bella-openapi 1.0.2.4__tar.gz → 1.0.3__tar.gz

{bella_openapi-1.0.2.4/src/bella_openapi.egg-info → bella_openapi-1.0.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: bella-openapi
-Version: 1.0.2.4
+Version: 1.0.3
 Summary: client for openapi service.
 Home-page: 
 Author: ['tangxiaolong', 'fanqiangwei', 'zhangxiaojia', 'liumin', 'wangyukun']

{bella_openapi-1.0.2.4 → bella_openapi-1.0.3}/setup.py

@@ -3,7 +3,7 @@ from setuptools import setup, find_packages
 
 SHORT = "client for openapi service."
 
-__version__ = "1.0.2.4"
+__version__ = "1.0.3"
 __author__ = ["tangxiaolong", "fanqiangwei", "zhangxiaojia", 'liumin', 'wangyukun']
 __email__ = ''
 readme_path = 'README.md'

{bella_openapi-1.0.2.4 → bella_openapi-1.0.3}/src/bella_openapi/__init__.py

@@ -3,19 +3,8 @@ from .log import operation_log, submit_log
 from .openapi_contexvar import trace_id_context, caller_id_context, request_url_context
 from .auth_billing import ErrorInfo, async_authenticate_decorator_args, authenticate_user, print_context, \
 get_context, set_context, clean_context, report
-from .entity import (
-    StandardDomTree,
-    StandardNode,
-    SourceFile,
-    StandardPosition,
-    StandardImage,
-    Cell,
-    StandardRow,
-    StandardBaseElement,
-    StandardElement,
-    StandardTableElement,
-    StandardImageElement
-)
+from .entity import StandardDomTree, StandardNode, SourceFile, StandardPosition, StandardImage, Cell, \
+    StandardRow, StandardBaseElement, StandardElement, StandardTableElement, StandardImageElement
 
 __all__ = ["validate_token", "operation_log",
            "support_model",

{bella_openapi-1.0.2.4 → bella_openapi-1.0.3}/src/bella_openapi/entity/standard_domtree.py

@@ -31,16 +31,7 @@ layout_type_mapping = {
     "TableNote": "Text",  # 目前实际解析出来没有
 }
 
-# 自研模型均用gpt-4计算（可能有误差，可忽略）
-def count_tokens(text: str, model: str = "gpt-4") -> int:
-    if not text:
-        return 0
-    encoding = tiktoken.encoding_for_model(model)
-    tokens = encoding.encode(text)
-    # 计算标记列表的长度，即标记的数量
-    token_count = len(tokens)
-    # 返回标记的数量
-    return token_count
+
 
 class SourceFile(BaseModel):
     id: str  # 文件ID，唯一标识符，类型为string
@@ -315,18 +306,16 @@ class StandardDomTree(BaseModel):
                 # 检查前一个节点
                 if i > 0:
                     prev_sibling = node.children[i - 1]
-                    if prev_sibling.element and prev_sibling.element.type == target_type:
-                        # 找到对应类型的前一个兄弟节点，合并节点
-                        if cls._merge_nodes(prev_sibling, current, target_type):
-                            merged = True
+                    # 找到对应类型的前一个兄弟节点，合并节点
+                    merged = ( prev_sibling.element and prev_sibling.element.type == target_type and
+                               cls._merge_nodes(prev_sibling, current, target_type))
 
                 # 如果没有与前一个节点合并，检查后一个节点
                 if not merged and i < len(node.children) - 1:
                     next_sibling = node.children[i + 1]
-                    if next_sibling.element and next_sibling.element.type == target_type:
-                        # 找到对应类型的后一个兄弟节点，合并节点
-                        if cls._merge_nodes(next_sibling, current, target_type):
-                            merged = True
+                    # 找到对应类型的后一个兄弟节点，合并节点
+                    merged = (next_sibling.element and next_sibling.element.type == target_type and
+                              cls._merge_nodes(next_sibling, current, target_type))
 
                 # 如果没有找到对应类型的兄弟节点，将当前节点类型改为 Text
                 if not merged:
@@ -357,23 +346,25 @@ class StandardDomTree(BaseModel):
         Returns:
             bool: 是否成功合并
         """
-        if node_type == 'Figure' and isinstance(target_node.element, StandardImageElement):
-            # 将 FigureName 的文本作为 Figure 的 name
+        # 定义节点类型与元素类型的映射
+        type_element_mapping = {
+            'Figure': StandardImageElement,
+            'Table': StandardTableElement
+        }
+
+        can_merge = (node_type in type_element_mapping and
+            isinstance(target_node.element, type_element_mapping[node_type]))
+
+        # 检查节点类型是否支持且目标节点元素类型匹配
+        if can_merge:
+            # 将源节点的文本作为目标节点的 name
             target_node.element.name = source_node.element.text
             # 更新 tokens 计数
             target_node.tokens += source_node.tokens
-            # 将 FigureName 的位置添加到 Figure 中
+            # 将源节点的位置添加到目标节点中
             target_node.element.positions += source_node.element.positions
-            return True
-        elif node_type == 'Table' and isinstance(target_node.element, StandardTableElement):
-            # 将 TableName 的文本作为 Table 的 name
-            target_node.element.name = source_node.element.text
-            # 更新 tokens 计数
-            target_node.tokens += source_node.tokens
-            # 将 Table 的位置添加到 Figure 中
-            target_node.element.positions += source_node.element.positions
-            return True
-        return False
+
+        return can_merge
 
     @classmethod
     def _from_domtree_node_to_base_info(cls, node: dict) -> Optional[StandardNode]:
@@ -440,7 +431,7 @@ class StandardDomTree(BaseModel):
                                       cell_data['end_col']],
                                 text=cell_text,
                                 # 目前只会有一个元素,且是Text类型，Path重新从头编号，相对cell是root
-                                nodes=[StandardNode(summary="", tokens=count_tokens(cell_text), path=[1], children=[],
+                                nodes=[StandardNode(summary="", tokens=cls.count_tokens(cell_text), path=[1], children=[],
                                                     element=StandardElement(
                                                         type='Text',
                                                         positions=[],
@@ -491,7 +482,7 @@ class StandardDomTree(BaseModel):
                     standard_node.children.append(standard_child)
 
         # 计算 token 数量：自身 text 的 token 数量 + 子节点 token 数量
-        tokens = count_tokens(text)
+        tokens = cls.count_tokens(text)
         for child in standard_node.children:
             tokens += child.tokens
 
@@ -500,3 +491,23 @@ class StandardDomTree(BaseModel):
 
         return standard_node
 
+    @classmethod
+    def count_tokens(cls, text: str) -> int:
+        """
+        计算文本的token数量
+
+        Args:
+            text: 要计算的文本
+
+        Returns:
+            int: token数量
+        """
+        model = "gpt-4" # 使用模型默认为gpt-4
+        if not text:
+            return 0
+        encoding = tiktoken.encoding_for_model(model)
+        tokens = encoding.encode(text)
+        # 计算标记列表的长度，即标记的数量
+        token_count = len(tokens)
+        # 返回标记的数量
+        return token_count

{bella_openapi-1.0.2.4 → bella_openapi-1.0.3/src/bella_openapi.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: bella-openapi
-Version: 1.0.2.4
+Version: 1.0.3
 Summary: client for openapi service.
 Home-page: 
 Author: ['tangxiaolong', 'fanqiangwei', 'zhangxiaojia', 'liumin', 'wangyukun']