@tus/server: add Content-Type and Content-Disposition headers on GetHandler.send response #655
Conversation
…andler.send response
…tion headers on GetHandler.send response"
🦋 Changeset detectedLatest commit: b4e7ccd The changes in this PR will be included in the next version bump. This PR includes changesets to release 1 package
Not sure what this means? Click here to learn what changesets are. Click here if you're a maintainer who wants to add another changeset to this PR |
jesusgoku
force-pushed
the
content-disposition
branch
from
3aba5a9 to
9790f70
Compare
| * See: https://datatracker.ietf.org/doc/html/rfc1341 (Page 6) | ||
| */ | ||
| reMimeType = | ||
| /^(?:application|audio|example|font|haptics|image|message|model|multipart|text|video|x-(?:[0-9A-Za-z!#$%&'*+.^_`|~-]+))\/([0-9A-Za-z!#$%&'*+.^_`|~-]+)((?:[ ]*;[ ]*[0-9A-Za-z!#$%&'*+.^_`|~-]+=(?:[0-9A-Za-z!#$%&'*+.^_`|~-]+|"(?:[^"\\]|\.)*"))*)$/ |
There was a problem hiding this comment.
Is it worth it to support parameters as well? We now have a very complex, slow regex with lots of backtracking. If it doesn't add much value I think I prefer to stay with the simple version? What do you think?
There was a problem hiding this comment.
I understand the caution, this last regex is definitely much more intimidating than the previous one, but it is not significantly slower, less than 2ms on average, and this last one analyzing the list of 2138 mime types registered by the IANA.
And answering your question, I think it is important, especially for text files, to preserve and transmit the information of the character set used, and this can be rendered properly when displayed inline.
If you like, we can change to an intermediate version that does not try to match on the parameters, and only looks at the mime type part, after that "anything can come".
Benchmarks
Simple Regex
Allow Parameters and RFC compliance Regex
Non Parameters match Regex
There was a problem hiding this comment.
Parameters in Content-Type are also relevant for some media files. Although I am not sure if necessary, they can contain information about the used codecs in audio/video files (see tus/tusd#1194).
Tusd currently does not support parameters, but I plan an changing this. Its implementation currently also uses a regular expression for checking the media type's validity, but my preferred solution is to replace it with Go's builtin media type parser. I'm not sure if such a method is easily available to you, but maybe it provides another perspective on this topic.
There was a problem hiding this comment.
@jesusgoku wow thanks for taking the time to actually test it. Sounds good to me.
jesusgoku
force-pushed
the
content-disposition
branch
from
9790f70 to
b4e7ccd
Compare
The tusd is the official implementation development in Go lang, and this adds
Content-TypeandContent-Dispositionheaders on response ofGetHandler, this allows browser for:References