ingitdb

package module
v1.32.0 Latest Latest
Warning

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

 Go to latest Published: Jun 15, 2026 License: MIT Imports: 21 Imported by: 0

README

Package ingitdb

inGitDB schema & config types.

Schema types

  • type Definition — top-level schema definition; holds all collection definitions
    • type CollectionDef — schema for one collection
    • type ViewDef — materialized view definition (ordering, columns, format, top-N limit)

Configuration types

  • package config
    • type RootConfig — parsed from .ingitdb.yaml; maps collection keys to paths

Documentation

Overview

specscore: feature/cli/resolve/auto-resolve/record-merge

Index

Constants

View Source 
const CollectionDefFileName = "definition.yaml"

CollectionDefFileName is the fixed file name for collection definitions inside the SchemaDir directory.

View Source 
const CollectionsDir = ".collections"

CollectionsDir is the shared-directory layout folder name. When a directory contains a CollectionsDir sub-folder, each non-$-prefixed sub-directory inside it is treated as a separate collection (ID = sub-directory name).

View Source 
const DefaultMarkdownContentField = "$content"

DefaultMarkdownContentField is the default name of the column that holds the Markdown body when a collection uses `format: markdown` and does not override `record_file.content_field`.

View Source 
const DefaultViewID = "$default_view"
View Source 
const IngitdbDir = "$ingitdb"
View Source 
const SchemaDir = ".collection"
View Source 
const SharedViewsDir = "$views"

SharedViewsDir is the reserved sub-folder name for named views inside a CollectionsDir/{name}/ directory (new layout).

Variables

This section is empty.

Functions

func ApplyLocaleToRead 

func ApplyLocaleToRead(data map[string]any, cols map[string]*ColumnDef) map[string]any

ApplyLocaleToRead transforms record data from file representation to application representation. For each column that has a Locale value set (e.g. column "title" with locale "en"), the paired map column (e.g. "titles") is inspected: the locale entry is extracted and exposed as the shortcut column ("title"), and that locale key is removed from the pair map to avoid duplication. The caller receives e.g. {"title": "Work", "titles": {"ru": "Работа"}}.

func ApplyLocaleToWrite 

func ApplyLocaleToWrite(data map[string]any, cols map[string]*ColumnDef) map[string]any

ApplyLocaleToWrite normalises record data before writing to file. For each column that has a Locale value set (e.g. column "title" with locale "en"):

  • The shortcut column ("title") is stored as-is in the file.
  • If the paired map column ("titles") contains an entry for the primary locale key ("en"), that entry is promoted to the shortcut column and removed from the map, so the value is never duplicated across both fields.
  • If the paired map becomes empty after removing the primary locale entry, it is dropped from the result to avoid writing a redundant empty map.

func EncodeListOfRecordsContent 

func EncodeListOfRecordsContent(rows []map[string]any, format RecordFormat, columnsOrder []string) ([]byte, error)

EncodeListOfRecordsContent serializes list rows back to the declared format, preserving record (insertion) order. Within each record, keys are emitted in columnsOrder first, then remaining keys alphabetically. JSONL writes one compact JSON object per line; JSON writes a pretty array; YAML writes a top-level sequence. csv and ingr keep their dedicated encoders.

func EncodeMapOfRecordsContent 

func EncodeMapOfRecordsContent(data map[string]map[string]any, format RecordFormat, recordsetName string, columnsOrder []string) ([]byte, error)

EncodeMapOfRecordsContent serializes a map of ID-keyed records into the declared format. For yaml/json/toml, the data is written as a top-level mapping (record ID -> field map). For INGR, the data is flattened into a list of records (each gets `$ID` injected as the first column) and written via the ingr-io writer.

recordsetName is used as the INGR header title (typically the collection ID). It is ignored for non-INGR formats. columnsOrder controls the column order in the INGR header; when empty, columns are emitted in alphabetical order with `$ID` first.

func EncodeRecordContentForCollection 

func EncodeRecordContentForCollection(value any, colDef *CollectionDef) ([]byte, error)

EncodeRecordContentForCollection serializes record content using the collection's declared format. It is the write-side counterpart of ParseRecordContentForCollection and the only path callers should use when writing records that may be in a format that requires schema access (csv today; possibly others later).

For csv, value MUST be []map[string]any — see encodeCSVForCollection. For the other six formats, value is passed through to marshalForFormat unchanged; callers that want a single-record map[string]any can keep using the schema-agnostic marshalForFormat directly.

func EvaluateFormula 

func EvaluateFormula(formula string, fields map[string]any) (any, error)

EvaluateFormula evaluates a computed-column formula as a single Starlark expression in a deterministic, side-effect-free sandbox.

Each entry of fields is bound as a top-level variable available to the expression. Supported Go input types are string, bool, int, int8, int16, int32, int64, uint, uint8, uint16, uint32, uint64, float32, and float64. The result is converted back to a Go-native value: string, bool, int64, or float64. A nil field value or a None result maps to/from Go nil.

The sandbox exposes no network, filesystem, clock, or randomness, and installs no load() loader, so evaluation has zero side effects and is deterministic: identical formula and fields always yield identical output. Evaluation is bounded by maxFormulaSteps. The compiled program is cached, so repeated reads of the same formula do not re-parse it.

func ParseListOfRecordsContent 

func ParseListOfRecordsContent(content []byte, format RecordFormat) ([]map[string]any, error)

ParseListOfRecordsContent parses a list-of-records file into ordered row maps. It handles a top-level YAML sequence, a top-level JSON array, and a JSONL stream (one JSON object per non-empty line). Empty content yields no rows. csv and ingr keep their dedicated parsers and are not handled here.

func ParseMapOfRecordsContent 

func ParseMapOfRecordsContent(content []byte, format RecordFormat) (map[string]map[string]any, error)

ParseMapOfRecordsContent parses content containing a map of ID-keyed records.

For yaml/json/toml, the file's top-level structure is a mapping from record ID to a per-record field map; we re-shape it into map[string]map[string]any.

For INGR, the file is a list of records where the reserved `$ID` column holds each record's key; we read the list and re-index by `$ID`. Records missing `$ID`, or with duplicate `$ID` values, are rejected as malformed.

func ParseRecordContent 

func ParseRecordContent(content []byte, format RecordFormat) (map[string]any, error)

ParseRecordContent parses record content in YAML or JSON format. For markdown-format collections use ParseRecordContentForCollection, which also has access to the column schema and content-field name.

func ParseRecordContentForCollection 

func ParseRecordContentForCollection(content []byte, colDef *CollectionDef) (map[string]any, error)

ParseRecordContentForCollection parses record content using the collection's declared format. For YAML and JSON this is equivalent to ParseRecordContent. For markdown records it parses YAML frontmatter, filters to columns declared in the schema, and exposes the body under the configured content_field column name.

func ResolveListRecordKey 

func ResolveListRecordKey(row map[string]any, colDef *CollectionDef) (string, bool)

ResolveListRecordKey resolves a list row's record key, in order: the collection's declared primary_key (composite values joined), else a "$id" field, else an "id" field. ok is false when none is available, meaning the list has no usable record key.

func Validate 

func Validate() func(*ReadOptions)

func ValidateCollectionID 

func ValidateCollectionID(id string) error

ValidateCollectionID validates a full collection ID. Allowed characters are alphanumeric, dot, and underscore; ID must start/end with alphanumeric.

func ValidateColumnType 

func ValidateColumnType(ct ColumnType) error

Types

type ChangeKind 

type ChangeKind string

ChangeKind describes how a file changed between two git refs. 

const (
	ChangeKindAdded    ChangeKind = "added"
	ChangeKindModified ChangeKind = "modified"
	ChangeKindDeleted  ChangeKind = "deleted"
	ChangeKindRenamed  ChangeKind = "renamed"
)

type ChangedFile 

type ChangedFile struct {
	Kind    ChangeKind
	Path    string // repo-relative path
	OldPath string // set only for ChangeKindRenamed
}

ChangedFile is one file that differed between two git refs.

type CollectionDef 

type CollectionDef struct {
	ID           string                `json:"-"` // Taken from dir name
	DirPath      string                `yaml:"-" json:"-"`
	Titles       map[string]string     `yaml:"titles,omitempty"`
	RecordFile   *RecordFileDef        `yaml:"record_file"`
	DataDir      string                `yaml:"data_dir,omitempty"`
	Columns      map[string]*ColumnDef `yaml:"columns"`
	ColumnsOrder []string              `yaml:"columns_order,omitempty"`
	// PrimaryKey lists the column names composing the source primary key,
	// in declared order. Persisted by CreateCollection so that
	// DescribeCollection can round-trip the real PK column names instead of
	// the synthesized "$key" placeholder. Omitted from older
	// definition.yaml files; callers should fall back to "$key" when empty.
	PrimaryKey  []string `yaml:"primary_key,omitempty"`
	DefaultView *ViewDef `yaml:"default_view,omitempty"`
	// SubCollections are not part of the collection definition file,
	// they are stored in the "subcollections" subdirectory as directories,
	// each containing their own .collection/definition.yaml.
	SubCollections map[string]*CollectionDef `yaml:"-" json:"-"`
	// Views are not part of the collection definition file,
	// they are stored in the "views" subdirectory.
	Views map[string]*ViewDef `yaml:"-" json:"-"`

	Readme *CollectionReadmeDef `yaml:"readme,omitempty" json:"readme,omitempty"`

	// ConflictResolution overrides the database-level conflict-resolution
	// settings for this collection. Nil fields inherit the database default.
	ConflictResolution *ConflictResolutionConfig `yaml:"conflict_resolution,omitempty" json:"conflict_resolution,omitempty"`
}

func (*CollectionDef) MarshalYAML 

func (c *CollectionDef) MarshalYAML() (interface{}, error)

MarshalYAML emits a CollectionDef as a YAML mapping with deterministic column ordering: columns named in ColumnsOrder appear in that order; any remaining columns follow in alphabetical order. The other fields keep their struct-declaration order.

This makes any yaml.Marshal of a CollectionDef diff-stable across runs and across machines.

func (*CollectionDef) Validate 

func (v *CollectionDef) Validate() error

type CollectionReadmeDef 

type CollectionReadmeDef struct {
	HideColumns        bool     `yaml:"hide_columns,omitempty" json:"hide_columns,omitempty"`
	HideSubcollections bool     `yaml:"hide_subcollections,omitempty" json:"hide_subcollections,omitempty"`
	HideViews          bool     `yaml:"hide_views,omitempty" json:"hide_views,omitempty"`
	HideTriggers       bool     `yaml:"hide_triggers,omitempty" json:"hide_triggers,omitempty"`
	DataPreview        *ViewDef `yaml:"data_preview,omitempty" json:"data_preview,omitempty"`
}

func (*CollectionReadmeDef) Validate 

func (r *CollectionReadmeDef) Validate() error

type CollectionsReader 

type CollectionsReader interface {
	ReadDefinition(dbPath string, opts ...ReadOption) (*Definition, error)
}

CollectionsReader loads the full database definition from disk. The default implementation wraps validator.ReadDefinition.

type ColumnDef 

type ColumnDef struct {
	Type       ColumnType        `yaml:"type"`
	Title      string            `yaml:"title,omitempty"`
	Titles     map[string]string `yaml:"titles,omitempty"`
	ValueTitle string            `yaml:"valueTitle,omitempty"`
	Required   bool              `yaml:"required,omitempty"`
	Length     int               `yaml:"length,omitempty"`
	MinLength  int               `yaml:"min_length,omitempty"`
	MaxLength  int               `yaml:"max_length,omitempty"`
	ForeignKey string            `yaml:"foreign_key,omitempty"`
	// Locale pairs this column with a map[locale]string column named <this_column_name>+"s".
	// For example, column "title" with locale "en" is paired with column "titles".
	// When reading, the locale value is extracted from the pair column and exposed as this column.
	// When writing, this column is stored as-is in the file; if the pair column contains an entry
	// for the primary locale key, that entry is promoted here and removed from the pair column.
	Locale string `yaml:"locale,omitempty"`
	// Format is an optional, free-form hint about the column's logical content
	// type. Well-known values include `markdown`, `html`, `json`, `jsonl`,
	// `yaml`, `uri`, `email`, `pdf`. inGitDB does not validate the value;
	// tooling may use it to choose a renderer or preview strategy.
	Format string `yaml:"format,omitempty"`
	// Formula declares this column as a computed (virtual) column. When set,
	// it must be a single Starlark expression that references only stored
	// (non-computed) sibling fields. Computed columns support only the
	// string, int, float, bool, and any declared types.
	//
	// Note: Starlark's `/` operator is float division, so for an int column
	// use integer division `//` (e.g. `total // count`) — `a / b` yields a
	// float and fails coercion into an int column unless the result is whole.
	Formula string `yaml:"formula,omitempty"`
}

func (*ColumnDef) Validate 

func (v *ColumnDef) Validate() error

type ColumnDefWithID 

type ColumnDefWithID struct {
	ID string `yaml:"id"`
	ColumnDef
}

type ColumnType 

type ColumnType string
const (
	ColumnTypeL10N     ColumnType = "map[locale]string"
	ColumnTypeString   ColumnType = "string"
	ColumnTypeInt      ColumnType = "int"
	ColumnTypeFloat    ColumnType = "float"
	ColumnTypeBool     ColumnType = "bool"
	ColumnTypeDate     ColumnType = "date"
	ColumnTypeTime     ColumnType = "time"
	ColumnTypeDateTime ColumnType = "datetime"
	ColumnTypeAny      ColumnType = "any"
)

type ConflictResolutionConfig 

type ConflictResolutionConfig struct {
	RecordMerge *RecordMergeConfig `yaml:"record_merge,omitempty"`
}

ConflictResolutionConfig groups conflict-resolution settings. It is set at the database level (in .ingitdb settings) and may be overridden per collection (in the collection's definition).

type Definition 

type Definition struct {
	Settings         Settings                  `yaml:"settings,omitempty"`
	RuntimeOverrides RuntimeOverrides          `yaml:"-"`
	Collections      map[string]*CollectionDef `yaml:"collections,omitempty"`
	Subscribers      map[string]*SubscriberDef `yaml:"subscribers,omitempty"`
}

type DiscordDef 

type DiscordDef struct {
	HandlerBase `yaml:",inline"`
	WebhookURL  string `yaml:"webhook_url"`
}

DiscordDef represents a Discord webhook subscriber handler

func (*DiscordDef) Validate 

func (d *DiscordDef) Validate() error

type EffectiveRecordMerge 

type EffectiveRecordMerge struct {
	Enabled    bool
	SameRecord bool
}

EffectiveRecordMerge is the resolved record-merge configuration after the database default and any per-collection override are applied.

func ResolveRecordMerge 

func ResolveRecordMerge(def *Definition, col *CollectionDef) EffectiveRecordMerge

ResolveRecordMerge computes the effective record-merge configuration for a collection: app defaults (enabled, not same-record) overlaid by the database-level config, then the per-collection override. Either argument may be nil.

type EmailDef 

type EmailDef struct {
	HandlerBase `yaml:",inline"`
	From        string   `yaml:"from,omitempty"`
	To          []string `yaml:"to"`
	SMTP        string   `yaml:"smtp"`
	Port        int      `yaml:"port,omitempty"`
	User        string   `yaml:"user,omitempty"`
	Pass        string   `yaml:"pass,omitempty"`
	Subject     string   `yaml:"subject,omitempty"`
}

EmailDef represents an email subscriber handler via SMTP

func (*EmailDef) Validate 

func (d *EmailDef) Validate() error

type GitHubActionDef 

type GitHubActionDef struct {
	HandlerBase `yaml:",inline"`
	Owner       string `yaml:"owner"`
	Repo        string `yaml:"repo"`
	Workflow    string `yaml:"workflow"`
	Ref         string `yaml:"ref"`
	Token       string `yaml:"token"`
}

GitHubActionDef represents a GitHub Actions workflow_dispatch subscriber handler

func (*GitHubActionDef) Validate 

func (d *GitHubActionDef) Validate() error

type GitLabCIDef 

type GitLabCIDef struct {
	HandlerBase `yaml:",inline"`
	ProjectID   string `yaml:"project_id"`
	Ref         string `yaml:"ref"`
	Token       string `yaml:"token"`
	Host        string `yaml:"host,omitempty"`
}

GitLabCIDef represents a GitLab CI pipeline trigger subscriber handler

func (*GitLabCIDef) Validate 

func (d *GitLabCIDef) Validate() error

type HandlerBase 

type HandlerBase struct {
	Name string `yaml:"name,omitempty"`
}

HandlerBase contains common fields for all subscriber handlers

type IRecordEntry 

type IRecordEntry interface {
	GetID() string
	GetData() map[string]any
}

func NewMapRecordEntry 

func NewMapRecordEntry[TKey comparable](id TKey, data map[string]any) IRecordEntry

type MaterializeResult 

type MaterializeResult struct {
	FilesCreated   int
	FilesUpdated   int
	FilesUnchanged int
	FilesDeleted   int
	Errors         []error
}

MaterializeResult summarises the outcome of a materialisation run.

type NtfyDef 

type NtfyDef struct {
	HandlerBase `yaml:",inline"`
	Topic       string `yaml:"topic"`
	Server      string `yaml:"server,omitempty"`
}

NtfyDef represents a ntfy.sh push notification subscriber handler

func (*NtfyDef) Validate 

func (d *NtfyDef) Validate() error

type ProgressEvent 

type ProgressEvent struct {
	Kind     ProgressKind
	TaskName string           // "validate" or "materialize"
	Scope    string           // collection or view ID
	ItemKey  string           // record key or output file name
	Done     int              // items completed so far
	Total    int              // total items; 0 = unknown
	Err      *ValidationError // non-nil only for ProgressKindError
}

ProgressEvent carries one progress update from a running task.

type ProgressKind 

type ProgressKind string

ProgressKind describes the type of progress update. 

const (
	ProgressKindStarted   ProgressKind = "started"
	ProgressKindItemDone  ProgressKind = "item_done"
	ProgressKindSkipped   ProgressKind = "skipped"
	ProgressKindCompleted ProgressKind = "completed"
	ProgressKindAborted   ProgressKind = "aborted"
	ProgressKindError     ProgressKind = "error"
)

type RSSDef 

type RSSDef struct {
	HandlerBase `yaml:",inline"`
	Output      string `yaml:"output"`
	Title       string `yaml:"title"`
	Link        string `yaml:"link"`
	Format      string `yaml:"format,omitempty"`
}

RSSDef represents an RSS or Atom feed generator subscriber handler

func (*RSSDef) Validate 

func (d *RSSDef) Validate() error

type ReadOption 

type ReadOption func(*ReadOptions)

type ReadOptions 

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

func NewReadOptions 

func NewReadOptions(o ...ReadOption) (opts ReadOptions)

func (*ReadOptions) IsValidationRequired 

func (o *ReadOptions) IsValidationRequired() bool

type RecordEntry 

type RecordEntry struct {
	ID   string // can be empty for list-type files
	Data map[string]any
}

RecordEntry is one parsed record.

func (RecordEntry) GetData 

func (r RecordEntry) GetData() map[string]any

func (RecordEntry) GetID 

func (r RecordEntry) GetID() string

type RecordFileDef 

type RecordFileDef struct {
	Name   string       `yaml:"name"`
	Format RecordFormat `yaml:"format"`

	// RecordType can have next values:
	// "map[string]any" - each record in a separate file
	// "[]map[string]any" - list of records
	// "map[$record_id]map[$field_name]any" - all records in one file; top-level keys are record IDs, second level is field names
	RecordType RecordType `yaml:"type"`

	// ContentField is the name of the column that maps to the Markdown body
	// in a `format: markdown` collection. When empty, the default
	// (`DefaultMarkdownContentField`, `$content`) is used.
	// It MUST only be set when Format is `markdown`.
	ContentField string `yaml:"content_field,omitempty"`

	// ExcludeRegex is an optional regular expression applied to the
	// basename of each candidate record file. Files whose basename matches
	// MUST be excluded from reads, validation, and record counts.
	//
	// Typical use: a record directory that legitimately contains a
	// README.md or .gitkeep alongside `{key}.md` records can set
	// `exclude_regex: '^README\.md$'` to keep those auxiliary files from
	// being treated as records.
	ExcludeRegex string `yaml:"exclude_regex,omitempty"`
}

func (RecordFileDef) GetRecordFileName 

func (rfd RecordFileDef) GetRecordFileName(record dal.Record) string

func (RecordFileDef) IsExcluded 

func (rfd RecordFileDef) IsExcluded(filename string) bool

IsExcluded reports whether a filename's basename matches the configured `exclude_regex`. Files for which IsExcluded returns true MUST be omitted from reads, validation, and record counts.

When ExcludeRegex is empty, IsExcluded always returns false. When the regex fails to compile (caller skipped Validate()), IsExcluded returns false rather than panicking — Validate() is the right place to surface compile errors.

func (RecordFileDef) RecordsBasePath 

func (rfd RecordFileDef) RecordsBasePath() string

RecordsBasePath returns "$records" when record_file.name contains {key}, causing inGitDB to store individual record files under a $records/ subdirectory. This keeps README.md visible at the top of the collection directory on GitHub.com.

func (RecordFileDef) ResolvedContentField 

func (rfd RecordFileDef) ResolvedContentField() string

ResolvedContentField returns the configured content_field name when set, otherwise the default (`$content`). Only meaningful when Format is `markdown`.

func (RecordFileDef) Validate 

func (rfd RecordFileDef) Validate() error

type RecordFormat 

type RecordFormat string
const (
	RecordFormatYAML     RecordFormat = "yaml"
	RecordFormatYML      RecordFormat = "yml"
	RecordFormatJSON     RecordFormat = "json"
	RecordFormatMarkdown RecordFormat = "markdown"
	RecordFormatTOML     RecordFormat = "toml"
	RecordFormatINGR     RecordFormat = "ingr"
	RecordFormatCSV      RecordFormat = "csv"
	RecordFormatJSONL    RecordFormat = "jsonl"
)

Recognized record file formats. Tooling MUST treat unrecognized values as unsupported. The values are the strings that appear in `record_file.format` in `.collection/definition.yaml`.

type RecordMergeConfig 

type RecordMergeConfig struct {
	// Enabled turns the default disjoint/non-divergent auto-merge on or off.
	// Defaults to true when unset at every scope.
	Enabled *bool `yaml:"enabled,omitempty"`
	// SameRecord enables opt-in merging of non-contested changes to the same
	// record. Defaults to false when unset at every scope.
	SameRecord *bool `yaml:"same_record,omitempty"`
}

RecordMergeConfig configures the record-aware auto-merge of data-row conflicts. Both fields are pointers so an explicit `false` at a narrower scope (e.g. a collection) can override an inherited `true` — a nil value means "inherit".

type RecordType 

type RecordType string
const (
	SingleRecord  RecordType = "map[string]any"
	ListOfRecords RecordType = "[]map[string]any"
	MapOfRecords  RecordType = "map[$record_id]map[$field_name]any"
)

type RecordsReader 

type RecordsReader interface {
	ReadRecords(
		ctx context.Context,
		dbPath string,
		col *CollectionDef,
		yield func(IRecordEntry) error,
	) error
}

RecordsReader streams records from one collection. Uses a yield callback to avoid goroutine leaks and keep allocation low.

type RuntimeOverrides 

type RuntimeOverrides struct {
	RecordsDelimiter *int `yaml:"-"`
}

RuntimeOverrides holds values set at runtime (e.g. CLI flags) that take highest priority over schema settings. Not persisted to YAML.

type SMSDef 

type SMSDef struct {
	HandlerBase `yaml:",inline"`
	Provider    string `yaml:"provider"`
	From        string `yaml:"from"`
	To          string `yaml:"to"`
	AccountSID  string `yaml:"account_sid,omitempty"`
	AuthToken   string `yaml:"auth_token"`
	APIKey      string `yaml:"api_key,omitempty"`
}

SMSDef represents an SMS subscriber handler via Twilio or Vonage

func (*SMSDef) Validate 

func (d *SMSDef) Validate() error

type Scanner 

type Scanner interface {
	// Scan walks dbPath, validates all records, and rebuilds all views.
	Scan(ctx context.Context, dbPath string, def *Definition) error
}

Scanner orchestrates the full pipeline: walk filesystem, invoke Validator and ViewBuilder.

type SearchIndexDef 

type SearchIndexDef struct {
	HandlerBase `yaml:",inline"`
	Provider    string `yaml:"provider"`
	Index       string `yaml:"index"`
	AppID       string `yaml:"app_id,omitempty"`
	APIKey      string `yaml:"api_key"`
	Host        string `yaml:"host,omitempty"`
}

SearchIndexDef represents a search index sync subscriber handler

func (*SearchIndexDef) Validate 

func (d *SearchIndexDef) Validate() error

type Settings 

type Settings struct {
	// RecordsDelimiter controls whether a "#-" line is written after each record in INGR output.
	// 0 = use project or app default (app default is 1 = enabled). 1 = enabled. -1 = disabled.
	RecordsDelimiter int `yaml:"records_delimiter,omitempty"`
	// ConflictResolution holds the database-level conflict-resolution
	// defaults (e.g. record-merge), overridable per collection.
	ConflictResolution *ConflictResolutionConfig `yaml:"conflict_resolution,omitempty"`
}

Settings holds database-level defaults that apply to all collections.

type Severity 

type Severity string

Severity is the level of a validation finding. 

const (
	SeverityError   Severity = "error"
	SeverityWarning Severity = "warning"
)

type SlackDef 

type SlackDef struct {
	HandlerBase `yaml:",inline"`
	WebhookURL  string `yaml:"webhook_url"`
}

SlackDef represents a Slack incoming webhook subscriber handler

func (*SlackDef) Validate 

func (d *SlackDef) Validate() error

type SubscriberDef 

type SubscriberDef struct {
	HandlerBase   `yaml:",inline"`
	For           *SubscriberFor    `yaml:"for"`
	Webhooks      []WebhookDef      `yaml:"webhooks,omitempty"`
	Emails        []EmailDef        `yaml:"emails,omitempty"`
	Telegrams     []TelegramDef     `yaml:"telegrams,omitempty"`
	WhatsApp      []WhatsAppDef     `yaml:"whatsapp,omitempty"`
	Slacks        []SlackDef        `yaml:"slacks,omitempty"`
	Discords      []DiscordDef      `yaml:"discords,omitempty"`
	GitHubActions []GitHubActionDef `yaml:"github_actions,omitempty"`
	GitLabCI      []GitLabCIDef     `yaml:"gitlab_ci,omitempty"`
	Ntfy          []NtfyDef         `yaml:"ntfy,omitempty"`
	SMS           []SMSDef          `yaml:"sms,omitempty"`
	SearchIndexes []SearchIndexDef  `yaml:"search_indexes,omitempty"`
	RSS           []RSSDef          `yaml:"rss,omitempty"`
}

SubscriberDef represents a single subscriber configuration block

func (*SubscriberDef) Validate 

func (s *SubscriberDef) Validate() error

type SubscriberFor 

type SubscriberFor struct {
	Paths  []string           `yaml:"paths,omitempty"`
	Events []TriggerEventType `yaml:"events,omitempty"`
}

SubscriberFor represents the selector for when a subscriber group should be triggered

func (*SubscriberFor) Validate 

func (s *SubscriberFor) Validate() error

type TelegramDef 

type TelegramDef struct {
	HandlerBase `yaml:",inline"`
	Token       string `yaml:"token"`
	ChatID      string `yaml:"chat_id"`
}

TelegramDef represents a Telegram subscriber handler

func (*TelegramDef) Validate 

func (d *TelegramDef) Validate() error

type TriggerDef 

type TriggerDef struct {
	On   []TriggerEventType        `yaml:"on"`
	Jobs map[string]*TriggerJobDef `yaml:"jobs"`
}

TriggerDef is the schema for a trigger workflow file (.collection/trigger_<name>.yaml). Modelled after GitHub Actions workflow syntax.

type TriggerEventType 

type TriggerEventType string

TriggerEventType is the lifecycle event that fires a trigger. 

const (
	TriggerEventCreated TriggerEventType = "created"
	TriggerEventUpdated TriggerEventType = "updated"
	TriggerEventDeleted TriggerEventType = "deleted"
)

type TriggerJobDef 

type TriggerJobDef struct {
	RunsOn string           `yaml:"runs-on"`
	Steps  []TriggerStepDef `yaml:"steps"`
}

TriggerJobDef is one job inside a trigger workflow. RunsOn must be "." (run in the same environment as ingitdb itself).

type TriggerStepDef 

type TriggerStepDef struct {
	Run string `yaml:"run"`
}

TriggerStepDef is a single shell step within a trigger job.

type ValidationError 

type ValidationError struct {
	Severity     Severity
	CollectionID string // e.g. "todo.tasks"
	FilePath     string // absolute path to offending file
	RecordKey    string // empty for list/map files where key is unknown
	FieldName    string // set for field-level errors
	Message      string
	Err          error // wrapped cause
}

ValidationError is one finding from data validation. All location fields (CollectionID…FieldName) are optional. Implements error.

func (ValidationError) Error 

func (v ValidationError) Error() string

Error implements the error interface.

type ValidationResult 

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

ValidationResult aggregates findings; mutex-protected for goroutine safety.

func (*ValidationResult) Append 

func (r *ValidationResult) Append(e ValidationError)

Append adds a finding to the result.

func (*ValidationResult) ErrorCount 

func (r *ValidationResult) ErrorCount() int

ErrorCount returns the total number of recorded findings.

func (*ValidationResult) Errors 

func (r *ValidationResult) Errors() []ValidationError

Errors returns a snapshot copy of all findings.

func (*ValidationResult) Filter 

func (r *ValidationResult) Filter(fn func(ValidationError) bool) []ValidationError

Filter returns findings that match the given predicate.

func (*ValidationResult) GetRecordCount 

func (r *ValidationResult) GetRecordCount(collectionID string) int

GetRecordCount returns the number of records validated for a collection.

func (*ValidationResult) GetRecordCounts 

func (r *ValidationResult) GetRecordCounts(collectionID string) (passed, total int)

GetRecordCounts returns the passed and total record counts for a collection.

func (*ValidationResult) HasErrors 

func (r *ValidationResult) HasErrors() bool

HasErrors reports whether any findings have been recorded.

func (*ValidationResult) SetRecordCount 

func (r *ValidationResult) SetRecordCount(collectionID string, count int)

SetRecordCount sets the number of records validated for a collection.

func (*ValidationResult) SetRecordCounts 

func (r *ValidationResult) SetRecordCounts(collectionID string, passed, total int)

SetRecordCounts sets the passed and total record counts for a collection.

type ViewDef 

type ViewDef struct {
	ID      string            `yaml:"-"`
	Titles  map[string]string `yaml:"titles,omitempty"`
	OrderBy string            `yaml:"order_by,omitempty"`

	// Formats TODO: Needs definition
	Formats []string `yaml:"formats,omitempty"`

	Columns []string `yaml:"columns,omitempty"`

	// How many records to include; 0 means all
	Top int `yaml:"top,omitempty"`

	// Where holds filtering condition
	Where string `yaml:"where,omitempty"`

	// Template path relative to the collection directory.
	/*
		Build in templates:
		  - md-table - renders a Markdown table
		  - md-list - renders a Markdown list
		  - JSON - renders JSON
		  - YAML - renders YAML
	*/
	Template string `yaml:"template,omitempty"`

	// Output file name relative to the collection directory.
	FileName string `yaml:"file_name,omitempty"`

	// RecordsVarName provides a custom Template variable name for the records slice. The default is "records".
	RecordsVarName string `yaml:"records_var_name,omitempty"`

	Format       string `yaml:"format,omitempty"`
	MaxBatchSize int    `yaml:"max_batch_size,omitempty"`
	IsDefault    bool   `yaml:"-" json:"-"`

	// IncludeHash controls whether a sha256 hash line is appended to INGR footer.
	IncludeHash bool `yaml:"include_hash,omitempty"`

	// RecordsDelimiter controls whether a "#-" line is written after each record in INGR output.
	// 0 = use project or app default (app default is 1 = enabled). 1 = enabled. -1 = disabled.
	RecordsDelimiter int `yaml:"records_delimiter,omitempty"`
}

func (*ViewDef) Validate 

func (v *ViewDef) Validate() error

Validate checks the view definition for consistency.

type WebhookDef 

type WebhookDef struct {
	HandlerBase `yaml:",inline"`
	URL         string            `yaml:"url"`
	Method      string            `yaml:"method,omitempty"`
	Headers     map[string]string `yaml:"headers,omitempty"`
}

WebhookDef represents a webhook subscriber handler

func (*WebhookDef) Validate 

func (d *WebhookDef) Validate() error

type WhatsAppDef 

type WhatsAppDef struct {
	HandlerBase `yaml:",inline"`
	From        string `yaml:"from"`
	To          string `yaml:"to"`
	AccountSID  string `yaml:"account_sid"`
	AuthToken   string `yaml:"auth_token"`
}

WhatsAppDef represents a WhatsApp Business API subscriber handler

func (*WhatsAppDef) Validate 

func (d *WhatsAppDef) Validate() error

Source Files

View all Source files

Directories

Path Synopsis
ingitdb module
internal
testutil
Package testutil provides shared test helpers for ingitdb-go tests.
 Package testutil provides shared test helpers for ingitdb-go tests.
Package markdown parses and serializes inGitDB Markdown record files — YAML frontmatter delimited by "---" lines followed by a body.
 Package markdown parses and serializes inGitDB Markdown record files — YAML frontmatter delimited by "---" lines followed by a body.
specscore: feature/cli/resolve/auto-resolve/record-merge
 specscore: feature/cli/resolve/auto-resolve/record-merge

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.