Downloading gzipped files in browsers #158
Comments
achingbrain
added a commit
to hugomrdias/playwright-test
that referenced
this issue
Feb 8, 2024
When loading a file that ends in `.gz`, `sirv` will set the `Content-Encoding` header to `gzip` which means browsers will unzip the content before handing it back to `fetch` or `XMLHttpRequest`. This PR adds a workaround to the asset server that sets the header to a garbage value if a file ending in `.gz` has been requested. It's necessary to use a garbage value because `sirv` will only set the header if it's not been set already, so we can't simply delete it. Refs: lukeed/sirv#158 Refs: ipfs/aegir#1462
6 tasks
hugomrdias
pushed a commit
to hugomrdias/playwright-test
that referenced
this issue
Feb 8, 2024
When loading a file that ends in `.gz`, `sirv` will set the `Content-Encoding` header to `gzip` which means browsers will unzip the content before handing it back to `fetch` or `XMLHttpRequest`. This PR adds a workaround to the asset server that sets the header to a garbage value if a file ending in `.gz` has been requested. It's necessary to use a garbage value because `sirv` will only set the header if it's not been set already, so we can't simply delete it. Refs: lukeed/sirv#158 Refs: ipfs/aegir#1462
|
@lukeed would you accept a PR that added a |
|
Hi there, thanks for the report Agree that this should be fixed, however don't think adding a new config option is the solution here. I'll have a window today where I can put together what I have in mind |
|
@achingbrain Can you do me a favor & see if the following change works for you? // circa L112
-- if (enc) headers['Content-Encoding'] = enc;
++ if (enc && ctype) headers['Content-Encoding'] = enc; |
|
Thanks for taking a look. That works, but only because mrmime doesn't currently map |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
If you try to download (for example) a
.tar.gzviasirvin a browser, because the file name ends in.gzthetoHeadersfunction setsContent-Encodingto"gzip".This makes browsers ungzip the file before making the bytes available to the user.
If the intention was to download a
.tar.gzfile things break because they actually get back the equivalent of a.tarfile.The text was updated successfully, but these errors were encountered: