add “check code syntax”

app.py

@@ -5,6 +5,7 @@ import json
 from typing import Dict, List, Optional, Tuple
 
 import gradio as gr
+from loguru import logger
 
 import sambanova_gradio
 from sambanova_gradio import get_fn
@@ -19,6 +20,54 @@ YOUR_API_TOKEN = os.getenv('SAMBANOVA_API_KEY')
 History = List[Tuple[str, str]]
 Messages = List[Dict[str, str]]
 
+from tree_sitter_languages import get_parser
+def check_code_error(code, language='python'):
+    language_ext = {'go': 'go', 
+                    'html': 'html', 
+                    'python': 'python', 
+                    'cpp': 'cpp', 
+                    # 'sql': 'sql', 
+                    'php': 'php', 
+                    'javascript': 'javascript', 
+                    # 'R': 'r', 
+                    'typescript': 'typescript', 
+                    'ruby': 'ruby', 
+                    'bash': 'bash', 
+                    'rust': 'rust', 
+                    # 'D': 'd', 
+                    'scala': 'scala', 
+                    'csharp': 'c_sharp', 
+                    'java': 'java',
+                    'json': 'json',
+                    'julia': 'julia',
+                    'css': 'css',
+                    'jsdoc': 'jsdoc',
+                    'swift': 'swift',
+                    'cli': 'cli',
+                    }    
+    if language not in language_ext.keys():
+        return {'result': True, 'error': 'no language in tree_sitter_languages'}
+    try:
+        parser = get_parser(language_ext[language])
+        code_bytes = code.encode('utf-8')
+        tree = parser.parse(code_bytes)
+        root_node = tree.root_node
+        
+        visit_list = [root_node]
+        ans = []
+        while visit_list:
+            node = visit_list.pop() 
+            if node.type == 'ERROR':
+                return {'result': True, 'error': True, 'root_node': root_node}
+            is_inner = True if node.children else False
+            ans.append([node, is_inner])
+            for child in node.children:
+                visit_list.append(child)
+
+        return {'result': True, 'error': False, 'root_node': root_node}
+    except Exception as e:
+        return {'result': False, 'error': str(e)}
+
 def history_to_messages(history: History, system: str) -> Messages:
     messages = [{'role': 'system', 'content': system}]
     for h in history:
@@ -33,8 +82,8 @@ def messages_to_history(messages: Messages) -> History:
         history.append([q['content'], r['content']])
     return history
 
-def remove_code_block(text):
-    pattern = r'```html\n(.+?)\n```'
+def remove_code_block(text, language='html'):
+    pattern = r'```' + language.lower() + '\n(.+?)\n```'
     match = re.search(pattern, text, re.DOTALL)
     if match:
         return match.group(1).strip()
@@ -75,8 +124,7 @@ def get_item_list(locale="zh", programming_language="", category="", difficulty=
     for sample in samples:
         labels = sample['labels']
         content = [sample['content']]
-        title = f"{labels['programming_language']}/{labels['category']}/{labels['difficulty']}"
-        sample_dict = [title, content, labels['programming_language']=='html']
+        sample_dict = [labels['programming_language'], f"/{labels['category']}/{labels['difficulty']}/", content]
         if (locale != "" and labels['locale'] == locale): 
             if programming_language == "":
                 item = labels['programming_language']
@@ -113,9 +161,21 @@ item_list_select, item_list_select_dropdown, item_samples = get_item_list(locale
 def get_input_examples(): 
     input_examples = []
     for demo in DEMO_LIST:
-        input_examples.append([demo['title'], demo['description'], True])
+        input_examples.append(['html', demo['title'], demo['description']])
     return input_examples
 
+def code_syntax_check_fun(code, language='html'):
+    check_result = check_code_error(code, language=language.lower())
+    if check_result['result'] == True:
+        if check_result['error'] == True:
+            return 'No pass'
+        elif check_result['error'] == False:
+            return 'Pass'
+        else:
+            return 'No check [no language in check_languages]'
+    else:
+        return check_result['error']
+    
 with gr.Blocks(css_paths="app.css", title="Qwen2.5 Coder Artifacts (+SambaNova+FullStackBench)") as demo:
     history_state = gr.State([])
     setting = gr.State({
@@ -123,14 +183,14 @@ with gr.Blocks(css_paths="app.css", title="Qwen2.5 Coder Artifacts (+SambaNova+F
     })
 
     with gr.Row():
-        use_system_prompt_example = gr.Checkbox(show_label=False, value=True, visible=False)
+        code_language_example = gr.Textbox(show_label=False, lines=1, container=False, visible=False)
         input_text_title= gr.TextArea(show_label=False, container=False, visible=False)
         input_text_example = gr.TextArea(show_label=False, container=False, visible=False)  
         with gr.Tab("Examples_HTML"):   
             input_examples = get_input_examples()
             gr.Examples(
                     examples=input_examples,
-                    inputs=[input_text_title, input_text_example, use_system_prompt_example],
+                    inputs=[code_language_example, input_text_title, input_text_example],
                     outputs=None,
                     label='',
                     examples_per_page=10,
@@ -165,7 +225,7 @@ with gr.Blocks(css_paths="app.css", title="Qwen2.5 Coder Artifacts (+SambaNova+F
             with gr.Row():
                 test_examples = gr.Examples(
                         examples=item_samples,
-                        inputs=[input_text_title, input_text_example, use_system_prompt_example],
+                        inputs=[code_language_example, input_text_title, input_text_example],
                         outputs=None,
                         label='Examples:',
                         examples_per_page=5,
@@ -234,7 +294,7 @@ with gr.Blocks(css_paths="app.css", title="Qwen2.5 Coder Artifacts (+SambaNova+F
                         """)
 
                         input = antd.InputTextarea(size="middle", allow_clear=True, placeholder="Please enter what kind of application you want")
-                        use_system_prompt = gr.Checkbox(value=True, label="use system prompt[for HTML]?")
+                        code_language = gr.Textbox(label="code language", lines=1)
                         btn = antd.Button("send", type="primary", size="large")                        
 
                         with antd.Flex(gap="small", wrap=True):
@@ -271,16 +331,6 @@ with gr.Blocks(css_paths="app.css", title="Qwen2.5 Coder Artifacts (+SambaNova+F
                     history_drawer.close(lambda: gr.update(
                         open=False), inputs=[], outputs=[history_drawer])
 
-                # with antd.Col(span=24, md=16):
-                #     with ms.Div(elem_classes="right_panel"):
-                #         gr.HTML('<div class="render_header"><span class="header_btn"></span><span class="header_btn"></span><span class="header_btn"></span></div>')
-                #         with antd.Tabs(active_key="empty", render_tab_bar="() => null") as state_tab:
-                #             with antd.Tabs.Item(key="empty"):
-                #                 empty = antd.Empty(description="empty input", elem_classes="right_content")
-                #             with antd.Tabs.Item(key="loading"):
-                #                 loading = antd.Spin(True, tip="coding...", size="large", elem_classes="right_content")
-                #             with antd.Tabs.Item(key="render"):
-                #                 sandbox = gr.HTML(elem_classes="html_content")
                 with antd.Col(span=24, md=16):
                     with ms.Div(elem_classes="right_panel"):
                         gr.HTML('<div class="render_header"><span class="header_btn"></span><span class="header_btn"></span><span class="header_btn"></span></div>')
@@ -293,8 +343,10 @@ with gr.Blocks(css_paths="app.css", title="Qwen2.5 Coder Artifacts (+SambaNova+F
                                 sandbox = gr.HTML(elem_classes="html_content")
                             with antd.Tabs.Item(key="render_other", label="Render_Code") as render_other:
                                 sandbox_other = legacy.Markdown()
+                    with antd.Row():
+                        code_syntax_check = gr.Textbox(label="check code syntax: ", lines=1, interactive=False)
 
-        def generation_code(query: Optional[str], use_system_prompt, _setting: Dict[str, str], _history: Optional[History]):
+        def generation_code(query: Optional[str], code_language, _setting: Dict[str, str], _history: Optional[History]):
             if query is None:
                 query = ''
             if _history is None:
@@ -302,7 +354,7 @@ with gr.Blocks(css_paths="app.css", title="Qwen2.5 Coder Artifacts (+SambaNova+F
 
             # Prepare the preprocess and postprocess functions
             def preprocess(message, history):
-                if use_system_prompt:
+                if code_language.lower() == 'html':
                     messages = [{'role': 'system', 'content': _setting['system']}]
                 else:
                     messages = [{'role': 'system', 'content': ""}]
@@ -339,7 +391,7 @@ with gr.Blocks(css_paths="app.css", title="Qwen2.5 Coder Artifacts (+SambaNova+F
             assistant_response = response_text
             local_history.append([query, assistant_response])
 
-            code = remove_code_block(assistant_response)
+            code = remove_code_block(assistant_response, language=code_language)
 
             if "```html" in assistant_response:
                 sandbox_val = send_to_sandbox(code)
@@ -354,30 +406,25 @@ with gr.Blocks(css_paths="app.css", title="Qwen2.5 Coder Artifacts (+SambaNova+F
                 render_other_val = gr.update(visible=True)
                 sandbox_other_val = content
 
+            code_syntax_check_val = code_syntax_check_fun(code, language=code_language)
             yield {
                 code_output: assistant_response,
                 history_state: local_history,
                 sandbox: sandbox_val,
-                state_tab:  state_tab_val,
-                code_drawer:  gr.update(open=False),
+                state_tab: state_tab_val,
+                code_drawer: gr.update(open=False),
                 render_html: render_html_val,
                 render_other: render_other_val,
                 sandbox_other: sandbox_other_val,
+                code_syntax_check: code_syntax_check_val,
             }
-            # yield {
-            #     code_output: assistant_response,
-            #     history_state: local_history,
-            #     sandbox: send_to_sandbox(code),
-            #     state_tab: gr.update(active_key="render"),
-            #     code_drawer: gr.update(open=False),
-            # }
-
-        def use_system_prompt_example_change(use_system_prompt):
-            return use_system_prompt            
-        use_system_prompt_example.change(
-                fn=use_system_prompt_example_change,
-                inputs=[use_system_prompt_example], 
-                outputs=[use_system_prompt], 
+
+        def code_language_example_change(code_language):
+            return code_language            
+        code_language_example.change(
+                fn=code_language_example_change,
+                inputs=[code_language_example], 
+                outputs=[code_language], 
             )
         def input_text_example_change(input_text_example):
             return input_text_example            
@@ -388,9 +435,9 @@ with gr.Blocks(css_paths="app.css", title="Qwen2.5 Coder Artifacts (+SambaNova+F
             )
         
         btn.click(generation_code,
-                  inputs=[input, use_system_prompt, setting, history_state],
+                  inputs=[input, code_language, setting, history_state],
                   outputs=[code_output, history_state, sandbox, state_tab, code_drawer,
-                           render_html, render_other, sandbox_other])
+                           render_html, render_other, sandbox_other, code_syntax_check])
 
         clear_btn.click(clear_history, inputs=[], outputs=[history_state])
 

requirements.txt

@@ -2,3 +2,6 @@ dashscope
 modelscope_studio~=1.0.0b
 sambanova_gradio
 openai
+tree-sitter
+tree-sitter-languages
+loguru