如何通过Monkey-patch改造OpenTimelineIO适配FCP XML导入KdenLive？

以下是针对你遇到的剩余错误的Monkey-patch修复代码，以及对前两个修复的标准验证版本：

错误3：'opentimelineio._otio.SerializableCollection' object has no attribute 'find_clips' #

SerializableCollection类本身未实现find_clips方法，当代码遍历项目结构时误将其当作可查找剪辑的容器调用该方法，导致报错。添加以下Monkey-patch为该类补充递归查找剪辑的方法：

import opentimelineio as otio

def serializable_collection_find_clips(self):
    clips = []
    for child in self:
        if hasattr(child, 'find_clips'):
            clips.extend(child.find_clips())
        elif isinstance(child, otio.schema.Clip):
            clips.append(child)
    return clips

otio.schema.SerializableCollection.find_clips = serializable_collection_find_clips

错误4：'NoneType' object has no attribute 'tag'（KdenLive导入时） #

该错误是因为处理XML元素时，某个预期节点为None却被调用了tag方法。以下是两种修复思路：

思路1：Monkey-patch KdenLive适配器的XML解析函数 #

假设错误出现在KdenLive适配器的XML解析环节，替换对应处理函数添加None检查：

from opentimelineio.adapters import kdenlive

# 替换为实际报错的解析函数名，比如_parse_track或_parse_clip
original_parse_func = kdenlive._parse_track

def patched_parse_func(element, *args, **kwargs):
    if element is None:
        return None  # 或返回符合逻辑的默认空结构
    return original_parse_func(element, *args, **kwargs)

kdenlive._parse_track = patched_parse_func

思路2：预处理OTIO结构，清理无效None元素 #

在读取FCPXML后先清理结构，避免None值进入导入流程：

def sanitize_otio_structure(item):
    if isinstance(item, otio.schema.SerializableCollection):
        item[:] = [child for child in item if child is not None]
        for child in item:
            sanitize_otio_structure(child)
    elif hasattr(item, 'children'):
        item.children = [child for child in item.children if child is not None]
        for child in item.children:
            sanitize_otio_structure(child)
    return item

# 使用示例
timeline = otio.adapters.read_from_file("your_project.fcpxml")
sanitized_timeline = sanitize_otio_structure(timeline)
otio.adapters.write_to_file(sanitized_timeline, "output.kdenlive")

前两个错误的标准修复验证 #

确保你之前的Monkey-patch符合以下标准实现：

错误1：fcpx_xml.py的_format_id_for_clip中'NoneType' object has no attribute 'find' #

import opentimelineio.adapters.fcpx_xml as fcpx_xml

original_format_id = fcpx_xml._format_id_for_clip

def patched_format_id(clip_xml, *args, **kwargs):
    if clip_xml is None:
        return ""
    ref = clip_xml.find('ref')
    if ref is None:
        return ""
    return original_format_id(clip_xml, *args, **kwargs)

fcpx_xml._format_id_for_clip = patched_format_id

错误2：fcpx_xml.py的_format_frame_duration中'NoneType' object has no attribute 'split' #

original_format_duration = fcpx_xml._format_frame_duration

def patched_format_duration(duration_str, *args, **kwargs):
    if not duration_str:
        return 0
    return original_format_duration(duration_str, *args, **kwargs)

fcpx_xml._format_frame_duration = patched_format_duration

注意：所有Monkey-patch代码需在调用OTIO读取FCPXML之前加载，确保方法替换生效。

内容的提问来源于stack exchange，提问作者ginjaemocoes