User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.4b) Gecko/20030414 Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.4b) Gecko/20030414 Patch follows Reproducible: Always Steps to Reproduce:
Created attachment 121677 [details] [diff] [review] Simple and easy patch Replace GetEntry with Available, no object creation/deletion, just the content-length. As Available is here called directly after opening the stream the returned value is always the real content-length.
Comment on attachment 121677 [details] [diff] [review] Simple and easy patch yup, i like it. this prevents the JAR channel from having to deal directly with nsIZipEntry :-) r=darin
Attachment #121677 - Flags: review?(darin) → review+
over to alfred
Assignee: darin → alfredkayser
Status: UNCONFIRMED → NEW
Ever confirmed: true
Comment on attachment 121677 [details] [diff] [review] Simple and easy patch sr=alecf I wonder if this will allow us to stop #including nsIZipEntry somewhere.
Attachment #121677 - Flags: superreview?(alecf) → superreview+
Created bug 203627 for nsIZipEntry removal investigation. How to apply the patch? Do I need to request CVS access, and then cvs it into the tree?
No longer blocks: 203627
Status: NEW → ASSIGNED
I'll check this in when the tree opens today.
Checked in. Sorry it took me so long.
Status: ASSIGNED → RESOLVED
Last Resolved: 15 years ago
Resolution: --- → FIXED
Quick question: this doesn't depend on uncompressed entries, like Java does?
No, this is about how to get the actual content length of the item, and 'Available' just returns that (when called directly after opening).
OK, I only ask because Java's InflaterInputStream.available() doesn't always.
You need to log in before you can comment on or make changes to this bug.