Python re.match整个文件应为类似字节的对象错误

Question

我已经看过How do I re.search or re.match on a whole file without reading it all into memory?和TypeError: sequence item 1: expected a bytes-like object, str found，但这两个问题的解决方案对我来说并不奏效

因此，本文的目标是在文本文件中对模式进行正则表达式搜索，特别是从示例文本文件

ParameterValue = Texture2D'/Game/Characters/Slashers/Bear/Textures/Outfit01/T_BEHead01_BC.T_BEHead01_BC' ParameterValue = Texture2D'/Game/Characters/Slashers/Bear/Textures/Outfit01/T_BEHead01_BC.T_BEHead01_BC'

我想提取“/Game/Characters/Slashers/Bear/Textures/ought01/T_behahead01_BC”和“/Game/Characters/Slashers/Bear/Textures/ought01/T_behahead01_BC”，我知道这是字符串正则表达式模式：

Texture2D\'(.*)\.

和捕获组的一部分。组（1）

相关代码部分为：

     with open(mytool.propstxt_path, 'r+') as f:
            data = mmap.mmap(f.fileno(), 0)
            match_obj = re.match('Texture2D\'(.*)\'', data)
            print(match_obj.group(1))

我使用mmap将整个文本文件读取到内存中，这样数据就像一个普通字符串一样，可以进行正则表达式匹配

但是，我遇到了一个问题：无法在这一行的类似字节的对象上使用类似字符串的模式：match_obj=re.match（'Texture2D'（.*）'，data）

因此，我尝试使用regex和compile将字符串模式转换为类似字节的模式。但不幸的是，这导致match_obj为nothing，因为打印它时返回None，所以我认为下面的代码段不正确

 with open(mytool.propstxt_path, 'r+') as f:
            data = mmap.mmap(f.fileno(), 0)
            regex = rb'Texture2D\'(.*)\''
            pattern = re.compile(regex)
            match_obj = re.match(pattern, data)#'Texture2D\'(.*)\''
            print(match_obj)

我还尝试在模式之前放置一个b以将其转换为字节，但这再次导致match_obj为None，没有属性。group（）

with open(mytool.propstxt_path, 'r+') as f:
            data = mmap.mmap(f.fileno(), 0)
            match_obj = re.match(b'Texture2D\'(.*)\'', data)
            print(match_obj.group(1))

我尝试的最后一件事是使用.decode（）将数据解码到UTF-8。执行以下操作，但返回错误mmap.mmap对象没有属性decode

with open(mytool.propstxt_path, 'r+') as f:
            data = mmap.mmap(f.fileno(), 0).decode('UTF-8')
            match_obj = re.match('Texture2D\'(.*)\'', data)
            print(match_obj.group(1))

我想尝试的最后一件事就是逐行读取整个文件，因为mmaps意味着您需要使用像regex这样的字节，我不知道在我的情况下该怎么做

我的全部代码如下所示，但警告是针对blender插件、创建面板等

#import all libraries including those needed for regex matching and mapping string to memory 
import bpy, re, mmap

class PathProps(bpy.types.PropertyGroup):
    propstxt_path: bpy.props.StringProperty(name="Select PropsTxt File ", description="Select a File", subtype="FILE_PATH")
    texloc_path: bpy.props.StringProperty(name="Select Exported Game Folder ", description="Select a Folder", subtype="DIR_PATH")

class DBDShaderScript_PT_main_panel(bpy.types.Panel):
    bl_label = "DBD Shader Maps"
    bl_idname = "DBDShaderScript_PT_main_panel"
    bl_space_type = 'VIEW_3D'
    bl_region_type = 'UI'
    bl_category = "DBD Shaders"
    
    def draw(self, context):
        layout = self.layout
        scene = context.scene
        
        mytool = scene.my_tool
        
        layout.prop(mytool, "propstxt_path")
        layout.prop(mytool, "texloc_path")
        
        layout.operator("dbdshaderscript.addbasic_operator")
        

class DBDShaderScript_OT_add_basic(bpy.types.Operator):
    bl_label = "Add Basic Shader Map (Every Material Except Hair)"
    bl_idname = "dbdshaderscript.addbasic_operator"
    def execute(self, context):
        
        #make reference to user inputted file name
        scene = context.scene
        mytool = scene.my_tool
        
        #find the selected material, s_mat stands for selected material
        s_mat = bpy.context.active_object.active_material
        s_mat.use_nodes = True
        
        #store new link function to variable
        link = s_mat.node_tree.links.new
        
        #assign Principled BSDF to a variable so can be referenced later
        #so that nodes can link to it
        principled_node = s_mat.node_tree.nodes.get('Principled BSDF')
        
        #start adding all nodes and respective links to shader map
        srgb_node =  s_mat.node_tree.nodes.new("ShaderNodeSeparateRGB")
        srgb_node.location = (-200,150) # x,y
        link(srgb_node.outputs[2], principled_node.inputs[4])
        
        node =  s_mat.node_tree.nodes.new("ShaderNodeValToRGB")
        node.location = (-500,50) # x,y
        link(node.outputs[0], principled_node.inputs[7])
        
        #node =  s_mat.node_tree.nodes.new("ShaderNodeSeparateRGB")
        #node.location = (-200,100) # x,y
        #link(node.outputs[2], principled_node.inputs[4])
         
        with open(mytool.propstxt_path, 'r+') as f:
            data = mmap.mmap(f.fileno(), 0)
            match_obj = re.match('Texture2D\'(.*)\'', data)
            print(match_obj.group(1))
        
        #add image texture nodes
        texImage = s_mat.node_tree.nodes.new('ShaderNodeTexImage')
        #example loading image
        #texImage.image = bpy.data.images.load("C:\\Users\\myName\\Downloads\\Textures\\Downloaded\\flooring5.jpg")
        
        #uncomment this
        #texImage.image = bpy.data.images.load(match_obj.group(1)[0])
        
        
        
        #make a link to principle BSDF node
        link(texImage.outputs[0], principled_node.inputs[0])
        
        return {"FINISHED"}


#register + unregister all classes with a loop
classes = [PathProps, DBDShaderScript_PT_main_panel, DBDShaderScript_OT_add_basic]
 
 
 
def register():
    for cls in classes:
        bpy.utils.register_class(cls)
        
        #create pointer to all properties
        bpy.types.Scene.my_tool = bpy.props.PointerProperty(type = PathProps)
 
def unregister():
    for cls in classes:
        bpy.utils.unregister_class(cls)
        del bpy.types.Scene.my_tool
 
 
if __name__ == "__main__":
    register()

Answer 1

好的，所以我解决了我的问题，不完全使用MMAP，因为它们要求按位正则表达式很烦人，我不喜欢。所以我所做的只是将整个文件读入一个字符串，当然这对较长的文件不起作用，但这对我的目的来说是很好的

这是我的解决方案，我刚换了这个

#with just means open and close file
with open(mytool.propstxt_path, 'r') as f:
    #read entire file to one string
    data = f.read()
    #find all matches through regex to the string Texture2D' with capture group 
    #any character zero to unlimited times and ending with '
    #also store capture groups into a list variable
    match_list = re.findall("Texture2D\'(.*)\'", data)