* Update the webauthn_credential_id_sequence in Postgres There is (yet) another problem with v210 in that Postgres will silently allow preset ID insertions ... but it will not update the sequence value. This PR simply adds a little step to the end of the v210 migration to update the sequence number. Users who have already migrated who find that they cannot insert new webauthn_credentials into the DB can either run: ```bash gitea doctor recreate-table webauthn_credential ``` or ```bash ./gitea doctor --run=check-db-consistency --fix ``` which will fix the bad sequence. Fix #19012 Signed-off-by: Andrew Thornton <art27@cantab.net>
6.4 KiB
date | title | slug | weight | toc | draft | menu | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2021-11-01T23:41:00+08:00 | Guidelines for Backend Development | guidelines-backend | 20 | false | false |
|
Guidelines for Backend Development
Table of Contents
{{< toc >}}
Background
Gitea uses Golang as the backend programming language. It uses many third-party packages and also write some itself.
For example, Gitea uses Chi as basic web framework. Xorm is an ORM framework that is used to interact with the database.
So it's very important to manage these packages. Please take the below guidelines before you start to write backend code.
Package Design Guideline
Packages List
To maintain understandable code and avoid circular dependencies it is important to have a good code structure. The Gitea backend is divided into the following parts:
build
: Scripts to help build Gitea.cmd
: All Gitea actual sub commands includes web, doctor, serv, hooks, admin and etc.web
will start the web service.serv
andhooks
will be invoked by Git or OpenSSH. Other sub commands could help to maintain Gitea.integrations
: Integration testsmodels
: Contains the data structures used by xorm to construct database tables. It also contains functions to query and update the database. Dependencies to other Gitea code should be avoided. You can make exceptions in cases such as logging.models/db
: Basic database operations. All othermodels/xxx
packages should depend on this package. TheGetEngine
function should only be invoked frommodels/
.models/fixtures
: Sample data used in unit tests and integration tests. Oneyml
file means one table which will be loaded into database when beginning the tests.models/migrations
: Stores database migrations between versions. PRs that change a database structure MUST also have a migration step.
modules
: Different modules to handle specific functionality in Gitea. Work in Progress: Some of them should be moved toservices
, in particular those that depend on models because they rely on the database.modules/setting
: Store all system configurations read from ini files and has been referenced by everywhere. But they should be used as function parameters when possible.modules/git
: Package to interactive withGit
command line or Gogit package.
public
: Compiled frontend files (javascript, images, css, etc.)routers
: Handling of server requests. As it uses other Gitea packages to serve the request, other packages (models, modules or services) must not depend on routers.routers/api
Contains routers for/api/v1
aims to handle RESTful API requests.routers/install
Could only respond when system is in INSTALL mode (INSTALL_LOCK=false).routers/private
will only be invoked by internal sub commands, especiallyserv
andhooks
.routers/web
will handle HTTP requests from web browsers or Git SMART HTTP protocols.
services
: Support functions for common routing operations or command executions. Usesmodels
andmodules
to handle the requests.templates
: Golang templates for generating the html output.
Package Dependencies
Since Golang don't support import cycles, we have to decide the package dependencies carefully. There are some levels between those packages. Below is the ideal package dependencies direction.
cmd
-> routers
-> services
-> models
-> modules
From left to right, left packages could depend on right packages, but right packages MUST not depend on left packages. The sub packages on the same level could depend on according this level's rules.
NOTICE
Why do we need database transactions outside of models
? And how?
Some actions should allow for rollback when database record insertion/update/deletion failed.
So services must be allowed to create a database transaction. Here is some example,
// servcies/repository/repo.go
func CreateXXXX() error {\
ctx, committer, err := db.TxContext()
if err != nil {
return err
}
defer committer.Close()
// do something, if return err, it will rollback automatically when `committer.Close()` is invoked.
if err := issues.UpdateIssue(ctx, repoID); err != nil {
// ...
}
// ......
return committer.Commit()
}
You should not use db.GetEngine(ctx)
in services
directly, but just write a function under models/
.
If the function will be used in the transaction, just let context.Context
as the function's first parameter.
// models/issues/issue.go
func UpdateIssue(ctx context.Context, repoID int64) error {
e := db.GetEngine(ctx)
// ......
}
Package Name
For the top level package, use a plural as package name, i.e. services
, models
, for sub packages, use singular,
i.e. servcies/user
, models/repository
.
Import Alias
Since there are some packages which use the same package name, it is possible that you find packages like modules/user
, models/user
, and services/user
. When these packages are imported in one Go file, it's difficult to know which package we are using and if it's a variable name or an import name. So, we always recommend to use import aliases. To differ from package variables which are commonly in camelCase, just use snake_case for import aliases.
i.e. import user_service "code.gitea.io/gitea/services/user"
Important Gotchas
- Never write
x.Update(exemplar)
without an explicitWHERE
clause:- This will cause all rows in the table to be updated with the non-zero values of the exemplar - including IDs.
- You should usually write
x.ID(id).Update(exemplar)
.
- If during a migration you are inserting into a table using
x.Insert(exemplar)
where the ID is preset:- You will need to
SET IDENTITY_INSERT `table` ON
for the MSSQL variant (the migration will fail otherwise) - However, you will also need to update the id sequence for postgres - the migration will silently pass here but later insertions will fail:
SELECT setval('table_name_id_seq', COALESCE((SELECT MAX(id)+1 FROM `table_name`), 1), false)
- You will need to
Future Tasks
Currently, we are creating some refactors to do the following things:
- Correct that codes which doesn't follow the rules.
- There are too many files in
models
, so we are moving some of them into a sub packagemodels/xxx
. - Some
modules
sub packages should be moved toservices
because they depend onmodels
.