hj/pleroma
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

Merge branch 'develop' into issue/1383

Browse source
This commit is contained in:
Maksim Pechnikov 2020-01-22 20:20:39 +03:00
commit cf19bf3c7c
485 changed files with 5044 additions and 1888 deletions
Download patch file Download diff file Expand all files Collapse all files

194
docs/API/admin_api.md
View file

@ -3,7 +3,7 @@
Authentication is required and the user must be an admin.
Configuration options:
* `[:auth, :enforce_oauth_admin_scope_usage]` — OAuth admin scope requirement toggle.
If `true`, admin actions explicitly demand admin OAuth scope(s) presence in OAuth token (client app must support admin scopes).
If `false` and token doesn't have admin scope(s), `is_admin` user flag grants access to admin-specific actions.
@ -665,27 +665,16 @@ Note: Available `:permission_group` is currently moderator and admin. 404 is ret
- 404 Not Found `"Not found"`
- On success: 200 OK `{}`
## `GET /api/pleroma/admin/config/migrate_to_db`
### Run mix task pleroma.config migrate_to_db
Copy settings on key `:pleroma` to DB.
- Params: none
- Response:
```json
{}
```
## `GET /api/pleroma/admin/config/migrate_from_db`
### Run mix task pleroma.config migrate_from_db
Copy all settings from DB to `config/prod.exported_from_db.secret.exs` with deletion from DB.
Copies all settings from database to `config/{env}.exported_from_db.secret.exs` with deletion from the table. Where `{env}` is the environment in which `pleroma` is running.
- Params: none
- Response:
- On failure:
- 400 Bad Request `"To use this endpoint you need to enable configuration from database."`
```json
{}
@ -693,20 +682,24 @@ Copy all settings from DB to `config/prod.exported_from_db.secret.exs` with dele
## `GET /api/pleroma/admin/config`
### List config settings
### Get list of merged default settings with saved in database.
List config settings only works with `:pleroma => :instance => :dynamic_configuration` setting to `true`.
**Only works when configuration from database is enabled.**
- Params: none
- Params:
- `only_db`: true (*optional*, get only saved in database settings)
- Response:
- On failure:
- 400 Bad Request `"To use this endpoint you need to enable configuration from database."`
- 400 Bad Request `"To use configuration from database migrate your settings to database."`
```json
{
configs: [
{
"group": string,
"key": string or string with leading `:` for atoms,
"value": string or {} or [] or {"tuple": []}
"group": ":pleroma",
"key": "Pleroma.Upload",
"value": []
}
]
}
@ -716,44 +709,107 @@ List config settings only works with `:pleroma => :instance => :dynamic_configur
### Update config settings
Updating config settings only works with `:pleroma => :instance => :dynamic_configuration` setting to `true`.
Module name can be passed as string, which starts with `Pleroma`, e.g. `"Pleroma.Upload"`.
Atom keys and values can be passed with `:` in the beginning, e.g. `":upload"`.
Tuples can be passed as `{"tuple": ["first_val", Pleroma.Module, []]}`.
`{"tuple": ["some_string", "Pleroma.Some.Module", []]}` will be converted to `{"some_string", Pleroma.Some.Module, []}`.
Keywords can be passed as lists with 2 child tuples, e.g.
`[{"tuple": ["first_val", Pleroma.Module]}, {"tuple": ["second_val", true]}]`.
**Only works when configuration from database is enabled.**
If value contains list of settings `[subkey: val1, subkey2: val2, subkey3: val3]`, it's possible to remove only subkeys instead of all settings passing `subkeys` parameter. E.g.:
{"group": "pleroma", "key": "some_key", "delete": "true", "subkeys": [":subkey", ":subkey3"]}.
Some modifications are necessary to save the config settings correctly:
Compile time settings (need instance reboot):
- all settings by this keys:
- strings which start with `Pleroma.`, `Phoenix.`, `Tesla.` or strings like `Oban`, `Ueberauth` will be converted to modules;
```
"Pleroma.Upload" -> Pleroma.Upload
"Oban" -> Oban
```
- strings starting with `:` will be converted to atoms;
```
":pleroma" -> :pleroma
```
- objects with `tuple` key and array value will be converted to tuples;
```
{"tuple": ["string", "Pleroma.Upload", []]} -> {"string", Pleroma.Upload, []}
```
- arrays with *tuple objects* will be converted to keywords;
```
[{"tuple": [":key1", "value"]}, {"tuple": [":key2", "value"]}] -> [key1: "value", key2: "value"]
```
Most of the settings will be applied in `runtime`, this means that you don't need to restart the instance. But some settings are applied in `compile time` and require a reboot of the instance, such as:
- all settings inside these keys:
- `:hackney_pools`
- `:chat`
- `Pleroma.Web.Endpoint`
- `Pleroma.Repo`
- part settings:
- `Pleroma.Captcha` -> `:seconds_valid`
- `Pleroma.Upload` -> `:proxy_remote`
- `:instance` -> `:upload_limit`
- partially settings inside these keys:
- `:seconds_valid` in `Pleroma.Captcha`
- `:proxy_remote` in `Pleroma.Upload`
- `:upload_limit` in `:instance`
- Params:
- `configs` => [
- `group` (string)
- `key` (string or string with leading `:` for atoms)
- `value` (string, [], {} or {"tuple": []})
- `delete` = true (optional, if parameter must be deleted)
- `subkeys` [(string with leading `:` for atoms)] (optional, works only if `delete=true` parameter is passed, otherwise will be ignored)
]
- `configs` - array of config objects
- config object params:
- `group` - string (**required**)
- `key` - string (**required**)
- `value` - string, [], {} or {"tuple": []} (**required**)
- `delete` - true (*optional*, if setting must be deleted)
- `subkeys` - array of strings (*optional*, only works when `delete=true` parameter is passed, otherwise will be ignored)
- Request (example):
*When a value have several nested settings, you can delete only some nested settings by passing a parameter `subkeys`, without deleting all settings by key.*
```
[subkey: val1, subkey2: val2, subkey3: val3] \\ initial value
{"group": ":pleroma", "key": "some_key", "delete": true, "subkeys": [":subkey", ":subkey3"]} \\ passing json for deletion
[subkey2: val2] \\ value after deletion
```
*Most of the settings can be partially updated through merge old values with new values, except settings value of which is list or is not keyword.*
Example of setting without keyword in value:
```elixir
config :tesla, :adapter, Tesla.Adapter.Hackney
```
List of settings which support only full update by key:
```elixir
@full_key_update [
{:pleroma, :ecto_repos},
{:quack, :meta},
{:mime, :types},
{:cors_plug, [:max_age, :methods, :expose, :headers]},
{:auto_linker, :opts},
{:swarm, :node_blacklist},
{:logger, :backends}
]
```
List of settings which support only full update by subkey:
```elixir
@full_subkey_update [
{:pleroma, :assets, :mascots},
{:pleroma, :emoji, :groups},
{:pleroma, :workers, :retries},
{:pleroma, :mrf_subchain, :match_actor},
{:pleroma, :mrf_keyword, :replace}
]
```
*Settings without explicit key must be sended in separate config object params.*
```elixir
config :quack,
level: :debug,
meta: [:all],
...
```
```json
{
configs: [
{"group": ":quack", "key": ":level", "value": ":debug"},
{"group": ":quack", "key": ":meta", "value": [":all"]},
...
]
}
```
- Request:
```json
{
configs: [
{
"group": "pleroma",
"group": ":pleroma",
"key": "Pleroma.Upload",
"value": [
{"tuple": [":uploader", "Pleroma.Uploaders.Local"]},
@ -763,7 +819,7 @@ Compile time settings (need instance reboot):
{"tuple": [":proxy_opts", [
{"tuple": [":redirect_on_failure", false]},
{"tuple": [":max_body_length", 1048576]},
{"tuple": [":http": [
{"tuple": [":http", [
{"tuple": [":follow_redirect", true]},
{"tuple": [":pool", ":upload"]},
]]}
@ -779,19 +835,53 @@ Compile time settings (need instance reboot):
```
- Response:
- On failure:
- 400 Bad Request `"To use this endpoint you need to enable configuration from database."`
```json
{
configs: [
{
"group": string,
"key": string or string with leading `:` for atoms,
"value": string or {} or [] or {"tuple": []}
"group": ":pleroma",
"key": "Pleroma.Upload",
"value": [...]
}
]
}
```
## ` GET /api/pleroma/admin/config/descriptions`
### Get JSON with config descriptions.
Loads json generated from `config/descriptions.exs`.
- Params: none
- Response:
```json
[{
"group": ":pleroma", // string
"key": "ModuleName", // string
"type": "group", // string or list with possible values,
"description": "Upload general settings", // string
"children": [
{
"key": ":uploader", // string or module name `Pleroma.Upload`
"type": "module",
"description": "Module which will be used for uploads",
"suggestions": ["module1", "module2"]
},
{
"key": ":filters",
"type": ["list", "module"],
"description": "List of filter modules for uploads",
"suggestions": [
"module1", "module2", "module3"
]
}
]
}]
```
## `GET /api/pleroma/admin/moderation_log`
### Get moderation log

2
docs/API/differences_in_mastoapi_responses.md
View file

@ -29,7 +29,7 @@ Has these additional fields under the `pleroma` object:
- `spoiler_text`: a map consisting of alternate representations of the `spoiler_text` property with the key being it's mimetype. Currently the only alternate representation supported is `text/plain`
- `expires_at`: a datetime (iso8601) that states when the post will expire (be deleted automatically), or empty if the post won't expire
- `thread_muted`: true if the thread the post belongs to is muted
- `emoji_reactions`: An object with all the emoji reactions with count. Contains no information about the reacting users, for that use the `emoji_reactions_by` endpoint.
- `emoji_reactions`: A list with emoji / reaction count tuples. Contains no information about the reacting users, for that use the `emoji_reactions_by` endpoint.
## Attachments

10
docs/API/pleroma_api.md
View file

@ -451,11 +451,11 @@ Emoji reactions work a lot like favourites do. They make it possible to react to
* Method: `GET`
* Authentication: optional
* Params: None
* Response: JSON, a map of emoji to account list mappings.
* Response: JSON, a list of emoji/account list tuples, sorted by emoji insertion date, in ascending order, e.g, the first emoji in the list is the oldest.
* Example Response:
```json
{
"😀" => [{"id" => "xyz.."...}, {"id" => "zyx..."}],
"🗡" => [{"id" => "abc..."}]
}
[
["😀", [{"id" => "xyz.."...}, {"id" => "zyx..."}]],
["☕", [{"id" => "abc..."}]]
]
```

79
docs/admin/config.md Normal file
View file

@ -0,0 +1,79 @@
# Configuring instance
You can configure your instance from admin interface. You need account with admin rights and little change in config file, which will allow settings configuration from database.
```elixir
config :pleroma, configurable_from_database: true
```
## How it works
Settings are stored in database and are applied in `runtime` after each change. Most of the settings take effect immediately, except some, which need instance reboot. These settings are needed in `compile time`, that's why settings are duplicated to the file.
File with duplicated settings is located in `config/{env}.exported_from_db.exs` if pleroma is runned from source. For prod env it will be `config/prod.exported_from_db.exs`.
For releases: `/etc/pleroma/prod.exported_from_db.secret.exs` or `PLEROMA_CONFIG_PATH/prod.exported_from_db.exs`.
## How to set it up
You need to migrate your existing settings to the database. This task will migrate only added by user settings.
For example you add settings to `prod.secret.exs` file, only these settings will be migrated to database. For release it will be `/etc/pleroma/config.exs` or `PLEROMA_CONFIG_PATH`.
You can do this with mix task (all config files will remain untouched):
```sh tab="OTP"
./bin/pleroma_ctl config migrate_to_db
```
```sh tab="From Source"
mix pleroma.config migrate_to_db
```
Now you can change settings in admin interface. After each save, settings from database are duplicated to the `config/{env}.exported_from_db.exs` file.
<span style="color:red">**ATTENTION**</span>
**<span style="color:red">Be careful while changing the settings. Every inaccurate configuration change can break the federation or the instance load.</span>**
*Compile time settings, which require instance reboot and can break instance loading:*
- all settings inside these keys:
- `:hackney_pools`
- `:chat`
- partially settings inside these keys:
- `:seconds_valid` in `Pleroma.Captcha`
- `:proxy_remote` in `Pleroma.Upload`
- `:upload_limit` in `:instance`
## How to dump settings from database to file
*Adding `-d` flag will delete migrated settings from database table.*
```sh tab="OTP"
./bin/pleroma_ctl config migrate_from_db [-d]
```
```sh tab="From Source"
mix pleroma.config migrate_from_db [-d]
```
## How to completely remove it
1. Truncate or delete all values from `config` table
```sql
TRUNCATE TABLE config;
```
2. Delete `config/{env}.exported_from_db.exs`.
For `prod` env:
```bash
cd /opt/pleroma
cp config/prod.exported_from_db.exs config/exported_from_db.back
rm -rf config/prod.exported_from_db.exs
```
*If you don't want to backup settings, you can skip step with `cp` command.*
3. Set configurable_from_database to `false`.
```elixir
config :pleroma, configurable_from_database: false
```
4. Restart pleroma instance
```bash
sudo service pleroma restart
```

6
docs/administration/CLI_tasks/config.md
View file

@ -18,11 +18,11 @@ mix pleroma.config migrate_to_db
## Transfer config from DB to `config/env.exported_from_db.secret.exs`
To delete transfered settings from database optional flag `-d` can be used. <env> is `prod` by default.
```sh tab="OTP"
./bin/pleroma_ctl config migrate_from_db <env>
./bin/pleroma_ctl config migrate_from_db [--env=<env>] [-d]
```
```sh tab="From Source"
mix pleroma.config migrate_from_db <env>
mix pleroma.config migrate_from_db [--env=<env>] [-d]
```

11
docs/configuration/cheatsheet.md
View file

@ -70,11 +70,6 @@ You shouldn't edit the base config directly to avoid breakages and merge conflic
* `account_field_value_length`: An account field value maximum length (default: `2048`).
* `external_user_synchronization`: Enabling following/followers counters synchronization for external users.
!!! danger
This is a Work In Progress, not usable just yet
* `dynamic_configuration`: Allow transferring configuration to DB with the subsequent customization from Admin api.
## Federation
### MRF policies
@ -355,7 +350,7 @@ Available caches:
* `proxy_url`: an upstream proxy to fetch posts and/or media with, (default: `nil`)
* `send_user_agent`: should we include a user agent with HTTP requests? (default: `true`)
* `user_agent`: what user agent should we use? (default: `:default`), must be string or `:default`
* `user_agent`: what user agent should we use? (default: `:default`), must be string or `:default`
* `adapter`: array of hackney options
@ -850,3 +845,7 @@ config :auto_linker,
## Custom Runtime Modules (`:modules`)
* `runtime_dir`: A path to custom Elixir modules (such as MRF policies).
## :configurable_from_database
Enable/disable configuration from database.