v2.1.16: 优化搜索功能，增加错误提示检测并简化正则表达式，更新JM发布页 (#102)

Author: hect0x7

src/jmcomic/__init__.py

@@ -2,6 +2,6 @@
 # 被依赖方 <--- 使用方
 # config <--- entity <--- toolkit <--- client <--- option <--- downloader
 
-__version__ = '2.1.15'
+__version__ = '2.1.16'
 
 from .api import *

src/jmcomic/jm_client_impl.py

@@ -418,6 +418,9 @@ def headers_key_ts(self):
             "accept-encoding": "gzip",
         }, key_ts
 
+    def debug_topic_request(self):
+        return 'api'
+
 
 class AsyncSaveImageClient(JmImageClient):
 

src/jmcomic/jm_config.py

@@ -17,11 +17,15 @@ def default_postman_constructor(session, **kwargs):
     return Postmans.new_postman(**kwargs)
 
 
+def default_raise_regex_error(msg, *_args, **_kwargs):
+    raise AssertionError(msg)
+
+
 class JmModuleConfig:
     # 网站相关
     PROT = "https://"
     JM_REDIRECT_URL = f'{PROT}jm365.work/3YeBdF'  # 永久網域，怕走失的小伙伴收藏起来
-    JM_PUB_URL = f'{PROT}jmcomic2.bet'
+    JM_PUB_URL = f'{PROT}jmcomic.ltd'
     JM_CDN_IMAGE_URL_TEMPLATE = PROT + 'cdn-msp.{domain}/media/photos/{photo_id}/{index:05}{suffix}'  # index 从1开始
     JM_IMAGE_SUFFIX = ['.jpg', '.webp', '.png', '.gif']
 
@@ -62,6 +66,9 @@ class JmModuleConfig:
     debug_executor = default_jm_debug
     # postman构造函数
     postman_constructor = default_postman_constructor
+    # 网页正则表达式解析失败时，执行抛出异常的函数，可以替换掉用于debug
+    raise_regex_error_executor = default_raise_regex_error
+
     # debug开关标记
     enable_jm_debug = True
 
@@ -163,7 +170,7 @@ def get_jmcomic_url(cls, postman=None):
         postman = postman or cls.new_postman(session=True)
 
         url = postman.with_redirect_catching().get(cls.JM_REDIRECT_URL)
-        cls.jm_debug('获取禁漫地址', f'[{cls.JM_REDIRECT_URL}] → [{url}]')
+        cls.jm_debug('获取禁漫URL', f'[{cls.JM_REDIRECT_URL}] → [{url}]')
         return url
 
     @classmethod
@@ -181,7 +188,10 @@ def get_jmcomic_domain_all(cls, postman=None):
             raise AssertionError(resp.text)
 
         from .jm_toolkit import JmcomicText
-        return JmcomicText.analyse_jm_pub_html(resp.text)
+        domain_list = JmcomicText.analyse_jm_pub_html(resp.text)
+
+        cls.jm_debug('获取禁漫全部域名', f'[{resp.url}] → {domain_list}')
+        return domain_list
 
     album_comment_headers = {
         'authority': '18comic.vip',

src/jmcomic/jm_toolkit.py

@@ -141,8 +141,12 @@ def match_field(field_key: str, pattern: Union[Pattern, List[Pattern]], text):
             field_value = match_field(field_name, pattern_value, html)
 
             if field_value is None:
-                write_text('./resp.txt', html)  # debug
-                raise AssertionError(f"文本没有匹配上字段：字段名为'{field_name}'，pattern: [{pattern_value.pattern}]")
+                JmModuleConfig.raise_regex_error_executor(
+                    f"文本没有匹配上字段：字段名为'{field_name}'，pattern: [{pattern_value.pattern}]",
+                    html,
+                    field_name,
+                    pattern_value
+                )
 
             # 保存字段
             field_dict[field_name] = field_value
@@ -167,9 +171,7 @@ def format_url(cls, path, domain=None):
 
 class JmSearchSupport:
     # 用来缩减html的长度
-    pattern_html_search_shorten_for = compile('<div class="well well-sm">([\s\S]*)'
-                                              '<div class="row">[\s\S]*'
-                                              '<div class="bot-per visible-xs visible-sm">')
+    pattern_html_search_shorten_for = compile('<div class="well well-sm">([\s\S]*)<div class="row">')
 
     # 用来提取搜索页面的的album的信息
     pattern_html_search_album_info_list = compile(
@@ -185,9 +187,24 @@ class JmSearchSupport:
     # 用来查找tag列表
     pattern_html_search_tag_list = compile('<a href=".*?">(.*?)</a>')
 
+    # 查找错误，例如 [错误，關鍵字過短，請至少輸入兩個字以上。]
+    pattern_html_search_error = compile('<fieldset>\n<legend>(.*?)</legend>\n<div class=.*?>\n(.*?)\n</div>\n</fieldset>')
+
     @classmethod
     def analyse_jm_search_html(cls, html: str) -> JmSearchPage:
-        html = cls.pattern_html_search_shorten_for.search(html)[0]
+        # 检查是否失败
+        match = cls.pattern_html_search_error.search(html)
+        if match is not None:
+            topic, reason = match[1], match[2]
+            JmModuleConfig.raise_regex_error_executor(f'{topic}: {reason}', html)
+
+        # 缩小文本范围
+        match = cls.pattern_html_search_shorten_for.search(html)
+        if match is None:
+            JmModuleConfig.raise_regex_error_executor('未匹配到搜索结果', html)
+        html = match[0]
+
+        # 提取结果
         album_info_list = cls.pattern_html_search_album_info_list.findall(html)
 
         for i, (album_id, title, *args) in enumerate(album_info_list):