api

package
v1.0.7 Latest Latest
Warning

This package is not in the latest version of its module.

 Go to latest Published: Jun 16, 2026 License: MIT Imports: 13 Imported by: 0

Documentation

Index

Constants

View Source 
const (
	ModeSession = "session"
	ModeJWT     = "jwt"
)

Transport modes mirror config.ModeSession / config.ModeJWT. Session is the default: works for ordinary accounts via Archery's AJAX endpoints + Django session cookie. JWT is the opt-in REST transport.

View Source 
const APIBasePath = "/api"

APIBasePath is the Archery REST API base path.

View Source 
const Code2FARequired = "E_2FA_REQUIRED"

Code2FARequired is the sentinel APIError.Code value set when a login is blocked on a missing 2FA code. The cmd layer maps it to output.E_2FA_REQUIRED + the human-required exit code; carrying it as a plain string keeps the api package free of an output dependency.

View Source 
const CodeValidation = "E_VALIDATION"

CodeValidation is the APIError.Code for a client-side validation failure (e.g. a wrong/expired 2FA code). It mirrors output.E_VALIDATION as a plain string.

View Source 
const WorkflowTypeSQLReview = 2

WorkflowTypeSQLReview is Archery's workflow_type for SQL上线申请 (SQL-review workflows), the only type the execute command targets.

Variables

This section is empty.

Functions

func SetClientOptions 

func SetClientOptions(o ClientOptions)

SetClientOptions updates global HTTP settings (from cmd flags).

func TagUntrusted 

func TagUntrusted(data map[string]any, fields ...string)

TagUntrusted adds a "_untrusted" key to data listing the names of fields that contain externally-sourced, user-generated content. Callers must pass only the inner data map (not the full envelope). Fields that do not exist in the map are silently skipped.

func WorkflowStatusCodeToInt added in v1.0.4 

func WorkflowStatusCodeToInt(code string) int

WorkflowStatusCodeToInt maps Archery's session status string codes to the integer status used by the REST serializer and the CLI's status badges, so both transports render the same label. Unknown codes map to -1.

Types

type APIError 

type APIError struct {
	StatusCode    int
	ErrorMessages []string
	Errors        map[string]string
	// Code, when non-empty, is an explicit error-code override the cmd layer uses
	// instead of deriving the code from StatusCode (e.g. Code2FARequired).
	Code string
}

APIError represents an error returned by the Archery API.

func (*APIError) Error 

func (e *APIError) Error() string

type ArchiveConfig 

type ArchiveConfig struct {
	ID            int    `json:"id"`
	Title         string `json:"title"`
	AuditWorkflow int    `json:"audit_workflow"`
	AuditUsers    string `json:"audit_users"`
	SourceHost    string `json:"source_host"`
	SourceDb      string `json:"source_db"`
	SourceTable   string `json:"source_table"`
	TargetHost    string `json:"target_host"`
	TargetDb      string `json:"target_db"`
	TargetTable   string `json:"target_table"`
	Mode          string `json:"mode"`
	Condition     string `json:"condition"`
	IsArchive     bool   `json:"is_archive"`
	ExecDate      string `json:"exec_date"`
	Status        int    `json:"status"`
}

ArchiveConfig represents an archive task configuration.

type ArchiveLog 

type ArchiveLog struct {
	ID          int    `json:"id"`
	ConfigID    int    `json:"config_id"`
	SQLText     string `json:"sql"`
	ExecTime    string `json:"exec_time"`
	Status      int    `json:"status"`
	RowsMoved   int    `json:"rows_moved"`
	ElapsedTime string `json:"elapsed_time"`
}

ArchiveLog represents a single archive execution log.

type AuthAPI 

type AuthAPI struct {
	// contains filtered or unexported fields
}

AuthAPI wraps authentication-related API calls.

func (*AuthAPI) Login 

func (a *AuthAPI) Login(ctx context.Context, username, password string) (string, string, error)

Login authenticates with username and password, storing tokens on success.

POST /api/auth/token/ Request: {"username": "...", "password": "..."} Response: {"access": "...", "refresh": "..."}

func (*AuthAPI) LoginWithSession 

func (a *AuthAPI) LoginWithSession(ctx context.Context, username, password string) (sessionID, csrfToken string, err error)

LoginWithSession authenticates via Archery's session-based AJAX login and returns the resulting Django cookies for caching.

Flow (verified against v1.8.5): 

GET  /login/          -> Set-Cookie csrftoken
POST /authenticate/   form username/password/csrfmiddlewaretoken
                      + headers X-CSRFToken, Referer -> {"status":0,"msg":"ok"} + sessionid

On success the cookie jar holds sessionid + (rotated) csrftoken, which are returned for persistence.

func (*AuthAPI) Refresh 

func (a *AuthAPI) Refresh(ctx context.Context, refreshToken string) (string, error)

Refresh obtains a new access token using the stored refresh token.

POST /api/auth/token/refresh/ Request: {"refresh": "..."} Response: {"access": "..."}

func (*AuthAPI) Verify 

func (a *AuthAPI) Verify(ctx context.Context, token string) error

Verify checks whether a token is still valid.

POST /api/auth/token/verify/ Request: {"token": "..."} Returns nil if valid, error otherwise.

type BinlogAPI added in v1.0.4 

type BinlogAPI struct {
	// contains filtered or unexported fields
}

BinlogAPI provides methods for the MySQL binlog endpoints.

These views are session-only: Archery exposes no REST/JWT equivalent under /api/v1/, so every method uses the session (AJAX/Django) transport regardless of the client's mode. Form fields and response shapes are taken from sql/binlog.py in the v1.8.5 source.

func (*BinlogAPI) List added in v1.0.4 

func (b *BinlogAPI) List(ctx context.Context, instanceName string) (*BinlogListResult, error)

List returns the binary logs of an instance. 

session: POST /binlog/list/ (instance_name)

func (*BinlogAPI) Parse added in v1.0.4 

func (b *BinlogAPI) Parse(ctx context.Context, p BinlogParseParams) (*BinlogParseResult, error)

Parse runs the my2sql engine over a binlog range and returns the first `num` statements (rollback statements when Rollback is set). 

session: POST /binlog/my2sql/

List fields use the suffixed keys the Django view reads with getlist: only_schemas, only_tables[], sql_type[]. Time fields are start_time and stop_time (the view has no end_time). Booleans are the literal string "true".

func (*BinlogAPI) Purge added in v1.0.4 

func (b *BinlogAPI) Purge(ctx context.Context, instanceID, binlogFile string) (*BinlogPurgeResult, error)

Purge runs `purge master logs to '<binlog>'` on an instance, deleting that file and all binlogs before it. 

session: POST /binlog/del_log/ (instance_id, binlog)

Note the view keys the instance by numeric id, not name.

type BinlogListResult added in v1.0.4 

type BinlogListResult struct {
	Status int              `json:"status"`
	Msg    string           `json:"msg"`
	Data   []map[string]any `json:"data"`
}

BinlogListResult is the decoded response of POST /binlog/list/.

The view runs `show binary logs;` and returns each row as a map keyed by the result columns (Log_name, File_size, and Encrypted on 8.0). Rows are kept as generic maps because the column set varies across MySQL versions.

type BinlogParseParams added in v1.0.4 

type BinlogParseParams struct {
	InstanceName string
	StartFile    string
	EndFile      string
	StartPos     int
	EndPos       int
	StartTime    string
	StopTime     string
	Schemas      []string // -> only_schemas (repeated)
	Tables       []string // -> only_tables[] (repeated)
	SQLTypes     []string // -> sql_type[] (repeated)
	Num          int
	Threads      int
	Rollback     bool
	SaveSQL      bool
}

BinlogParseParams collects the my2sql request fields. Zero-valued numeric fields are sent as empty strings (see emptyIfZero) so the view applies its own defaults (num=30, threads=4) without 500ing on int(None); an empty sql_type list lets the view default to all DML.

type BinlogParseResult 

type BinlogParseResult struct {
	Status int             `json:"status"`
	Msg    string          `json:"msg"`
	Data   json.RawMessage `json:"data"`
}

BinlogParseResult is the decoded response of POST /binlog/my2sql/.

data is kept raw because its shape depends on status: the success path sets it to a list of parsed rows, while the arg-check failure path returns {"status":1,"data":{}} (an object). Rows() decodes the list only when the call succeeded, so an error envelope never causes an unmarshal failure.

func (*BinlogParseResult) Rows added in v1.0.4 

func (r *BinlogParseResult) Rows() ([]BinlogParseRow, error)

Rows decodes the parsed statements from a successful response. It returns an empty slice for any non-list data payload (e.g. the {} returned on error).

type BinlogParseRow added in v1.0.4 

type BinlogParseRow struct {
	SQL        string `json:"sql"`
	BinlogInfo string `json:"binlog_info,omitempty"`
}

BinlogParseRow is a single parsed statement. The my2sql engine returns only the SQL text; the binlog2sql engine also fills BinlogInfo (position metadata).

type BinlogPurgeResult added in v1.0.4 

type BinlogPurgeResult struct {
	Status int    `json:"status"`
	Msg    string `json:"msg"`
}

BinlogPurgeResult is the decoded response of POST /binlog/del_log/.

The view returns status 0 on success, 1 when no binlog was given, and 2 when the `purge master logs` statement failed.

type Client 

type Client struct {
	Auth      *AuthAPI
	Workflows *WorkflowAPI
	Binlog    *BinlogAPI
	Users     *UserAPI
	// contains filtered or unexported fields
}

Client wraps the Archery API HTTP client.

func NewClient 

func NewClient(host string) *Client

NewClient creates a new API client using global HTTP options. The transport defaults to session mode; callers switch to JWT via SetMode(ModeJWT).

func (*Client) APIPath 

func (c *Client) APIPath(resource string) string

APIPath returns "/api" + the relative resource path.

func (*Client) CookieJar 

func (c *Client) CookieJar() http.CookieJar

CookieJar returns the underlying cookie jar for session-based auth.

func (*Client) Delete 

func (c *Client) Delete(ctx context.Context, path string) error

Delete sends a DELETE request with Bearer token auth.

func (*Client) ExportSessionCookies added in v1.0.4 

func (c *Client) ExportSessionCookies() (sessionID, csrfToken string)

ExportSessionCookies returns the current sessionid and csrftoken from the jar, for persistence after a successful login. Either value may be empty.

func (*Client) Get 

func (c *Client) Get(ctx context.Context, path string) ([]byte, error)

Get sends a GET request with Bearer token auth. Returns body bytes only.

func (*Client) GetWithPagination 

func (c *Client) GetWithPagination(ctx context.Context, path string) ([]byte, Pagination, error)

GetWithPagination sends a GET request and returns body + pagination metadata from the response headers. Useful for list endpoints.

func (*Client) Host 

func (c *Client) Host() string

Host returns the underlying Archery host (without trailing slash).

func (*Client) InjectSessionCookies added in v1.0.4 

func (c *Client) InjectSessionCookies(sessionID, csrfToken string)

InjectSessionCookies seeds the cookie jar with cached Django cookies so that session requests can skip the interactive form login. Empty values are ignored. When a sessionid is present the client is marked ready, avoiding a redundant /authenticate/ round-trip on the first call.

func (*Client) InternalDelete 

func (c *Client) InternalDelete(ctx context.Context, path string) error

InternalDelete sends a DELETE request with session cookie auth.

func (*Client) InternalGet 

func (c *Client) InternalGet(ctx context.Context, path string) ([]byte, error)

InternalGet sends a GET request with session cookie auth.

func (*Client) InternalPost 

func (c *Client) InternalPost(ctx context.Context, path string, form url.Values) ([]byte, error)

InternalPost sends a POST request with session cookie auth.

func (*Client) InternalPut 

func (c *Client) InternalPut(ctx context.Context, path string, form url.Values) ([]byte, error)

InternalPut sends a PUT request with session cookie auth.

func (*Client) Mode added in v1.0.4 

func (c *Client) Mode() string

Mode returns the current transport mode.

func (*Client) Post 

func (c *Client) Post(ctx context.Context, path string, body any) ([]byte, error)

Post sends a POST request with Bearer token auth.

func (*Client) Put 

func (c *Client) Put(ctx context.Context, path string, body any) ([]byte, error)

Put sends a PUT request with Bearer token auth.

func (*Client) RefreshToken 

func (c *Client) RefreshToken() string

RefreshToken returns the current refresh token.

func (*Client) SessionGet added in v1.0.4 

func (c *Client) SessionGet(ctx context.Context, path string) ([]byte, error)

SessionGet sends a GET request with Django session cookie auth. Canonical name for the default (session) transport; InternalGet is kept as an alias used throughout the command layer.

func (*Client) SessionPost added in v1.0.4 

func (c *Client) SessionPost(ctx context.Context, path string, form url.Values) ([]byte, error)

SessionPost sends a form-encoded POST with Django session cookie auth.

func (*Client) SessionPostExpectRedirect added in v1.0.4 

func (c *Client) SessionPostExpectRedirect(ctx context.Context, path string, form url.Values) (int, error)

SessionPostExpectRedirect posts a form to a Django view that signals success with a 302 redirect to the new object's detail page, and returns the object id parsed from the Location header (e.g. /detail/42/ -> 42). Archery's submit views render error.html with HTTP 200 on a validation failure instead of redirecting, so a 200 (non-redirect) is reported as an error carrying the page's message hint. This does not follow the redirect: the id is in the Location header, and following it would lose it.

func (*Client) SetMode added in v1.0.4 

func (c *Client) SetMode(mode string)

SetMode selects the transport: ModeSession (default) or ModeJWT. Any other value falls back to session.

func (*Client) SetOTP added in v1.0.5 

func (c *Client) SetOTP(otp string)

SetOTP supplies a 6-digit 2FA code used to complete login for accounts with two-factor authentication enabled. Empty means "no code provided": ensureSession then surfaces an E_2FA_REQUIRED error when the account demands one.

func (*Client) SetSessionCredentials 

func (c *Client) SetSessionCredentials(username, password string)

SetSessionCredentials supplies the username/password used to establish a Django session for internal-mode (legacy web endpoint) commands. JWT-only configs (keyring) leave these empty; ensureSession then returns a clear error directing the caller to provide credentials.

func (*Client) SetTokens 

func (c *Client) SetTokens(accessToken, refreshToken string)

SetTokens sets the access and refresh tokens for REST-mode requests.

type ClientOptions 

type ClientOptions struct {
	Timeout            time.Duration
	InsecureSkipVerify bool
}

ClientOptions configures HTTP transport settings.

func CurrentClientOptions 

func CurrentClientOptions() ClientOptions

CurrentClientOptions returns the global HTTP settings (for tests/diagnostics).

type DataDictionaryColumn 

type DataDictionaryColumn struct {
	ColumnName    string `json:"column_name"`
	ColumnType    string `json:"column_type"`
	ColumnComment string `json:"column_comment"`
	IsNullable    string `json:"is_nullable"`
	ColumnKey     string `json:"column_key"`
	ColumnDefault any    `json:"column_default"`
}

DataDictionaryColumn describes a single column in a data dictionary table.

type DataDictionaryTable 

type DataDictionaryTable struct {
	ID          int    `json:"id"`
	Name        string `json:"name"`
	Comment     string `json:"comment"`
	InstanceID  int    `json:"instance_id"`
	DBName      string `json:"db_name"`
	TableName   string `json:"table_name"`
	IsPublished bool   `json:"is_published"`
}

DataDictionaryTable represents a table entry in the data dictionary.

type DataDictionaryTableInfo 

type DataDictionaryTableInfo struct {
	TableName string                 `json:"table_name"`
	Comment   string                 `json:"comment"`
	Columns   []DataDictionaryColumn `json:"columns"`
}

DataDictionaryTableInfo holds detailed column info for a dictionary table.

type Group 

type Group struct {
	ID   int    `json:"id"`
	Name string `json:"name"`
}

Group represents an Archery permission group.

type GroupRow added in v1.0.4 

type GroupRow struct {
	ID   int
	Name string
}

GroupRow is a Django auth group (permission group).

type Instance 

type Instance struct {
	ID           json.Number `json:"id"`
	InstanceName string      `json:"instance_name"`
	DbType       string      `json:"db_type"`
	Host         string      `json:"host"`
	Port         int         `json:"port"`
	User         string      `json:"user"`
	Password     string      `json:"password"`
	DBName       string      `json:"db_name"`
	Charset      string      `json:"charset"`
	IsActive     bool        `json:"is_active"`
	Environment  string      `json:"environment"`
	// Archery's instance_tag is a ManyToMany relation serialized as an array,
	// not a string (empty M2M -> []).
	InstanceTag []any `json:"instance_tag"`
}

Instance represents a database instance configured in Archery.

type InternalPage 

type InternalPage[T any] struct {
	Total int `json:"total"`
	Rows  []T `json:"rows"`
}

InternalPage holds the pagination payload for internal endpoints.

type InternalPaginatedResponse 

type InternalPaginatedResponse[T any] struct {
	Status int             `json:"status"`
	Msg    string          `json:"msg"`
	Data   InternalPage[T] `json:"data"`
}

InternalPaginatedResponse wraps Archery internal paginated responses that use {status, msg, data: {total, rows}} format.

type ListResult 

type ListResult[T any] struct {
	Page       PaginatedResponse[T]
	Pagination Pagination
}

ListResult holds the parsed response and pagination metadata for a list call.

type LockInfo 

type LockInfo struct {
	LockTrxID  string `json:"lock_trx_id"`
	LockMode   string `json:"lock_mode"`
	LockType   string `json:"lock_type"`
	LockTable  string `json:"lock_table"`
	LockIndex  string `json:"lock_index"`
	LockSpace  int    `json:"lock_space"`
	LockPage   int    `json:"lock_page"`
	LockRec    int    `json:"lock_rec"`
	LockData   string `json:"lock_data"`
	TrxID      string `json:"trx_id"`
	TrxState   string `json:"trx_state"`
	TrxStarted string `json:"trx_started"`
	TrxQuery   string `json:"trx_query"`
	TrxWait    int    `json:"trx_wait"`
}

LockInfo represents an InnoDB lock entry.

type PaginatedResponse 

type PaginatedResponse[T any] struct {
	Count    int `json:"count"`
	Next     any `json:"next"`
	Previous any `json:"previous"`
	Results  []T `json:"results"`
}

PaginatedResponse wraps DRF paginated list responses.

type Pagination 

type Pagination struct {
	// Count is the total number of items (from X-Total-Count header or JSON "count").
	Count int
	// Page is the current page number (from X-Page header, 1-based).
	Page int
	// PerPage is the page size (from X-Per-Page header).
	PerPage int
	// NextPage is the next page number (0 if last page).
	NextPage int
	// PrevPage is the previous page number (0 if first page).
	PrevPage int
	// Link is the raw Link header value (RFC 5988).
	Link string
}

Pagination holds page metadata from DRF response headers or response body.

Archery uses Django REST Framework. Depending on the paginator configured server-side, pagination info may arrive as:

  • HTTP headers: X-Total-Count, Link (RFC 5988)
  • JSON body: {"count": N, "next": "url", "previous": "url", "results": [...]}

This struct captures both sources. The JSON-body fields (Count, Next, Previous) come from PaginatedResponse; the header fields are populated by extractPagination when the caller uses GetWithPagination.

type ProcessInfo 

type ProcessInfo struct {
	ID      int    `json:"id"`
	User    string `json:"user"`
	Host    string `json:"host"`
	DB      string `json:"db"`
	Command string `json:"command"`
	Time    int    `json:"time"`
	State   string `json:"state"`
	Info    string `json:"info"`
}

ProcessInfo represents a MySQL process list entry.

type QueryLog 

type QueryLog struct {
	ID         int    `json:"id"`
	Username   string `json:"username"`
	DbUser     string `json:"db_user"`
	SQLText    string `json:"sqllog"`
	EffectRows int    `json:"effect_row"`
	CostTime   string `json:"cost_time"`
	Instance   string `json:"instance_name"`
	ExecTime   string `json:"exec_time"`
}

QueryLog represents a single query log entry.

type QueryResult 

type QueryResult struct {
	Rows         []map[string]any `json:"rows"`
	ColumnNames  []string         `json:"column_names"`
	RowCount     int              `json:"row_count"`
	AffectedRows int              `json:"affected_rows"`
	IsExecute    bool             `json:"is_execute"`
	ErrorMessage string           `json:"error_message"`
}

QueryResult holds the result of a SQL query execution.

type ResourceGroup 

type ResourceGroup struct {
	ID    int    `json:"id"`
	Name  string `json:"name"`
	Users []int  `json:"users"`
}

ResourceGroup represents a resource group for instance access control.

type ResourceGroupObjectType added in v1.0.4 

type ResourceGroupObjectType string

ResourceGroupObjectType selects which relation a resourcegroup-add mutation targets: users or instances. Archery's view encodes these as the string "0" and "1" respectively. 

const (
	ResourceGroupObjectUser     ResourceGroupObjectType = "0"
	ResourceGroupObjectInstance ResourceGroupObjectType = "1"
)

type ResourceGroupRow added in v1.0.4 

type ResourceGroupRow struct {
	ID          int
	Name        string
	ActiveUsers int
	BindUsers   int
}

ResourceGroupRow is a resource group used for instance access control. The session projection carries only the id/name (and ding webhook); the REST serializer additionally reports active/bind user counts, left zero in session mode.

type SQLCheckResult 

type SQLCheckResult struct {
	Level        string `json:"level"`
	Message      string `json:"message"`
	AffectedRows int    `json:"affected_rows"`
	SQL          string `json:"sql"`
	ErrLevel     int    `json:"errlevel"`
	StageStatus  string `json:"stagestatus,omitempty"`
}

SQLCheckResult is a single item from the SQL check response. Level/Message are the REST serializer's fields; ErrLevel/StageStatus come from the session ReviewResult and drive auto-review classification (errlevel: 0 ok, 1 warning, 2 error). When the session transport fills the result, Level is derived from ErrLevel so existing badge rendering keeps working.

type SQLWorkflow 

type SQLWorkflow struct {
	ID           int    `json:"id"`
	Title        string `json:"workflow_name"`
	Status       int    `json:"status"`
	StatusCode   string `json:"-"`
	Engineer     string `json:"engineer"`
	InstanceID   int    `json:"instance_id"`
	InstanceName string `json:"-"`
	DBName       string `json:"db_name"`
	GroupID      int    `json:"group_id"`
	GroupName    string `json:"-"`
	CreateDate   string `json:"create_date"`
}

SQLWorkflow is a single row from the workflow list endpoint. It is the common shape both transports map into: the REST serializer fills the numeric fields (Status, InstanceID, GroupID), while the session projection fills StatusCode (the string status), InstanceName and GroupName. The unset side stays zero.

type SQLWorkflowAuditEntry 

type SQLWorkflowAuditEntry struct {
	User       string `json:"audit_user"`
	Action     string `json:"audit_type"`
	Remark     string `json:"remark"`
	CreateDate string `json:"create_date"`
}

SQLWorkflowAuditEntry is a single audit log entry within a workflow detail.

type SQLWorkflowDetail 

type SQLWorkflowDetail struct {
	ID         int                     `json:"id"`
	Title      string                  `json:"workflow_name"`
	Status     int                     `json:"status"`
	Engineer   string                  `json:"engineer"`
	InstanceID int                     `json:"instance_id"`
	DBName     string                  `json:"db_name"`
	GroupID    int                     `json:"group_id"`
	SQLContent string                  `json:"sql_content"`
	IsBackup   bool                    `json:"is_backup"`
	DemandURL  string                  `json:"demand_url"`
	CreateDate string                  `json:"create_date"`
	AuditLog   []SQLWorkflowAuditEntry `json:"audit_log"`

	// 以下字段来自 session 路径的组合查询（detail_content + 列表），不来自单一 JSON 端点。
	// StatusCode 是字符串状态码（如 workflow_exception），从工单列表取，
	// 避免 detail 把缺省 status 误报成 0。
	StatusCode   string `json:"-"`
	InstanceName string `json:"-"`
	GroupName    string `json:"-"`
	// Result 是执行/审核结果行（含 stage、errlevel、errormessage、affected_rows），
	// 来自 /sqlworkflow/detail_content/，是「执行失败看原因」的关键数据。
	Result []reviewResultRow `json:"-"`
}

SQLWorkflowDetail is the full detail of an SQL workflow.

type SlowQuery 

type SlowQuery struct {
	ID           int     `json:"id"`
	SQLText      string  `json:"sql_text"`
	MySQLTotal   float64 `json:"mysql_total"`
	Count        int     `json:"count"`
	MaxQueryTime float64 `json:"max_query_time"`
	MysqlMaxTime float64 `json:"mysql_max_time"`
}

SlowQuery represents a slow query entry from the analysis system.

type SlowQueryHistory 

type SlowQueryHistory struct {
	ID           int     `json:"id"`
	SlowQueryID  int     `json:"slowquery_id"`
	SQLText      string  `json:"sql_text"`
	StartTime    string  `json:"start_time"`
	StopTime     string  `json:"stop_time"`
	QueryTimeAvg float64 `json:"query_time_avg"`
	LockTimeAvg  float64 `json:"lock_time_avg"`
	RowsExamined int     `json:"rows_examined_avg"`
	RowsSent     int     `json:"rows_sent_avg"`
}

SlowQueryHistory represents historical slow query data.

type TableSpace 

type TableSpace struct {
	TableName string `json:"table_name"`
	Engine    string `json:"engine"`
	RowFormat string `json:"row_format"`
	Rows      int    `json:"table_rows"`
	DataSize  string `json:"data_length"`
	IndexSize string `json:"index_length"`
	TotalSize string `json:"total_size"`
}

TableSpace represents table space usage information.

type TransactionInfo 

type TransactionInfo struct {
	TrxID           string `json:"trx_id"`
	TrxState        string `json:"trx_state"`
	TrxStarted      string `json:"trx_started"`
	TrxQuery        string `json:"trx_query"`
	TrxRowsLocked   int    `json:"trx_rows_locked"`
	TrxRowsModified int    `json:"trx_rows_modified"`
}

TransactionInfo represents a running transaction.

type User 

type User struct {
	ID       int    `json:"id"`
	Username string `json:"username"`
	Display  string `json:"display"`
	Email    string `json:"email"`
	IsActive bool   `json:"is_active"`
	IsStaff  bool   `json:"is_staff"`
	Groups   []int  `json:"groups"`
}

User represents an Archery user.

type UserAPI added in v1.0.4 

type UserAPI struct {
	// contains filtered or unexported fields
}

UserAPI provides methods for the user / group / resource-group endpoints.

Each method branches on the client transport mode. Session mode (the default) targets Archery's AJAX/Django endpoints that ordinary accounts can reach; JWT mode targets the REST API under /api/v1/. The session shapes are taken from sql/user.py and sql/resource_group.py in the v1.8.5 source.

func (*UserAPI) AddResourceGroupRelation added in v1.0.4 

func (u *UserAPI) AddResourceGroupRelation(ctx context.Context, groupID int, objectType ResourceGroupObjectType, objectIDs []int) error

AddResourceGroupRelation associates users or instances with a resource group. 

session: POST /group/addrelation/ (group_id, object_type, object_info)

object_info is a JSON-encoded array of strings; the view splits each element on "," and takes the leading integer as the object id (matching the web UI's "id,name" rows). This is a session-only operation: there is no REST/JWT route for it in v1.8.5, so the call uses the session transport regardless of mode.

func (*UserAPI) ListGroups added in v1.0.4 

func (u *UserAPI) ListGroups(ctx context.Context, limit, offset int) ([]GroupRow, int, error)

ListGroups returns Django auth (permission) groups via the REST endpoint. 

jwt: GET /api/v1/user/group/?limit&offset

Session mode has no JSON route in v1.8.5: the web page /group/ renders HTML, not a list payload. The caller gates session mode before reaching here.

func (*UserAPI) ListResourceGroups added in v1.0.4 

func (u *UserAPI) ListResourceGroups(ctx context.Context, limit, offset int, search string) ([]ResourceGroupRow, int, error)

ListResourceGroups returns resource groups. 

session: POST /group/group/ (limit, offset, search)
jwt:     GET  /api/v1/user/resourcegroup/?limit&offset&search

func (*UserAPI) ListUsers added in v1.0.4 

func (u *UserAPI) ListUsers(ctx context.Context) ([]UserRow, error)

ListUsers returns all users. 

session: GET /user/list/ (no params; returns the full list)
jwt:     GET /api/v1/user/?limit&offset&search

The session view (user.lists) does no server-side pagination or search, so those filters are applied client-side by the caller.

type UserRow added in v1.0.4 

type UserRow struct {
	ID          int
	Username    string
	Display     string
	Email       string
	IsActive    bool
	IsStaff     bool
	IsSuperuser bool
}

UserRow is the common shape both transports map into. The session projection (user.lists) fills IsSuperuser; the REST serializer does not expose it, so the unset side stays zero/false.

type Workflow 

type Workflow struct {
	ID      int               `json:"id"`
	Name    string            `json:"name"`
	IsUsed  bool              `json:"is_used"`
	Content []WorkflowContent `json:"content"`
}

Workflow represents an Archery workflow definition (template).

type WorkflowAPI 

type WorkflowAPI struct {
	// contains filtered or unexported fields
}

WorkflowAPI provides methods for the SQL workflow endpoints.

Each method branches on the client transport mode. Session mode (the default) targets Archery's AJAX/Django endpoints, which ordinary accounts can reach; JWT mode targets the REST API under /api/v1/. The session shapes are taken from sql/sql_workflow.py in the v1.8.5 source.

func (*WorkflowAPI) Audit 

func (w *WorkflowAPI) Audit(ctx context.Context, req WorkflowAuditRequest) error

Audit approves a workflow (audit pass). Cancel/reject goes through Cancel. 

session: POST /passed/ (workflow_id + audit_remark; requires auditor)
jwt:     POST /api/v1/workflow/audit/

func (*WorkflowAPI) Cancel 

func (w *WorkflowAPI) Cancel(ctx context.Context, req WorkflowCancelRequest) error

Cancel terminates a workflow. 

session: POST /cancel/ (workflow_id + cancel_remark)

The session field is cancel_remark and the view requires it to be present (a nil value renders "终止原因不能为空"). There is no REST cancel endpoint.

func (*WorkflowAPI) Detail 

func (w *WorkflowAPI) Detail(ctx context.Context, id int) (*SQLWorkflowDetail, error)

Detail fetches workflow details. v1.8.5 has no REST detail endpoint (/api/v1/workflow/<id>/ -> 404), so both modes read the session view. 

GET /sqlworkflow/detail_content/?workflow_id=<id>

The view returns {"rows":[ReviewResult...]} — the SQL check / execution result rows, not workflow metadata (no such metadata endpoint exists in v1.8.5). SQL content is reconstructed from the rows. List rows carry the metadata (title/status/engineer) when the caller has them.

func (*WorkflowAPI) Execute 

func (w *WorkflowAPI) Execute(ctx context.Context, req WorkflowExecuteRequest) error

Execute starts execution of an approved workflow. 

session: POST /execute/ (workflow_id + mode; requires execute permission)
jwt:     POST /api/v1/workflow/execute/

func (*WorkflowAPI) List 

func (w *WorkflowAPI) List(ctx context.Context, params WorkflowListParams) (*ListResult[SQLWorkflow], error)

List returns a paginated list of SQL workflows. 

session: POST /sqlworkflow_list/ (or /sqlworkflow_list_audit/ when Audit)
jwt:     GET  /api/v1/workflow/

func (*WorkflowAPI) SQLCheck 

func (w *WorkflowAPI) SQLCheck(ctx context.Context, req WorkflowSQLCheckRequest) ([]SQLCheckResult, error)

SQLCheck runs SQL syntax and risk checks without creating a workflow. 

session: POST /simplecheck/ (instance_name + db_name + sql_content)
jwt:     POST /api/v1/workflow/sqlcheck/

func (*WorkflowAPI) Submit 

func (w *WorkflowAPI) Submit(ctx context.Context, req WorkflowSubmitRequest) (*WorkflowSubmitResponse, error)

Submit creates a new SQL workflow. 

session: POST /autoreview/ (names; is_backup as "True"/"False"; 302 redirect)
jwt:     POST /api/v1/workflow/

type WorkflowAudit 

type WorkflowAudit struct {
	ID           int    `json:"id"`
	Title        string `json:"title"`
	Workflow     int    `json:"workflow"`
	WorkflowType int    `json:"workflow_type"`
	Status       int    `json:"status"`
	CurrentStep  int    `json:"current_step"`
	CreateUser   string `json:"create_user"`
	CreateDate   string `json:"create_date"`
	InstanceId   int    `json:"instance_id"`
}

WorkflowAudit represents a pending workflow audit entry.

type WorkflowAuditRequest 

type WorkflowAuditRequest struct {
	WorkflowID int    `json:"workflow_id"`
	Action     string `json:"audit_type"`
	// Archery's AuditWorkflowSerializer names this audit_remark and treats it as
	// required; sending "remark" left it empty and the API rejected the audit.
	Remark string `json:"audit_remark"`
}

WorkflowAuditRequest is the payload for POST /api/v1/workflow/audit/.

type WorkflowCancelRequest 

type WorkflowCancelRequest struct {
	WorkflowID int    `json:"workflow_id"`
	Remark     string `json:"remark,omitempty"`
}

WorkflowCancelRequest is the payload for POST /sqlworkflow/cancel/.

type WorkflowContent 

type WorkflowContent struct {
	Step            int    `json:"step"`
	Assignee        int    `json:"assignee"`
	AuditAuthGroups string `json:"audit_auth_groups"`
}

WorkflowContent is a single step in a workflow definition.

type WorkflowExecuteRequest 

type WorkflowExecuteRequest struct {
	WorkflowID   int    `json:"workflow_id"`
	WorkflowType int    `json:"workflow_type"`
	Engineer     string `json:"engineer,omitempty"`
	Mode         string `json:"mode"`
}

WorkflowExecuteRequest is the payload for POST /api/v1/workflow/execute/.

Archery's ExecuteWorkflowSerializer requires workflow_type and, for SQL-review workflows (type 2), engineer + mode — not just workflow_id. WorkflowType defaults to 2 (SQL上线申请); Engineer is the authenticated operator.

type WorkflowListParams 

type WorkflowListParams struct {
	Status     string
	Engineer   string
	InstanceID int
	GroupID    int
	DBName     string
	Search     string
	StartDate  string
	EndDate    string
	Limit      int
	Offset     int
	Audit      bool
}

WorkflowListParams holds query parameters for the workflow list endpoint.

Session and REST transports accept overlapping but not identical filters. The session endpoint (/sqlworkflow_list/) filters on navStatus (the string status code), group_id, instance_id, search (matches engineer/title) and a start/end date range; it has no engineer or db_name filter. REST keeps the engineer/db_name filters it already supported. Audit narrows the list to the current user's pending-audit scope (/sqlworkflow_list_audit/).

type WorkflowSQLCheckRequest 

type WorkflowSQLCheckRequest struct {
	InstanceID   int    `json:"instance_id"`
	InstanceName string `json:"-"`
	DBName       string `json:"db_name"`
	SQLContent   string `json:"sql_content"`
}

WorkflowSQLCheckRequest is the payload for the SQL pre-check. Session (/simplecheck/) uses instance_name; REST uses instance_id. The command fills the identifier the active transport needs.

type WorkflowSubmitRequest 

type WorkflowSubmitRequest struct {
	Name         string   `json:"workflow_name"`
	InstanceID   int      `json:"instance_id"`
	InstanceName string   `json:"-"`
	DBName       string   `json:"db_name"`
	SQLContent   string   `json:"sql_content"`
	GroupID      int      `json:"group_id,omitempty"`
	GroupName    string   `json:"-"`
	IsBackup     bool     `json:"is_backup"`
	DemandURL    string   `json:"demand_url,omitempty"`
	CCUsers      []string `json:"-"`
	RunDateStart string   `json:"-"`
	RunDateEnd   string   `json:"-"`
}

WorkflowSubmitRequest is the payload for submitting a workflow.

The two transports key on different identifiers. The session endpoint (/autoreview/) takes names — InstanceName and GroupName — plus is_backup as the string "True"/"False"; the REST endpoint takes the numeric InstanceID and GroupID. The command fills whichever the active mode needs; the API layer reads the matching set. CCUsers and the RunDate* window are session-only optional fields (carbon-copy notify recipients, timed execution).

type WorkflowSubmitResponse 

type WorkflowSubmitResponse struct {
	ID int `json:"id"`
}

WorkflowSubmitResponse is returned after a successful workflow submission.

Source Files

View all Source files

Jump to

Keyboard shortcuts

? : This menu
/ : Search site
f or F : Jump to
y or Y : Canonical URL
go.dev uses cookies from Google to deliver and enhance the quality of its services and to analyze traffic. Learn more.