disabled auto file mime type detection, rely on extension

This commit is contained in:
João 2025-01-09 18:24:26 -03:00
parent 9d9ed0716f
commit 47c21a5edc

View File

@ -273,7 +273,8 @@ class _DocumentConversionInput(BaseModel):
formats: list[InputFormat] = [] formats: list[InputFormat] = []
if isinstance(obj, Path): if isinstance(obj, Path):
mime = filetype.guess_mime(str(obj)) # mime = filetype.guess_mime(str(obj)) # We're having too much conflicts with documents being worngly classified as ZIP
mime = None
if mime is None: if mime is None:
ext = obj.suffix[1:] ext = obj.suffix[1:]
mime = _DocumentConversionInput._mime_from_extension(ext) mime = _DocumentConversionInput._mime_from_extension(ext)
@ -359,7 +360,7 @@ class _DocumentConversionInput(BaseModel):
mime = FormatToMimeType[InputFormat.PPTX][0] mime = FormatToMimeType[InputFormat.PPTX][0]
elif ext in FormatToExtensions[InputFormat.XLSX]: elif ext in FormatToExtensions[InputFormat.XLSX]:
mime = FormatToMimeType[InputFormat.XLSX][0] mime = FormatToMimeType[InputFormat.XLSX][0]
return mime return mime
@staticmethod @staticmethod