refactor: table输出格式重构

llm_web_kit/extractor/html/recognizer/table.py

@@ -100,17 +100,25 @@ def to_content_list_node(self, base_url: str, parsed_content: HtmlElement, raw_h
         # 使用传入的 raw_html_segment 或将 parsed_content 转换为字符串
         if table_type:
             cc_table_type = DocElementType.COMPLEX_TABLE
+            d = {
+                'type': cc_table_type,
+                'content': {
+                    'html': html_content,
+                    'table_nest_level': table_nest_level,
+                    "caption": [],
+                    "footnote": []
+                }
+            }
         else:
             cc_table_type = DocElementType.SIMPLE_TABLE
-        d = {
-            'type': cc_table_type,
-            # 'raw_content': raw_html_segment,
-            'content': {
-                'html': html_content,
-                'is_complex': table_type,
-                'table_nest_level': table_nest_level
+            d = {
+                'type': cc_table_type,
+                'content': {
+                    'html': html_content,
+                    "caption": [],
+                    "footnote": []
+                }
             }
-        }
         return d
 
     def __is_contain_cc_html(self, cc_html: HtmlElement) -> bool:

tests/llm_web_kit/extractor/html/recognizer/assets/recognizer/table_to_content_list_complex_res.json

@@ -2,7 +2,8 @@
     "type": "complex_table",
     "content": {
         "html": "<table><caption>ফেব্রুয়ারি ২০২৪</caption><thead><tr><th>সোম</th><th>মঙ্গল</th><th>বুধ</th><th>বৃহ</th><th>শুক্র</th><th>শনি</th><th>রবি</th></tr></thead><tfoot><tr><td colspan=\\\"3\\\">« জানুয়ারি</td><td></td><td colspan=\\\"3\\\"></td></tr></tfoot><tbody><tr><td colspan=\\\"3\\\"></td><td>১</td><td>২</td><td>৩</td><td>৪</td></tr><tr><td>৫</td><td>৬</td><td>৭</td><td>৮</td><td>৯</td><td>১০</td><td>১১</td></tr><tr><td>১২</td><td>১৩</td><td>১৪</td><td>১৫</td><td>১৬</td><td>১৭</td><td>১৮</td></tr><tr><td>১৯</td><td>২০</td><td>২১</td><td>২২</td><td>২৩</td><td>২৪</td><td>২৫</td></tr><tr><td>২৬</td><td>২৭</td><td>২৮</td><td>২৯</td><td colspan=\\\"3\\\"></td></tr></tbody></table>",
-        "is_complex": true,
-        "table_nest_level": null
+        "table_nest_level": null,
+        "caption": [],
+        "footnote": []
     }
 }

tests/llm_web_kit/extractor/html/recognizer/assets/recognizer/table_to_content_list_simple_res.json

@@ -1 +1 @@
-{"type": "simple_table", "content": {"html": "<table><tr><td>1</td><td>2</td></tr><tr><td>3</td><td>4</td></tr></table>", "is_complex": false}}
+{"type": "simple_table", "content": {"html": "<table><tr><td>1</td><td>2</td></tr><tr><td>3</td><td>4</td></tr></table>", "caption": [], "footnote": []}}

tests/llm_web_kit/extractor/html/recognizer/test_table.py

@@ -131,7 +131,6 @@ def test_table_to_content_list_node_simple(self):
             expect = base_dir.joinpath(test_case['expected'][0])
             expect_json = expect.read_text(encoding='utf-8')
             assert result['type'] == json.loads(expect_json)['type']
-            assert result['content']['is_complex'] == json.loads(expect_json)['content']['is_complex']
             self.assertTrue(result['content']['html'].startswith('<table>'))
             self.assertTrue(result['content']['html'].endswith('</table>'))
 

tests/llm_web_kit/extractor/test_extractor_chain.py

@@ -112,13 +112,11 @@ def test_html_pipeline(self):
         # 然后是simple table
         html_content = html_content_list[4]
         self.assertEqual(html_content['type'], DocElementType.SIMPLE_TABLE)
-        self.assertEqual(html_content['content']['is_complex'], False)
         assert html_content['content']['html'].startswith('<table')
 
         # 然后是complex table
         html_content = html_content_list[5]
         self.assertEqual(html_content['type'], DocElementType.COMPLEX_TABLE)
-        self.assertEqual(html_content['content']['is_complex'], True)
 
         # 然后是list
         html_content = html_content_list[6]
@@ -541,8 +539,8 @@ def test_table_span_error(self):
         test_data = self.data_json[31]
         input_data = DataJson(test_data)
         result = chain.extract(input_data)
-        result_flag = result.get_content_list()._get_data()[0][0]['content']['is_complex']
-        assert result_flag is True
+        result_flag = result.get_content_list()._get_data()[0][0]['type']
+        assert result_flag == "complex_table"
 
     def test_table_colspan_error(self):
         """测试table的colspan标签为字符串引起的异常错误."""
@@ -551,8 +549,8 @@ def test_table_colspan_error(self):
         test_data = self.data_json[32]
         input_data = DataJson(test_data)
         result = chain.extract(input_data)
-        result_flag = result.get_content_list()._get_data()[0][15]['content']['is_complex']
-        assert result_flag is False
+        result_flag = result.get_content_list()._get_data()[0][15]['type']
+        assert result_flag == "simple_table"
 
     def test_table_colspan_percent_err(self):
         """测试table的colspan标签为百分数引起的异常错误."""
@@ -561,8 +559,8 @@ def test_table_colspan_percent_err(self):
         test_data = self.data_json[33]
         input_data = DataJson(test_data)
         result = chain.extract(input_data)
-        result_flag = result.get_content_list()._get_data()[0][0]['content']['is_complex']
-        assert result_flag is True
+        result_flag = result.get_content_list()._get_data()[0][0]['type']
+        assert result_flag == "complex_table"
 
     def test_table_colspan_str_error(self):
         """测试table的colspan标签为字符串引起的异常错误."""
@@ -571,8 +569,8 @@ def test_table_colspan_str_error(self):
         test_data = self.data_json[34]
         input_data = DataJson(test_data)
         result = chain.extract(input_data)
-        result_flag = result.get_content_list()._get_data()[0][28]['content']['is_complex']
-        assert result_flag is False
+        result_flag = result.get_content_list()._get_data()[0][28]['type']
+        assert result_flag == "simple_table"
 
     def test_table_invalid_percent(self):
         """测试table的colspan标签为百分数引起的异常错误."""
@@ -581,8 +579,8 @@ def test_table_invalid_percent(self):
         test_data = self.data_json[35]
         input_data = DataJson(test_data)
         result = chain.extract(input_data)
-        result_flag = result.get_content_list()._get_data()[0][0]['content']['is_complex']
-        assert result_flag is False
+        result_flag = result.get_content_list()._get_data()[0][0]['type']
+        assert result_flag == "simple_table"
 
     def test_maigc_html(self):
         """测试magic-html."""

tests/llm_web_kit/input/assets/content_json.json

@@ -83,8 +83,8 @@
         "type": "simple_table",
         "content": {
           "html": "<table><tr><td>1.1</td><td>2.1</td></tr><tr><td>3.1</td><td>4.1</td></tr></table>",
-          "is_complex": false,
-          "table_nest_level": "1"
+          "caption": [],
+          "footnote": []
         }
       },
       {
@@ -100,8 +100,9 @@
         "type": "complex_table",
         "content": {
           "html": "<table><tr><td rowspan=\"2\">1</td><td>2</td><td>3</td></tr><tr><td colspan=\"2\">4</td></tr><tr><td>5</td><td>6</td><td>7</td></tr></table>",
-          "is_complex": true,
-          "table_nest_level": "1"
+          "table_nest_level": "1",
+          "caption": [],
+          "footnote": []
         }
       },
       {