Add code/cube3d/training/count_ldr_main_section.py

code/cube3d/training/count_ldr_main_section.py

@@ -0,0 +1,138 @@
+import os
+
+def checkEncoding(filepath):
+    with open(filepath, "rb") as encode_check:
+        encoding = encode_check.readline(3)
+    if encoding == b"\xfe\xff\x00":
+        return "utf_16_be"
+    elif encoding == b"\xff\xfe0":
+        return "utf_16_le"
+    else:
+        return "utf_8"
+
+def readTextFile(filepath):
+    if os.path.exists(filepath):
+        file_encoding = checkEncoding(filepath)
+        try:
+            with open(filepath, "rt", encoding=file_encoding) as f_in:
+                return f_in.readlines()
+        except:
+            with open(filepath, "rt", encoding="latin_1") as f_in:
+                return f_in.readlines()
+    return None
+
+def count_valid_lines(filepath):
+    """
+    统计LDR文件main_section中以"1 "开头的有效行数
+    返回值: 有效行数，如果文件不存在或读取失败则返回-1
+    """
+    if not os.path.isfile(filepath):
+        return -1
+        
+    lines = readTextFile(filepath)
+    if lines is None:
+        return -1
+
+    # 统计main_section中以"1 "开头的有效行数
+    valid_line_count = 0
+    startLine = 0
+    endLine = 0
+    lineCount = 0
+    foundEnd = False
+    in_main_section = False
+    
+    # 首先确定main_section的范围
+    for line in lines:
+        parameters = line.strip().split()
+        if len(parameters) > 2:
+            if parameters[0] == "0" and parameters[1] == "FILE":
+                if not foundEnd:
+                    endLine = lineCount
+                    if endLine > startLine:
+                        # 标记main_section结束
+                        foundEnd = True
+                        break
+                # 标记main_section开始
+                startLine = lineCount
+                foundEnd = False
+                in_main_section = True
+
+            if parameters[0] == "0" and parameters[1] == "NOFILE":
+                endLine = lineCount
+                foundEnd = True
+                in_main_section = False
+                break
+        lineCount += 1
+
+    # 如果没找到结束标记，设置结束行为文件末尾
+    if not foundEnd:
+        endLine = len(lines)
+        in_main_section = True  # 整个文件视为main_section
+
+    # 统计main_section范围内以"1 "开头的行
+    for i in range(startLine, endLine):
+        if lines[i].startswith("1 "):
+            valid_line_count += 1
+    
+    return valid_line_count
+
+def filter_ldr_by_valid_lines(input_dir):
+    """
+    筛选目录下LDR文件，仅打印有效行数(以"1 "开头) >300或<10的文件名
+    """
+    if not os.path.isdir(input_dir):
+        print(f"错误：目录不存在 → {input_dir}")
+        return
+
+    ldr_files = [f for f in os.listdir(input_dir) if f.lower().endswith('.ldr')]
+    if not ldr_files:
+        print(f"提示：在 {input_dir} 中未找到任何.ldr文件")
+        return
+
+    print(f"=== 开始筛选 {input_dir} 下的LDR文件 ===")
+    print(f"总计找到 {len(ldr_files)} 个LDR文件，仅显示符合条件的文件：\n")
+
+    # 分类存储符合条件的文件
+    less_than_10 = []  # 有效行数<10的文件
+    more_than_300 = []  # 有效行数>300的文件
+    failed_files = []   # 读取失败的文件
+
+    for ldr_file in ldr_files:
+        file_path = os.path.join(input_dir, ldr_file)
+        valid_count = count_valid_lines(file_path)
+
+        if valid_count == -1:
+            failed_files.append(ldr_file)
+        else:
+            if valid_count < 30:
+                less_than_10.append((ldr_file, valid_count))
+            elif valid_count > 300:
+                more_than_300.append((ldr_file, valid_count))
+
+    # 打印筛选结果
+    if less_than_10:
+        print("【1】有效行数（1开头）< 30 的文件：")
+        for file, count in less_than_10:
+            print(f"  - {file} → 有效行数：{count}")
+        print()
+
+    if more_than_300:
+        print("【2】有效行数（1开头）> 300 的文件：")
+        for file, count in more_than_300:
+            print(f"  - {file} → 有效行数：{count}")
+        print()
+
+    if failed_files:
+        print(f"【3】读取失败的文件（共{len(failed_files)}个）：")
+        for file in failed_files:
+            print(f"  - {file}")
+
+    print(f"\n=== 筛选完成 ===")
+    print(f"符合条件的文件总数：{len(less_than_10) + len(more_than_300)} 个")
+
+if __name__ == "__main__":
+    # 请修改为你的LDR文件所在目录
+    LDR_DIRECTORY = "/public/home/wangshuo/gap/assembly/data/car_1k/subset/ldr/"
+    
+    # 执行筛选
+    filter_ldr_by_valid_lines(LDR_DIRECTORY)