This commit prefers the use of 'defer' for fs.Unlock (and fs.RUnlock)
because it is more idiomatic Go and reduces repetition in the code,
lending to a cleaner code base.
It also switches a few uses of the lock to read-only locks, which should
improve performance of those functions dramatically in certain contexts.
This API takes input XML input in following form.
```
<?xml version="1.0" encoding="UTF-8"?>
<Delete>
<Quiet>true</Quiet>
<Object>
<Key>Key</Key>
</Object>
<Object>
<Key>Key</Key>
</Object>
...
</Delete>
```
and responds the list of successful deletes, list of errors
for all the deleted objects.
```
<?xml version="1.0" encoding="UTF-8"?>
<DeleteResult xmlns="http://s3.amazonaws.com/doc/2006-03-01/">
<Deleted>
<Key>sample1.txt</Key>
</Deleted>
<Error>
<Key>sample2.txt</Key>
<Code>AccessDenied</Code>
<Message>Access Denied</Message>
</Error>
</DeleteResult>
```
Golang 1.6 is default version for the build now.
Additionally set 'GODEBUG=cgocheck=0' for now, until
we fix the erasure coding package.
Readmore here https://tip.golang.org/doc/go1.6#cgo
- Move fs-definitions.go and break them into fs-datatypes.go, fs-bucket-acl.go
and fs-utils.go
- Move api-definitions.go to api-response.go, where they should be.
- Move web-definitions to its related handlers.
Minor optimization.
- Add 1000 entries buffered channel for walkerCh.
- Reset marker after the lexical order has reached and
compare only if the marker is set.
- Fixes initiating parallel uploads, and configs being quickly
re-written by another incoming request.
- Parallel uploads work smoothly now and return expected behavior.
Fail createObject() if a file already exists and one attempts
to create a prefix/directory by same name.
Send an approriate error back to the client as 409 Conflict.
Currently the server would set 'application/octet-stream' for all
objects, set this value based on the file extension transparently.
This is useful in case of minio browser to facilitate displaying
proper icons for the different mime data types.
Incoming request params in presigned can come in different order
for different implementations. Rather than verifying a full string
we should verify individual params instead.
This patch fixes an incompatibility issue with AWS SDK Java.
Fixes#1059 - Thanks to @notnoopci for reporting this problem.
listObjects was returning inconsistent results, i.e missing
entries during recursive and non-recursive listing. This led
to 'mc mirror' copying contents repeatedly consisdering
these files to be missing on the destination.
This patch addresses this problem - fixes#1056
HMAC is a much simpler implementation, providing the same
benefits as RSA, avoids additional steps and keeps the code
simpler.
This patch also additionally
- Implements PutObjectURL API.
- GetObjectURL, PutObjectURL take TargetHost as another
argument for generating URL's for proper target destination.
- Adds experimental TLS support for JSON RPC calls.
Without this change listObjects() goes into an infinite loop for
files which have special characters i.e "++" encoded with "%2B%2B".
We have to unescape and convert them to their native representation
before being used internally.
Fixes#1052
Existing code
```
{
if os.IsNotExist(e) {
e = os.MkdirAll(objectDir, 0700)
if e != nil {
return "", probe.NewError(e)
}
}
return "", probe.NewError(e) ---> Error was here.
}
```
For a successful 'MkdirAll' it would still return an empty uploadID,
but the 'error' would be nil. This would succeed the request but
client would fail.
Fix is to re-arrange the logic. Thanks to Alexander Neumann @fd0, for
reporting this problem.
- limit list buckets to limit only 100 buckets, all uppercase buckets
are now lowercase and work transparently with all calls.
- Change disk.Stat to disk.GetInfo and return back disk.Info{} struct.
- Introduce new ioutils package which implements ReadDirN(path, n),
ReadDirNamesN(path, n)
Brendan Ashworth