estel/dify
1
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity

chore: the consistency of MultiModalPromptMessageContent (#11721)

Browse Source
This commit is contained in:
非法操作
2024-12-17 15:01:38 +08:00
committed by GitHub
parent 78c3051585
commit c9b4029ce7
14 changed files with 108 additions and 99 deletions
Download Patch File Download Diff File Expand all files Collapse all files

55
api/core/file/file_manager.py
View File

@@ -42,33 +42,31 @@ def to_prompt_message_content(
*,
image_detail_config: ImagePromptMessageContent.DETAIL | None = None,
):
match f.type:
case FileType.IMAGE:
image_detail_config = image_detail_config or ImagePromptMessageContent.DETAIL.LOW
if dify_config.MULTIMODAL_SEND_IMAGE_FORMAT == "url":
data = _to_url(f)
else:
data = _to_base64_data_string(f)
if f.extension is None:
raise ValueError("Missing file extension")
if f.mime_type is None:
raise ValueError("Missing file mime_type")
return ImagePromptMessageContent(data=data, detail=image_detail_config, format=f.extension.lstrip("."))
case FileType.AUDIO:
data = _to_base64_data_string(f)
if f.extension is None:
raise ValueError("Missing file extension")
return AudioPromptMessageContent(data=data, format=f.extension.lstrip("."))
case FileType.VIDEO:
if dify_config.MULTIMODAL_SEND_VIDEO_FORMAT == "url":
data = _to_url(f)
else:
data = _to_base64_data_string(f)
if f.extension is None:
raise ValueError("Missing file extension")
return VideoPromptMessageContent(data=data, format=f.extension.lstrip("."))
case FileType.DOCUMENT:
data = _to_base64_data_string(f)
return DocumentPromptMessageContent(encode_format="base64", data=data, format=f.extension.lstrip("."))
case _:
raise ValueError(f"file type {f.type} is not supported")
params = {
"base64_data": _get_encoded_string(f) if dify_config.MULTIMODAL_SEND_FORMAT == "base64" else "",
"url": _to_url(f) if dify_config.MULTIMODAL_SEND_FORMAT == "url" else "",
"format": f.extension.removeprefix("."),
"mime_type": f.mime_type,
}
if f.type == FileType.IMAGE:
params["detail"] = image_detail_config or ImagePromptMessageContent.DETAIL.LOW
prompt_class_map = {
FileType.IMAGE: ImagePromptMessageContent,
FileType.AUDIO: AudioPromptMessageContent,
FileType.VIDEO: VideoPromptMessageContent,
FileType.DOCUMENT: DocumentPromptMessageContent,
}
try:
return prompt_class_map[f.type](**params)
except KeyError:
raise ValueError(f"file type {f.type} is not supported")
def download(f: File, /):
@@ -122,11 +120,6 @@ def _get_encoded_string(f: File, /):
return encoded_string
def _to_base64_data_string(f: File, /):
encoded_string = _get_encoded_string(f)
return f"data:{f.mime_type};base64,{encoded_string}"
def _to_url(f: File, /):
if f.transfer_method == FileTransferMethod.REMOTE_URL:
if f.remote_url is None: