1E(1) User Commands E(1)
2
6 E - manual page for E 2.5-DEBUG Avongrove
7 (a8acce0b281f27282ba15f0a8541e54662223340)
8
10 eprover [options] [files]
11
13 E 2.5-DEBUG "Avongrove"
14 Read a set of first-order clauses and formulae and try to refute it.
16
18 -h
19 --help
21 Print a short description of program usage and options.
23 -V
25 --version
27 Print the version number of the prover. Please include this with
29 all bug reports (if any).
30 -v
32 --verbose[=<arg>]
34 Verbose comments on the progress of the program. This differs
36 from the output level (below) in that technical information is
37 printed to stderr, while the output level determines which logi‐
38 cal manipulations of the clauses are printed to stdout. The
39 short form or the long form without the optional argument is
40 equivalent to --verbose=1.
41 -o <arg>
43 --output-file=<arg>
45 Redirect output into the named file.
47 -s
49 --silent
51 Equivalent to --output-level=0.
53 -l <arg>
55 --output-level=<arg>
57 Select an output level, greater values imply more verbose out‐
59 put. Level 0 produces nearly no output, level 1 will output each
60 clause as it is processed, level 2 will output generating infer‐
61 ences, level 3 will give a full protocol including rewrite steps
62 and level 4 will include some internal clause renamings. Levels
63 >= 2 also imply PCL2 or TSTP formats (which can be post-pro‐
64 cessed with suitable tools).
65 -p
67 --proof-object[=<arg>]
69 Generate (and print, in case of success) an internal proof ob‐
71 ject. Level 0 will not print a proof object, level 1 will build
72 asimple, compact proof object that only contains inference rules
73 and dependencies, level 2 will build a proof object where infer‐
74 ences are unambiguously described by giving inference positions,
75 and level 3 will expand this to a proof object where all inter‐
76 mediate results are explicit. This feature is under development,
77 so far only level 0 and 1 are operational. By default The proof
78 object will be provided in TPTP-3 or LOP syntax, depending on
79 input format and explicit settings. The following option will
80 suppress normal output of the proof object in favour of a
81 graphial representation. The short form or the long form without
82 the optional argument is equivalent to --proof-object=1.
83 --proof-graph[=<arg>]
85 Generate (and print, in case of success) an internal proof ob‐
87 ject in the form of a GraphViz dot graph. The optional argument
88 can be 1 (nodes are labelled with just the name of the
89 clause/formula), 2 (nodes are labelled with the TPTP clause/for‐
90 mula) or 3 (nodes also labelled with source/inference record.
91 The option without the optional argument is equivalent to
92 --proof-graph=3.
93 -d
95 --full-deriv
97 Include all derived formuas/clauses in the proof graph/proof ob‐
99 ject, not just the ones contributing to the actual proof.
100 --force-deriv[=<arg>]
102 Force output of the derivation even in cases where the prover
104 terminates in an indeterminate state. By default, the derivia‐
105 tion of all processed clauses is included in the derivation ob‐
106 ject. With value 2, derivation of all clauses will be printed
107 The option without the optional argument is equivalent to
108 --force-deriv=1.
109 --record-gcs
111 Record given-clause selection as separate (pseudo-)inferences
113 and preserve the form of given clauses evaluated and selected
114 via archiving for analysis and possibly machine learning.
115 --training-examples[=<arg>]
117 Generate and process training examples from the proof search ob‐
119 ject. Implies --record-gcs. The argument is a binary or of the
120 desired processig. Bit zero prints positive exampels. Bit 1
121 prints negative examples. Additional selectors will be added
122 later. The option without the optional argument is equivalent to
123 --training-examples=1.
124 --pcl-terms-compressed
126 Print terms in the PCL output in shared representation.
128 --pcl-compact
130 Print PCL steps without additional spaces for formatting (safes
132 disk space for large protocols).
133 --pcl-shell-level[=<arg>]
135 Determines level to which clauses and formulas are suppressed in
137 the output. Level 0 will print all, level 1 will only print ini‐
138 tial clauses/formulas, level 2 will print no clauses or axioms.
139 All levels will still print the dependency graph. The option
140 without the optional argument is equivalent to
141 --pcl-shell-level=1.
142 --print-statistics
144 Print the inference statistics (only relevant for output level
146 0, otherwise they are printed automatically.
147 -0
149 --print-detailed-statistics
151 Print data about the proof state that is potentially expensive
153 to collect. Includes number of term cells and number of rewrite
154 steps.
155 -S
157 --print-saturated[=<arg>]
159 Print the (semi-) saturated clause sets after terminating the
161 saturation process. The argument given describes which parts
162 should be printed in which order. Legal characters are
163 'teigEIGaA', standing for type declarations, processed positive
164 units, processed negative units, processed non-units, unpro‐
165 cessed positive units, unprocessed negative units, unprocessed
166 non-units, and two types of additional equality axioms, respec‐
167 tively. Equality axioms will only be printed if the original
168 specification contained real equality. In this case, 'a' re‐
169 quests axioms in which a separate substitutivity axiom is given
170 for each argument position of a function or predicate symbol,
171 while 'A' requests a single substitutivity axiom (covering all
172 positions) for each symbol. The short form or the long form
173 without the optional argument is equivalent to --print-satu‐
174 rated=eigEIG.
175 --print-sat-info
177 Print additional information (clause number, weight, etc) as a
179 comment for clauses from the semi-saturated end system.
180 --filter-saturated[=<arg>]
182 Filter the
184 (semi-) saturated clause sets after terminating the
185 saturation process. The argument is a string describing which
187 operations to take (and in which order). Options are 'u' (remove
188 all clauses with more than one literal), 'c' (delete all but one
189 copy of identical clauses, 'n', 'r', 'f' (forward contraction,
190 unit-subsumption only, no rewriting, rewriting with rules only,
191 full rewriting, respectively), and 'N', 'R' and 'F' (as their
192 lower case counterparts, but with non-unit-subsumption enabled
193 as well). The option without the optional argument is equivalent
194 to --filter-saturated=Fc.
195 --prune
197 Stop after relevancy pruning, SInE pruning, and output of the
199 initial clause- and formula set. This will automatically set
200 output level to 4 so that the pruned problem specification is
201 printed. Note that the desired pruning methods must still be
202 specified (e.g. '--sine=Auto').
203 --cnf
205 Convert the input problem into clause normal form and print it.
207 This is (nearly) equivalent to '--print-saturated=eigEIG --pro‐
208 cessed-clauses-limit=0' and will by default perform some usually
209 useful simplifications. You can additionally specify e.g.
210 '--no-preprocessing' if you want just the result of CNF transla‐
211 tion.
212 --print-pid
214 Print the process id of the prover as a comment after option
216 processing.
217 --print-version
219 Print the version number of the prover as a comment after option
221 processing. Note that unlike -version, the prover will not ter‐
222 minate, but proceed normally.
223 --error-on-empty
225 Return with an error code if the input file contains no clauses.
227 Formally, the empty clause set (as an empty conjunction of
228 clauses) is trivially satisfiable, and E will treat any empty
229 input set as satisfiable. However, in composite systems this is
230 more often a sign that something went wrong. Use this option to
231 catch such bugs.
232 -m <arg>
234 --memory-limit=<arg>
236 Limit the memory the prover may use. The argument is the allowed
238 amount of memory in MB. If you use the argument 'Auto', the sys‐
239 tem will try to figure out the amount of physical memory of your
240 machine and claim most of it. This option may not work every‐
241 where, due to broken and/or strange behaviour of setrlimit() in
242 some UNIX implementations, and due to the fact that I know of no
243 portable way to figure out the physical memory in a machine.
244 Both the option and the 'Auto' version do work under all tested
245 versions of Solaris and GNU/Linux. Due to problems with limit
246 data types, it is currently impossible to set a limit of more
247 than 2 GB (2048 MB).
248 --cpu-limit[=<arg>]
250 Limit the cpu time the prover should run. The optional argument
252 is the CPU time in seconds. The prover will terminate immedi‐
253 ately after reaching the time limit, regardless of internal
254 state. This option may not work everywhere, due to broken and/or
255 strange behaviour of setrlimit() in some UNIX implementations.
256 It does work under all tested versions of Solaris, HP-UX,
257 MacOS-X, and GNU/Linux. As a side effect, this option will in‐
258 hibit core file writing. Please note that if you use both
259 --cpu-limit and --soft-cpu-limit, the soft limit has to be
260 smaller than the hard limit to have any effect. The option
261 without the optional argument is equivalent to --cpu-limit=300.
262 --soft-cpu-limit[=<arg>]
264 Limit the cpu time the prover should spend in the main satura‐
266 tion phase. The prover will then terminate gracefully, i.e. it
267 will perform post-processing, filtering and printing of unpro‐
268 cessed clauses, if these options are selected. Note that for
269 some filtering options (in particular those which perform full
270 subsumption), the post-processing time may well be larger than
271 the saturation time. This option is particularly useful if you
272 want to use E as a preprocessor or lemma generator in a larger
273 system. The option without the optional argument is equivalent
274 to --soft-cpu-limit=290.
275 -R
277 --resources-info
279 Give some information about the resources used by the prover.
281 You will usually get CPU time information. On systems returning
282 more information with the rusage() system call, you will also
283 get information about memory consumption.
284 -C <arg>
286 --processed-clauses-limit=<arg>
288 Set the maximal number of clauses to process (i.e. the number of
290 traversals of the main-loop).
291 -P <arg>
293 --processed-set-limit=<arg>
295 Set the maximal size of the set of processed clauses. This dif‐
297 fers from the previous option in that redundant and back-simpli‐
298 fied processed clauses are not counted.
299 -U <arg>
301 --unprocessed-limit=<arg>
303 Set the maximal size of the set of unprocessed clauses. This is
305 a termination condition, not something to use to control the
306 deletion of bad clauses. Compare --delete-bad-limit.
307 -T <arg>
309 --total-clause-set-limit=<arg>
311 Set the maximal size of the set of all clauses. See previous op‐
313 tion.
314 --generated-limit=<arg>
316 Set the maximal number of generated clauses before the proof
318 search stops. This is a reasonable (though not great) estimate
319 of the work done.
320 --tb-insert-limit=<arg>
322 Set the maximal number of of term bank term top insertions. This
324 is a reasonable (though not great) estimate of the work done.
325 --answers[=<arg>]
327 Set the maximal number of answers to print for existentially
329 quantified questions. Without this option, the prover terminates
330 after the first answer found. If the value is different from 1,
331 the prover is no longer guaranteed to terminate, even if there
332 is a finite number of answers. The option without the optional
333 argument is equivalent to --answers=2147483647.
334 --conjectures-are-questions
336 Treat all conjectures as questions to be answered. This is a
338 wart necessary because CASC-J6 has categories requiring answers,
339 but does not yet support the 'question' type for formulas.
340 -n
342 --eqn-no-infix
344 In LOP, print equations in prefix notation equal(x,y).
346 -e
348 --full-equational-rep
350 In LOP. print all literals as equations, even non-equational
352 ones.
353 --lop-in
355 Set E-LOP as the input format. If no input format is selected by
357 this or one of the following options, E will guess the input
358 format based on the first token. It will almost always correctly
359 recognize TPTP-3, but it may misidentify E-LOP files that use
360 TPTP meta-identifiers as logical symbols.
361 --pcl-out
363 Set PCL as the proof object output format.
365 --tptp-in
367 Set TPTP-2 as the input format (but note that includes are still
369 handled according to TPTP-3 semantics).
370 --tptp-out
372 Print TPTP format instead of E-LOP. Implies --eqn-no-infix and
374 will ignore --full-equational-rep.
375 --tptp-format
377 Equivalent to --tptp-in and --tptp-out.
379 --tptp2-in
381 Synonymous with --tptp-in.
383 --tptp2-out
385 Synonymous with --tptp-out.
387 --tptp2-format
389 Synonymous with --tptp-format.
391 --tstp-in
393 Set TPTP-3 as the input format (Note that TPTP-3 syntax is still
395 under development, and the version in E may not be fully con‐
396 forming at all times. E works on all TPTP 6.3.0 FOF and CNF
397 files (including includes).
398 --tstp-out
400 Print output clauses in TPTP-3 syntax. In particular, for output
402 levels >=2, write derivations as TPTP-3 derivations.
403 --tstp-format
405 Equivalent to --tstp-in and --tstp-out.
407 --tptp3-in
409 Synonymous with --tstp-in.
411 --tptp3-out
413 Synonymous with --tstp-out.
415 --tptp3-format
417 Synonymous with --tstp-format.
419 --auto
421 Automatically determine settings for proof search. This is
423 equivalent to -xAuto -tAuto --sine=Auto.
424 --satauto
426 Automatically determine settings for proof/saturation search.
428 This is equivalent to -xAuto -tAuto.
429 --autodev
431 Automatically determine settings for proof search (development
433 version). This is equivalent to -xAutoDev -tAutoDev
434 --sine=Auto.
435 --satautodev
437 Automatically determine settings for proof/saturation search
439 (development version). This is equivalent to -xAutoDev -tAu‐
440 toDev.
441 --auto-schedule
443 Use the (experimental) strategy scheduling. This will try sev‐
445 eral different fully specified search strategies (aka
446 "Auto-Modes"), one after the other, until a proof or saturation
447 is found, or the time limit is exceeded.
448 --satauto-schedule
450 Use the (experimental) strategy scheduling without SInE, thus
452 maintaining completeness.
453 --no-preprocessing
455 Do not perform preprocessing on the initial clause set. Prepro‐
457 cessing currently removes tautologies and orders terms, literals
458 and clauses in a certain ("canonical") way before anything else
459 happens. Unless limited by one of the following options, it will
460 also unfold equational definitions.
461 --eq-unfold-limit=<arg>
463 During preprocessing, limit unfolding (and removing) of equa‐
465 tional definitions to those where the expanded definition is at
466 most the given limit bigger (in terms of standard weight) than
467 the defined term.
468 --eq-unfold-maxclauses=<arg>
470 During preprocessing, don't try unfolding of equational defini‐
472 tions if the problem has more than this limit of clauses.
473 --no-eq-unfolding
475 During preprocessing, abstain from unfolding (and removing)
477 equational definitions.
478 --sine[=<arg>]
480 Apply SInE to prune the unprocessed axioms with the specified
482 filter. 'Auto' will automatically pick a filter. The option
483 without the optional argument is equivalent to --sine=Auto.
484 --rel-pruning-level[=<arg>]
486 Perform relevancy pruning up to the given level on the unpro‐
488 cessed axioms. The option without the optional argument is
489 equivalent to --rel-pruning-level=3.
490 --presat-simplify
492 Before proper saturation do a complete interreduction of the
494 proof state.
495 --ac-handling[=<arg>]
497 Select AC handling mode, i.e. determine what to do with redun‐
499 dant AC tautologies. The default is equivalent to 'DiscardAll',
500 the other possible values are 'None' (to disable AC handling),
501 'KeepUnits', and 'KeepOrientable'. The option without the op‐
502 tional argument is equivalent to --ac-handling=KeepUnits.
503 --ac-non-aggressive
505 Do AC resolution on negative literals only on processing (by de‐
507 fault, AC resolution is done after clause creation). Only effec‐
508 tive if AC handling is not disabled.
509 -W <arg>
511 --literal-selection-strategy=<arg>
513 Choose a strategy for selection of negative literals. There are
515 two special values for this option: NoSelection will select no
516 literal (i.e. perform normal superposition) and NoGeneration
517 will inhibit all generating inferences. For a list of the other
518 (hopefully self-documenting) values run 'eprover -W none'. There
519 are two variants of each strategy. The one prefixed with 'P'
520 will allow paramodulation into maximal positive literals in ad‐
521 dition to paramodulation into maximal selected negative liter‐
522 als.
523 --no-generation
525 Don't perform any generating inferences (equivalent to --lit‐
527 eral-selection-strategy=NoGeneration).
528 --select-on-processing-only
530 Perform literal selection at processing time only (i.e. select
532 only in the _given clause_), not before clause evaluation. This
533 is relevant because many clause selection heuristics give spe‐
534 cial consideration to maximal or selected literals.
535 -i
537 --inherit-paramod-literals
539 Always select the negative literals a previous inference
541 paramodulated into (if possible). If no such literal exists, se‐
542 lect as dictated by the selection strategy.
543 -j
545 --inherit-goal-pm-literals
547 In a goal (all negative clause), always select the negative lit‐
549 erals a previous inference paramodulated into (if possible). If
550 no such literal exists, select as dictated by the selection
551 strategy.
552 --inherit-conjecture-pm-literals
554 In a conjecture-derived clause, always select the negative lit‐
556 erals a previous inference paramodulated into (if possible). If
557 no such literal exists, select as dictated by the selection
558 strategy.
559 --selection-pos-min=<arg>
561 Set a lower limit for the number of positive literals a clause
563 must have to be eligible for literal selection.
564 --selection-pos-max=<arg>
566 Set a upper limit for the number of positive literals a clause
568 can have to be eligible for literal selection.
569 --selection-neg-min=<arg>
571 Set a lower limit for the number of negative literals a clause
573 must have to be eligible for literal selection.
574 --selection-neg-max=<arg>
576 Set a upper limit for the number of negative literals a clause
578 can have to be eligible for literal selection.
579 --selection-all-min=<arg>
581 Set a lower limit for the number of literals a clause must have
583 to be eligible for literal selection.
584 --selection-all-max=<arg>
586 Set an upper limit for the number of literals a clause must have
588 to be eligible for literal selection.
589 --selection-weight-min=<arg>
591 Set the minimum weight a clause must have to be eligible for
593 literal selection.
594 --prefer-initial-clauses
596 Always process all initial clauses first.
598 -x <arg>
600 --expert-heuristic=<arg>
602 Select one of the clause selection heuristics. Currently at
604 least available: Auto, Weight, StandardWeight, RWeight, FIFO,
605 LIFO, Uniq, UseWatchlist. For a full list check HEURIS‐
606 TICS/che_proofcontrol.c. Auto is recommended if you only want to
607 find a proof. It is special in that it will also set some addi‐
608 tional options. To have optimal performance, you also should
609 specify -tAuto to select a good term ordering. LIFO is unfair
610 and will make the prover incomplete. Uniq is used internally and
611 is not very useful in most cases. You can define more heuristics
612 using the option -H (see below).
613 --filter-orphans-limit[=<arg>]
615 Orphans are unprocessed clauses where one of the parents has
617 been removed by back-simolification. They are redundant and usu‐
618 ally removed lazily (i.e. only when they are selected for pro‐
619 cessing). With this option you can select a limit on back-sim‐
620 plified clauses after which orphans will be eagerly deleted.
621 The option without the optional argument is equivalent to --fil‐
622 ter-orphans-limit=100.
623 --forward-contract-limit[=<arg>]
625 Set a limit on the number of processed clauses after which the
627 unprocessed clause set will be re-simplified and reweighted.
628 The option without the optional argument is equivalent to --for‐
629 ward-contract-limit=80000.
630 --delete-bad-limit[=<arg>]
632 Set the number of storage units after which bad clauses are
634 deleted without further consideration. This causes the prover to
635 be potentially incomplete, but will allow you to limit the maxi‐
636 mum amount of memory used fairly well. The prover will tell you
637 if a proof attempt failed due to the incompleteness introduced
638 by this option. It is recommended to set this limit signifi‐
639 cantly higher than --filter-limit or --filter-copies-limit. If
640 you select -xAuto and set a memory limit, the prover will deter‐
641 mine a good value automatically. The option without the optional
642 argument is equivalent to --delete-bad-limit=1500000.
643 --assume-completeness
645 There are various way (e.g. the next few options) to configure
647 the prover to be strongly incomplete in the general case. E will
648 detect when such an option is selected and return corresponding
649 exit states (i.e. it will not claim satisfiability just because
650 it ran out of unprocessed clauses). If you _know_ that for your
651 class of problems the selected strategy is still complete, use
652 this option to tell the system that this is the case.
653 --assume-incompleteness
655 This option instructs the prover to assume incompleteness (typi‐
657 cally because the axiomatization already is incomplete because
658 axioms have been filtered before they are handed to the system.
659 --disable-eq-factoring
661 Disable equality factoring. This makes the prover incomplete for
663 general non-Horn problems, but helps for some specialized
664 classes. It is not necessary to disable equality factoring for
665 Horn problems, as Horn clauses are not factored anyways.
666 --disable-paramod-into-neg-units
668 Disable paramodulation into negative unit clause. This makes the
670 prover incomplete in the general case, but helps for some spe‐
671 cialized classes.
672 --condense
674 Enable condensing for the given clause. Condensing replaces a
676 clause by a more general factor (if such a factor exists).
677 --condense-aggressive
679 Enable condensing for the given and newly generated clauses.
681 --disable-given-clause-fw-contraction
683 Disable simplification and subsumption of the newly selected
685 given clause (clauses are still simplified when they are gener‐
686 ated). In general, this breaks some basic assumptions of the
687 DISCOUNT loop proof search procedure. However, there are some
688 problem classes in which this simplifications empirically never
689 occurs. In such cases, we can save significant overhead. The op‐
690 tion _should_ work in all cases, but is not expected to improve
691 things in most cases.
692 --simul-paramod
694 Use simultaneous paramodulation to implement superposition. De‐
696 fault is to use plain paramodulation.
697 --oriented-simul-paramod
699 Use simultaneous paramodulation for oriented from-literals. This
701 is an experimental feature.
702 --supersimul-paramod
704 Use supersimultaneous paramodulation to implement superposition.
706 Default is to use plain paramodulation.
707 --oriented-supersimul-paramod
709 Use supersimultaneous paramodulation for oriented from-literals.
711 This is an experimental feature.
712 --split-clauses[=<arg>]
714 Determine which clauses should be subject to splitting. The ar‐
716 gument is the binary 'OR' of values for the desired classes:
717 1: Horn clauses
719 2: Non-Horn clauses
721 4: Negative clauses
723 8: Positive clauses
725 16: Clauses with both positive and negative literals
727 Each set bit adds that class to the set of clauses which will be
729 split. The option without the optional argument is equivalent
730 to --split-clauses=7.
731 --split-method=<arg>
733 Determine how to treat ground literals in splitting. The argu‐
735 ment is either '0' to denote no splitting of ground literals
736 (they are all assigned to the first split clause produced), '1'
737 to denote that all ground literals should form a single new
738 clause, or '2', in which case ground literals are treated as
739 usual and are all split off into individual clauses.
740 --split-aggressive
742 Apply splitting to new clauses (after simplification) and before
744 evaluation. By default, splitting (if activated) is only per‐
745 formed on selected clauses.
746 --split-reuse-defs
748 If possible, reuse previous definitions for splitting.
750 -t <arg>
752 --term-ordering=<arg>
754 Select an ordering type (currently Auto, LPO, LPO4, KBO or
756 KBO6). -tAuto is suggested, in particular with -xAuto. KBO and
757 KBO6 are different implementations of the same ordering, KBO6 is
758 usually faster and has had more testing. Similarly, LPO4 is a
759 new, equivalent but superior implementation of LPO.
760 -w <arg>
762 --order-weight-generation=<arg>
764 Select a method for the generation of weights for use with the
766 term ordering. Run 'eprover -w none' for a list of options.
767 --order-weights=<arg>
769 Describe a (partial) assignments of weights to function symbols
771 for term orderings (in particular, KBO). You can specify a list
772 of weights of the form 'f1:w1,f2:w2, ...'. Since a total weight
773 assignment is needed, E will _first_ apply any weight generation
774 scheme specified (or the default one), and then modify the
775 weights as specified. Note that E performs only very basic san‐
776 ity checks, so you probably can specify weights that break KBO
777 constraints.
778 -G <arg>
780 --order-precedence-generation=<arg>
782 Select a method for the generation of a precedence for use with
784 the term ordering. Run 'eprover -G none' for a list of options.
785 --prec-pure-conj[=<arg>]
787 Set a weight for symbols that occur in conjectures only to de‐
789 terminewhere to place it in the precedence. This value is used
790 for a roughpre-order, the normal schemes only sort within sym‐
791 bols with the sameoccurance modifier. The option without the op‐
792 tional argument is equivalent to --prec-pure-conj=10.
793 --prec-conj-axiom[=<arg>]
795 Set a weight for symbols that occur in both conjectures and ax‐
797 iomsto determine where to place it in the precedence. This value
798 is used for a rough pre-order, the normal schemes only sort
799 within symbols with the same occurance modifier. The option
800 without the optional argument is equivalent to --prec-conj-ax‐
801 iom=5.
802 --prec-pure-axiom[=<arg>]
804 Set a weight for symbols that occur in axioms only to determine
806 where to place it in the precedence. This value is used for a
807 rough pre-order, the normal schemes only sort within symbols
808 with the same occurance modifier. The option without the op‐
809 tional argument is equivalent to --prec-pure-axiom=2.
810 --prec-skolem[=<arg>]
812 Set a weight for Skolem symbols to determine where to place it
814 in the precedence. This value is used for a rough pre-order, the
815 normal schemes only sort within symbols with the same occurance
816 modifier. The option without the optional argument is equivalent
817 to --prec-skolem=2.
818 --prec-defpred[=<arg>]
820 Set a weight for introduced predicate symbols (usually via defi‐
822 nitional CNF or clause splitting) to determine where to place it
823 in the precedence. This value is used for a rough pre-order, the
824 normal schemes only sort within symbols with the same occurance
825 modifier. The option without the optional argument is equivalent
826 to --prec-defpred=2.
827 -c <arg>
829 --order-constant-weight=<arg>
831 Set a special weight > 0 for constants in the term ordering. By
833 default, constants are treated like other function symbols.
834 --precedence[=<arg>]
836 Describe a (partial) precedence for the term ordering used for
838 the proof attempt. You can specify a comma-separated list of
839 precedence chains, where a precedence chain is a list of func‐
840 tion symbols (which all have to appear in the proof problem),
841 connected by >, <, or =. If this option is used in connection
842 with --order-precedence-generation, the partial ordering will be
843 completed using the selected method, otherwise the prover runs
844 with a non-ground-total ordering. The option without the op‐
845 tional argument is equivalent to --precedence=.
846 --lpo-recursion-limit[=<arg>]
848 Set a depth limit for LPO comparisons. Most comparisons do not
850 need more than 10 or 20 levels of recursion. By default, recur‐
851 sion depth is limited to 1000 to avoid stack overflow problems.
852 If the limit is reached, the prover assumes that the terms are
853 uncomparable. Smaller values make the comparison attempts
854 faster, but less exact. Larger values have the opposite effect.
855 Values up to 20000 should be save on most operating systems. If
856 you run into segmentation faults while using LPO or LPO4, first
857 try to set this limit to a reasonable value. If the problem per‐
858 sists, send a bug report ;-) The option without the optional ar‐
859 gument is equivalent to --lpo-recursion-limit=100.
860 --restrict-literal-comparisons
862 Make all literals uncomparable in the term ordering (i.e. do not
864 use the term ordering to restrict paramodulation, equality reso‐
865 lution and factoring to certain literals. This is necessary to
866 make Set-of-Support-strategies complete for the non-equational
867 case (It still is incomplete for the equational case, but pretty
868 useless anyways).
869 --literal-comparison=<arg>
871 Modify how literal comparisons are done. 'None' is equivalent to
873 the previous option, 'Normal' uses the normal lifting of the
874 term ordering, 'TFOEqMax' uses the equivalent of a transfinite
875 ordering deciding on the predicate symbol and making equational
876 literals maximal, and 'TFOEqMin' modifies this by making equa‐
877 tional symbols minimal.
878 --sos-uses-input-types
880 If input is TPTP format, use TPTP conjectures for initializing
882 the Set of Support. If not in TPTP format, use E-LOP queries
883 (clauses of the form ?-l(X),...,m(Y)). Normally, all negative
884 clauses are used. Please note that most E heuristics do not use
885 this information at all, it is currently only useful for certain
886 parameter settings (including the SimulateSOS priority func‐
887 tion).
888 --destructive-er
890 Allow destructive equality resolution inferences on pure-vari‐
892 able literals of the form X!=Y, i.e. replace the original clause
893 with the result of an equality resolution inference on this lit‐
894 eral.
895 --strong-destructive-er
897 Allow destructive equality resolution inferences on literals of
899 the form X!=t (where X does not occur in t), i.e. replace the
900 original clause with the result of an equality resolution infer‐
901 ence on this literal. Unless I am brain-dead, this maintains
902 completeness, although the proof is rather tricky.
903 --destructive-er-aggressive
905 Apply destructive equality resolution to all newly generated
907 clauses, not just to selected clauses. Implies --destructive-er.
908 --forward-context-sr
910 Apply contextual simplify-reflect with processed clauses to the
912 given clause.
913 --forward-context-sr-aggressive
915 Apply contextual simplify-reflect with processed clauses to new
917 clauses. Implies --forward-context-sr.
918 --backward-context-sr
920 Apply contextual simplify-reflect with the given clause to pro‐
922 cessed clauses.
923 -g
925 --prefer-general-demodulators
927 Prefer general demodulators. By default, E prefers specialized
929 demodulators. This affects in which order the rewrite index is
930 traversed.
931 -F <arg>
933 --forward-demod-level=<arg>
935 Set the desired level for rewriting of unprocessed clauses. A
937 value of 0 means no rewriting, 1 indicates to use rules (ori‐
938 entable equations) only, 2 indicates full rewriting with rules
939 and instances of unorientable equations. Default behavior is 2.
940 --strong-rw-inst
942 Instantiate unbound variables in matching potential demodulators
944 with a small constant terms.
945 -u
947 --strong-forward-subsumption
949 Try multiple positions and unit-equations to try to equationally
951 subsume a single new clause. Default is to search for a single
952 position.
953 --satcheck-proc-interval[=<arg>]
955 Enable periodic SAT checking at the given interval of main loop
957 non-trivial processed clauses. The option without the optional
958 argument is equivalent to --satcheck-proc-interval=5000.
959 --satcheck-gen-interval[=<arg>]
961 Enable periodic SAT checking whenever the total proof state size
963 increases by the given limit. The option without the optional
964 argument is equivalent to --satcheck-gen-interval=10000.
965 --satcheck-ttinsert-interval[=<arg>]
967 Enable periodic SAT checking whenever the number of term tops
969 insertions matches the given limit (which grows exponentially).
970 The option without the optional argument is equivalent to
971 --satcheck-ttinsert-interval=5000000.
972 --satcheck[=<arg>]
974 Set the grounding strategy for periodic SAT checking. Note that
976 to enable SAT checking, it is also necessary to set the interval
977 with one of the previous two options. The option without the op‐
978 tional argument is equivalent to --satcheck=FirstConst.
979 --satcheck-decision-limit[=<arg>]
981 Set the number of decisions allowed for each run of the SAT
983 solver. If the option is not given, the built-in value is 10000.
984 Use -1 to allow unlimited decision. The option without the op‐
985 tional argument is equivalent to --satcheck-decision-limit=100.
986 --satcheck-normalize-const
988 Use the current normal form (as recorded in the termbank rewrite
990 cache) of the selected constant as the term for the grounding
991 substitution.
992 --satcheck-normalize-unproc
994 Enable re-simplification (heuristic re-revaluation) of unpro‐
996 cessed clauses before grounding for SAT checking.
997 --watchlist[=<arg>]
999 Give the name for a file containing clauses to be watched for
1001 during the saturation process. If a clause is generated that
1002 subsumes a watchlist clause, the subsumed clause is removed from
1003 the watchlist. The prover will terminate when the watchlist is
1004 empty. If you want to use the watchlist for guiding the proof,
1005 put the empty clause onto the list and use the built-in clause
1006 selection heuristic 'UseWatchlist' (or build a heuristic your‐
1007 self using the priority functions 'PreferWatchlist' and 'Defer‐
1008 Watchlist'). Use the argument 'Use inline watchlist type' (or no
1009 argument) and the special clause type 'watchlist' if you want to
1010 put watchlist clauses into the normal input stream. This is only
1011 supported for TPTP input formats. The option without the op‐
1012 tional argument is equivalent to --watchlist='Use inline watch‐
1013 list type'.
1014 --static-watchlist[=<arg>]
1016 This is identical to the previous option, but subsumed clauses
1018 willnot be removed from the watchlist (and hence the prover will
1019 not terminate if all watchlist clauses have been subsumed. This
1020 may be more useful for heuristic guidance. The option without
1021 the optional argument is equivalent to --static-watchlist='Use
1022 inline watchlist type'.
1023 --no-watchlist-simplification
1025 By default, the watchlist is brought into normal form with re‐
1027 spect to the current processed clause set and certain simplifi‐
1028 cations. This option disables simplification for the watchlist.
1029 --conventional-subsumption
1031 Equivalent to --subsumption-indexing=None.
1033 --subsumption-indexing=<arg>
1035 Determine choice of indexing for (most) subsumption operations.
1037 Choices are 'None' for naive subsumption, 'Direct' for direct
1038 mapped FV-Indexing, 'Perm' for permuted FV-Indexing and 'Per‐
1039 mOpt' for permuted FV-Indexing with deletion of (suspected)
1040 non-informative features. Default behaviour is 'Perm'.
1041 --fvindex-featuretypes=<arg>
1043 Select the feature types used for indexing. Choices are "None"
1045 to disable FV-indexing, "AC" for AC compatible features (the de‐
1046 fault) (literal number and symbol counts), "SS" for set subsump‐
1047 tion compatible features (symbol depth), and "All" for all fea‐
1048 tures.Unless you want to measure the effects of the different
1049 features, I suggest you stick with the default.
1050 --fvindex-maxfeatures[=<arg>]
1052 Set the maximum initial number of symbols for feature computa‐
1054 tion. Depending on the feature selection, a value of X here
1055 will convert into 2X+2 features (for set subsumption features),
1056 2X+4 features (for AC-compatible features) or 4X+6 features (if
1057 all features are used, the default). Note that the actually used
1058 set of features may be smaller than this if the signature does
1059 not contain enough symbols.For the Perm and PermOpt version,
1060 this is _also_ used to set the maximum depth of the feature vec‐
1061 tor index. Yes, I should probably make this into two separate
1062 options. If you select a small value here, you should probably
1063 not use "Direct" for the --subsumption-indexing option. The op‐
1064 tion without the optional argument is equivalent to --fvin‐
1065 dex-maxfeatures=200.
1066 --fvindex-slack[=<arg>]
1068 Set the number of slots reserved in the index for function sym‐
1070 bols that may be introduced into the signature later, e.g. by
1071 splitting. If no new symbols are introduced, this just wastes
1072 time and memory. If PermOpt is chosen, the slackness slots will
1073 be deleted from the index anyways, but will still waste (a lit‐
1074 tle) time in computing feature vectors. The option without the
1075 optional argument is equivalent to --fvindex-slack=0.
1076 --rw-bw-index[=<arg>]
1078 Select fingerprint function for backwards rewrite index. "NoIn‐
1080 dex" will disable paramodulation indexing. For a list of the
1081 other values run 'eprover --pm-index=none'. FPX functions will
1082 use a fingerprint of X positions, the letters disambiguate be‐
1083 tween different fingerprints with the same sample size. The op‐
1084 tion without the optional argument is equivalent to --rw-bw-in‐
1085 dex=FP7.
1086 --pm-from-index[=<arg>]
1088 Select fingerprint function for the index for paramodulation
1090 from indexed clauses. "NoIndex" will disable paramodulation in‐
1091 dexing. For a list of the other values run 'eprover --pm-in‐
1092 dex=none'. FPX functionswill use a fingerprint of X positions,
1093 the letters disambiguate between different fingerprints with the
1094 same sample size. The option without the optional argument is
1095 equivalent to --pm-from-index=FP7.
1096 --pm-into-index[=<arg>]
1098 Select fingerprint function for the index for paramodulation
1100 into the indexed clauses. "NoIndex" will disable paramodulation
1101 indexing. For a list of the other values run 'eprover --pm-in‐
1102 dex=none'. FPX functionswill use a fingerprint of X positions,
1103 the letters disambiguate between different fingerprints with the
1104 same sample size. The option without the optional argument is
1105 equivalent to --pm-into-index=FP7.
1106 --fp-index[=<arg>]
1108 Select fingerprint function for all fingerprint indices. See
1110 above. The option without the optional argument is equivalent to
1111 --fp-index=FP7.
1112 --fp-no-size-constr
1114 Disable usage of size constraints for matching with fingerprint
1116 indexing.
1117 --pdt-no-size-constr
1119 Disable usage of size constraints for matching with perfect dis‐
1121 crimination trees indexing.
1122 --pdt-no-age-constr
1124 Disable usage of age constraints for matching with perfect dis‐
1126 crimination trees indexing.
1127 --detsort-rw
1129 Sort set of clauses eliminated by backward rewriting using a to‐
1131 tal syntactic ordering.
1132 --detsort-new
1134 Sort set of newly generated and backward simplified clauses us‐
1136 ing a total syntactic ordering.
1137 -D <arg>
1139 --define-weight-function=<arg>
1141 Define
1143 a weight function (see manual for details). Later definitions
1144 override previous definitions.
1146 -H <arg>
1148 --define-heuristic=<arg>
1150 Define a clause selection heuristic (see manual for details).
1152 Later definitions override previous definitions.
1153 --free-numbers
1155 Treat numbers (strings of decimal digits) as normal free func‐
1157 tion symbols in the input. By default, number now are supposed
1158 to denote domain constants and to be implicitly different from
1159 each other.
1160 --free-objects
1162 Treat object identifiers (strings in double quotes) as normal
1164 free function symbols in the input. By default, object identi‐
1165 fiers now represent domain objects and are implicitly different
1166 from each other (and from numbers, unless those are declared to
1167 be free).
1168 --definitional-cnf[=<arg>]
1170 Tune the clausification algorithm to introduces definitions for
1172 subformulae to avoid exponential blow-up. The optional argument
1173 is a fudge factor that determines when definitions are intro‐
1174 duced. 0 disables definitions completely. The default works
1175 well. The option without the optional argument is equivalent to
1176 --definitional-cnf=24.
1177 --old-cnf[=<arg>]
1179 As the previous option, but use the classical, well-tested
1181 clausification algorithm as opposed to the newewst one which
1182 avoides some algorithmic pitfalls and hence works better on some
1183 exotic formulae. The two may produce slightly different (but eq‐
1184 uisatisfiable) clause normal forms. The option without the op‐
1185 tional argument is equivalent to --old-cnf=24.
1186 --miniscope-limit[=<arg>]
1188 Set the limit of sub-formula-size to miniscope. The build-inde‐
1190 fault is 256. Only applies to the new (default) clausification
1191 algorithm The option without the optional argument is equivalent
1192 to --miniscope-limit=2147483648.
1193 --print-types
1195 Print the type of every term. Useful for debugging purposes.
1197 --app-encode
1199 Encodes terms in the proof state using applicative encoding,
1201 prints encoded input problem and exits.
1202
1204 Report bugs to <schulz@eprover.org>. Please include the following, if
1205 possible:
1206 * The version of the package as reported by eprover --version.
1208 * The operating system and version.
1210 * The exact command line that leads to the unexpected behaviour.
1212 * A description of what you expected and what actually happend.
1214 * If possible all input files necessary to reproduce the bug.
1216
1218 Copyright 1998-2020 by Stephan Schulz, schulz@eprover.org, and the E
1219 contributors (see DOC/CONTRIBUTORS).
1220 This program is a part of the distribution of the equational theorem
1222 prover E. You can find the latest version of the E distribution as well
1223 as additional information at http://www.eprover.org
1224 This program is free software; you can redistribute it and/or modify it
1226 under the terms of the GNU General Public License as published by the
1227 Free Software Foundation; either version 2 of the License, or (at your
1228 option) any later version.
1229 This program is distributed in the hope that it will be useful, but
1231 WITHOUT ANY WARRANTY; without even the implied warranty of MER‐
1232 CHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General
1233 Public License for more details.
1234 You should have received a copy of the GNU General Public License along
1236 with this program (it should be contained in the top level directory of
1237 the distribution in the file COPYING); if not, write to the Free Soft‐
1238 ware Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
1239 02111-1307 USA
1240 The original copyright holder can be contacted via email or as
1242 Stephan Schulz DHBW Stuttgart Fakultaet Technik Informatik Rote‐
1244 buehlplatz 41 70178 Stuttgart Germany
1245E 2.5-DEBUG Avongrove (a8acce0b281Jfa2n7u2a8r2yba21052f10a8541e54662223340) E(1)