diff --git a/gen_index.py b/gen_index.py
index 8efa237..7b3e3dd 100644
--- a/gen_index.py
+++ b/gen_index.py
@@ -1,40 +1,184 @@
 # Generate nav part of mkdocs.yaml
 # from markdown file titles in file/
 import os
+import re
 import yaml
 
 # base: index_base.yaml (with everything except nav)
 # result: mkdocs.yaml 
 
+authorname_chars = 'abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789-_'
+
+def tokenize(text):
+    token_specification = [
+        ('SECTION_START', r'\{section\}'),
+        ('SECTION_END', r'\{section end\}'),
+        ('HEADER', r'(?<!#)# [^\r\n]*\r?\n'),
+        ('CONTENT', r'(?:[^#\{]+|(?:##+ [^\r\n]*\r?\n))+')
+    ]
+    tok_regex = '|'.join(f'(?P<{pair[0]}>{pair[1]})' for pair in token_specification)
+    get_token = re.compile(tok_regex)
+    
+    tokens = []
+    for match in get_token.finditer(text):
+        kind = match.lastgroup
+        value = match.group()
+        tokens.append((kind, value))
+    return tokens
+
+def parse(tokens):
+    stack = []
+    ast = []
+    current_section = None
+
+    for kind, value in tokens:
+        if kind == 'SECTION_START':
+            new_section = []
+            section_node = {'section': new_section}
+            if current_section is not None:
+                current_section.append(section_node)
+            else:
+                ast.append(section_node)
+            stack.append(current_section)
+            current_section = new_section
+        elif kind == 'SECTION_END':
+            if not stack:
+                raise SyntaxError("Unmatched section end.")
+            current_section = stack.pop()
+        else:
+            if current_section is not None:
+                current_section.append((kind, value))
+            else:
+                ast.append((kind, value))
+    
+    if stack:
+        raise SyntaxError("Unmatched section start.")
+    
+    return ast
+
 def get_markdown_files(directory):
     """Get a list of all markdown files in the specified directory."""
-    return [f for f in os.listdir(directory) if f.endswith('.md')]
+    if __debug__:
+        return [f for f in os.listdir(directory) if f.endswith('.md')]
+    return [f for f in os.listdir(directory) if f.endswith('.md') and f != 'guide_template.md']
+
+def match_guide_info(content):
+    """Match the author name, date, and author nickname in the content."""
+    title = re.search(r'title: (.+)', content).group(1)
+    date = re.search(r'date: (.+)', content).group(1)
+    author = re.search(r'author: (.+)', content).group(1)
+    author_nickname = re.search(r'author_nickname: (.+)', content).group(1)
+    if any([c not in authorname_chars for c in author]):
+        raise ValueError('Invalid author name')
+    return title, date, author, author_nickname
+
+def create_page(content, file_dir, title = "", date = "", author_nickname = ""):
+    if title != "":
+        # create mkdocs !!!note block at the beginning
+        content = f"!!! note\n    - 作者: {author_nickname}\n    - 日期: {date}\n\n" + content
+    # create file
+    with open(file_dir, 'w', encoding='utf-8') as f:
+        f.write(content)
+
+def append_content_to_file(content, file_dir):
+    with open(file_dir, 'a', encoding='utf-8') as f:
+        f.write(content)
+
+def generate_by_ast(ast, directory, title, date, author_nickname, page_id: list, section_id: list):
+    for node in ast:
+        if isinstance(node, dict):
+            section_id.append(section_id[-1] + 1)
+            page_id.append(0)
+            if section_id[-1] != 0:
+                dir_with_section = directory + f's{section_id[-1]}/'
+                os.mkdir(dir_with_section)
+            else:
+                dir_with_section = directory
+            generate_by_ast(node['section'], dir_with_section, title, date, author_nickname, page_id, section_id)
+            section_id.pop()
+            page_id.pop()
+        else:
+            kind, value = node
+            if kind == 'HEADER':
+                # TODO: 匹配 section 标题, 有些时候 (在 {section 底下的时候)
+                # header 不是新建页面的标题, 而是 nav 一个层级的标题
+                page_id[-1] += 1
+                if not value.isspace():
+                    create_page(value, directory + f'p{page_id[-1]}.md', title, date, author_nickname)
+            elif kind == 'CONTENT':
+                if page_id[-1] != 0:
+                    append_content_to_file(value, directory + f'p{page_id[-1]}.md')
+    # TODO: 生成 nav
+            
+def split_content_to_files(content, title, date, author, author_nickname, directory):
+    # the following code may be extremely ugly
+    # because I haven't learned complier theory
+    # A better way is to implement a parser then generate AST
+    
+    """
+    create dir docs/guide/authorname/
+    create sub dir for {section} tag
+    create a markdown file for every level 1 header
+    
+    single page be like:  docs/guide/authoname/p{page id}.md
+    section be like: docs/guide/authoname/s{section id}/p{page id}.md
+    id starts from 1
+    """
+    author_dir = directory + author + '/'
+    if not os.path.exists(author_dir):
+        os.mkdir(author_dir)
+    tokens = tokenize(content)
+    ast = parse(tokens)
+    if __debug__:
+        print(ast)
+    page_id, section_id = [0], [0]
+    current_page = None
+    generate_by_ast(ast, author_dir, title, date, author_nickname, page_id, section_id)
+    return 114514
+    
+
+def generate_single_file_nav(directory, file):
+    with open(directory + file, 'r', encoding='utf-8') as f:
+        content = f.read()
+        title, date, author, author_nickname = match_guide_info(content)
+        if __debug__:
+            print(title, date, author, author_nickname)
+        content = content.split('---', 2)[2].strip()
+        return split_content_to_files(content, title, date, author, author_nickname, directory)
+        
 
 def generate_nav_section(files, directory):
     """Generate the nav section from markdown file titles."""
     nav = []
     for file in files:
-        title = os.path.splitext(file)[0]  # Remove .md extension
-        nav.append({title: os.path.join(directory, file)})
+        nav.append(generate_single_file_nav(directory, file))
     return nav
 
-def merge_with_base_nav(base_file, nav_section):
+def merge_with_base_nav(base_file, guide_section):
     """Merge the base YAML content with the generated nav section."""
-    with open(base_file, 'r') as f:
+    # dump file
+    with open(base_file, 'r', encoding='utf-8') as f:
         base_content = yaml.safe_load(f)
-
-    base_content['nav'] = nav_section
+        # extract '经验分享' section
+        guide_index = base_content['nav'].index({'经验分享': 'leave this empty'})
+        if __debug__:
+            print(base_content['nav'][guide_index])
+        base_content['nav'][guide_index]['经验分享'] = guide_section
+        if __debug__:
+            print(base_content['nav'][guide_index])
     
     return base_content
 
 def save_yaml(content, output_file):
     """Save the content to a YAML file."""
-    with open(output_file, 'w') as f:
-        yaml.dump(content, f, default_flow_style=False)
+    with open(output_file, 'w', encoding='utf-8') as f:
+        yaml.dump(content, f, default_flow_style=False, allow_unicode=True)
 
 def main():
     base_file = 'index_base.yaml'
     output_file = 'mkdocs.yaml'
+    if __debug__:
+        output_file = 'mkdocs_temp.yaml'
     file_dir = 'docs/guide/'
     markdown_files = get_markdown_files(file_dir)
     nav_section = generate_nav_section(markdown_files, file_dir)
@@ -42,6 +186,6 @@ def main():
     save_yaml(merged_content, output_file)
 
 if __name__ == '__main__':
-    print("Finished generating nav section in mkdocs.yaml")
-    # still working in progress...
-    # main()
\ No newline at end of file
+    # main()
+    # still in progress
+    print("Finished generating nav section in mkdocs.yaml")
\ No newline at end of file