use search_json instead of search_regex and parse_json

simplify Array constructor replacement by using backreferences

Co-authored-by: Simon Sawicki <accounts@grub4k.xyz>

test/test_utils.py

@@ -1200,6 +1200,12 @@ class TestUtil(unittest.TestCase):
         self.assertEqual(js_to_json('`${name}"${name}"`', {'name': '5'}), '"5\\"5\\""')
         self.assertEqual(js_to_json('`${name}`', {}), '"name"')
 
+    def test_js_to_json_map_array_constructors(self):
+        self.assertEqual(json.loads(js_to_json('new Map([["a", 5]])')), {'a': 5})
+        self.assertEqual(json.loads(js_to_json('Array(5, 10)')), [5, 10])
+        self.assertEqual(json.loads(js_to_json('new Array(15,5)')), [15, 5])
+        self.assertEqual(json.loads(js_to_json('new Map([Array(5, 10),new Array(15,5)])')), {'5': 10, '15': 5})
+
     def test_extract_attributes(self):
         self.assertEqual(extract_attributes('<e x="y">'), {'x': 'y'})
         self.assertEqual(extract_attributes("<e x='y'>"), {'x': 'y'})

yt_dlp/extractor/iprima.py

@@ -139,11 +139,9 @@ class IPrimaIE(InfoExtractor):
                 nuxt_data, (..., 'content', 'additionals', 'videoPlayId', {str}), get_all=False)
 
         if not video_id:
-            nuxt_data = self._parse_json(
-                self._search_regex(
-                    r'(?s)<script[^>]+\bid=["\']__NUXT_DATA__["\'][^>]+>(.+?)</script>',
-                    webpage, 'nuxt data'),
-                'nuxt data')
+            nuxt_data = self._search_json(
+                r'<script[^>]+\bid=["\']__NUXT_DATA__["\'][^>]*>',
+                webpage, 'nuxt data', None, end_pattern=r'</script>', contains_pattern=r'(?s:.+?)')
 
             video_id = traverse_obj(nuxt_data, lambda _, v: re.fullmatch(r'p\d+', v), get_all=False)
 

yt_dlp/utils/_utils.py

@@ -3234,15 +3234,8 @@ def js_to_json(code, vars={}, *, strict=False):
     def create_map(mobj):
         return json.dumps(dict(json.loads(js_to_json(mobj.group(1) or '[]', vars=vars))))
 
-    def create_array(mobj):
-        return mobj.group(1) + js_to_json(f'[{mobj.group(2)}]', vars=vars) + mobj.group(3)
-
+    code = re.sub(r'(?:new\s+)?Array\((.*?)\)', r'[\g<1>]', code)
     code = re.sub(r'new Map\((\[.*?\])?\)', create_map, code)
-    while True:
-        old_code = code
-        code = re.sub(r'^(.*?)(?:new\s+)?Array\((.*?)\)(.*?)$', create_array, old_code)
-        if old_code == code:
-            break
     if not strict:
         code = re.sub(r'new Date\((".+")\)', r'\g<1>', code)
         code = re.sub(r'new \w+\((.*?)\)', lambda m: json.dumps(m.group(0)), code)