nullopt;
} catch (...) { } catch (...) {
const std::string msg = "Failed to parse command-line arguments: unknown error"; const std::string msg =
"Failed to parse command-line arguments: unknown error";
if (logger) { if (logger) {
logger->Log(LogLevel::Error, PipelinePhase::Startup, msg); logger->Log({.level = LogLevel::Error,
.phase = PipelinePhase::Startup,
.message = msg});
} else { } else {
std::cerr << msg << std::endl; std::cerr << msg << std::endl;
} }

View File

@@ -3,13 +3,16 @@
* @brief BiergartenDataGenerator::GenerateBreweries() implementation. * @brief BiergartenDataGenerator::GenerateBreweries() implementation.
*/ */
#include "services/logging/logger.h" #include <chrono>
#include "biergarten_pipeline_orchestrator.h" #include "biergarten_pipeline_orchestrator.h"
#include "services/logging/logger.h"
void BiergartenPipelineOrchestrator::GenerateBreweries( void BiergartenPipelineOrchestrator::GenerateBreweries(
std::span<const EnrichedCity> cities) { std::span<const EnrichedCity> cities) {
logger_->Log(LogLevel::Info, PipelinePhase::BreweryAndBeerGeneration, logger_->Log({.level = LogLevel::Info,
"=== SAMPLE BREWERY GENERATION ==="); .phase = PipelinePhase::BreweryAndBeerGeneration,
.message = "=== SAMPLE BREWERY GENERATION ==="});
generated_breweries_.clear(); generated_breweries_.clear();
size_t skipped_count = 0; size_t skipped_count = 0;
@@ -29,33 +32,38 @@ void BiergartenPipelineOrchestrator::GenerateBreweries(
} catch (const std::exception& export_exception) { } catch (const std::exception& export_exception) {
++export_failed_count; ++export_failed_count;
logger_->Log(LogLevel::Warn, PipelinePhase::BreweryAndBeerGeneration, logger_->Log(
{.level = LogLevel::Warn,
.phase = PipelinePhase::BreweryAndBeerGeneration,
.message =
std::string("[Pipeline] Generated brewery for '") + std::string("[Pipeline] Generated brewery for '") +
location.city + "' (" + location.country + location.city + "' (" + location.country +
") but SQLite export failed: " + ") but SQLite export failed: " + export_exception.what()});
export_exception.what());
} }
} catch (const std::exception& e) { } catch (const std::exception& e) {
++skipped_count; ++skipped_count;
logger_->Log(LogLevel::Warn, PipelinePhase::BreweryAndBeerGeneration, logger_->Log({.level = LogLevel::Warn,
std::string("[Pipeline] Skipping city '") + location.city + .phase = PipelinePhase::BreweryAndBeerGeneration,
" (" + location.country + "): brewery generation failed: " + .message = std::string("[Pipeline] Skipping city '") +
e.what()); location.city + " (" + location.country +
"): brewery generation failed: " + e.what()});
} }
} }
if (skipped_count > 0) { if (skipped_count > 0) {
logger_->Log(LogLevel::Warn, PipelinePhase::BreweryAndBeerGeneration, logger_->Log({.level = LogLevel::Warn,
std::string("[Pipeline] Skipped ") + .phase = PipelinePhase::BreweryAndBeerGeneration,
.message = std::string("[Pipeline] Skipped ") +
std::to_string(skipped_count) + std::to_string(skipped_count) +
" city/cities due to generation errors"); " city/cities due to generation errors"});
} }
if (export_failed_count > 0) { if (export_failed_count > 0) {
logger_->Log(LogLevel::Warn, PipelinePhase::Teardown, logger_->Log({.level = LogLevel::Warn,
std::string("[Pipeline] Failed to export ") + .phase = PipelinePhase::Teardown,
.message = std::string("[Pipeline] Failed to export ") +
std::to_string(export_failed_count) + std::to_string(export_failed_count) +
" generated brewery/breweries to SQLite"); " generated brewery/breweries to SQLite"});
} }
} }

View File

@@ -3,10 +3,11 @@
* @brief BiergartenDataGenerator::LogResults() implementation. * @brief BiergartenDataGenerator::LogResults() implementation.
*/ */
#include "services/logging/logger.h" #include <chrono>
#include <sstream>
#include "biergarten_pipeline_orchestrator.h" #include "biergarten_pipeline_orchestrator.h"
#include <sstream> #include "services/logging/logger.h"
void BiergartenPipelineOrchestrator::LogResults() const { void BiergartenPipelineOrchestrator::LogResults() const {
std::ostringstream msg; std::ostringstream msg;
@@ -15,17 +16,18 @@ void BiergartenPipelineOrchestrator::LogResults() const {
for (const auto& [location, brewery] : generated_breweries_) { for (const auto& [location, brewery] : generated_breweries_) {
msg << index << ". city=\"" << location.city << "\" country=\"" msg << index << ". city=\"" << location.city << "\" country=\""
<< location.country << "\" state=\"" << location.state_province << location.country << "\" state=\"" << location.state_province
<< "\" iso3166_2=" << location.iso3166_2 << " lat=" << "\" iso3166_2=" << location.iso3166_2 << " lat=" << location.latitude
<< location.latitude << " lon=" << location.longitude << "\n"; << " lon=" << location.longitude << "\n";
msg << " brewery_name_en=\"" << brewery.name_en << "\"\n"; msg << " brewery_name_en=\"" << brewery.name_en << "\"\n";
msg << " brewery_description_en=\"" << brewery.description_en msg << " brewery_description_en=\"" << brewery.description_en << "\"\n";
<< "\"\n";
msg << " brewery_name_local=\"" << brewery.name_local << "\"\n"; msg << " brewery_name_local=\"" << brewery.name_local << "\"\n";
msg << " brewery_description_local=\"" << brewery.description_local msg << " brewery_description_local=\"" << brewery.description_local
<< "\"\n"; << "\"\n";
++index; ++index;
} }
logger_->Log(LogLevel::Debug, PipelinePhase::Teardown, msg.str()); logger_->Log({.level = LogLevel::Info,
.phase = PipelinePhase::Teardown,
.message = msg.str()});
} }

View File

@@ -3,9 +3,8 @@
* @brief BiergartenDataGenerator::QueryCitiesWithCountries() implementation. * @brief BiergartenDataGenerator::QueryCitiesWithCountries() implementation.
*/ */
#include "services/logging/logger.h"
#include <algorithm> #include <algorithm>
#include <chrono>
#include <filesystem> #include <filesystem>
#include <format> #include <format>
#include <iterator> #include <iterator>
@@ -13,16 +12,18 @@
#include "biergarten_pipeline_orchestrator.h" #include "biergarten_pipeline_orchestrator.h"
#include "json_handling/json_loader.h" #include "json_handling/json_loader.h"
#include "services/logging/logger.h"
std::vector<Location> BiergartenPipelineOrchestrator::QueryCitiesWithCountries() { std::vector<Location>
logger_->Log(LogLevel::Info, PipelinePhase::Startup, BiergartenPipelineOrchestrator::QueryCitiesWithCountries() {
"=== GEOGRAPHIC DATA OVERVIEW ==="); logger_->Log({.level = LogLevel::Info,
.phase = PipelinePhase::Startup,
.message = "=== GEOGRAPHIC DATA OVERVIEW ==="});
const std::filesystem::path locations_path = "locations.json"; const std::filesystem::path locations_path = "locations.json";
auto all_locations = JsonLoader::LoadLocations(locations_path, logger_); auto all_locations = JsonLoader::LoadLocations(locations_path, logger_);
const size_t sample_count = std::min( const size_t sample_count = std::min(
static_cast<size_t>(application_options_.pipeline.location_count), static_cast<size_t>(application_options_.pipeline.location_count),
all_locations.size()); all_locations.size());
@@ -38,9 +39,13 @@ std::vector<Location> BiergartenPipelineOrchestrator::QueryCitiesWithCountries()
std::ranges::sample(all_locations, std::back_inserter(sampled_locations), std::ranges::sample(all_locations, std::back_inserter(sampled_locations),
sample_count_signed, random_generator); sample_count_signed, random_generator);
logger_->Log(LogLevel::Info, PipelinePhase::Startup, logger_->Log({.level = LogLevel::Info,
std::format(" Locations available: {}", all_locations.size())); .phase = PipelinePhase::Startup,
logger_->Log(LogLevel::Info, PipelinePhase::Startup, .message = std::format(" Locations available: {}",
std::format(" Sampled locations: {}", sampled_locations.size())); all_locations.size())});
logger_->Log({.level = LogLevel::Info,
.phase = PipelinePhase::Startup,
.message = std::format(" Sampled locations: {}",
sampled_locations.size())});
return sampled_locations; return sampled_locations;
} }

View File

@@ -3,11 +3,11 @@
* @brief BiergartenDataGenerator::Run() implementation. * @brief BiergartenDataGenerator::Run() implementation.
*/ */
#include "services/logging/logger.h" #include <chrono>
#include <utility> #include <utility>
#include "biergarten_pipeline_orchestrator.h" #include "biergarten_pipeline_orchestrator.h"
#include "services/logging/logger.h"
bool BiergartenPipelineOrchestrator::Run() { bool BiergartenPipelineOrchestrator::Run() {
try { try {
@@ -30,18 +30,21 @@ bool BiergartenPipelineOrchestrator::Run() {
.region_context = std::move(region_context)}); .region_context = std::move(region_context)});
} catch (const std::exception& exception) { } catch (const std::exception& exception) {
++skipped_count; ++skipped_count;
logger_->Log(LogLevel::Warn, PipelinePhase::UserGeneration, logger_->Log(
std::string("[Pipeline] Skipping city '") + city.city + {.level = LogLevel::Warn,
" (" + city.country + "): context lookup failed: " + .phase = PipelinePhase::UserGeneration,
exception.what()); .message = std::string("[Pipeline] Skipping city '") + city.city +
" (" + city.country +
"): context lookup failed: " + exception.what()});
} }
} }
if (skipped_count > 0) { if (skipped_count > 0) {
logger_->Log(LogLevel::Warn, PipelinePhase::UserGeneration, logger_->Log({.level = LogLevel::Warn,
std::string("[Pipeline] Skipped ") + .phase = PipelinePhase::UserGeneration,
.message = std::string("[Pipeline] Skipped ") +
std::to_string(skipped_count) + std::to_string(skipped_count) +
" city/cities due to context lookup errors"); " city/cities due to context lookup errors"});
} }
this->GenerateBreweries(enriched); this->GenerateBreweries(enriched);
@@ -49,9 +52,11 @@ bool BiergartenPipelineOrchestrator::Run() {
this->LogResults(); this->LogResults();
return true; return true;
} catch (const std::exception& e) { } catch (const std::exception& e) {
logger_->Log(LogLevel::Error, PipelinePhase::Teardown, logger_->Log(
std::string("Pipeline execution failed with error: ") + {.level = LogLevel::Error,
e.what()); .phase = PipelinePhase::Teardown,
.message =
std::string("Pipeline execution failed with error: ") + e.what()});
return false; return false;
} }
} }

View File

@@ -4,6 +4,7 @@
* inference, and validates structured JSON output for brewery records. * inference, and validates structured JSON output for brewery records.
*/ */
#include <chrono>
#include <format> #include <format>
#include <optional> #include <optional>
#include <stdexcept> #include <stdexcept>
@@ -99,9 +100,11 @@ BreweryResult LlamaGenerator::GenerateBrewery(
raw = this->Infer(system_prompt, user_prompt, max_tokens, raw = this->Infer(system_prompt, user_prompt, max_tokens,
kBreweryJsonGrammar); kBreweryJsonGrammar);
if (logger_) { if (logger_) {
logger_->Log(LogLevel::Debug, PipelinePhase::BreweryAndBeerGeneration, logger_->Log(
std::string("LlamaGenerator: raw output (attempt ") + {.level = LogLevel::Debug,
std::to_string(attempt + 1) + "): " + raw); .phase = PipelinePhase::BreweryAndBeerGeneration,
.message = std::string("LlamaGenerator: raw output (attempt ") +
std::to_string(attempt + 1) + "): " + raw});
} }
// Validate output: parse JSON and check required fields // Validate output: parse JSON and check required fields
@@ -114,10 +117,12 @@ BreweryResult LlamaGenerator::GenerateBrewery(
// Success: return parsed brewery data // Success: return parsed brewery data
if (logger_) { if (logger_) {
logger_->Log(LogLevel::Info, PipelinePhase::BreweryAndBeerGeneration, logger_->Log(
std::string( {.level = LogLevel::Info,
"LlamaGenerator: successfully generated brewery data on attempt ") + .phase = PipelinePhase::BreweryAndBeerGeneration,
std::to_string(attempt + 1)); .message = std::string("LlamaGenerator: successfully generated "
"brewery data on attempt ") +
std::to_string(attempt + 1)});
} }
return brewery; return brewery;
@@ -127,9 +132,12 @@ BreweryResult LlamaGenerator::GenerateBrewery(
last_error = *validation_error; last_error = *validation_error;
if (logger_) { if (logger_) {
logger_->Log(LogLevel::Warn, PipelinePhase::BreweryAndBeerGeneration, logger_->Log(
{.level = LogLevel::Warn,
.phase = PipelinePhase::BreweryAndBeerGeneration,
.message =
std::string("LlamaGenerator: malformed brewery JSON (attempt ") + std::string("LlamaGenerator: malformed brewery JSON (attempt ") +
std::to_string(attempt + 1) + "): " + *validation_error); std::to_string(attempt + 1) + "): " + *validation_error});
} }
// Update prompt with error details to guide LLM toward correct output. // Update prompt with error details to guide LLM toward correct output.
@@ -148,10 +156,13 @@ BreweryResult LlamaGenerator::GenerateBrewery(
// All retry attempts exhausted: log failure and throw exception // All retry attempts exhausted: log failure and throw exception
if (logger_) { if (logger_) {
logger_->Log(LogLevel::Error, PipelinePhase::BreweryAndBeerGeneration, logger_->Log(
{.level = LogLevel::Error,
.phase = PipelinePhase::BreweryAndBeerGeneration,
.message =
std::string("LlamaGenerator: malformed brewery response after ") + std::string("LlamaGenerator: malformed brewery response after ") +
std::to_string(max_attempts) + " attempts: " + std::to_string(max_attempts) +
(last_error.empty() ? raw : last_error)); " attempts: " + (last_error.empty() ? raw : last_error)});
} }
throw std::runtime_error("LlamaGenerator: malformed brewery response"); throw std::runtime_error("LlamaGenerator: malformed brewery response");
} }

View File

@@ -6,6 +6,7 @@
*/ */
#include <algorithm> #include <algorithm>
#include <chrono>
#include <memory> #include <memory>
#include <stdexcept> #include <stdexcept>
#include <string> #include <string>
@@ -171,11 +172,12 @@ std::string LlamaGenerator::InferFormatted(const std::string& formatted_prompt,
if (token_count > prompt_budget) { if (token_count > prompt_budget) {
if (logger_) { if (logger_) {
logger_->Log( logger_->Log(
LogLevel::Warn, PipelinePhase::BreweryAndBeerGeneration, {.level = LogLevel::Warn,
std::string("LlamaGenerator: prompt too long (") + .phase = PipelinePhase::BreweryAndBeerGeneration,
.message = std::string("LlamaGenerator: prompt too long (") +
std::to_string(token_count) + ") tokens, truncating to " + std::to_string(token_count) + ") tokens, truncating to " +
std::to_string(prompt_budget) + std::to_string(prompt_budget) +
" tokens to fit n_batch/n_ctx limits"); " tokens to fit n_batch/n_ctx limits"});
} }
prompt_tokens.resize(static_cast<size_t>(prompt_budget)); prompt_tokens.resize(static_cast<size_t>(prompt_budget));
token_count = prompt_budget; token_count = prompt_budget;

View File

@@ -5,6 +5,7 @@
*/ */
#include <algorithm> #include <algorithm>
#include <chrono>
#include <stdexcept> #include <stdexcept>
#include <string> #include <string>
#include <utility> #include <utility>
@@ -48,7 +49,9 @@ void LlamaGenerator::Load(const std::string& model_path) {
context_ = std::move(loaded_context); context_ = std::move(loaded_context);
if (logger_) { if (logger_) {
logger_->Log(LogLevel::Info, PipelinePhase::Startup, logger_->Log({.level = LogLevel::Info,
std::string("[LlamaGenerator] Loaded model: ") + model_path); .phase = PipelinePhase::Startup,
.message = std::string("[LlamaGenerator] Loaded model: ") +
model_path});
} }
} }

View File

@@ -4,11 +4,14 @@
* initializes shared infrastructure, and executes the pipeline entry flow. * initializes shared infrastructure, and executes the pipeline entry flow.
*/ */
#include <spdlog/fmt/fmt.h>
#include <spdlog/spdlog.h> #include <spdlog/spdlog.h>
#include <boost/di.hpp> #include <boost/di.hpp>
#include <boost/program_options.hpp> #include <boost/program_options.hpp>
#include <chrono>
#include <exception> #include <exception>
#include <format>
#include <memory> #include <memory>
#include <optional> #include <optional>
#include <string> #include <string>
@@ -42,11 +45,19 @@ int main(const int argc, char** argv) {
spdlog::set_level(spdlog::level::debug); spdlog::set_level(spdlog::level::debug);
spdlog::set_pattern("│ %Y-%m-%d %H:%M:%S.%e │ %^%-7l%$ │ %v"); spdlog::set_pattern("│ %Y-%m-%d %H:%M:%S.%e │ %^%-7l%$ │ %v");
BoundedChannel<LogEntry> log_channel(kLogMaxCount); BoundedChannel<LogEntry> log_channel(kLogMaxCount);
auto log_dispatcher = std::make_unique<LogDispatcher>(log_channel);
std::thread log_thread([&log_dispatcher] { log_dispatcher->Run(); });
auto log_dispatcher = //
std::make_unique<LogDispatcher>(log_channel);
std::shared_ptr<ILogger> log_producer = std::shared_ptr<ILogger> log_producer =
std::make_shared<LogProducer>(log_channel); std::make_shared<LogProducer>(log_channel);
std::thread log_thread([&log_dispatcher] { log_dispatcher->Run(); });
auto shutdown = [&](const int exit_code) {
log_channel.Close();
log_thread.join();
return exit_code;
};
try { try {
Timer timer; Timer timer;
@@ -54,15 +65,15 @@ int main(const int argc, char** argv) {
const LlamaBackendState llama_backend_state; const LlamaBackendState llama_backend_state;
#endif #endif
log_producer->Log(LogLevel::Info, PipelinePhase::Startup, log_producer->Log({.level = LogLevel::Info,
"STARTING PIPELINE"); .phase = PipelinePhase::Startup,
.message = "STARTING PIPELINE"});
const std::optional<ApplicationOptions> parsed_options = const std::optional<ApplicationOptions> parsed_options =
ParseArguments(argc, argv, log_producer); ParseArguments(argc, argv, log_producer);
if (!parsed_options.has_value()) { if (!parsed_options.has_value()) {
log_channel.Close(); return shutdown(0);
log_thread.join();
return 0;
} }
const auto options = *parsed_options; const auto options = *parsed_options;
@@ -70,41 +81,79 @@ int main(const int argc, char** argv) {
const auto sampling = const auto sampling =
options.generator.sampling.value_or(SamplingOptions{}); options.generator.sampling.value_or(SamplingOptions{});
// -----------------------------------------------------------------------
// Prompt directory
// Conditionally constructed before the injector; moved into LlamaGenerator.
// -----------------------------------------------------------------------
std::unique_ptr<IPromptDirectory> prompt_directory; std::unique_ptr<IPromptDirectory> prompt_directory;
if (!options.generator.use_mocked) { if (!options.generator.use_mocked) {
try { try {
prompt_directory = std::make_unique<PromptDirectory>( prompt_directory = std::make_unique<PromptDirectory>(
options.pipeline.prompt_dir, log_producer); options.pipeline.prompt_dir, log_producer);
} catch (const std::exception& dir_error) { } catch (const std::exception& dir_error) {
log_producer->Log( log_producer->Log({.level = LogLevel::Error,
LogLevel::Error, PipelinePhase::Startup, .phase = PipelinePhase::Startup,
std::string("Invalid --prompt-dir: ") + dir_error.what()); .message = std::format("Invalid --prompt-dir: {}",
log_channel.Close(); dir_error.what())});
log_thread.join();
return 1; return shutdown(1);
} }
} }
// -----------------------------------------------------------------------
// Dependency injection
// -----------------------------------------------------------------------
const auto injector = di::make_injector( const auto injector = di::make_injector(
di::bind<ILogger>().to(log_producer), di::bind<ILogger>().to(log_producer),
di::bind<ApplicationOptions>().to(options), di::bind<ApplicationOptions>().to(options),
di::bind<std::string>().to(model_path), di::bind<std::string>().to(model_path),
di::bind<WebClient>().to<HttpWebClient>(),
di::bind<IExportService>().to<SqliteExportService>(), di::bind<IExportService>().to<SqliteExportService>(),
di::bind<IPromptFormatter>().to<Gemma4JinjaPromptFormatter>(), di::bind<IPromptFormatter>().to([options, log_producer] {
if (options.generator.use_mocked) {
{
log_producer->Log(
{.level = LogLevel::Info,
.phase = PipelinePhase::Startup,
.message = "Prompt formatter: none (mock mode)"});
}
return std::unique_ptr<IPromptFormatter>(nullptr);
}
{
log_producer->Log(
{.level = LogLevel::Info,
.phase = PipelinePhase::Startup,
.message = "Prompt formatter: Gemma4JinjaPromptFormatter"});
}
return std::unique_ptr<IPromptFormatter>(
std::make_unique<Gemma4JinjaPromptFormatter>());
}),
di::bind<WebClient>().to([options, log_producer] {
if (options.generator.use_mocked) {
{
log_producer->Log({.level = LogLevel::Info,
.phase = PipelinePhase::Startup,
.message = "Web client: none (mock mode)"});
}
return std::unique_ptr<WebClient>(nullptr);
}
{
log_producer->Log({.level = LogLevel::Info,
.phase = PipelinePhase::Startup,
.message = "Web client: HttpWebClient"});
}
return std::unique_ptr<WebClient>(
std::make_unique<HttpWebClient>(log_producer));
}),
di::bind<IEnrichmentService>().to( di::bind<IEnrichmentService>().to(
[options, &log_producer]( [options, &log_producer](
const auto& inj) -> std::unique_ptr<IEnrichmentService> { const auto& inj) -> std::unique_ptr<IEnrichmentService> {
// if (options.generator.use_mocked) { if (options.generator.use_mocked) {
// return std::make_unique<MockEnrichmentService>(); {
// } log_producer->Log({.level = LogLevel::Info,
.phase = PipelinePhase::Startup,
.message = "Enrichment: mock"});
}
return std::make_unique<MockEnrichmentService>();
}
{
log_producer->Log({.level = LogLevel::Info,
.phase = PipelinePhase::Startup,
.message = "Enrichment: Wikipedia"});
}
return std::make_unique<WikipediaEnrichmentService>( return std::make_unique<WikipediaEnrichmentService>(
inj.template create<std::unique_ptr<WebClient>>(), inj.template create<std::unique_ptr<WebClient>>(),
log_producer); log_producer);
@@ -113,57 +162,53 @@ int main(const int argc, char** argv) {
[&options, &model_path, &sampling, &prompt_directory, [&options, &model_path, &sampling, &prompt_directory,
&log_producer](const auto& inj) -> std::unique_ptr<DataGenerator> { &log_producer](const auto& inj) -> std::unique_ptr<DataGenerator> {
if (options.generator.use_mocked) { if (options.generator.use_mocked) {
log_producer->Log( {
LogLevel::Info, PipelinePhase::Startup, log_producer->Log({.level = LogLevel::Info,
"Using MockGenerator (no model path provided)"); .phase = PipelinePhase::Startup,
.message = "Generator: mock"});
}
return std::make_unique<MockGenerator>(); return std::make_unique<MockGenerator>();
} }
{
log_producer->Log( log_producer->Log(
LogLevel::Info, PipelinePhase::Startup, {.level = LogLevel::Info,
"Using LlamaGenerator: " + model_path + .phase = PipelinePhase::Startup,
" (temperature=" + std::to_string(sampling.temperature) + .message = fmt::format(
", top-p=" + std::to_string(sampling.top_p) + "Generator: LlamaGenerator | model={} | temp={:.2f} "
", top-k=" + std::to_string(sampling.top_k) + "top_p={:.2f} top_k={} n_ctx={} seed={}",
", n_ctx=" + std::to_string(sampling.n_ctx) + model_path, sampling.temperature, sampling.top_p,
", seed=" + std::to_string(sampling.seed) + ")"); sampling.top_k, sampling.n_ctx, sampling.seed)});
}
return std::make_unique<LlamaGenerator>( return std::make_unique<LlamaGenerator>(
options, model_path, log_producer, options, model_path, log_producer,
inj.template create<std::unique_ptr<IPromptFormatter>>(), inj.template create<std::unique_ptr<IPromptFormatter>>(),
std::move(prompt_directory)); std::move(prompt_directory));
})); }));
// -----------------------------------------------------------------------
// Pipeline execution
// -----------------------------------------------------------------------
const auto orchestrator = const auto orchestrator =
injector.create<std::unique_ptr<BiergartenPipelineOrchestrator>>(); injector.create<std::unique_ptr<BiergartenPipelineOrchestrator>>();
if (!orchestrator->Run()) { if (!orchestrator->Run()) {
log_producer->Log(LogLevel::Error, PipelinePhase::Teardown, log_producer->Log({.level = LogLevel::Error,
"Pipeline execution failed"); .phase = PipelinePhase::Teardown,
log_channel.Close(); .message = "Pipeline execution failed"});
log_thread.join(); return shutdown(1);
return 1;
} }
log_producer->Log(LogLevel::Info, PipelinePhase::Teardown, log_producer->Log({.level = LogLevel::Info,
"Pipeline executed successfully in " + .phase = PipelinePhase::Teardown,
std::to_string(timer.Elapsed()) + " ms"); .message = fmt::format("Pipeline complete in {} ms",
timer.Elapsed())});
log_channel.Close(); return shutdown(0);
log_thread.join();
return 0;
} catch (const std::exception& exception) { } catch (const std::exception& exception) {
// Attempt to use the logging infrastructure; if channel/dispatcher are
// compromised this is a best-effort fallback.
if (log_producer) { if (log_producer) {
log_producer->Log( log_producer->Log({.level = LogLevel::Error,
LogLevel::Error, PipelinePhase::Teardown, .phase = PipelinePhase::Teardown,
std::string("Unhandled fatal error in main: ") + exception.what()); .message = fmt::format("Unhandled fatal error: {}",
exception.what())});
} }
log_channel.Close(); return shutdown(1);
log_thread.join();
return 1;
} }
} }

View File

@@ -14,15 +14,16 @@
using namespace boost; using namespace boost;
std::string WikipediaEnrichmentService::FetchExtract(std::string_view query) { std::string WikipediaEnrichmentService::FetchExtract(std::string_view query) {
const std::string cache_key(query); const std::string cache_key(query);
// 1. Cache Lookup // 1. Cache Lookup
if (const auto cache_it = this->extract_cache_.find(cache_key); if (const auto cache_it = this->extract_cache_.find(cache_key);
cache_it != this->extract_cache_.end()) { cache_it != this->extract_cache_.end()) {
if (logger_) { if (logger_) {
logger_->Log(LogLevel::Debug, PipelinePhase::UserGeneration, logger_->Log({.level = LogLevel::Debug,
std::string("Wikipedia: Cache hit for ") + cache_key + "!"); .phase = PipelinePhase::UserGeneration,
.message = std::string("Wikipedia: Cache hit for ") +
cache_key + "!"});
} }
return cache_it->second; return cache_it->second;
} }
@@ -33,7 +34,6 @@ std::string WikipediaEnrichmentService::FetchExtract(std::string_view query) {
"api.php?action=query&titles={}&prop=extracts&explaintext=1&format=json", "api.php?action=query&titles={}&prop=extracts&explaintext=1&format=json",
encoded); encoded);
const std::string body = this->client_->Get(url); const std::string body = this->client_->Get(url);
{ {
using namespace std::literals::chrono_literals; using namespace std::literals::chrono_literals;
@@ -46,9 +46,11 @@ std::string WikipediaEnrichmentService::FetchExtract(std::string_view query) {
if (ec) { if (ec) {
if (logger_) { if (logger_) {
logger_->Log(LogLevel::Warn, PipelinePhase::UserGeneration, logger_->Log(
std::string("WikipediaService: JSON parse error for '") + {.level = LogLevel::Warn,
std::string(query) + "': " + ec.message()); .phase = PipelinePhase::UserGeneration,
.message = std::string("WikipediaService: JSON parse error for '") +
std::string(query) + "': " + ec.message()});
} }
return {}; return {};
} }
@@ -57,9 +59,12 @@ std::string WikipediaEnrichmentService::FetchExtract(std::string_view query) {
const json::object* obj = doc.if_object(); const json::object* obj = doc.if_object();
if (obj == nullptr) { if (obj == nullptr) {
if (logger_) { if (logger_) {
logger_->Log(LogLevel::Warn, PipelinePhase::UserGeneration, logger_->Log(
{.level = LogLevel::Warn,
.phase = PipelinePhase::UserGeneration,
.message =
std::string("WikipediaService: Expected root object for '") + std::string("WikipediaService: Expected root object for '") +
std::string(query) + "'"); std::string(query) + "'"});
} }
return {}; return {};
} }
@@ -72,9 +77,12 @@ std::string WikipediaEnrichmentService::FetchExtract(std::string_view query) {
if ((pages_ptr == nullptr) || !pages_ptr->is_object()) { if ((pages_ptr == nullptr) || !pages_ptr->is_object()) {
if (logger_) { if (logger_) {
logger_->Log(LogLevel::Warn, PipelinePhase::UserGeneration, logger_->Log(
{.level = LogLevel::Warn,
.phase = PipelinePhase::UserGeneration,
.message =
std::string("WikipediaService: Missing query.pages for '") + std::string("WikipediaService: Missing query.pages for '") +
std::string(query) + "'"); std::string(query) + "'"});
} }
return {}; return {};
} }
@@ -83,9 +91,11 @@ std::string WikipediaEnrichmentService::FetchExtract(std::string_view query) {
if (pages.empty()) { if (pages.empty()) {
if (logger_) { if (logger_) {
logger_->Log(LogLevel::Warn, PipelinePhase::UserGeneration, logger_->Log(
std::string("WikipediaService: No pages returned for '") + {.level = LogLevel::Warn,
std::string(query) + "'"); .phase = PipelinePhase::UserGeneration,
.message = std::string("WikipediaService: No pages returned for '") +
std::string(query) + "'"});
} }
this->extract_cache_.emplace(cache_key, ""); this->extract_cache_.emplace(cache_key, "");
return {}; return {};
@@ -97,9 +107,12 @@ std::string WikipediaEnrichmentService::FetchExtract(std::string_view query) {
if (!page_val.is_object()) { if (!page_val.is_object()) {
if (logger_) { if (logger_) {
logger_->Log(LogLevel::Warn, PipelinePhase::UserGeneration, logger_->Log(
{.level = LogLevel::Warn,
.phase = PipelinePhase::UserGeneration,
.message =
std::string("WikipediaService: Unexpected page format for '") + std::string("WikipediaService: Unexpected page format for '") +
std::string(query) + "'"); std::string(query) + "'"});
} }
return {}; return {};
} }
@@ -109,9 +122,10 @@ std::string WikipediaEnrichmentService::FetchExtract(std::string_view query) {
// Handle 404/Missing status // Handle 404/Missing status
if (page.contains("missing")) { if (page.contains("missing")) {
if (logger_) { if (logger_) {
logger_->Log(LogLevel::Warn, PipelinePhase::UserGeneration, logger_->Log({.level = LogLevel::Warn,
std::string("WikipediaService: Page '") + std::string(query) + .phase = PipelinePhase::UserGeneration,
"' does not exist"); .message = std::string("WikipediaService: Page '") +
std::string(query) + "' does not exist"});
} }
this->extract_cache_.emplace(cache_key, ""); this->extract_cache_.emplace(cache_key, "");
return {}; return {};
@@ -121,9 +135,12 @@ std::string WikipediaEnrichmentService::FetchExtract(std::string_view query) {
if ((extract_ptr == nullptr) || !extract_ptr->is_string()) { if ((extract_ptr == nullptr) || !extract_ptr->is_string()) {
if (logger_) { if (logger_) {
logger_->Log(LogLevel::Warn, PipelinePhase::UserGeneration, logger_->Log(
{.level = LogLevel::Warn,
.phase = PipelinePhase::UserGeneration,
.message =
std::string("WikipediaService: No extract string found for '") + std::string("WikipediaService: No extract string found for '") +
std::string(query) + "'"); std::string(query) + "'"});
} }
this->extract_cache_.emplace(cache_key, ""); this->extract_cache_.emplace(cache_key, "");
return {}; return {};
@@ -132,10 +149,11 @@ std::string WikipediaEnrichmentService::FetchExtract(std::string_view query) {
// 4. Success // 4. Success
std::string extract(extract_ptr->as_string()); std::string extract(extract_ptr->as_string());
if (logger_) { if (logger_) {
logger_->Log(LogLevel::Info, PipelinePhase::UserGeneration, logger_->Log({.level = LogLevel::Info,
std::string("WikipediaService: Fetched ") + .phase = PipelinePhase::UserGeneration,
.message = std::string("WikipediaService: Fetched ") +
std::to_string(extract.size()) + " chars for '" + std::to_string(extract.size()) + " chars for '" +
std::string(query) + "'"); std::string(query) + "'"});
} }
this->extract_cache_.insert_or_assign(cache_key, extract); this->extract_cache_.insert_or_assign(cache_key, extract);

View File

@@ -10,12 +10,14 @@
#include "services/enrichment/wikipedia_service.h" #include "services/enrichment/wikipedia_service.h"
std::string WikipediaEnrichmentService::GetLocationContext(const Location& loc) { std::string WikipediaEnrichmentService::GetLocationContext(
const Location& loc) {
using namespace std::literals::chrono_literals; using namespace std::literals::chrono_literals;
if (!this->client_) { if (!this->client_) {
if (logger_) { if (logger_) {
logger_->Log(LogLevel::Warn, PipelinePhase::UserGeneration, logger_->Log({.level = LogLevel::Warn,
"Wikipedia client is nullptr."); .phase = PipelinePhase::UserGeneration,
.message = "Wikipedia client is nullptr."});
} }
return {}; return {};
} }
@@ -48,17 +50,20 @@ std::string WikipediaEnrichmentService::GetLocationContext(const Location& loc)
append_extract(FetchExtract(brewing_query)); append_extract(FetchExtract(brewing_query));
append_extract(FetchExtract(beer_query)); append_extract(FetchExtract(beer_query));
if (logger_) { if (logger_) {
logger_->Log(LogLevel::Info, PipelinePhase::UserGeneration, logger_->Log({.level = LogLevel::Info,
std::string("Done fetching for ") + location_query + .phase = PipelinePhase::UserGeneration,
". Sleeping for 10 seconds."); .message = std::string("Done fetching for ") +
location_query + ". Sleeping for 10 seconds."});
} }
std::this_thread::sleep_for(10s); std::this_thread::sleep_for(10s);
} catch (const std::runtime_error& e) { } catch (const std::runtime_error& e) {
if (logger_) { if (logger_) {
logger_->Log(LogLevel::Debug, PipelinePhase::UserGeneration, logger_->Log(
std::string("WikipediaService lookup failed for '") + {.level = LogLevel::Debug,
location_query + "': " + e.what()); .phase = PipelinePhase::UserGeneration,
.message = std::string("WikipediaService lookup failed for '") +
location_query + "': " + e.what()});
} }
} }
return result; return result;

View File

@@ -16,10 +16,4 @@
LogProducer::LogProducer(BoundedChannel<LogEntry>& channel) LogProducer::LogProducer(BoundedChannel<LogEntry>& channel)
: channel_(channel) {} : channel_(channel) {}
void LogProducer::Log(LogLevel level, PipelinePhase phase, void LogProducer::Log(const LogEntry& entry) { channel_.Send(entry); }
const std::string_view message) {
channel_.Send(LogEntry{.timestamp = std::chrono::system_clock::now(),
.level = level,
.phase = phase,
.message = std::string(message)});
}

View File

@@ -6,6 +6,7 @@
#include "services/prompting/prompt_directory.h" #include "services/prompting/prompt_directory.h"
#include <chrono>
#include <filesystem> #include <filesystem>
#include <fstream> #include <fstream>
#include <stdexcept> #include <stdexcept>
@@ -48,9 +49,12 @@ PromptDirectory::PromptDirectory(const std::filesystem::path& prompt_dir,
} }
if (logger_) { if (logger_) {
logger_->Log(LogLevel::Info, PipelinePhase::Startup, logger_->Log(
{.level = LogLevel::Info,
.phase = PipelinePhase::Startup,
.message =
std::string("[PromptDirectory] Resolved prompt directory: ") + std::string("[PromptDirectory] Resolved prompt directory: ") +
prompt_dir_.string()); prompt_dir_.string()});
} }
} }
@@ -84,10 +88,11 @@ std::string PromptDirectory::Load(std::string_view key) {
} }
if (logger_) { if (logger_) {
logger_->Log(LogLevel::Info, PipelinePhase::Startup, logger_->Log({.level = LogLevel::Info,
std::string("[PromptDirectory] Loaded prompt '") + key_str + .phase = PipelinePhase::Startup,
"' from '" + file_path.string() + "' (" + .message = std::string("[PromptDirectory] Loaded prompt '") +
std::to_string(content.size()) + " chars)"); key_str + "' from '" + file_path.string() + "' (" +
std::to_string(content.size()) + " chars)"});
} }
cache_.emplace(key_str, content); cache_.emplace(key_str, content);

View File

@@ -7,6 +7,7 @@
#include <httplib.h> #include <httplib.h>
#include <chrono>
#include <regex> #include <regex>
#include <stdexcept> #include <stdexcept>
#include <string> #include <string>
@@ -40,27 +41,26 @@ std::string HttpWebClient::Get(const std::string& url) {
client.set_follow_location(true); client.set_follow_location(true);
client.set_connection_timeout(kConnectionTimeoutSeconds); client.set_connection_timeout(kConnectionTimeoutSeconds);
client.set_read_timeout(kReadTimeoutSeconds); client.set_read_timeout(kReadTimeoutSeconds);
client.set_default_headers({ client.set_default_headers({{"Accept", "application/json"},
{"Accept", "application/json"}, {"User-Agent", "biergarten-pipeline/1.0"}});
{"User-Agent", "biergarten-pipeline/1.0"}
});
const httplib::Result result = client.Get(path); const httplib::Result result = client.Get(path);
if (!result) { if (!result) {
throw std::runtime_error( throw std::runtime_error("[HttpWebClient] Request failed for URL: " + url +
"[HttpWebClient] Request failed for URL: " + url +
"" + httplib::to_string(result.error())); "" + httplib::to_string(result.error()));
} }
if (result->status < kSuccessMin || result->status >= kSuccessMax) { if (result->status < kSuccessMin || result->status >= kSuccessMax) {
if (logger_) { if (logger_) {
logger_->Log(LogLevel::Error, PipelinePhase::UserGeneration, logger_->Log(
std::string("[HttpWebClient] Request failed for URL: ") + {.level = LogLevel::Error,
url); .phase = PipelinePhase::UserGeneration,
.message =
std::string("[HttpWebClient] Request failed for URL: ") + url});
} }
throw std::runtime_error( throw std::runtime_error("[HttpWebClient] HTTP " +
"[HttpWebClient] HTTP " + std::to_string(result->status) + std::to_string(result->status) +
" for URL: " + url); " for URL: " + url);
} }