diff --git a/ci/simple-tests/proof_stat.human.exp-out b/ci/simple-tests/proof_stat.human.exp-out index 901c8b382..472312ed8 100644 --- a/ci/simple-tests/proof_stat.human.exp-out +++ b/ci/simple-tests/proof_stat.human.exp-out @@ -1,10 +1,10 @@ Proof check results for proof_stat.v -4 opaque proofs recognized: 2 successful 2 FAILING +4 opaque proofs recognized: 1 successful 3 FAILING FAIL a1_equal_4 OK b_equal_6 FAIL b2_equal_6 - OK use_admit + FAIL use_admit diff --git a/ci/simple-tests/proof_stat.tap.exp-out b/ci/simple-tests/proof_stat.tap.exp-out index 1735d29f7..86bd533ba 100644 --- a/ci/simple-tests/proof_stat.tap.exp-out +++ b/ci/simple-tests/proof_stat.tap.exp-out @@ -3,6 +3,6 @@ TAP version 13 not ok 1 a1_equal_4 ok 2 b_equal_6 not ok 3 b2_equal_6 -ok 4 use_admit +not ok 4 use_admit diff --git a/ci/simple-tests/proof_stat.v b/ci/simple-tests/proof_stat.v index 415b18bc2..ffdccda66 100644 --- a/ci/simple-tests/proof_stat.v +++ b/ci/simple-tests/proof_stat.v @@ -22,7 +22,7 @@ Lemma b2_equal_6 : b = 2 * 3. (* FAIL *) Proof using. (* this proof should fail *) Qed. -Lemma use_admit : 0 = 1. +Lemma use_admit : 0 = 1. (* FAIL *) Proof using. (* this proof succeeds but should count as failing *) admit. Admitted. diff --git a/coq/coq-syntax.el b/coq/coq-syntax.el index e68bb8f88..fbb5398f9 100644 --- a/coq/coq-syntax.el +++ b/coq/coq-syntax.el @@ -1136,6 +1136,12 @@ different." :type 'string :group 'coq) +(defcustom coq-omit-cheating-regexp "Admitted" + "Value for `proof-omit-cheating-regexp'. +Very similar to `coq-omit-proof-admit-command', but without the dot." + :type 'regexp + :group 'coq) + ;; ----- keywords for font-lock. (defvar coq-keywords-kill-goal diff --git a/coq/coq.el b/coq/coq.el index 4d4b5a33e..fdafe2dd4 100644 --- a/coq/coq.el +++ b/coq/coq.el @@ -1964,7 +1964,8 @@ at `proof-assistant-settings-cmds' evaluation time.") proof-script-definition-end-regexp coq-definition-end-regexp proof-script-proof-admit-command coq-omit-proof-admit-command proof-script-cmd-prevents-proof-omission #'coq-cmd-prevents-proof-omission - proof-script-cmd-force-next-proof-kept coq-cmd-force-next-proof-kept) + proof-script-cmd-force-next-proof-kept coq-cmd-force-next-proof-kept + proof-omit-cheating-regexp coq-omit-cheating-regexp) ;; proof-check-report/proof-check-annotate config (setq diff --git a/doc/PG-adapting.texi b/doc/PG-adapting.texi index d63478a38..865a3083d 100644 --- a/doc/PG-adapting.texi +++ b/doc/PG-adapting.texi @@ -1289,9 +1289,12 @@ The other material is asserted in the usual way and opaque proofs it first tries to assert them in the usual way too. If this succeeds the proof is considered valid. Otherwise the proof is replaced with @code{proof-script-proof-admit-command} and the proof is -considered invalid. To associate theorem names with opaque proofs, the -function @code{proof-get-proof-info-fn} is called, which is identical -to @code{proof-tree-get-proof-info}, @xref{Proof Tree Elisp +considered invalid. In order to consider Admitted proofs as invalid +ones, proofs whose last command matches +@code{proof-omit-cheating-regexp} count as invalid. To associate +theorem names with opaque proofs, the function +@code{proof-get-proof-info-fn} is called, which is identical to +@code{proof-tree-get-proof-info}, @xref{Proof Tree Elisp configuration}. To enable proof status statistics, the omit-proofs feature must be @@ -1328,6 +1331,17 @@ This function takes a state as argument as returned by the proof assistant back to the same state. @end defvar +@c TEXI DOCSTRING MAGIC: proof-omit-cheating-regexp +@defvar proof-omit-cheating-regexp +Regular expression matching proof closing commands for incomplete proofs.@* +If set, this regular expression is applied to the last command of +opaque proofs. If it matches the proofs counts as invalid for the +proof-status statistics and annotation feature. For Coq this is +used to mark Admitted proofs as invalid. + +This option can be left at @samp{nil}. +@end defvar + @node Safe (state-preserving) commands @section Safe (state-preserving) commands diff --git a/doc/ProofGeneral.texi b/doc/ProofGeneral.texi index e5aaeb23d..acaed3617 100644 --- a/doc/ProofGeneral.texi +++ b/doc/ProofGeneral.texi @@ -2267,7 +2267,8 @@ in the current buffer, i.e., it generates an overview that shows which of the opaque proofs in the current buffer are currently valid and which are failing. When used interactively, the proof status is shown in the buffer @code{*proof-check-report*} (as long as -@code{proof-check-report-buffer} is not changed). +@code{proof-check-report-buffer} is not changed). Note that incomplete +proofs, i.e., Admitted proofs for Coq, count as invalid. The command @code{proof-check-annotate} (menu @code{Proof-General -> Annotate Failing Proofs}) modifies the current buffer and places @@ -2285,7 +2286,10 @@ only work for Coq. Generate an overview about valid and invalid proofs.@* This command completely processes the current buffer and generates an overview about all the opaque proofs in it and -whether their proof scripts are valid or invalid. +whether their proof scripts are valid or invalid. Note that +proofs closed with a cheating command (see +@samp{@code{proof-omit-cheating-regexp}}), i.e., Admitted for Coq, count as +invalid. This command makes sense for a development process where invalid proofs are permitted and vos compilation and the omit proofs @@ -5318,12 +5322,12 @@ The command @code{proof-check-report} (menu @code{Proof-General -> Check Opaque Proofs}) generates the proof status of all opaque proofs in the current buffer, i.e., it generates an overview that shows which of the opaque proofs in the current buffer are currently valid and -which are failing. This command is useful for a development process -where invalid proofs are permitted and vos compilation (@xref{Quick -and inconsistent compilation}) and the omit proofs feature -(@xref{Omitting proofs for speed}) are used to work at the most -interesting or challenging point instead of on the first invalid -proof. +which are failing, where Admitted proofs count as failing. This +command is useful for a development process where invalid proofs are +permitted and vos compilation (@xref{Quick and inconsistent +compilation}) and the omit proofs feature (@xref{Omitting proofs for +speed}) are used to work at the most interesting or challenging point +instead of on the first invalid proof. The command @code{proof-check-annotate} (menu @code{Proof-General -> Annotate Failing Proofs}) can then be used to consistently annotate diff --git a/generic/pg-user.el b/generic/pg-user.el index ed86fc062..07ff560fb 100644 --- a/generic/pg-user.el +++ b/generic/pg-user.el @@ -735,7 +735,9 @@ errors are silently replaced by `proof-script-proof-admit-command'. The result is a list of proof status results, one for each 'proof chunk in the same order. Each proof-status result is a list of 4 elements as follows. -- 1st: proof status as `t' or `nil'. +- 1st: proof status as `t' or `nil'. Proofs closed with a match + of `proof-omit-cheating-regexp', if defined, count as failing, + i.e., their status is `nil'. - 2nd: the name of the proof as reported by `proof-get-proof-info-fn'. - 3rd: Position of the start of the span containing the theorem @@ -762,7 +764,8 @@ proof-status result is a list of 4 elements as follows. (nth 2 last-item) (cons 'empty-action-list (nth 3 last-item)))) (vanillas-rev-updated (cons new-last-item (cdr vanillas-rev))) - error) + error + cheated) ;; if this is a proof chunk the next must be no-proof or must not exist (cl-assert (or (not (eq this-mode 'proof)) (or (eq next-mode 'no-proof) (eq next-mode nil))) @@ -813,8 +816,15 @@ proof-status result is a list of 4 elements as follows. (proof-set-locked-end last-span-end) (cl-assert (not (cadr (funcall proof-get-proof-info-fn))) nil "proof-check: still in proof after admitting")) + (when (and proof-omit-cheating-regexp + ;; check if somebody cheated - cheated proofs + ;; should count as failing + (proof-string-match + proof-omit-cheating-regexp + (mapconcat #'identity (nth 1 last-item) " "))) + (setq cheated t)) (push - (cons (not error) + (cons (not (or error cheated)) (cons (cadr current-proof-state-and-name) proof-start-points)) proof-results))) @@ -863,7 +873,10 @@ This function does not (re-)compile required files." "Generate an overview about valid and invalid proofs. This command completely processes the current buffer and generates an overview about all the opaque proofs in it and -whether their proof scripts are valid or invalid. +whether their proof scripts are valid or invalid. Note that +proofs closed with a cheating command (see +`proof-omit-cheating-regexp'), i.e., Admitted for Coq, count as +invalid. This command makes sense for a development process where invalid proofs are permitted and vos compilation and the omit proofs diff --git a/generic/proof-config.el b/generic/proof-config.el index 1c5ede807..aa79b85c4 100644 --- a/generic/proof-config.el +++ b/generic/proof-config.el @@ -738,6 +738,17 @@ without surrounding space." :type 'boolean :group 'proof-script) +(defcustom proof-omit-cheating-regexp nil + "Regular expression matching proof closing commands for incomplete proofs. +If set, this regular expression is applied to the last command of +opaque proofs. If it matches the proofs counts as invalid for the +proof-status statistics and annotation feature. For Coq this is +used to mark Admitted proofs as invalid. + +This option can be left at `nil'." + :type 'regexp + :group 'proof-script) + ;; proof-omit-proofs-option is in proof-useropts as user option (defcustom proof-script-proof-start-regexp nil