* fix(list): honor --status when combined with --ready (#5832) `bd list --status X --ready` silently dropped --status and returned the unfiltered ready set (typically open issues). BuildListFilter pinned StatusOpen whenever ReadyFlag was set, and ReadyFilterFromIssueFilter hard-coded the same pin, so a custom status or in_progress selector never reached GetReadyWork. --ready without --status still defaults to open. An explicit --status is now the ready-work status selector, matching how other list filters already compose with --ready. `--status all` remains the no-filter spelling and keeps the open default. * fix(list): status-aware --ready footer and honor-path contract docs (#5832) The query intersection from the prior commit is unchanged. Round-1 review asked the ReadyFlag/Statuses contract comments and the text footer to describe the honor path: default --ready is still open, an explicit --status is named instead of "excludes in_progress".
156 lines
6.5 KiB
Go
156 lines
6.5 KiB
Go
package workapi
|
|
|
|
import (
|
|
"fmt"
|
|
"strings"
|
|
|
|
"github.com/steveyegge/beads/internal/types"
|
|
"github.com/steveyegge/beads/internal/utils"
|
|
"github.com/steveyegge/beads/issueops"
|
|
)
|
|
|
|
// DefaultReadyLimit is the default number of rows a ready-work query returns
|
|
// when the caller does not ask for a specific limit. Every frontend registers
|
|
// its limit knob from this constant so the surfaces cannot drift apart; 0
|
|
// still means unlimited.
|
|
const DefaultReadyLimit = 100
|
|
|
|
// LimitOr resolves a request's optional limit: nil takes the surface's shared
|
|
// default, and an explicit 0 stays 0, which means unlimited at both storage
|
|
// seams. The pointer is what keeps "unset" and "explicitly unlimited"
|
|
// distinguishable — collapsing them would make one constant unable to serve
|
|
// both surfaces.
|
|
func LimitOr(limit *int, fallback int) int {
|
|
if limit == nil {
|
|
return fallback
|
|
}
|
|
return *limit
|
|
}
|
|
|
|
// BuildReadyFilter turns a ready request into the storage-level work filter.
|
|
// It is the single definition of what `bd ready` means: open issues only (an
|
|
// empty WorkFilter would default to open plus in_progress), issue-type alias
|
|
// expansion, label and exclude-type normalization, limit defaulting, and the
|
|
// assignee rule that lets Unassigned win over a stale Assignee.
|
|
//
|
|
// Labels, LabelsAny and ExcludeLabels are normalized here, so a caller can
|
|
// pass raw user input. A caller that has to decide something from the label
|
|
// sets - the CLI's directory-label default is the one such case - reads them
|
|
// back off the returned filter, where they are already normalized, rather than
|
|
// normalizing its own copy: a value it then puts on the filter itself is its
|
|
// own to shape, and running it through here would change it.
|
|
func BuildReadyFilter(in issueops.ReadyRequest) (types.WorkFilter, error) {
|
|
filter := types.WorkFilter{
|
|
// Open only, not in_progress - the same set `bd list --ready` shows
|
|
// when --status is omitted (an explicit --status is an intersection;
|
|
// GH#5832).
|
|
Status: types.StatusOpen,
|
|
Type: utils.NormalizeIssueType(in.IssueType),
|
|
Limit: LimitOr(in.Limit, DefaultReadyLimit),
|
|
// Carried for the callers that consume this filter as a VALUE and run
|
|
// their own query (cmd/bd's proxied `bd ready`), where the seam beneath
|
|
// them renders it. Both implementations of issueops.Reader take it back
|
|
// off and skip in the shared page epilogue; see BuildListFilter and
|
|
// FinishPageAt.
|
|
Offset: in.Offset,
|
|
Unassigned: in.Unassigned,
|
|
SortPolicy: types.SortPolicy(in.Sort),
|
|
Labels: utils.NormalizeLabels(in.Labels),
|
|
LabelsAny: utils.NormalizeLabels(in.LabelsAny),
|
|
ExcludeLabels: utils.NormalizeLabels(in.ExcludeLabels),
|
|
LabelPattern: in.LabelPattern,
|
|
LabelRegex: in.LabelRegex,
|
|
IncludeDeferred: in.IncludeDeferred,
|
|
IncludeEphemeral: in.IncludeEphemeral,
|
|
ExcludeTypes: normalizeExcludeTypes(in.ExcludeTypes),
|
|
HasMetadataKey: in.HasMetadataKey,
|
|
Lite: in.Brief,
|
|
}
|
|
|
|
if in.Priority != nil {
|
|
p := *in.Priority
|
|
filter.Priority = &p
|
|
}
|
|
if in.Assignee != "" && !in.Unassigned {
|
|
a := in.Assignee
|
|
filter.Assignee = &a
|
|
}
|
|
if in.ParentID != "" {
|
|
pid := in.ParentID
|
|
filter.ParentID = &pid
|
|
}
|
|
if in.MolType != nil {
|
|
filter.MolType = in.MolType
|
|
}
|
|
if len(in.MetadataFields) > 0 {
|
|
filter.MetadataFields = in.MetadataFields
|
|
}
|
|
|
|
if err := ValidateMetadataFilters(in.MetadataFields, in.HasMetadataKey); err != nil {
|
|
return filter, err
|
|
}
|
|
|
|
if !filter.SortPolicy.IsValid() {
|
|
// A deterministic request-validation failure, so it matches
|
|
// ErrValidation: every role whose filter vocabulary this builds —
|
|
// Reader.Ready, ReadyClaimer.ClaimNext and BatchCloser's ClaimNext —
|
|
// promises a caller can classify one with errors.Is rather than by
|
|
// reading prose. The wrap is %.0w rather than a "%w: " prefix because
|
|
// this text is what `bd ready --sort bogus` prints verbatim behind
|
|
// "Error: "; prefixing it would change user-visible copy to say
|
|
// something the reader already knows.
|
|
return filter, fmt.Errorf("invalid sort policy '%s'. Valid values: hybrid, priority, oldest%.0w", in.Sort, issueops.ErrValidation)
|
|
}
|
|
return filter, nil
|
|
}
|
|
|
|
// BuildReadyCountFilter turns a ready request into the storage-level filter a
|
|
// COUNT of the ready set runs against: BuildReadyFilter's filter with the page
|
|
// removed, and the single definition of what `bd ready`'s published total means.
|
|
//
|
|
// It refuses a request carrying a page. issueops.ReadyCounter.CountReady
|
|
// promises its answer equals len(Reader.Ready(r with Limit=0).Items), and a
|
|
// Limit would make that "how many of the first N" while an Offset would
|
|
// subtract the rows it skipped from the size of a set that still holds them.
|
|
//
|
|
// The zeroed limit is set on a LOCAL copy: a nil Limit means the shared ready
|
|
// default at BuildReadyFilter, so an unlimited count has to say so explicitly.
|
|
func BuildReadyCountFilter(in issueops.ReadyRequest) (types.WorkFilter, error) {
|
|
if in.Limit != nil {
|
|
return types.WorkFilter{}, fmt.Errorf("%w: a ready count does not take a limit", issueops.ErrValidation)
|
|
}
|
|
if in.Offset != 0 {
|
|
return types.WorkFilter{}, fmt.Errorf("%w: a ready count does not take an offset", issueops.ErrValidation)
|
|
}
|
|
unlimited := 0
|
|
counted := in
|
|
counted.Limit = &unlimited
|
|
// Brief is CARRIED, not refused and not cleared, which is the opposite of
|
|
// what ClaimNext does with it. A count reads no field of any row, so the
|
|
// projection cannot make the number wrong; and the count is not always
|
|
// cheap enough for that to be the end of it. The unit-of-work seam has no
|
|
// COUNT(*) over the ready predicate and sizes the set by running the
|
|
// unbounded page and taking its length (uow/ready_counter.go), so clearing
|
|
// the field here would hydrate every heavy column of the whole ready set to
|
|
// answer `bd ready --brief`, which is the cost the projection exists to
|
|
// avoid and larger than the page it was asked for. Carrying it also keeps
|
|
// the count filter what this builder says it is: the listing's filter with
|
|
// the PAGE removed, and nothing else removed.
|
|
return BuildReadyFilter(counted)
|
|
}
|
|
|
|
// normalizeExcludeTypes splits comma-separated exclusions and expands type
|
|
// aliases, so --exclude-type=mr,epic and --exclude-type mr --exclude-type epic
|
|
// mean the same thing.
|
|
func normalizeExcludeTypes(raw []string) []types.IssueType {
|
|
var out []types.IssueType
|
|
for _, group := range raw {
|
|
for _, t := range strings.Split(group, ",") {
|
|
t = strings.TrimSpace(t)
|
|
if t != "" {
|
|
out = append(out, types.IssueType(utils.NormalizeIssueType(t)))
|
|
}
|
|
}
|
|
}
|
|
return out
|
|
}
|