mirror of
https://github.com/aaronpo97/the-biergarten-app.git
synced 2026-10-15 01:54:43 +00:00
Move brewery reverse geocoding from C# seeding to C++ pipeline (#249)
This commit is contained in:
@@ -13,10 +13,12 @@ BiergartenPipelineOrchestrator::BiergartenPipelineOrchestrator(
|
||||
std::unique_ptr<DataGenerator> generator,
|
||||
std::unique_ptr<IExportService> exporter,
|
||||
std::unique_ptr<ICuratedDataService> curated_data_service,
|
||||
std::unique_ptr<IAddressService> address_service,
|
||||
const ApplicationOptions& application_options)
|
||||
: logger_(std::move(logger)),
|
||||
context_service_(std::move(context_service)),
|
||||
generator_(std::move(generator)),
|
||||
exporter_(std::move(exporter)),
|
||||
curated_data_service_(std::move(curated_data_service)),
|
||||
address_service_(std::move(address_service)),
|
||||
application_options_(application_options) {}
|
||||
@@ -7,14 +7,21 @@
|
||||
#include <format>
|
||||
#include <optional>
|
||||
#include <random>
|
||||
#include <string>
|
||||
#include <string_view>
|
||||
|
||||
#include "biergarten_pipeline_orchestrator.h"
|
||||
#include "biergarten_pipeline_orchestrator/coords.h"
|
||||
#include "services/address/address_service.h"
|
||||
#include "services/logging/logger.h"
|
||||
|
||||
using biergarten_pipeline_orchestrator_internal::Coords;
|
||||
using biergarten_pipeline_orchestrator_internal::GetRandomCoordsWithinRange;
|
||||
|
||||
namespace {
|
||||
constexpr std::string_view kPlaceholderAddressLine1 = "Address unavailable";
|
||||
} // namespace
|
||||
|
||||
void BiergartenPipelineOrchestrator::GenerateBreweries(
|
||||
std::span<const EnrichedCity> cities) {
|
||||
logger_->Log({.level = LogLevel::Info,
|
||||
@@ -39,10 +46,32 @@ void BiergartenPipelineOrchestrator::GenerateBreweries(
|
||||
const Coords brewery_coords = GetRandomCoordsWithinRange(
|
||||
city_centre, kMaxDistanceFromCentreKm, rng);
|
||||
|
||||
const std::optional<AddressLookupResult> geocoded =
|
||||
address_service_->ReverseGeocode(brewery_coords.longitude,
|
||||
brewery_coords.latitude);
|
||||
|
||||
if (!geocoded.has_value()) {
|
||||
logger_->Log(
|
||||
{.level = LogLevel::Warn,
|
||||
.phase = PipelinePhase::BreweryAndBeerGeneration,
|
||||
.message = std::format(
|
||||
"[Pipeline] Could not reverse geocode brewery in '{}' "
|
||||
"({}, {}); using a placeholder address.",
|
||||
enriched_city.location.city, brewery_coords.latitude,
|
||||
brewery_coords.longitude)});
|
||||
}
|
||||
|
||||
return BreweryRecord{
|
||||
.address = BreweryAddress{.city = enriched_city.location,
|
||||
.longitude = brewery_coords.longitude,
|
||||
.latitude = brewery_coords.latitude},
|
||||
.address =
|
||||
BreweryAddress{
|
||||
.city = enriched_city.location,
|
||||
.longitude = brewery_coords.longitude,
|
||||
.latitude = brewery_coords.latitude,
|
||||
.address_line1 = geocoded.has_value()
|
||||
? geocoded->address_line1
|
||||
: std::string(kPlaceholderAddressLine1),
|
||||
.postal_code =
|
||||
geocoded.has_value() ? geocoded->postal_code : ""},
|
||||
.brewery = brewery};
|
||||
} catch (const std::exception& e) {
|
||||
++skipped_count;
|
||||
|
||||
@@ -166,6 +166,26 @@ int main(const int argc, char** argv) {
|
||||
inj.template create<std::unique_ptr<WebClient>>(),
|
||||
log_producer);
|
||||
}),
|
||||
di::bind<IAddressService>().to(
|
||||
[options, &log_producer](
|
||||
const auto& inj) -> std::unique_ptr<IAddressService> {
|
||||
if (options.generator.mode == GeneratorMode::kMock) {
|
||||
log_producer->Log({.level = LogLevel::Info,
|
||||
.phase = PipelinePhase::Startup,
|
||||
.message = "Address service: mock"});
|
||||
|
||||
return std::make_unique<MockAddressService>();
|
||||
}
|
||||
|
||||
log_producer->Log(
|
||||
{.level = LogLevel::Info,
|
||||
.phase = PipelinePhase::Startup,
|
||||
.message = "Address service: Nominatim"});
|
||||
|
||||
return std::make_unique<NominatimAddressService>(
|
||||
inj.template create<std::unique_ptr<WebClient>>(),
|
||||
log_producer);
|
||||
}),
|
||||
di::bind<DataGenerator>().to([&options, &model_path, &sampling,
|
||||
&prompt_directory,
|
||||
&log_producer](const auto& inj)
|
||||
|
||||
@@ -0,0 +1,133 @@
|
||||
/**
|
||||
* @file services/address/nominatim_address_service.cc
|
||||
* @brief NominatimAddressService implementation.
|
||||
*/
|
||||
|
||||
#include "services/address/nominatim_address_service.h"
|
||||
|
||||
#include <boost/json.hpp>
|
||||
#include <chrono>
|
||||
#include <format>
|
||||
#include <stdexcept>
|
||||
#include <string>
|
||||
#include <thread>
|
||||
#include <utility>
|
||||
|
||||
using namespace boost;
|
||||
|
||||
NominatimAddressService::NominatimAddressService(
|
||||
std::unique_ptr<WebClient> client, std::shared_ptr<ILogger> logger)
|
||||
: client_(std::move(client)), logger_(std::move(logger)) {}
|
||||
|
||||
std::optional<AddressLookupResult> NominatimAddressService::ReverseGeocode(
|
||||
double longitude, double latitude) {
|
||||
if (!client_) {
|
||||
if (logger_) {
|
||||
logger_->Log({.level = LogLevel::Warn,
|
||||
.phase = PipelinePhase::Enrichment,
|
||||
.message = "Nominatim client is nullptr."});
|
||||
}
|
||||
return std::nullopt;
|
||||
}
|
||||
|
||||
const std::string url = std::format(
|
||||
"https://nominatim.openstreetmap.org/reverse?format=jsonv2&lat={}&"
|
||||
"lon={}&zoom=18&addressdetails=1",
|
||||
latitude, longitude);
|
||||
|
||||
std::string body;
|
||||
try {
|
||||
body = client_->Get(url);
|
||||
} catch (const std::runtime_error& e) {
|
||||
if (logger_) {
|
||||
logger_->Log({.level = LogLevel::Warn,
|
||||
.phase = PipelinePhase::Enrichment,
|
||||
.message = std::format(
|
||||
"NominatimAddressService: request failed for "
|
||||
"({}, {}): {}",
|
||||
latitude, longitude, e.what())});
|
||||
}
|
||||
return std::nullopt;
|
||||
}
|
||||
|
||||
{
|
||||
using namespace std::literals::chrono_literals;
|
||||
std::this_thread::sleep_for(1s);
|
||||
}
|
||||
|
||||
system::error_code ec;
|
||||
const json::value doc = json::parse(body, ec);
|
||||
|
||||
if (ec) {
|
||||
if (logger_) {
|
||||
logger_->Log({.level = LogLevel::Warn,
|
||||
.phase = PipelinePhase::Enrichment,
|
||||
.message = std::format(
|
||||
"NominatimAddressService: JSON parse error for "
|
||||
"({}, {}): {}",
|
||||
latitude, longitude, ec.message())});
|
||||
}
|
||||
return std::nullopt;
|
||||
}
|
||||
|
||||
const json::object* obj = doc.if_object();
|
||||
if ((obj == nullptr) || obj->contains("error")) {
|
||||
if (logger_) {
|
||||
logger_->Log(
|
||||
{.level = LogLevel::Warn,
|
||||
.phase = PipelinePhase::Enrichment,
|
||||
.message = std::format("NominatimAddressService: no address "
|
||||
"found for ({}, {})",
|
||||
latitude, longitude)});
|
||||
}
|
||||
return std::nullopt;
|
||||
}
|
||||
|
||||
const json::value* address_ptr = obj->if_contains("address");
|
||||
if ((address_ptr == nullptr) || !address_ptr->is_object()) {
|
||||
if (logger_) {
|
||||
logger_->Log(
|
||||
{.level = LogLevel::Warn,
|
||||
.phase = PipelinePhase::Enrichment,
|
||||
.message = std::format("NominatimAddressService: missing "
|
||||
"'address' for ({}, {})",
|
||||
latitude, longitude)});
|
||||
}
|
||||
return std::nullopt;
|
||||
}
|
||||
|
||||
const json::object& address = address_ptr->get_object();
|
||||
|
||||
const auto extract_string = [&address](std::string_view key) -> std::string {
|
||||
const json::value* value_ptr = address.if_contains(key);
|
||||
if ((value_ptr == nullptr) || !value_ptr->is_string()) {
|
||||
return {};
|
||||
}
|
||||
return std::string(value_ptr->get_string());
|
||||
};
|
||||
|
||||
const std::string house_number = extract_string("house_number");
|
||||
const std::string road = extract_string("road");
|
||||
const std::string postal_code = extract_string("postcode");
|
||||
|
||||
std::string address_line1 = road;
|
||||
if (!house_number.empty()) {
|
||||
address_line1 = address_line1.empty() ? house_number
|
||||
: house_number + " " + road;
|
||||
}
|
||||
|
||||
if (address_line1.empty()) {
|
||||
if (logger_) {
|
||||
logger_->Log(
|
||||
{.level = LogLevel::Warn,
|
||||
.phase = PipelinePhase::Enrichment,
|
||||
.message = std::format("NominatimAddressService: no usable "
|
||||
"street address for ({}, {})",
|
||||
latitude, longitude)});
|
||||
}
|
||||
return std::nullopt;
|
||||
}
|
||||
|
||||
return AddressLookupResult{.address_line1 = address_line1,
|
||||
.postal_code = postal_code};
|
||||
}
|
||||
@@ -184,6 +184,20 @@ uint64_t SqliteExportService::ProcessRecord(const BreweryRecord& brewery) {
|
||||
sqlite_export_service_internal::kBreweryAddressLatitudeBindIndex,
|
||||
.value = brewery.address.latitude,
|
||||
.action = "Failed to bind SQLite brewery address latitude"});
|
||||
sqlite_export_service_internal::Bind(
|
||||
insert_brewery_address_stmt_,
|
||||
sqlite_export_service_internal::BoundParam<std::string_view>{
|
||||
.index =
|
||||
sqlite_export_service_internal::kBreweryAddressLine1BindIndex,
|
||||
.value = brewery.address.address_line1,
|
||||
.action = "Failed to bind SQLite brewery address line 1"});
|
||||
sqlite_export_service_internal::Bind(
|
||||
insert_brewery_address_stmt_,
|
||||
sqlite_export_service_internal::BoundParam<std::string_view>{
|
||||
.index = sqlite_export_service_internal::
|
||||
kBreweryAddressPostalCodeBindIndex,
|
||||
.value = brewery.address.postal_code,
|
||||
.action = "Failed to bind SQLite brewery address postal code"});
|
||||
|
||||
sqlite_export_service_internal::StepStatement(
|
||||
db_handle_, insert_brewery_address_stmt_,
|
||||
|
||||
Reference in New Issue
Block a user