Commit graph

624 commits

Author SHA1 Message Date
b4d2f5c1a6 Each test should test with a unique test file name 2016-10-12 08:38:07 +02:00
600059be82 Add error output 2016-10-12 08:38:07 +02:00
27e816fc69 Do not simply unwrap, but return error in case of error 2016-10-12 08:38:07 +02:00
9a8a2f1c29 Add hook tests for hooks that error with an error that does not abort the action 2016-10-09 18:26:16 +02:00
7050067675 Merge pull request #761 from matthiasbeyer/libimagstore/store-tests-more-succeeding-hook-tests
libimagstore/store-tests: more succeeding hook tests
2016-10-09 13:29:48 +02:00
5202c5112a Add flag for _update() whether precense should be modified
This is a bugfix for an very particular issue.

Here's what happens:

If we create() an FileLockEntry and then update() it, we are running
into a problem because update() calls _update() which changes the
precense status of a FileLockEntry.

Because update() is _consuming_, the FileLockEntry gets drop()ed
afterwards. This _again_ causes _update() to be called, but with a new
presence status, which is not true in this moment (as the FileLockEntry
is still borrowed, but we already marked it as present).

This patch is a short-term fix. The real problem is, that our Store
interface is consuming. If the Store interface would be non-consuming,
this issue wouldn't happen, as the drop() call would not happen.

I'm rather sure that this patch will not be reverted in the process of
rewriting the Store interface to be non-consuming. But we never know.
2016-10-07 21:59:37 +02:00
7b11e7dabb Add testing implementation for Drop for FileLockEntry 2016-10-07 21:59:37 +02:00
d375a6d2c6 Add output to test helper, so we can see in the trace whats happening 2016-10-07 21:59:37 +02:00
ae66b00f5f store tests: Add erroring-hook-tests 2016-10-07 21:59:37 +02:00
095ae19416 [CHERRY-PICK] Add flag for _update() whether precense should be modified
This is a bugfix for an very particular issue.

Here's what happens:

If we create() an FileLockEntry and then update() it, we are running
into a problem because update() calls _update() which changes the
precense status of a FileLockEntry.

Because update() is _consuming_, the FileLockEntry gets drop()ed
afterwards. This _again_ causes _update() to be called, but with a new
presence status, which is not true in this moment (as the FileLockEntry
is still borrowed, but we already marked it as present).

This patch is a short-term fix. The real problem is, that our Store
interface is consuming. If the Store interface would be non-consuming,
this issue wouldn't happen, as the drop() call would not happen.

I'm rather sure that this patch will not be reverted in the process of
rewriting the Store interface to be non-consuming. But we never know.
2016-10-07 21:18:49 +02:00
cf50ddae33 [CHERRY-PICK] Add testing implementation for Drop for FileLockEntry 2016-10-07 21:18:49 +02:00
485d280236 Bugfix: The StoreEntry should know the _new_ StoreId
When moving a entry in the store, we also should tell the StoreEntry
the new id.
2016-10-07 21:18:49 +02:00
c72291159e Add comment/documentation for Store::move_by_id() 2016-10-07 17:53:11 +02:00
b189bf7b8c Add check if entry is present
If we try to rename an entry that is borrowed, we fail, as renaming an
borrowed entry might result in some _really_ ugly bugs.
2016-10-07 17:32:35 +02:00
1e83ad7bbd Add test for hook execution for each hook position 2016-10-07 17:32:35 +02:00
d9f4898a3a Abstract testing of hook execution in helper function 2016-10-07 17:32:35 +02:00
dd621a08a6 Merge pull request #772 from matthiasbeyer/libimagstore/store-tests-erroring-hooks
libimagstore/store-tests: erroring hooks
2016-10-06 18:30:24 +02:00
1244a6666f Add assert to Store::get() the new ID 2016-10-03 12:34:07 +02:00
6cc1a90870 libimagstore: Add copyright notice to all files 2016-10-01 17:35:06 +02:00
1526c0b2dc Make hook configurable whether it succeeds or not 2016-09-30 13:39:33 +02:00
f16c09a981 Do not check whether old or new id exists/does not exist 2016-09-30 13:20:37 +02:00
886eed3ff4 Rename: SucceedingHook -> TestHook 2016-09-22 08:47:29 +02:00
4f83b22b98 Fix Store::move_by_id(): Move in cache as well 2016-09-22 08:42:19 +02:00
caa214f1bd Fix negation error
We check whether the old key already exists.

If it does _not_ exist, the entry is borrowed, from my understanding.
I'm not sure, though.
2016-09-22 08:41:12 +02:00
d5a275fec0 Add test: Store::move_by_id() 2016-09-22 08:41:01 +02:00
61201082c8 Fix: FileAbstraction::remove_file(): Add implementation 2016-09-21 10:20:16 +02:00
babf74e1e5 Fix Store::get() to not check FS but internal hashmap 2016-09-21 10:20:16 +02:00
4f2019a20a Extend test_store_create_delete_get()
to actually test:

1. get      -> Should return Ok(None)
2. create   -> Should return Ok(())
3. get      -> Should return Ok(Some(_))
4. delete   -> Should return Ok(())
5. get      -> Should return Ok(None)
2016-09-21 09:52:37 +02:00
513a9bd066 Merge pull request #751 from matthiasbeyer/libimagstore/store-tests-succeeding-hook-tests
libimagstore/store: tests succeeding hook tests
2016-09-19 22:39:23 +02:00
72a95ee5da Rewrite config validity checker to return Result<()>
And add more detailed error kinds for config errors
2016-09-19 21:52:28 +02:00
83ebe88022 Add PreCreate hook test 2016-09-19 21:51:14 +02:00
4155924f85 Add test for Store::retrieve_for_module() 2016-09-19 11:25:36 +02:00
fd41fe5998 Fix Store::delete() for nonexistent IDs
If the ID does not exist, we should return an error instead of doing
nothing, shouldn't we?
2016-09-19 11:07:38 +02:00
8345ff8248 Add test to delete non-existent entries 2016-09-19 11:03:38 +02:00
52b2a4589d Add test to get non-existent entries 2016-09-19 11:01:56 +02:00
de6b641201 Merge branch 'master' into libimagstorestdhook/git-integration
This is a integration-merge, because the libimagstorestdhook/git branch
got really long and rebasing is not an option, as it would mainly
destroy history.

Using an integration branch is less hassle, really.
2016-09-19 10:44:57 +02:00
85097554e9 Add hook implementation for succeeding hook tests 2016-09-19 10:38:46 +02:00
0cdeeb4c55 Fix test backend implementation
We must return an error if the file is not found here.

This is because if we unwrap() here, we panic if the store action was
Store::retrieve() and there wasn't a file there. We then unwrap() on
None and panic because of this, causing all other tests to panic as
well because the Mutex gets corrupted.

The store handles FileNotFound errors on its own, so it is safe to
return the error here.
2016-09-19 10:37:28 +02:00
07ff755ed7 Minify error handling in Store::_update with try!() 2016-09-18 14:52:55 +02:00
b288e938f9 Minify error handling in Store::update() 2016-09-18 14:52:00 +02:00
41c7cd2363 [CHERRY-PICK] Add newline before header
The lastest release of the `toml-rs` crate (2.1) removes leading spaces
before arrays and tables, causing our tests to fail.

This fixes it.
2016-09-09 17:22:20 +02:00
d5647069cb Add test that Store::retrieve() does ::create() underneath 2016-09-09 13:07:46 +02:00
c9994c33b6 Add test whether store-internal hashmap gets actually filled on Store::create() 2016-09-09 13:07:45 +02:00
e023a856f1 Add Store testing 2016-09-09 13:07:32 +02:00
ff5a4b46c7 Remove StoreId::is_file() 2016-09-09 12:56:29 +02:00
21714660df Merge pull request #729 from matthiasbeyer/fix-header-missing-newline
Add newline before header
2016-09-09 12:54:28 +02:00
7b827ff149 Add newline before header
The lastest release of the `toml-rs` crate (2.1) removes leading spaces
before arrays and tables, causing our tests to fail.

This fixes it.
2016-09-09 11:52:12 +02:00
88ff94fb6d Remove StoreId::is_dir() 2016-09-08 15:05:25 +02:00
3252f9b751 Move Update-Hook execution into Store._update() 2016-09-08 14:23:10 +02:00
d1fee32c7d Just do not execute mutable hooks, but let pass the nonmutables 2016-09-07 22:05:30 +02:00
af4a0bfaf2 Export result-helper here 2016-09-06 15:31:55 +02:00
5f04d29345 Adapt store to not use StoreId::into() but new StoreId::into_pathbuf() 2016-09-06 15:31:55 +02:00
a313cede06 Add StoreId::into_pathbuf() 2016-09-06 15:31:55 +02:00
b8f612632d Add error kind if StoreId has no base part 2016-09-06 15:31:29 +02:00
84b279de95 Remove Into<PathBuf> for StoreId 2016-09-06 15:31:29 +02:00
aa7e050436 Merge pull request #707 from matthiasbeyer/libimagstore/remove-storeid-storified
Remove StoreId::storified()
2016-09-06 12:31:01 +02:00
d7447a64c8 Remove StoreId::storified() 2016-09-06 09:37:25 +02:00
364ad01179 Use StoreId::from_full_path() in GlobStoreIdIterator::next() 2016-09-05 18:22:55 +02:00
8362b077e2 Add StoreId::from_full_path() 2016-09-05 18:21:34 +02:00
f60afa581c Add error type for failed build of StoreId object from full path 2016-09-05 18:21:21 +02:00
e4e5b52171 Fix Iterator impl for GlobStoreIdIterator
This patch fixes the `impl Iterator for GlobStoreIdIterator` which used
the glob() result to fetch the files from the FS, but glob() returns the
absolute pathes (to filesystem root).

We have to strip the `store_path` prefix and use the local part for
building the StoreId object.
2016-09-05 18:12:29 +02:00
ccffeb91a2 Remove StoreId::is_in_collection(), add StoreId::local()
Having a ::is_in_collection() is a nice thing, though it is _way_ better
if we just give the user of the `StoreId` object access to the local
part of the ID.

Using this new function, one can do all the actions one might need on
the Path for the actual entry without nasty copying or such.

`StoreId::is_in_collection()` can be replaced by
`StoreId::local().starts_with()` and everything is fine, as we do not
have to move a `PathBuf` object into the function anymore.
2016-09-04 10:40:38 +02:00
253c8a6f54 Adapt tests for new StoreId::is_in_collection() interface 2016-09-03 19:32:27 +02:00
a77cc2bab2 Change the interface of StoreId::is_in_collection()
to use &PathBuf rather than &[&str] as argument.
2016-09-02 17:15:15 +02:00
85f79ae3e5 Replace .map().all(|x| x) by calling .all() in the first place 2016-09-02 17:01:40 +02:00
4896c14086 Merge pull request #681 from matthiasbeyer/rewrite-storeid-type-cleanup-libimagentrylink
libimagentrylink cleanup
2016-09-02 08:25:10 +02:00
b2806ab592 Add StoreId::is_in_collection()
The concept of this function was introduced in

    37076f028c6cde0924b820154f3464f53ef65268

but here added to StoreId as function.
2016-08-31 09:20:08 +02:00
22d6815abd Remove Store::get_all_versions() as we do not longer have the concept of store-entry versions 2016-08-31 08:26:54 +02:00
82f08c6021 Remove version argument from module_entry_path_mod!() macro 2016-08-30 11:40:56 +02:00
ec2eb096e7 Remove id part from macro 2016-08-30 11:40:56 +02:00
fedb75ec98 Impl IntoStoreId for PathBuf, using StoreId::new_baseless() 2016-08-28 18:41:56 +02:00
5462714fcc Remove libimagstore::storeid::build_entry_path() 2016-08-28 18:41:56 +02:00
d252623b75 Add StoreId::components() to get components of StoreId "id" part 2016-08-28 18:41:55 +02:00
197940ac33 Fixup Store test implementations 2016-08-28 18:41:55 +02:00
d538a5b380 Fixup StoreId test 2016-08-28 18:41:55 +02:00
fd1a96fa74 Fix storification of StoreId in CRUD functions 2016-08-28 18:41:55 +02:00
4c435d9347 Add StoreId::with_base() 2016-08-28 18:41:54 +02:00
784f1cc4c6 Fix interface of IntoStoreId trait to return Result<StoreId> and impls 2016-08-28 18:41:54 +02:00
40f1efa258 Fix for new StoreId interface in Iterator for Walk impl 2016-08-28 18:41:54 +02:00
a110ecc2ec Fix GlobStoreIdIterator implementation for new StoreId interface 2016-08-28 18:41:54 +02:00
151877d95d Fix impl Drop for Store for new StoreId interface 2016-08-28 18:41:54 +02:00
d5723791a6 Add StoreId::without_base() 2016-08-28 18:41:54 +02:00
ab23d04e04 Add StoreId::new_baseless() 2016-08-28 18:41:54 +02:00
d1f07cd087 Add check in StoreId::new() whether the local part is absolute 2016-08-28 18:41:54 +02:00
ccc2b6b735 Add error kind for StoreId::new() with id-part that is absolute 2016-08-28 18:41:54 +02:00
5b23059d93 Fix scoping 2016-08-28 18:41:54 +02:00
3c1be0fbe9 Fix use of store ids when passing to FileAbstraction::* 2016-08-28 18:41:54 +02:00
111cb4f29a Add StoreId::to_str() -> String 2016-08-28 18:41:54 +02:00
aa6f220659 Fix GlobStoreIterator type implementation
... to use new StoreId interface
2016-08-28 18:41:53 +02:00
3f88c27834 Remove unused imports 2016-08-28 18:41:53 +02:00
9cf7897286 Fixup Walk type to handle StoreId properly 2016-08-28 18:41:53 +02:00
1de09f1606 Rewrite StoreId::storified() 2016-08-28 18:41:53 +02:00
1af3ab018b Use StoreId::new() in macro 2016-08-28 18:41:53 +02:00
fe73a53d44 Add StoreId::new() 2016-08-28 18:41:53 +02:00
8dff5685fd Make StoreId.base optional 2016-08-28 18:41:53 +02:00
ad92b05fb4 Remove IntoStoreId for PathBuf 2016-08-28 18:41:53 +02:00
adf213bbc7 Remove From<PathBuf> for StoreId 2016-08-28 18:41:53 +02:00
cf84df3ce7 Remove From<String> for StoreId 2016-08-28 18:41:53 +02:00
bcd917055d Remove AsRef<Path> for StoreId 2016-08-28 18:41:53 +02:00
30a76d57f6 Remove Borrow<Path> for StoreId 2016-08-28 18:41:53 +02:00
8c5a427a01 Rewrite Borrow<Path> for StoreId 2016-08-28 18:41:52 +02:00
55953cfe5b Rewrite AsRef<Path> for StoreId 2016-08-28 18:41:52 +02:00
1d165f6fc2 Remove Deref for StoreId but provide basic functions within the type 2016-08-28 18:41:52 +02:00
69b3f6bf87 Rewrite Display for StoreId 2016-08-28 18:41:52 +02:00
fb1df533c5 Rewrite Into<PathBuf> for StoreId 2016-08-28 18:41:52 +02:00
6841b7c340 Rewrite StoreId::storified() 2016-08-28 18:41:52 +02:00
264d651f1d Make StoreId be two parts: base and id 2016-08-28 18:41:52 +02:00
Marcel Müller
820fa287a0
Remove duplication 2016-08-25 09:30:47 +02:00
Marcel Müller
a82ef32142
Add missing FileAbstractions call 2016-08-24 16:58:48 +02:00
Marcel Müller
60a315e553
Rename lazyfile file to file_abstraction 2016-08-24 16:58:47 +02:00
Marcel Müller
c8e296e221
Move all abstractions to LazyFile and change it to FileAbstraction 2016-08-24 16:58:47 +02:00
Marcel Müller
1c71326ede
Add simple test 2016-08-24 16:58:47 +02:00
Marcel Müller
74dcd1f141
Add implementation of Test Lazyfile 2016-08-24 16:58:47 +02:00
Marcel Müller
bc1af19750
Add FileNot{Written,Seeked} errors 2016-08-24 16:58:47 +02:00
Marcel Müller
1d1ad65705
Split up LazyFile for test/prod 2016-08-24 16:58:46 +02:00
Marcel Müller
363a1d246a
Use Read instead of File for from_file
Also change it to `from_reader`
2016-08-24 16:58:46 +02:00
Marcel Müller
ef4455d9f3
Comment out tests 2016-08-24 16:58:43 +02:00
ddbc38e756 Merge pull request #659 from asuivelentine/libimagstore/unused-import
removed unused import
2016-08-23 21:36:03 +02:00
Kai Sickeler
5148bc1a6e removed unused import 2016-08-23 19:58:50 +02:00
e37316f3f4 Fix possible w-lock panic
For a detailed explanation on the issue this commit tries to solve, have
a look at

    7b5d37a039

which explains the issue in detail.
2016-08-02 18:59:17 +02:00
84ae516dc3 Fix possible w-lock panic
for a detailed explanation of the issue this commit tries to solve have
a look at

    7b5d37a039

which explains the issue in detail.
2016-08-02 18:59:17 +02:00
ea17755ad4 Fix possible w-lock panic
for a detailed explanation have a look at

    7b5d37a039

Which explains the issue this commit solves in more detail.
2016-08-01 20:28:38 +02:00
7b5d37a039 Fix: Make sure we release the Write-Lock
This was suggested by Marcel Müller, who also debugged this issue and
found that this snippet was the error.

The problem was, that we had the write-lock when starting the hooks. If
the hook runs into an Err() mapping part where the Entry could already
be drop()ed, Rust tries to be smart and drops the object. As we are
still in the creation code of the entry, this paniced as we still hold
the W-Lock and the drop() tryies to call _update() on the entry, which
also tries to W-lock it.

With the new additional scope, the W-lock gets dropped early and we do
not have this problem anymore.

Signed-off-by: Matthias Beyer <mail@beyermatthias.de>
Suggested-by: Marcel Müller <neikos@neikos.email>
Reported-by: Matthias Beyer <mail@beyermatthias.de>
2016-08-01 20:00:31 +02:00
Kai Sickeler
b6437a0b0f Implement Debug for FileLockEntry, do not derive it 2016-07-27 19:26:45 +02:00
0ba67f7873 Check whether implictely creating the store path is allowed 2016-07-25 17:02:38 +02:00
4d650a63f7 Add error kind for implicit create denied 2016-07-25 17:02:38 +02:00
442b1537be Add helper to check config for implicit-create 2016-07-25 17:02:38 +02:00
cdf6bc6fa5 Make warning more verbose, tell what kind of variable cannot be fetched 2016-07-18 15:44:34 +02:00
ce766427da Merge pull request #561 from matthiasbeyer/libimagstore/verify
Libimagstore/verify
2016-07-17 13:03:54 +02:00
6b1fdfbc1d Add Store::verify() 2016-07-17 00:59:04 +02:00
bb8dad256c Fix comment how config has to look 2016-07-16 22:17:05 +02:00
2a95d4f79d Merge pull request #553 from matthiasbeyer/libimagstore/custom-hook-error-data-interface
Add interface for custom error data type
2016-07-16 19:15:39 +02:00
b81750ba09 Add interface for custom error data type 2016-07-16 16:52:02 +02:00
01c21e1d95 Remove "Sync" trait bound - we don't need to sync between threads here
(and we do not have the possibility to, actually).
2016-07-16 16:18:37 +02:00
6c181d32b6 Merge pull request #548 from matthiasbeyer/libimagstore/storeid-hook-exec-in-mutable-hook-exec
Execute StoreIdAccess accessors in MutableHookDataAccessor impl for A…
2016-07-16 11:23:52 +02:00
76d23577c4 Merge pull request #546 from matthiasbeyer/libimagstore/fixed-and-more-hook-error-wrapping
Add more error levels in hook execution error wrapping
2016-07-16 11:23:38 +02:00
083b9da1e9 Merge pull request #545 from matthiasbeyer/libimagstore/fix-aspect-config-getting
Fix AspectConfig::get_for() implementation
2016-07-16 11:23:20 +02:00
c6be35554b Merge pull request #547 from matthiasbeyer/libimagstore/aspect-execution-output
Add error output in Aspect execution code
2016-07-16 11:21:41 +02:00
ed4744449c Execute StoreIdAccess accessors in MutableHookDataAccessor impl for Aspect as well
We have no other way than doing it this way, and it doesn't hurt, does
it?
2016-07-16 01:15:48 +02:00
6a85760feb Debug output in MutableHookDatsaAccessor impl for Aspect 2016-07-16 01:15:08 +02:00
8de32622b9 Add more error levels in hook execution error wrapping 2016-07-16 00:57:33 +02:00
ded6a3f240 Add error output in Aspect execution code 2016-07-16 00:57:20 +02:00
c8edb2102f Fix AspectConfig::get_for() implementation 2016-07-16 00:33:54 +02:00
32617d0a16 Add debug config when getting configuration for a hook 2016-07-16 00:32:58 +02:00
33771f3903 Add debugging output when setting hook configuration 2016-07-16 00:32:54 +02:00
fc72b3f53a Add more debug output in Store::new() 2016-07-16 00:32:50 +02:00
John Sirois
8ef7b4e0c6 Cleanup HookError tracing logic.
Use `Result::or_else` instead of matching and no-oping the `Ok` case
manually and extract a helper to centralize repeated trace logic.
2016-07-14 20:37:16 -06:00
John Sirois
7f57e5e234 Move from a helper function to a typeclass.
This introduces the `FoldResut` trait to move from `func(receiver, ...)`
style to `receiver.func(...)` style. Also add a means to pass the
default result explicitly.
2016-07-14 07:28:54 -06:00
John Sirois
2c40b8734e Add a fold_ok utility.
Add a utility that folds an iterator into a result and uses it to reduce
boilerplate in the codebase.
2016-07-13 12:43:04 -06:00
c6f9d91599 Merge pull request #513 from matthiasbeyer/libimagstore/hooks/optionally-deny-mutable-hooks
Libimagstore/hooks/optionally deny mutable hooks
2016-07-07 21:52:08 +02:00