124 lines
9.3 KiB
PHP
124 lines
9.3 KiB
PHP
<?php
|
|
|
|
// Private, idempotent geometry rehearsal. This tool cannot target an active database.
|
|
use Illuminate\Contracts\Http\Kernel;
|
|
use Illuminate\Support\Facades\DB;
|
|
use Illuminate\Support\Str;
|
|
|
|
$root = dirname(__DIR__);
|
|
$directory = $root.'/storage/app/private/investmoz-v2/portal-audit';
|
|
$proof = json_decode(file_get_contents($directory.'/../opportunity-resources-backup-proof.json'), true, 512, JSON_THROW_ON_ERROR);
|
|
$target = $proof['restored_target'];
|
|
if (! ($proof['restore_verified'] ?? false) || ! preg_match('/^desk_restore_\d{8}_\d{6}_test$/D', $target['database']) || (int) $target['port'] !== 55441) {
|
|
throw new RuntimeException('Verified isolated QA restore required.');
|
|
}
|
|
foreach (['APP_ENV' => 'testing', 'APP_KEY' => 'base64:AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA=', 'DB_CONNECTION' => 'pgsql', 'DB_HOST' => '127.0.0.1', 'DB_PORT' => '55441', 'DB_DATABASE' => $target['database'], 'DB_USERNAME' => 'postgres', 'DB_PASSWORD' => '', 'DB_URL' => '', 'CACHE_STORE' => 'array', 'SESSION_DRIVER' => 'array'] as $key => $value) {
|
|
putenv($key.'='.$value);
|
|
$_ENV[$key] = $_SERVER[$key] = $value;
|
|
}
|
|
require $root.'/vendor/autoload.php';
|
|
$app = require $root.'/bootstrap/app.php';
|
|
$app->make(Kernel::class)->bootstrap();
|
|
$identity = DB::selectOne('SELECT current_database() AS database,inet_server_port() AS port');
|
|
if (! app()->environment('testing') || $identity->database !== $target['database'] || (int) $identity->port !== 55441) {
|
|
throw new RuntimeException('Non-QA target rejected before writes.');
|
|
}
|
|
$action = $argv[1] ?? 'preview';
|
|
$level = (int) ($argv[2] ?? 1);
|
|
if (! in_array($level, [1, 2, 3], true)) {
|
|
throw new RuntimeException('Administrative levels 1-3 only.');
|
|
}
|
|
if (! in_array($action, ['preview', 'import'], true)) {
|
|
throw new RuntimeException('Only preview/import are supported.');
|
|
}
|
|
$file = $directory.'/moz-adm'.$level.'-ine-20190607.geojson';
|
|
$metadataFile = $directory.'/moz-adm'.$level.'-ine-metadata.html';
|
|
$sourceHash = hash_file('sha256', $file);
|
|
$sourceScan = null;
|
|
if ($action === 'import') {
|
|
$proofs = json_decode(file_get_contents($directory.'/geometry-source-scans.json'), true, 512, JSON_THROW_ON_ERROR);
|
|
foreach ($proofs['files'] as $scan) {
|
|
if ($scan['file'] === basename($file) && $scan['sha256'] === $sourceHash) {
|
|
$sourceScan = ($scan['scan']['result'] ?? null) === 'clean' ? $scan['scan'] : ($scan['gis_scan'] ?? null);
|
|
}
|
|
}
|
|
if (($sourceScan['result'] ?? null) !== 'clean' || strtotime($proofs['captured_at']) < time() - 86400) {
|
|
throw new RuntimeException('Current clean scan tied to the retained original hash required before QA import.');
|
|
}
|
|
}
|
|
$metadata = file_get_contents($metadataFile);
|
|
if (! str_contains($metadata, 'Mozambique INE - Instituto Nacional de Estatistica') || ! str_contains($metadata, '2019-06-07')) {
|
|
throw new RuntimeException('Source identity/version absent from retained metadata.');
|
|
}
|
|
$source = json_decode(file_get_contents($file), true, 512, JSON_THROW_ON_ERROR);
|
|
if (($source['type'] ?? '') !== 'FeatureCollection' || count($source['features'] ?? []) !== [1 => 11, 2 => 159, 3 => 411][$level] || ($source['exceededTransferLimit'] ?? false)) {
|
|
throw new RuntimeException('Incomplete/unexpected provincial dataset.');
|
|
}
|
|
$normalize = fn (string $value) => strtolower(trim(Str::ascii($value)));
|
|
$locations = DB::table('geo_locations')->where('location_type', [1 => 'province', 2 => 'district', 3 => 'admin_post'][$level])->whereNull('deleted_at')->get(['id', 'name', 'parent_id', 'code', 'source_code']);
|
|
$results = [];
|
|
foreach ($source['features'] as $feature) {
|
|
$properties = $feature['properties'];
|
|
$name = $properties['ADM'.$level.'_PT'];
|
|
$parent = $level > 1 ? DB::table('geo_geometries')->where('source_system', 'ine-cod-ab')->where('external_id', '2019-06-07:adm'.($level - 1).':'.$properties['ADM'.($level - 1).'_PCODE'])->whereNull('deleted_at')->value('geo_location_id') : null;
|
|
$matches = $locations->filter(fn ($location) => $normalize($location->name) === $normalize($name) && ($level === 1 || ($parent !== null && $location->parent_id === $parent)))->values();
|
|
$results[] = ['pcode' => $properties['ADM'.$level.'_PCODE'], 'source_name' => $name, 'source_parent_code' => $level > 1 ? $properties['ADM'.($level - 1).'_PCODE'] : null,
|
|
'candidates' => $matches->all(), 'matched_id' => count($matches) === 1 ? $matches[0]->id : null,
|
|
'match_basis' => count($matches) === 1 ? 'unique_exact_name_and_matched_parent_after_case_and_diacritics' : 'unresolved_no_implicit_translation_or_parent_mapping',
|
|
'geometry_type' => $feature['geometry']['type'], 'feature' => $feature];
|
|
}
|
|
$inserted = 0;
|
|
$reused = 0;
|
|
if ($action === 'import') {
|
|
DB::transaction(function () use (&$results, &$inserted, &$reused, $sourceHash, $metadataFile, $level, $sourceScan) {
|
|
DB::select('SELECT pg_advisory_xact_lock(91026, 20190607)');
|
|
foreach ($results as &$row) {
|
|
if ($row['matched_id'] === null) {
|
|
continue;
|
|
}
|
|
$geometry = json_encode($row['feature']['geometry'], JSON_THROW_ON_ERROR);
|
|
$type = strtolower($row['geometry_type']);
|
|
$valid = DB::selectOne('SELECT ST_IsValid(g) AS valid, ST_IsEmpty(g) AS empty, GeometryType(g) AS type FROM (SELECT ST_SetSRID(ST_GeomFromGeoJSON(?),4326) AS g) x', [$geometry]);
|
|
if (! $valid->valid || $valid->empty || ! in_array($type, ['polygon', 'multipolygon'], true)) {
|
|
$row['matched_id'] = null;
|
|
$row['match_basis'] = 'invalid_geometry_no_automatic_repair';
|
|
|
|
continue;
|
|
}
|
|
$external = '2019-06-07:adm'.$level.':'.$row['pcode'];
|
|
$existing = DB::table('geo_geometries')->where('source_system', 'ine-cod-ab')->where('external_id', $external)->first();
|
|
if ($existing) {
|
|
$saved = json_decode($existing->properties, true, 512, JSON_THROW_ON_ERROR);
|
|
if ($existing->geo_location_id !== $row['matched_id'] || ($saved['source_sha256'] ?? null) !== $sourceHash || $existing->deleted_at !== null) {
|
|
throw new RuntimeException('Identity/version conflict; nothing overwritten.');
|
|
}
|
|
$reused++;
|
|
|
|
continue;
|
|
}
|
|
$trace = ['authority' => 'Mozambique INE - Instituto Nacional de Estatistica', 'provider' => 'UNOCHA; exemplar hosted by World Bank', 'source_version' => '2019-06-07',
|
|
'source_url' => 'https://services.arcgis.com/iQ1dY19aHwbSDYIF/arcgis/rest/services/Mozambique/FeatureServer/'.($level + 2),
|
|
'source_sha256' => $sourceHash, 'metadata_sha256' => hash_file('sha256', $metadataFile), 'source_properties' => $row['feature']['properties'],
|
|
'match_basis' => $row['match_basis'], 'qa_only' => true, 'purpose' => 'administrative_boundary', 'legal_currentness' => 'not_determined', 'investment_site' => false, 'source_scan' => $sourceScan];
|
|
DB::insert('INSERT INTO geo_geometries(id,source_system,external_id,geo_location_id,geometry_type,geom,properties,data_classification) VALUES (?,?,?,?,?,ST_SetSRID(ST_GeomFromGeoJSON(?),4326),?::jsonb,?)', [(string) Str::uuid(), 'ine-cod-ab', $external, $row['matched_id'], $type, $geometry, json_encode($trace, JSON_THROW_ON_ERROR), 'internal']);
|
|
$inserted++;
|
|
}
|
|
if ($inserted > 0) {
|
|
DB::table('audit_events')->insert(['source_system' => 'portal-audit-qa', 'entity_type' => 'geo_geometries', 'action' => 'official_geometry_qa_import', 'data_classification' => 'internal', 'after_data' => json_encode(['level' => $level, 'source_sha256' => $sourceHash, 'inserted' => $inserted, 'published' => false], JSON_THROW_ON_ERROR)]);
|
|
}
|
|
});
|
|
}
|
|
foreach ($results as &$row) {
|
|
unset($row['feature']);
|
|
}
|
|
$saved = DB::select("SELECT external_id,geo_location_id,data_classification,ST_IsValid(geom) AS valid,ST_SRID(geom) AS srid,ST_NPoints(geom) AS vertices,properties->>'source_sha256' AS source_sha256 FROM geo_geometries WHERE source_system='ine-cod-ab' ORDER BY external_id");
|
|
$result = ['captured_at' => gmdate('c'), 'action' => $action, 'level' => $level, 'target' => $target, 'source_sha256' => $sourceHash, 'metadata_sha256' => hash_file('sha256', $metadataFile), 'source_features' => count($source['features']),
|
|
'mappings' => $results, 'inserted' => $inserted, 'reused' => $reused, 'recovered_geometries' => $saved, 'active_writes' => 0, 'published' => false, 'visual_verified' => false,
|
|
'remaining' => ['Unmatched names/parents listed per feature; no fuzzy mapping', 'No locality polygons in the identified official exemplar', 'Private QA import does not populate public map layers or investment locations']];
|
|
$destination = $directory.'/geometry-qa-adm'.$level.'-'.$action.'.json';
|
|
if (is_file($destination)) {
|
|
copy($destination, $directory.'/geometry-qa-adm'.$level.'-'.$action.'-'.gmdate('Ymd_His').'-'.bin2hex(random_bytes(3)).'.json');
|
|
}
|
|
file_put_contents($destination, json_encode($result, JSON_PRETTY_PRINT | JSON_UNESCAPED_UNICODE | JSON_THROW_ON_ERROR));
|
|
echo json_encode(['action' => $action, 'level' => $level, 'features' => count($results), 'matched' => count(array_filter($results, fn ($row) => $row['matched_id'] !== null)), 'inserted' => $inserted, 'reused' => $reused, 'persisted_total' => count($saved), 'target' => $target, 'active_writes' => 0], JSON_THROW_ON_ERROR).PHP_EOL;
|