调试代码Pair-wise-Similarity-module-master

第一步：运行.py文件生成json文件

问题一：json.decoder.JSONDecodeError: Invalid \escape: line 1 column 31616 (char 31615)

解决：

for dataset in dataset_list:
    with open(datasetmap[dataset] + ".csv", "r") as lines:
        for i, line in enumerate(lines):
            if i == 0:
                continue

            print(f"Processing line {i}: {line.strip()}")

            fid, _, label = re.split(',|\.', line.strip())
            label = label.replace('\n', '')

            if label not in filelists[dataset]:
                folderlist.append(label)
                filelists[dataset][label] = []
                print(f"Checking directory: {os.path.join(data_path, label)}")

                # 检查路径是否存在
                if not exists(os.path.join(data_path, label)):
                    print(f"Directory does not exist: {join(data_path, label)}")
                    continue

                fnames = listdir(os.path.join(data_path, label))
                print(f"Found fnames: {fnames}")

                fname_number = [int(re.split('_|\.', fname)[1]) for fname in fnames if re.split('_|\.', fname)[1].isdigit()]
                sorted_fnames = [fname for fname, _ in sorted(zip(fnames, fname_number), key=lambda f_tuple: f_tuple[1])]
                print(f"Sorted fnames: {sorted_fnames}")

            if fid[-5:].isdigit():
                fid = int(fid[-5:]) - 1
            else:
                print(f"Invalid fid: {fid}")
                continue

            print("Adjusted fid:", fid)
            if fid < 0 or fid >= len(sorted_fnames):
                print(f"Index {fid} is out of range for sorted_fnames")
                continue

            fname = join(data_path, label, sorted_fnames[fid])
            print(f"Appending filename: {fname}")
            filelists[dataset][label].append(fname)

    for key, filelist in filelists[dataset].items():
        cl += 1
        random.shuffle(filelist)
        filelists_flat[dataset] += filelist
        labellists_flat[dataset] += np.repeat(cl, len(filelist)).tolist()

问题二：OSError: [Errno 22] Invalid argument: 'E:\\桌面文件\\Pair-wise-Similarity-module-master\\filelists\\miniImagenet\\train\n07584110\n07584110_9665.JPEG'

解决：\n是转义字符，会导致路径识别错误，我要把json文件中所有的 \ 变成 \\ 。

# 假设json文件路径为
json_file_path = 'E:/code/Pair-wise-Similarity-module-master/filelists\miniImagenet/val.json'

# 读取JSON文件
with open(json_file_path, 'r', encoding='GB2312') as f:
    data = json.load(f)

# 将所有的单斜杠`\`替换成双反斜杠`\\`
modified_data = json.dumps(data).replace('\\', '\\\\')

# 写回到JSON文件中
with open(json_file_path, 'w', encoding='utf-8') as f:
    f.write(modified_data)

print(f"JSON文件 {json_file_path} 中的内容已经修改完成，所有的单斜杠'\\'都被替换成了双反斜杠'\\\\'。")

问题三：SyntaxError: Non-UTF-8 code starting with '\xbc' in file E:\桌面文件\Pair-wise-Similarity-module-master\filelists\miniImagenet\json.py on line 6, but no encoding declared

解决：文件是其他编码类型，需要先测试出文件编码类型，然后修改

encoding='GB2312'

import chardet

# 检测文件是什么类型的编码
with open('E:/code/Pair-wise-Similarity-module-master/filelists\miniImagenet/val.json', 'rb') as f:
    result = chardet.detect(f.read())  # 读取一定量的数据进行编码检测

print(result['encoding'])  # 打印检测到的编码