Migrations currently uses the default Xorm mapper which is
not the same as the mapper Gitea actually uses.
This means that there is a difference between the struct
parsing and mapping to database tables in migrations as
compared to normal Sync2.
This was the cause for the catastrophic problem in v168 -
untagged fields are not mapped in the same way in migrations
as compared to outside of migrations.
This is also likely the cause of some weird subtle failures
in other migrations as any untagged field may not be being
mapped exactly the same way.
This PR suggests that we ensure that the mapper is set at
the start of the migrations code - but also enforces a strict
clean mapper between each migration.
Signed-off-by: Andrew Thornton <art27@cantab.net>
Co-authored-by: zeripath <art27@cantab.net>
* Enhance Ghost comment mitigation Settings (#14392)
* refactor models.DeleteComment and delete related reactions too
* use deleteComment for UserDeleteWithCommentsMaxDays in DeleteUser
* Resolve Fixme & fix potential deadlock
* rm refactor
* make diff eaven less
* Prevent panic on fuzzer provided string
The fuzzer has found that providing a <body> tag with an attribute to
PostProcess causes a panic. This PR removes any rendered html or body
tags from the output.
Signed-off-by: Andrew Thornton <art27@cantab.net>
* Placate lint
* placate lint again
Signed-off-by: Andrew Thornton <art27@cantab.net>
* minor cleanup
Signed-off-by: Andrew Thornton <art27@cantab.net>
Co-authored-by: zeripath <art27@cantab.net>
The incorrect use of filepath instead of path means that
it is possible to cause a stackoverflow on Windows
Signed-off-by: Andrew Thornton <art27@cantab.net>
Co-authored-by: zeripath <art27@cantab.net>
Co-authored-by: Lauris BH <lauris@nix.lv>
Backport #14131
Unfortunately every connection to postgres requires that the search path is
set appropriately.
This PR shadows the postgres driver to ensure that as soon as a connection
is open, the search_path is set appropriately.
Fix#14088
Signed-off-by: Andrew Thornton <art27@cantab.net>
* Fix bug of link query order on markdown render
* Fix bluemonday bug and fix one wrong test
Co-authored-by: 6543 <6543@obermui.de>
Co-authored-by: 6543 <6543@obermui.de>
The design is very flexible, but not implemented correctly.
This commit fixes several issues:
* Costom storage type stated in https://docs.gitea.io/en-us/config-cheat-sheet/#storage-storage
not working
* [storage.attachments], [storage.minio] section not respected
Signed-off-by: 胡玮文 <huww98@outlook.com>
Co-authored-by: 胡玮文 <huww98@outlook.com>
* Fix panic in BasicAuthDecode
If the string does not contain ":" that function would run into an
`index out of range [1] with length 1` error. prevent that.
* Update BasicAuthDecode()
Co-authored-by: 6543 <6543@obermui.de>
Co-authored-by: 6543 <6543@obermui.de>
Co-authored-by: zeripath <art27@cantab.net>
* Don't use simpleMDE editor on mobile devices
simpleMDE doesn't work properly on mobile devices -- We've replaced it with the slightly more working easyMDE in 1.14 but since that change can't be backported to 1.13 we will just disable the editor on mobile here.
* make isMobile function per code review -- disable simpleMDE for code review and replies
* Fix issue with plain text and wiki
Co-authored-by: silverwind <me@silverwind.io>
* Update view.tmpl
Added rendering of emoji to project label
* Add RenderEmojiPlain to the title and remove has-emoji
Co-authored-by: zeripath <art27@cantab.net>
Co-authored-by: Rakshith Ravi <rakshith.ravi@gmx.com>
Co-authored-by: zeripath <art27@cantab.net>
Backport #13981#13882 has revealed that the refname of an action is actually only a
refname pattern and necessarily a branch. For examplem pushing to
refs/heads/master will result in action with refname refs/heads/master
but pushing to master will result in a refname master.
The simplest solution to providing a fix here is to trim the prefix
therefore this PR proposes this.
Signed-off-by: Andrew Thornton <art27@cantab.net>
Co-authored-by: a1012112796 <1012112796@qq.com>
Co-authored-by: a1012112796 <1012112796@qq.com>
`ctx.Error` requires that templates are available for this to
render the error page otherwise there will be a panic at this
time.
This was fixed in #13164 but was not completely backported.
Fix#13971
Signed-off-by: Andrew Thornton <art27@cantab.net>
Backport #13966
`user.HashPassword` may potentially - and in fact now likely does - change
the `passwd_hash_algo` therefore whenever the `passwd` is updated, this
also needs to be updated.
Fix#13832
Thanks @fblaese for the hint
Signed-off-by: Andrew Thornton <art27@cantab.net>
The default branch in wikis must be master - therefore forcibly set the HEAD
to master.
Fix#13846
Signed-off-by: Andrew Thornton <art27@cantab.net>
Co-authored-by: 6543 <6543@obermui.de>
Co-authored-by: zeripath <art27@cantab.net>
