analyze_mule_flow.py

import os
import yaml
import xml.etree.ElementTree as ET

directory_path = "/path/to/src/main/mule"
global_config_path = "/path/to/src/main/resources/properties/global-config.yaml"

def load_global_config(path):
  # YAMLファイルの読み込み
  with open(path, 'r') as file:
      data = yaml.safe_load(file)
  return data

config_yaml = load_global_config(global_config_path)

def get_nested_element(data, key):
    keys = key.split('.')
    result = data
    for k in keys:
        result = result.get(k)
        if result is None:
            break
    return result

def cache_xml_structures(directory):
    xml_structures = {}

    for filename in os.listdir(directory):
        if filename.endswith(".xml"):
            file_path = os.path.join(directory, filename)
            try:
                tree = ET.parse(file_path)
                xml_structures[filename] = tree.getroot()
            except ET.ParseError as e:
                print(f"Error parsing {filename}: {e}")
    
    return xml_structures

def get_http_request_urls(filename, root, flow_name, nest = 3):
    tab_repeated = '\t' * nest
    filepath = directory_path + '/' + filename
    namespaces = {
        'mule': 'http://www.mulesoft.org/schema/mule/core',
        'http': 'http://www.mulesoft.org/schema/mule/http',
    }

    # 指定された名前のflow要素を探す
    xpath_flow = f".//mule:flow[@name='{flow_name}']"
    flow_elements = root.findall(xpath_flow, namespaces)

    # 該当するflow内のhttp:request要素のurl属性を取得する
    for flow_element in flow_elements:
        print(f"{tab_repeated}File: {filepath}")
        http_requests = flow_element.findall(".//http:request", namespaces)
        print(f"{tab_repeated}Flow: {flow_name}")
        # Find HTTP requests within the flow
        flow_refs = flow_element.findall('.//mule:flow-ref', namespaces)
        for flow_ref in flow_refs:
          name = flow_ref.get('name')
          print(f"{tab_repeated}Flow Ref: {name}")
          for filename, root in xml_structures.items():
            get_http_request_urls(filename, root, name, nest + 1)

        for request in http_requests:
            path_key = request.get('path')
            path = get_nested_element(config_yaml, path_key.replace('${', '').replace('}', ''))
            print(f"{tab_repeated}URL in Flow '{flow_name}': {path}")

def extract_structure_from_router(xml_structures):
    namespaces = {
        'apikit': 'http://www.mulesoft.org/schema/mule/mule-apikit',
    }
    flow_namespaces = {
        'mule': 'http://www.mulesoft.org/schema/mule/core',
    }
    for filename, root in xml_structures.items():
        # print(f"File: {filename}")
        routers = root.findall('.//apikit:router', namespaces)
        for router in routers:
            router_name = router.get('config-ref')
            print(f"Router: {router_name}")

            flows = root.findall('.//mule:flow', flow_namespaces)
            for flow in flows:
                flow_name = flow.get('name')
                print(f"\tFlow: {flow_name}")
                # Find HTTP requests within the flow
                flow_refs = flow.findall('.//mule:flow-ref', flow_namespaces)
                for flow_ref in flow_refs:
                    name = flow_ref.get('name')
                    print(f"\t\tFlow Ref: {name}")
                    for filename, root in xml_structures.items():
                      get_http_request_urls(filename, root, name)

# ディレクトリパスを指定して実行する例
xml_structures = cache_xml_structures(directory_path)
extract_structure_from_router(xml_structures)