Serialization is slow

[petr-stefan-szn]

Hi,

I have a project that creates some calendar events and outputs them in ical representation. Some of them can include attachments as base64 encoded binary data in ATTACH property. Here the serialization is really slow, like several seconds for 10 MB pdf file.

I tracked down the issue to the foldOneLine function (https://github.com/py-vobject/vobject/blob/master/vobject/base.py#L898). If I use something like this, the same output is generated in a fraction of a second.

def customFoldOneLine(outbuf, input, line_length=75):
    chunk_size = line_length - 1
    for start in range(0, len(input), chunk_size):
        prefix = "\r\n " if start > 0 else ""
        outbuf.write(f"{prefix}{input[start:start + chunk_size]}")
    outbuf.write("\r\n")


vobject.base.foldOneLine = customFoldOneLine

I don't think the old complex logic is needed anymore in python3.

[rsb-23]

Have a look at my version.

@petr-stefan-szn
your custom code may break byte chars which cause troubles.

[petr-stefan-szn]

Thanks @rsb-23, I looked at this more and found out that the output should be no more that 75 octets (not characters), so that's one problem with my snippet.

You're right that we should not split multibyte characters in half, byt according to RFC5545 it should not break anytning:

When parsing a content line, folded lines MUST first be unfolded
according to the unfolding procedure described above.

Note: It is possible for very simple implementations to generate
improperly folded lines in the middle of a UTF-8 multi-octet
sequence. For this reason, implementations need to unfold lines
in such a way to properly restore the original sequence.