The hint tactic.

The hint tactic tries the kitchen sink: it runs every tactic registered via the register_hint tac command on the current goal, and reports which ones succeed.

Future work

It would be nice to run the tactics in parallel.

opaque Mathlib.Tactic.Hint.hintExtension : Lean.SimplePersistentEnvExtension (Lean.TSyntax `tactic) (List (Lean.TSyntax `tactic))

An environment extension for registering hint tactics.

def Mathlib.Tactic.Hint.addHint (stx : Lean.TSyntax `tactic) : Lean.CoreM Unit

Register a new hint tactic.

Equations

• Mathlib.Tactic.Hint.addHint stx = Lean.modifyEnv fun (env : Lean.Environment) => Lean.PersistentEnvExtension.addEntry Mathlib.Tactic.Hint.hintExtension env stx

def Mathlib.Tactic.Hint.getHints : Lean.CoreM (List (Lean.TSyntax `tactic))

Return the list of registered hint tactics.

Equations

• Mathlib.Tactic.Hint.getHints = do let __do_lift ← Lean.getEnv pure (Mathlib.Tactic.Hint.hintExtension.getState __do_lift)

def Mathlib.Tactic.Hint.registerHintStx : Lean.ParserDescr

Register a tactic for use with the hint tactic, e.g. register_hint simp_all.

Equations

• One or more equations did not get rendered due to their size.

def Mathlib.Tactic.Hint.suggestion (tac : Lean.TSyntax `tactic) (msgs : Lean.MessageLog := { hadErrors := false, unreported := { root := Lean.PersistentArrayNode.node (Array.mkEmpty Lean.PersistentArray.branching.toNat), tail := Array.mkEmpty Lean.PersistentArray.branching.toNat, size := 0, shift := Lean.PersistentArray.initShift, tailOff := 0 }, reportedKinds := ∅ }) : Lean.Elab.Tactic.TacticM Lean.Meta.Tactic.TryThis.Suggestion

Construct a suggestion for a tactic.

• Check the passed MessageLog for an info message beginning with "Try this: ".
• If found, use that as the suggestion.
• Otherwise use the provided syntax.
• Also, look for remaining goals and pretty print them after the suggestion.

Equations

• One or more equations did not get rendered due to their size.

def Mathlib.Tactic.Hint.withMessageLog (t : Lean.Elab.Tactic.TacticM Unit) : Lean.Elab.Tactic.TacticM Lean.MessageLog

Run a tactic, returning any new messages rather than adding them to the message log.

Equations

• One or more equations did not get rendered due to their size.

def Mathlib.Tactic.Hint.withoutInfoTrees (t : Lean.Elab.Tactic.TacticM Unit) : Lean.Elab.Tactic.TacticM Unit

Run a tactic, but revert any changes to info trees. We use this to inhibit the creation of widgets by subsidiary tactics.

Equations

• One or more equations did not get rendered due to their size.

def Mathlib.Tactic.Hint.hint (stx : Lean.Syntax) : Lean.Elab.Tactic.TacticM Unit

Run all tactics registered using register_hint. Print a "Try these:" suggestion for each of the successful tactics.

If one tactic succeeds and closes the goal, we don't look at subsequent tactics.

Equations

• One or more equations did not get rendered due to their size.

def Mathlib.Tactic.Hint.hintStx : Lean.ParserDescr

The hint tactic tries every tactic registered using register_hint tac, and reports any that succeed.

Equations

• Mathlib.Tactic.Hint.hintStx = Lean.ParserDescr.node `Mathlib.Tactic.Hint.hintStx 1024 (Lean.ParserDescr.nonReservedSymbol "hint" false)