Backport changes from the Quamotion repository by qmfrederik · Pull Request #18 · DiscUtils/DiscUtils

qmfrederik · 2017-05-01T12:25:42Z

A couple of PRs were merged into quamotion/discutils before discutils/discutils was forked; this PR adds those changes back.

Fixes #1

@LordMike Can you have a first look at this and see if there's anything you object to?

Thanks!

open lvm logical volume check lvm metadata checksum

* initial XFS implementation * read XFS AG inode B+tree info * add xfs B+tree header * xfs B+tree inode structure * xfs: read inodes * xfs read local & extent file content * xfs: Block Directory structures * - fixed block directory offsets - fixed filename encoding - added B+tree extents - added support for large directories and heavily fragmented files * cleanup

* support disk usage for Ntfs * support disk usage for Fat * support disk usage for Ext * parse jbd2 journal superblock to exclude journal size from available size * use ulong for fs size * - support ext4 64Bit block groups - use ext BlockDescriptor to calc free blocks * add FS Size to base class remove marker interface * add size support for Nfs3 * add size support for NativeFileSystem * fix CLS warning * add missing file * cleanup

LordMike

Only deletion is the core function to get a base / relative path. That has to keep functioning.

The rest seems fine (glanced across it - it already passed review elsewhere).

LordMike · 2017-05-01T20:34:18Z

-                merged = @"\\" + merged;
-            }
-            else if (basePath.StartsWith(@"\", StringComparison.Ordinal))
+            if (basePath.StartsWith(@"\") && merged[1].Equals(':'))


We risk exceptions here. Right? - the merged[1] part.

Sure, added a merged.Length > 2 check (to also account for the merged.SubString part on the next line)

This function in all seems way shorter than the original. Are we certain it produces the same results?

It's a lot shorter because it uses Path.GetFullPath to do most of the work instead of doing the work in custom code - so that's what you're seeing here. The only difference with Path.GetFullPath is that the root is always \ instead of e.g. C:\.

The original documentation said it would return relative paths in some situations.. This would always return full paths?

If the base path is relative, the return value will also be relative. I had a look, this code is mainly used by the Vhd/Vhdx parts which comes from Windows land so Path.GetFullPath should be OK here, plus it was used by @zivillian when he wrote the new code so I think it's safe to guess this is OK.

LordMike · 2017-05-01T20:35:23Z

        public static bool Is8Dot3Char(char ch)
        {
-            return (ch >= 'A' && ch <= 'Z') || (ch >= '0' && ch <= '9') || "_^$~!#%�-{}()@'`&".IndexOf(ch) != -1;
+            return (ch >= 'A' && ch <= 'Z') || (ch >= '0' && ch <= '9') || "_^$~!#%£-{}()@'`&".IndexOf(ch) != -1;


What changed here?

Most likely the line ending

My bad, it's the encoding of the £ sign which has changed. You can see it in the GitHub notification e-mails, in the original code the £ is not printed, whereas in the updated code it is.

VS complained that some of the files had inconsistent line endings and also saved everything as UTF-8, so I guess that's why it shows up in the Github diff now.

I noticed the mail part too - didn't think of it.
The file is UTF-8 now, and the sign is correctly set?

Would suck to suddenly skip a certain character because the source file used a different symbol for it ... :)

Yes, it's UTF-8 now (with BOM) and the £ is uniquely defined in UTF-8. The reason it was mangled in the original e-mail is because the £ sign does not appear in ASCII, so it was encoded using one of the Windows-specific codepages. Should be all good now.

LordMike · 2017-05-01T20:37:40Z

        {
            MemoryStream ms = new MemoryStream();
-            XdrDataWriter writer = StartCallMessage(ms, _client.Credentials, 1);
+            XdrDataWriter writer = StartCallMessage(ms, _client.Credentials, NfsProc3.GetAttr);


Great. We love enums 👍

Credits go to @zivillian for that - rather impressive additions to the library!

Includes PRs: #17, #18, #21, #22, #23, #27, #30, #31, #33, #34, #35, #36, #38, #39, #40, #41, #48, #51, #52, #55, #60

Bianco Veigel and others added 5 commits May 1, 2017 20:09

LVM2 header & metadata parsing

559874c

open lvm logical volume check lvm metadata checksum

Fix VHD relative path resolution for diff disks.

127e0bd

Add KDNET BCD well known elements

f97f7dc

qmfrederik mentioned this pull request May 1, 2017

Unit Test seem to be broken quamotion/discutils#15

Closed

LordMike requested changes May 1, 2017

View reviewed changes

Be more defensive in Utilities.ResolveRelativePath

c3acd88

LordMike approved these changes May 1, 2017

View reviewed changes

LordMike merged commit 178f491 into DiscUtils:master May 1, 2017

LordMike added a commit that referenced this pull request Aug 16, 2017

Release 0.13.0-alpha

f3efdf9

Includes PRs: #17, #18, #21, #22, #23, #27, #30, #31, #33, #34, #35, #36, #38, #39, #40, #41, #48, #51, #52, #55, #60

qmfrederik deleted the quamotion branch December 22, 2017 15:46

Conversation

qmfrederik commented May 1, 2017

Uh oh!

LordMike left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants