自动推广
This commit is contained in:
wangjinlei
@@ -4,31 +4,21 @@ namespace app\api\controller;
|
||||
|
||||
use think\Cache;
|
||||
use think\Db;
|
||||
use GuzzleHttp\Client;
|
||||
use think\Validate;
|
||||
use app\common\ExpertFinderService;
|
||||
|
||||
class ExpertFinder extends Base
|
||||
{
|
||||
private $httpClient;
|
||||
private $ncbiBaseUrl = 'https://eutils.ncbi.nlm.nih.gov/entrez/eutils/';
|
||||
private $service;
|
||||
|
||||
public function __construct(\think\Request $request = null)
|
||||
{
|
||||
parent::__construct($request);
|
||||
$this->httpClient = new Client([
|
||||
'timeout' => 180,
|
||||
'connect_timeout' => 15,
|
||||
'verify' => false,
|
||||
]);
|
||||
$this->service = new ExpertFinderService();
|
||||
}
|
||||
|
||||
/**
|
||||
* Main search endpoint
|
||||
* Params:
|
||||
* keyword - search term (e.g. "biomedical engineering")
|
||||
* page - page number, default 1
|
||||
* per_page - articles per page, default 100, max 100
|
||||
* min_year - earliest publication year, default current-3
|
||||
* source - "pubmed" (fast, email from affiliation) or "pmc" (slower, structured email)
|
||||
*/
|
||||
public function search()
|
||||
{
|
||||
@@ -49,16 +39,12 @@ class ExpertFinder extends Base
|
||||
}
|
||||
|
||||
try {
|
||||
if ($source === 'pmc') {
|
||||
$result = $this->searchViaPMC($keyword, $perPage, $minYear, $page);
|
||||
} else {
|
||||
$result = $this->searchViaPubMed($keyword, $perPage, $minYear, $page);
|
||||
}
|
||||
$result = $this->service->searchExperts($keyword, $perPage, $minYear, $page, $source);
|
||||
} catch (\Exception $e) {
|
||||
return jsonError('Search failed: ' . $e->getMessage());
|
||||
}
|
||||
|
||||
$saveResult = $this->saveExperts($result['experts'], $keyword, $source);
|
||||
$saveResult = $this->service->saveExperts($result['experts'], $keyword, $source);
|
||||
$result['saved_new'] = $saveResult['inserted'];
|
||||
$result['saved_exist'] = $saveResult['existing'];
|
||||
|
||||
@@ -69,16 +55,6 @@ class ExpertFinder extends Base
|
||||
|
||||
/**
|
||||
* Get experts from local database
|
||||
* Params:
|
||||
* field - filter by field keyword (searches t_expert_field)
|
||||
* major_id - filter by major_id (searches t_expert_field)
|
||||
* state - filter by state (0-5), -1 for all
|
||||
* keyword - search name/email/affiliation
|
||||
* no_recent - if 1, exclude experts promoted within last N days (default 30)
|
||||
* recent_days - days threshold for no_recent filter, default 30
|
||||
* page - page number, default 1
|
||||
* per_page - items per page, default 20
|
||||
* min_experts - auto-fetch threshold, default 50
|
||||
*/
|
||||
public function getList()
|
||||
{
|
||||
@@ -171,9 +147,6 @@ class ExpertFinder extends Base
|
||||
|
||||
/**
|
||||
* Update expert state
|
||||
* Params:
|
||||
* expert_id - single id or comma-separated ids
|
||||
* state - new state (0-5)
|
||||
*/
|
||||
public function updateState()
|
||||
{
|
||||
@@ -195,9 +168,6 @@ class ExpertFinder extends Base
|
||||
|
||||
/**
|
||||
* Delete expert (soft: set state=5 blacklist, or hard delete)
|
||||
* Params:
|
||||
* expert_id - single id or comma-separated ids
|
||||
* hard - 1 for hard delete, default 0 (blacklist)
|
||||
*/
|
||||
public function deleteExpert()
|
||||
{
|
||||
@@ -221,7 +191,6 @@ class ExpertFinder extends Base
|
||||
|
||||
/**
|
||||
* Export search results to Excel
|
||||
* Same params as search(), exports current page results
|
||||
*/
|
||||
public function export()
|
||||
{
|
||||
@@ -240,11 +209,7 @@ class ExpertFinder extends Base
|
||||
|
||||
if (!$cached) {
|
||||
try {
|
||||
if ($source === 'pmc') {
|
||||
$cached = $this->searchViaPMC($keyword, $perPage, $minYear, $page);
|
||||
} else {
|
||||
$cached = $this->searchViaPubMed($keyword, $perPage, $minYear, $page);
|
||||
}
|
||||
$cached = $this->service->searchExperts($keyword, $perPage, $minYear, $page, $source);
|
||||
Cache::set($cacheKey, $cached, 3600);
|
||||
} catch (\Exception $e) {
|
||||
return jsonError('Search failed: ' . $e->getMessage());
|
||||
@@ -276,16 +241,82 @@ class ExpertFinder extends Base
|
||||
|
||||
// ==================== Cron / Auto Fetch ====================
|
||||
|
||||
/**
|
||||
* Daily cron: auto-fetch experts for every journal's fields via queue
|
||||
*/
|
||||
public function dailyFetchAll()
|
||||
{
|
||||
$journalId = intval($this->request->param('journal_id', 0));
|
||||
$perPage = max(10, intval($this->request->param('per_page', 200)));
|
||||
$source = $this->request->param('source', 'pubmed');
|
||||
$minYear = intval($this->request->param('min_year', date('Y') - 3));
|
||||
|
||||
if ($journalId) {
|
||||
$journals = Db::name('journal')->field("journal_id,issn,title")->where('journal_id', $journalId)->select();
|
||||
} else {
|
||||
$journals = Db::name('journal')->field("journal_id,issn,title")->where('state', 0)->select();
|
||||
}
|
||||
|
||||
if (empty($journals)) {
|
||||
return jsonSuccess(['msg' => 'No active journals found', 'queued' => 0]);
|
||||
}
|
||||
|
||||
$queued = 0;
|
||||
$skipped = 0;
|
||||
$details = [];
|
||||
$todayStart = strtotime(date('Y-m-d'));
|
||||
foreach ($journals as $journal) {
|
||||
$issn = trim($journal['issn'] ?? '');
|
||||
if (empty($issn)) continue;
|
||||
|
||||
$majors = Db::name('major_to_journal')
|
||||
->alias('mtj')
|
||||
->join('t_major m', 'm.major_id = mtj.major_id', 'left')
|
||||
->where('mtj.journal_issn', $issn)
|
||||
->where('mtj.mtj_state', 0)
|
||||
->where("m.pid", "<>", 0)
|
||||
->where('m.major_state', 0)
|
||||
->column('m.major_title');
|
||||
|
||||
$majors = array_unique(array_filter($majors));
|
||||
if (empty($majors)) continue;
|
||||
foreach ($majors as $keyword) {
|
||||
$keyword = trim($keyword);
|
||||
if (empty($keyword)) continue;
|
||||
|
||||
$fetchLog = $this->service->getFetchLog($keyword, $source);
|
||||
if ($fetchLog['last_time'] >= $todayStart) {
|
||||
$skipped++;
|
||||
continue;
|
||||
}
|
||||
|
||||
$delay = $queued * 10;
|
||||
\think\Queue::later($delay, 'app\api\job\FetchExperts@fire', [
|
||||
'field' => $keyword,
|
||||
'source' => $source,
|
||||
'per_page' => $perPage,
|
||||
'min_year' => $minYear,
|
||||
'journal_id' => $journal['journal_id'],
|
||||
], 'FetchExperts');
|
||||
|
||||
$queued++;
|
||||
$details[] = [
|
||||
'journal' => $journal['title'] ?? $journal['journal_id'],
|
||||
'keyword' => $keyword,
|
||||
'delay_s' => $delay,
|
||||
];
|
||||
}
|
||||
}
|
||||
|
||||
return jsonSuccess([
|
||||
'queued' => $queued,
|
||||
'skipped' => $skipped,
|
||||
'details' => $details,
|
||||
]);
|
||||
}
|
||||
|
||||
/**
|
||||
* Cron job: daily fetch experts for given keywords
|
||||
* Params:
|
||||
* keywords - comma-separated keywords (e.g. "biomedical engineering,tissue engineering")
|
||||
* source - pubmed or pmc, default pubmed
|
||||
* per_page - articles per page, default 100
|
||||
* min_year - default current-3
|
||||
*
|
||||
* Uses cache to remember which page was last fetched per keyword,
|
||||
* so each cron run fetches the next page automatically.
|
||||
*/
|
||||
public function cronFetch()
|
||||
{
|
||||
@@ -303,39 +334,14 @@ class ExpertFinder extends Base
|
||||
$report = [];
|
||||
|
||||
foreach ($keywords as $kw) {
|
||||
if (empty($kw)) {
|
||||
continue;
|
||||
}
|
||||
|
||||
$fetchLog = $this->getFetchLog($kw, $source);
|
||||
$page = $fetchLog['last_page'] + 1;
|
||||
if (empty($kw)) continue;
|
||||
|
||||
try {
|
||||
if ($source === 'pmc') {
|
||||
$result = $this->searchViaPMC($kw, $perPage, $minYear, $page);
|
||||
} else {
|
||||
$result = $this->searchViaPubMed($kw, $perPage, $minYear, $page);
|
||||
}
|
||||
|
||||
$saveResult = $this->saveExperts($result['experts'], $kw, $source);
|
||||
|
||||
$nextPage = $result['has_more'] ? $page : 0;
|
||||
$totalPages = isset($result['total_pages']) ? $result['total_pages'] : 0;
|
||||
$this->updateFetchLog($kw, $source, $nextPage, $totalPages);
|
||||
|
||||
$report[] = [
|
||||
'keyword' => $kw,
|
||||
'page' => $page,
|
||||
'experts_found' => $result['total'],
|
||||
'saved_new' => $saveResult['inserted'],
|
||||
'saved_exist' => $saveResult['existing'],
|
||||
'field_enriched' => $saveResult['field_enriched'],
|
||||
'has_more' => $result['has_more'],
|
||||
];
|
||||
$result = $this->service->doFetchForField($kw, $source, $perPage, $minYear);
|
||||
$report[] = $result;
|
||||
} catch (\Exception $e) {
|
||||
$report[] = [
|
||||
'keyword' => $kw,
|
||||
'page' => $page,
|
||||
'error' => $e->getMessage(),
|
||||
];
|
||||
}
|
||||
@@ -357,7 +363,7 @@ class ExpertFinder extends Base
|
||||
}
|
||||
Cache::set($lockKey, 1, 300);
|
||||
|
||||
\think\Queue::push('app\api\job\FetchExperts', [
|
||||
\think\Queue::push('app\api\job\FetchExperts@fire', [
|
||||
'field' => $field,
|
||||
'source' => 'pubmed',
|
||||
'per_page' => 100,
|
||||
@@ -365,459 +371,17 @@ class ExpertFinder extends Base
|
||||
], 'FetchExperts');
|
||||
}
|
||||
|
||||
/**
|
||||
* Internal method: run a fetch for a single keyword (used by both cron and queue job)
|
||||
*/
|
||||
public function doFetchForField($field, $source = 'pubmed', $perPage = 100, $minYear = null)
|
||||
public function mytest()
|
||||
{
|
||||
if ($minYear === null) {
|
||||
$minYear = date('Y') - 3;
|
||||
}
|
||||
|
||||
$fetchLog = $this->getFetchLog($field, $source);
|
||||
$page = $fetchLog['last_page'] + 1;
|
||||
|
||||
if ($source === 'pmc') {
|
||||
$result = $this->searchViaPMC($field, $perPage, $minYear, $page);
|
||||
} else {
|
||||
$result = $this->searchViaPubMed($field, $perPage, $minYear, $page);
|
||||
}
|
||||
|
||||
$saveResult = $this->saveExperts($result['experts'], $field, $source);
|
||||
|
||||
$nextPage = $result['has_more'] ? $page : 0;
|
||||
$totalPages = isset($result['total_pages']) ? $result['total_pages'] : 0;
|
||||
$this->updateFetchLog($field, $source, $nextPage, $totalPages);
|
||||
|
||||
return [
|
||||
'keyword' => $field,
|
||||
'page' => $page,
|
||||
'experts_found' => $result['total'],
|
||||
'saved_new' => $saveResult['inserted'],
|
||||
'saved_exist' => $saveResult['existing'],
|
||||
'field_enriched' => $saveResult['field_enriched'],
|
||||
'has_more' => $result['has_more'],
|
||||
];
|
||||
}
|
||||
|
||||
// ==================== PubMed Search ====================
|
||||
|
||||
private function searchViaPubMed($keyword, $perPage, $minYear, $page = 1)
|
||||
{
|
||||
set_time_limit(600);
|
||||
|
||||
$searchResult = $this->esearch('pubmed', $keyword, $perPage, $minYear, $page);
|
||||
$ids = $searchResult['ids'];
|
||||
$totalArticles = $searchResult['total'];
|
||||
|
||||
if (empty($ids)) {
|
||||
return $this->buildPagedResult([], 0, 0, $totalArticles, $page, $perPage, 'pubmed');
|
||||
}
|
||||
|
||||
$allAuthors = [];
|
||||
$batches = array_chunk($ids, 50);
|
||||
foreach ($batches as $batch) {
|
||||
$xml = $this->efetchWithRetry('pubmed', $batch);
|
||||
if ($xml) {
|
||||
$authors = $this->parsePubMedXml($xml);
|
||||
$allAuthors = array_merge($allAuthors, $authors);
|
||||
}
|
||||
usleep(400000);
|
||||
}
|
||||
|
||||
$experts = $this->aggregateExperts($allAuthors);
|
||||
|
||||
return $this->buildPagedResult($experts, count($experts), count($ids), $totalArticles, $page, $perPage, 'pubmed');
|
||||
}
|
||||
|
||||
// ==================== PMC Search ====================
|
||||
|
||||
private function searchViaPMC($keyword, $perPage, $minYear, $page = 1)
|
||||
{
|
||||
set_time_limit(600);
|
||||
|
||||
$searchResult = $this->esearch('pmc', $keyword, $perPage, $minYear, $page);
|
||||
$ids = $searchResult['ids'];
|
||||
$totalArticles = $searchResult['total'];
|
||||
|
||||
if (empty($ids)) {
|
||||
return $this->buildPagedResult([], 0, 0, $totalArticles, $page, $perPage, 'pmc');
|
||||
}
|
||||
|
||||
$allAuthors = [];
|
||||
$batches = array_chunk($ids, 5);
|
||||
foreach ($batches as $batch) {
|
||||
$xml = $this->efetchWithRetry('pmc', $batch);
|
||||
if ($xml) {
|
||||
$authors = $this->parsePMCXml($xml);
|
||||
$allAuthors = array_merge($allAuthors, $authors);
|
||||
}
|
||||
usleep(500000);
|
||||
}
|
||||
|
||||
$experts = $this->aggregateExperts($allAuthors);
|
||||
|
||||
return $this->buildPagedResult($experts, count($experts), count($ids), $totalArticles, $page, $perPage, 'pmc');
|
||||
}
|
||||
|
||||
// ==================== NCBI API Calls ====================
|
||||
|
||||
private function esearch($db, $keyword, $perPage, $minYear, $page = 1)
|
||||
{
|
||||
$term = $keyword . ' AND ' . $minYear . ':' . date('Y') . '[pdat]';
|
||||
$retstart = ($page - 1) * $perPage;
|
||||
|
||||
$response = $this->httpClient->get($this->ncbiBaseUrl . 'esearch.fcgi', [
|
||||
'query' => [
|
||||
'db' => $db,
|
||||
'term' => $term,
|
||||
'retstart' => $retstart,
|
||||
'retmax' => $perPage,
|
||||
'retmode' => 'json',
|
||||
'sort' => 'relevance',
|
||||
],
|
||||
$data = $this->request->post();
|
||||
$rule = new Validate([
|
||||
"field" => "require"
|
||||
]);
|
||||
|
||||
$data = json_decode($response->getBody()->getContents(), true);
|
||||
$ids = $data['esearchresult']['idlist'] ?? [];
|
||||
$total = intval($data['esearchresult']['count'] ?? 0);
|
||||
|
||||
return ['ids' => $ids, 'total' => $total];
|
||||
}
|
||||
|
||||
private function efetch($db, $ids)
|
||||
{
|
||||
$response = $this->httpClient->post($this->ncbiBaseUrl . 'efetch.fcgi', [
|
||||
'form_params' => [
|
||||
'db' => $db,
|
||||
'id' => implode(',', $ids),
|
||||
'retmode' => 'xml',
|
||||
],
|
||||
]);
|
||||
|
||||
return $response->getBody()->getContents();
|
||||
}
|
||||
|
||||
private function efetchWithRetry($db, $ids, $maxRetries = 3)
|
||||
{
|
||||
for ($attempt = 1; $attempt <= $maxRetries; $attempt++) {
|
||||
try {
|
||||
return $this->efetch($db, $ids);
|
||||
} catch (\Exception $e) {
|
||||
if ($attempt === $maxRetries) {
|
||||
if (count($ids) > 1) {
|
||||
$half = ceil(count($ids) / 2);
|
||||
$firstHalf = array_slice($ids, 0, $half);
|
||||
$secondHalf = array_slice($ids, $half);
|
||||
$xml1 = $this->efetchWithRetry($db, $firstHalf, 2);
|
||||
$xml2 = $this->efetchWithRetry($db, $secondHalf, 2);
|
||||
return $this->mergeXml($xml1, $xml2);
|
||||
}
|
||||
return null;
|
||||
}
|
||||
sleep($attempt * 2);
|
||||
}
|
||||
if (!$rule->check($data)) {
|
||||
return jsonError($rule->getError());
|
||||
}
|
||||
return null;
|
||||
}
|
||||
|
||||
private function mergeXml($xml1, $xml2)
|
||||
{
|
||||
if (empty($xml1)) return $xml2;
|
||||
if (empty($xml2)) return $xml1;
|
||||
return $xml1 . "\n" . $xml2;
|
||||
}
|
||||
|
||||
// ==================== PubMed XML Parsing ====================
|
||||
|
||||
private function parsePubMedXml($xmlString)
|
||||
{
|
||||
$results = [];
|
||||
|
||||
libxml_use_internal_errors(true);
|
||||
$xml = simplexml_load_string($xmlString);
|
||||
if ($xml === false) {
|
||||
return $results;
|
||||
}
|
||||
|
||||
foreach ($xml->PubmedArticle as $article) {
|
||||
$citation = $article->MedlineCitation;
|
||||
$articleData = $citation->Article;
|
||||
|
||||
$title = $this->xmlNodeToString($articleData->ArticleTitle);
|
||||
$pmid = (string) $citation->PMID;
|
||||
|
||||
$journal = '';
|
||||
if (isset($articleData->Journal->Title)) {
|
||||
$journal = (string) $articleData->Journal->Title;
|
||||
}
|
||||
|
||||
if (!isset($articleData->AuthorList->Author)) {
|
||||
continue;
|
||||
}
|
||||
|
||||
foreach ($articleData->AuthorList->Author as $author) {
|
||||
$lastName = (string) ($author->LastName ?? '');
|
||||
$foreName = (string) ($author->ForeName ?? '');
|
||||
$fullName = trim($foreName . ' ' . $lastName);
|
||||
|
||||
if (empty($fullName)) {
|
||||
continue;
|
||||
}
|
||||
|
||||
$email = '';
|
||||
$affiliation = '';
|
||||
|
||||
if (isset($author->AffiliationInfo)) {
|
||||
foreach ($author->AffiliationInfo as $affInfo) {
|
||||
$affText = (string) $affInfo->Affiliation;
|
||||
if (empty($affiliation)) {
|
||||
$affiliation = $affText;
|
||||
}
|
||||
if (empty($email)) {
|
||||
$email = $this->extractEmailFromText($affText);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
if (empty($email)) {
|
||||
continue;
|
||||
}
|
||||
|
||||
$results[] = [
|
||||
'name' => $fullName,
|
||||
'email' => strtolower($email),
|
||||
'affiliation' => $this->cleanAffiliation($affiliation),
|
||||
'article_title' => $title,
|
||||
'article_id' => $pmid,
|
||||
'journal' => $journal,
|
||||
];
|
||||
}
|
||||
}
|
||||
|
||||
return $results;
|
||||
}
|
||||
|
||||
// ==================== PMC XML Parsing ====================
|
||||
|
||||
private function parsePMCXml($xmlString)
|
||||
{
|
||||
$results = [];
|
||||
|
||||
libxml_use_internal_errors(true);
|
||||
$xml = simplexml_load_string($xmlString);
|
||||
if ($xml === false) {
|
||||
return $results;
|
||||
}
|
||||
|
||||
$articles = $xml->article ?? $xml->children();
|
||||
|
||||
foreach ($articles as $article) {
|
||||
if ($article->getName() !== 'article') {
|
||||
continue;
|
||||
}
|
||||
|
||||
$front = $article->front;
|
||||
if (!$front) {
|
||||
continue;
|
||||
}
|
||||
|
||||
$articleMeta = $front->{'article-meta'};
|
||||
if (!$articleMeta) {
|
||||
continue;
|
||||
}
|
||||
|
||||
$title = $this->xmlNodeToString($articleMeta->{'title-group'}->{'article-title'} ?? null);
|
||||
$pmcId = '';
|
||||
if (isset($articleMeta->{'article-id'})) {
|
||||
foreach ($articleMeta->{'article-id'} as $idNode) {
|
||||
if ((string) $idNode['pub-id-type'] === 'pmc') {
|
||||
$pmcId = (string) $idNode;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
$journal = '';
|
||||
if (isset($front->{'journal-meta'}->{'journal-title'})) {
|
||||
$journal = (string) $front->{'journal-meta'}->{'journal-title'};
|
||||
} elseif (isset($front->{'journal-meta'}->{'journal-title-group'}->{'journal-title'})) {
|
||||
$journal = (string) $front->{'journal-meta'}->{'journal-title-group'}->{'journal-title'};
|
||||
}
|
||||
|
||||
$correspEmails = [];
|
||||
if (isset($articleMeta->{'author-notes'})) {
|
||||
$this->extractEmailsFromNode($articleMeta->{'author-notes'}, $correspEmails);
|
||||
}
|
||||
|
||||
$affiliationMap = [];
|
||||
if (isset($articleMeta->{'contrib-group'})) {
|
||||
foreach ($articleMeta->{'contrib-group'}->children() as $child) {
|
||||
if ($child->getName() === 'aff') {
|
||||
$affId = (string) ($child['id'] ?? '');
|
||||
$affText = $this->xmlNodeToString($child);
|
||||
if ($affId) {
|
||||
$affiliationMap[$affId] = $affText;
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
if (isset($front->{'article-meta'}->{'aff'})) {
|
||||
foreach ($front->{'article-meta'}->{'aff'} as $aff) {
|
||||
$affId = (string) ($aff['id'] ?? '');
|
||||
$affText = $this->xmlNodeToString($aff);
|
||||
if ($affId) {
|
||||
$affiliationMap[$affId] = $affText;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
if (!isset($articleMeta->{'contrib-group'})) {
|
||||
continue;
|
||||
}
|
||||
|
||||
foreach ($articleMeta->{'contrib-group'}->contrib as $contrib) {
|
||||
$contribType = (string) ($contrib['contrib-type'] ?? '');
|
||||
if ($contribType !== 'author') {
|
||||
continue;
|
||||
}
|
||||
|
||||
$nameNode = $contrib->name;
|
||||
if (!$nameNode) {
|
||||
continue;
|
||||
}
|
||||
$surname = (string) ($nameNode->surname ?? '');
|
||||
$givenNames = (string) ($nameNode->{'given-names'} ?? '');
|
||||
$fullName = trim($givenNames . ' ' . $surname);
|
||||
|
||||
if (empty($fullName)) {
|
||||
continue;
|
||||
}
|
||||
|
||||
$email = '';
|
||||
if (isset($contrib->email)) {
|
||||
$email = strtolower(trim((string) $contrib->email));
|
||||
}
|
||||
|
||||
$affiliation = '';
|
||||
if (isset($contrib->xref)) {
|
||||
foreach ($contrib->xref as $xref) {
|
||||
if ((string) $xref['ref-type'] === 'aff') {
|
||||
$rid = (string) $xref['rid'];
|
||||
if (isset($affiliationMap[$rid])) {
|
||||
$affiliation = $affiliationMap[$rid];
|
||||
break;
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
if (empty($affiliation) && isset($contrib->aff)) {
|
||||
$affiliation = $this->xmlNodeToString($contrib->aff);
|
||||
}
|
||||
|
||||
$isCorresponding = false;
|
||||
if (isset($contrib->xref)) {
|
||||
foreach ($contrib->xref as $xref) {
|
||||
if ((string) $xref['ref-type'] === 'corresp') {
|
||||
$isCorresponding = true;
|
||||
}
|
||||
}
|
||||
}
|
||||
if ((string) ($contrib['corresp'] ?? '') === 'yes') {
|
||||
$isCorresponding = true;
|
||||
}
|
||||
|
||||
if (empty($email) && $isCorresponding && !empty($correspEmails)) {
|
||||
$email = $correspEmails[0];
|
||||
}
|
||||
|
||||
if (empty($email)) {
|
||||
$extracted = $this->extractEmailFromText($affiliation);
|
||||
if ($extracted) {
|
||||
$email = $extracted;
|
||||
}
|
||||
}
|
||||
|
||||
if (empty($email)) {
|
||||
continue;
|
||||
}
|
||||
|
||||
$results[] = [
|
||||
'name' => $fullName,
|
||||
'email' => strtolower($email),
|
||||
'affiliation' => $this->cleanAffiliation($affiliation),
|
||||
'article_title' => $title,
|
||||
'article_id' => $pmcId,
|
||||
'journal' => $journal,
|
||||
];
|
||||
}
|
||||
}
|
||||
|
||||
return $results;
|
||||
}
|
||||
|
||||
// ==================== Pagination ====================
|
||||
|
||||
private function buildPagedResult($experts, $expertCount, $articlesScanned, $totalArticles, $page, $perPage, $source)
|
||||
{
|
||||
$totalPages = $totalArticles > 0 ? ceil($totalArticles / $perPage) : 0;
|
||||
|
||||
return [
|
||||
'experts' => $experts,
|
||||
'total' => $expertCount,
|
||||
'articles_scanned' => $articlesScanned,
|
||||
'total_articles' => $totalArticles,
|
||||
'page' => $page,
|
||||
'per_page' => $perPage,
|
||||
'total_pages' => $totalPages,
|
||||
'has_more' => $page < $totalPages,
|
||||
'source' => $source,
|
||||
];
|
||||
}
|
||||
|
||||
// ==================== Aggregation ====================
|
||||
|
||||
private function aggregateExperts($authorRecords)
|
||||
{
|
||||
$map = [];
|
||||
|
||||
foreach ($authorRecords as $record) {
|
||||
$key = strtolower(trim($record['email']));
|
||||
if (empty($key)) {
|
||||
continue;
|
||||
}
|
||||
|
||||
if (!isset($map[$key])) {
|
||||
$map[$key] = [
|
||||
'name' => $record['name'],
|
||||
'email' => $record['email'],
|
||||
'affiliation' => $record['affiliation'],
|
||||
'paper_count' => 0,
|
||||
'papers' => [],
|
||||
];
|
||||
}
|
||||
|
||||
$map[$key]['paper_count']++;
|
||||
|
||||
if (count($map[$key]['papers']) < 10) {
|
||||
$map[$key]['papers'][] = [
|
||||
'title' => $record['article_title'],
|
||||
'article_id' => $record['article_id'],
|
||||
'journal' => $record['journal'],
|
||||
];
|
||||
}
|
||||
|
||||
if (empty($map[$key]['affiliation']) && !empty($record['affiliation'])) {
|
||||
$map[$key]['affiliation'] = $record['affiliation'];
|
||||
}
|
||||
}
|
||||
|
||||
$experts = array_values($map);
|
||||
|
||||
usort($experts, function ($a, $b) {
|
||||
return $b['paper_count'] - $a['paper_count'];
|
||||
});
|
||||
|
||||
return $experts;
|
||||
$res = $this->service->doFetchForField($data['field'], "pubmed", 100, date('Y') - 3);
|
||||
return jsonSuccess($res);
|
||||
}
|
||||
|
||||
// ==================== Excel Export ====================
|
||||
@@ -880,188 +444,4 @@ class ExpertFinder extends Base
|
||||
'count' => count($experts),
|
||||
]);
|
||||
}
|
||||
|
||||
// ==================== Database Storage ====================
|
||||
|
||||
private function saveExperts($experts, $field, $source)
|
||||
{
|
||||
$inserted = 0;
|
||||
$existing = 0;
|
||||
$fieldEnrich = 0;
|
||||
|
||||
foreach ($experts as $expert) {
|
||||
$email = strtolower(trim($expert['email']));
|
||||
if (empty($email)) {
|
||||
continue;
|
||||
}
|
||||
|
||||
$exists = Db::name('expert')->where('email', $email)->find();
|
||||
|
||||
if ($exists) {
|
||||
$existing++;
|
||||
$fieldEnrich += $this->enrichExpertField($exists['expert_id'], $field);
|
||||
continue;
|
||||
}
|
||||
|
||||
$insert = [
|
||||
'name' => mb_substr($expert['name'], 0, 255),
|
||||
'email' => mb_substr($email, 0, 128),
|
||||
'affiliation' => mb_substr($expert['affiliation'], 0, 128),
|
||||
'source' => mb_substr($source, 0, 128),
|
||||
'ctime' => time(),
|
||||
'ltime' => 0,
|
||||
'state' => 0,
|
||||
];
|
||||
|
||||
try {
|
||||
$expertId = Db::name('expert')->insertGetId($insert);
|
||||
$this->enrichExpertField($expertId, $field);
|
||||
$inserted++;
|
||||
} catch (\Exception $e) {
|
||||
$existing++;
|
||||
}
|
||||
}
|
||||
|
||||
return ['inserted' => $inserted, 'existing' => $existing, 'field_enriched' => $fieldEnrich];
|
||||
}
|
||||
|
||||
// ==================== Fetch Log (t_expert_fetch) ====================
|
||||
|
||||
private function getFetchLog($field, $source)
|
||||
{
|
||||
$log = Db::name('expert_fetch')
|
||||
->where('field', $field)
|
||||
->where('source', $source)
|
||||
->find();
|
||||
|
||||
if (!$log) {
|
||||
return ['last_page' => 0, 'total_pages' => 0, 'last_time' => 0];
|
||||
}
|
||||
|
||||
return $log;
|
||||
}
|
||||
|
||||
private function updateFetchLog($field, $source, $lastPage, $totalPages)
|
||||
{
|
||||
$exists = Db::name('expert_fetch')
|
||||
->where('field', $field)
|
||||
->where('source', $source)
|
||||
->find();
|
||||
|
||||
if ($exists) {
|
||||
Db::name('expert_fetch')
|
||||
->where('expert_fetch_id', $exists['expert_fetch_id'])
|
||||
->update([
|
||||
'last_page' => $lastPage,
|
||||
'total_pages' => $totalPages,
|
||||
'last_time' => time(),
|
||||
]);
|
||||
} else {
|
||||
Db::name('expert_fetch')->insert([
|
||||
'field' => mb_substr($field, 0, 128),
|
||||
'source' => mb_substr($source, 0, 128),
|
||||
'last_page' => $lastPage,
|
||||
'total_pages' => $totalPages,
|
||||
'last_time' => time(),
|
||||
]);
|
||||
}
|
||||
}
|
||||
|
||||
private function enrichExpertField($expertId, $field)
|
||||
{
|
||||
$field = trim($field);
|
||||
if (empty($field)) {
|
||||
return 0;
|
||||
}
|
||||
|
||||
$exists = Db::name('expert_field')
|
||||
->where('expert_id', $expertId)
|
||||
->where('field', $field)
|
||||
->where('state', 0)
|
||||
->find();
|
||||
|
||||
if ($exists) {
|
||||
return 0;
|
||||
}
|
||||
|
||||
Db::name('expert_field')->insert([
|
||||
'expert_id' => $expertId,
|
||||
'major_id' => 0,
|
||||
'field' => mb_substr($field, 0, 128),
|
||||
'state' => 0,
|
||||
]);
|
||||
|
||||
return 1;
|
||||
}
|
||||
|
||||
// ==================== Helper Methods ====================
|
||||
|
||||
private function extractEmailFromText($text)
|
||||
{
|
||||
if (empty($text)) {
|
||||
return '';
|
||||
}
|
||||
|
||||
if (preg_match('/[Ee]lectronic address:\s*([^\s;,]+@[^\s;,]+)/', $text, $m)) {
|
||||
return strtolower(trim($m[1], '.'));
|
||||
}
|
||||
|
||||
if (preg_match('/[Ee]-?mail:\s*([^\s;,]+@[^\s;,]+)/', $text, $m)) {
|
||||
return strtolower(trim($m[1], '.'));
|
||||
}
|
||||
|
||||
if (preg_match('/\b([a-zA-Z0-9._%+\-]+@[a-zA-Z0-9.\-]+\.[a-zA-Z]{2,})\b/', $text, $m)) {
|
||||
return strtolower(trim($m[1], '.'));
|
||||
}
|
||||
|
||||
return '';
|
||||
}
|
||||
|
||||
private function extractEmailsFromNode($node, &$emails)
|
||||
{
|
||||
if ($node === null) {
|
||||
return;
|
||||
}
|
||||
|
||||
foreach ($node->children() as $child) {
|
||||
if ($child->getName() === 'email') {
|
||||
$email = strtolower(trim((string) $child));
|
||||
if (!empty($email) && !in_array($email, $emails)) {
|
||||
$emails[] = $email;
|
||||
}
|
||||
}
|
||||
$this->extractEmailsFromNode($child, $emails);
|
||||
}
|
||||
|
||||
$text = (string) $node;
|
||||
if (preg_match_all('/\b([a-zA-Z0-9._%+\-]+@[a-zA-Z0-9.\-]+\.[a-zA-Z]{2,})\b/', $text, $matches)) {
|
||||
foreach ($matches[1] as $email) {
|
||||
$email = strtolower(trim($email, '.'));
|
||||
if (!in_array($email, $emails)) {
|
||||
$emails[] = $email;
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
private function cleanAffiliation($text)
|
||||
{
|
||||
$text = preg_replace('/\s*[Ee]lectronic address:\s*[^\s;,]+@[^\s;,]+/', '', $text);
|
||||
$text = preg_replace('/\s*[Ee]-?mail:\s*[^\s;,]+@[^\s;,]+/', '', $text);
|
||||
$text = preg_replace('/\s*\b[a-zA-Z0-9._%+\-]+@[a-zA-Z0-9.\-]+\.[a-zA-Z]{2,}\b/', '', $text);
|
||||
$text = trim($text, " \t\n\r\0\x0B.,;");
|
||||
return $text;
|
||||
}
|
||||
|
||||
private function xmlNodeToString($node)
|
||||
{
|
||||
if ($node === null) {
|
||||
return '';
|
||||
}
|
||||
|
||||
$xml = $node->asXML();
|
||||
$text = strip_tags($xml);
|
||||
$text = html_entity_decode($text, ENT_QUOTES | ENT_XML1, 'UTF-8');
|
||||
return trim(preg_replace('/\s+/', ' ', $text));
|
||||
}
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user