1

application/api/controller/Base.php

@@ -1047,6 +1047,162 @@ class Base extends Controller
         return $this->production_article_obj->insertGetId($insert);
     }
 
+    public function markUnusedReferencesForArticle(int $articleId)
+    {
+        if ($articleId <= 0) return;
+
+        $production = Db::name('production_article')
+            ->where('article_id', $articleId)
+            ->where('state', 0)
+            ->field('p_article_id')
+            ->find();
+        $pArticleId = intval($production['p_article_id'] ?? 0);
+        if ($pArticleId <= 0) return;
+
+        // 1) 收集已使用的 p_refer_id（与 convertArticleMainCitationsToMycite 一致：正文 + 表格 + 图）
+        $usedIds = [];
+
+        // t_article_main：type=0 等为正文；type=2 时引用在 article_main_table，content 可能为空，仍扫描不增加成本
+        $mains = Db::name('article_main')
+            ->where('article_id', $articleId)
+            ->whereIn('state', [0, 2])
+            ->field('content')
+            ->select();
+        foreach ($mains as $row) {
+            $usedIds = array_merge($usedIds, $this->extractMyciteIds((string)($row['content'] ?? '')));
+        }
+
+        // t_article_main_table：type=2 对应表格数据（含 url）
+        $tables = Db::name('article_main_table')
+            ->where('article_id', $articleId)
+            ->where('state', 0)
+            ->field('table_data,html_data,title,note,url')
+            ->select();
+        foreach ($tables as $row) {
+            $usedIds = array_merge($usedIds, $this->extractMyciteIdsFromTableDataJson((string)($row['table_data'] ?? '')));
+            $usedIds = array_merge($usedIds, $this->extractMyciteIds((string)($row['html_data'] ?? '')));
+            $usedIds = array_merge($usedIds, $this->extractMyciteIds((string)($row['title'] ?? '')));
+            $usedIds = array_merge($usedIds, $this->extractMyciteIds((string)($row['note'] ?? '')));
+            $usedIds = array_merge($usedIds, $this->extractMyciteIds((string)($row['url'] ?? '')));
+        }
+
+        // t_article_main_image：type=1 图注等
+        $images = Db::name('article_main_image')
+            ->where('article_id', $articleId)
+            ->where('state', 0)
+            ->field('title,note,url')
+            ->select();
+        foreach ($images as $row) {
+            $usedIds = array_merge($usedIds, $this->extractMyciteIds((string)($row['title'] ?? '')));
+            $usedIds = array_merge($usedIds, $this->extractMyciteIds((string)($row['note'] ?? '')));
+            $usedIds = array_merge($usedIds, $this->extractMyciteIds((string)($row['url'] ?? '')));
+        }
+
+        $usedIds = array_values(array_unique($usedIds));
+
+        // 2) 标记：先全置 0，再把用到的置 1
+        try {
+            Db::name('production_article_refer')
+                ->where('p_article_id', $pArticleId)
+                ->where('state', 0)
+                ->update(['is_used' => 0, 'update_time' => time()]);
+
+            if (!empty($usedIds)) {
+                Db::name('production_article_refer')
+                    ->where('p_article_id', $pArticleId)
+                    ->where('state', 0)
+                    ->whereIn('p_refer_id', $usedIds)
+                    ->update(['is_used' => 1, 'update_time' => time()]);
+            }
+        } catch (\Exception $e) {
+            // 工具方法：不影响主流程，忽略异常（可按需改为记录日志）
+        }
+    }
+    public function extractMyciteIds(string $text): array
+    {
+        if ($text === '') return [];
+        $ids = [];
+        if (preg_match_all('/<\s*mycite\b[^>]*\bdata-id\s*=\s*(["\'])(.*?)\1[^>]*>/iu', $text, $m)) {
+            foreach ($m[2] as $raw) {
+                $raw = trim((string)$raw);
+                if ($raw === '') continue;
+                $parts = preg_split('/\s*,\s*/', $raw);
+                foreach ($parts as $p) {
+                    $p = trim((string)$p);
+                    if ($p === '') continue;
+                    $v = intval($p);
+                    if ($v > 0) $ids[] = $v;
+                }
+            }
+        }
+        $ids = array_values(array_unique($ids));
+        sort($ids);
+        return $ids;
+    }
+
+    /**
+     * table_data：二维数组 JSON [[{text,colspan,rowspan},...],...]；支持双重 JSON 字符串编码。
+     *
+     * @return array|null
+     */
+    protected function decodeTableDataJsonToArray(string $raw): ?array
+    {
+        $raw = trim($raw);
+        if ($raw === '') {
+            return null;
+        }
+        if (preg_match('/^\xEF\xBB\xBF/', $raw)) {
+            $raw = substr($raw, 3);
+        }
+        $decoded = json_decode($raw, true);
+        if (json_last_error() !== JSON_ERROR_NONE) {
+            return null;
+        }
+        if (is_array($decoded)) {
+            return $decoded;
+        }
+        if (is_string($decoded)) {
+            $decoded2 = json_decode($decoded, true);
+            if (json_last_error() === JSON_ERROR_NONE && is_array($decoded2)) {
+                return $decoded2;
+            }
+        }
+        return null;
+    }
+
+    /**
+     * table_data 为 JSON：递归收集所有字符串里的 <mycite>；非合法 JSON 时按整串解析。
+     *
+     * @return int[]
+     */
+    protected function extractMyciteIdsFromTableDataJson(string $tableDataJson): array
+    {
+        $tableDataJson = trim($tableDataJson);
+        if ($tableDataJson === '') {
+            return [];
+        }
+        $decoded = $this->decodeTableDataJsonToArray($tableDataJson);
+        if ($decoded === null) {
+            return $this->extractMyciteIds($tableDataJson);
+        }
+        $ids = [];
+        $walk = function ($node) use (&$walk, &$ids) {
+            if (is_string($node)) {
+                $ids = array_merge($ids, $this->extractMyciteIds($node));
+                return;
+            }
+            if (is_array($node)) {
+                foreach ($node as $v) {
+                    $walk($v);
+                }
+            }
+        };
+        $walk($decoded);
+        $ids = array_values(array_unique($ids));
+        sort($ids);
+        return $ids;
+    }
+
 
 }