Skip to content

Latest commit

 

History

History
126 lines (82 loc) · 3.23 KB

TODO.md

File metadata and controls

126 lines (82 loc) · 3.23 KB
Raw

TODO

  • is it really useful to compute a canonical type/term?

  • add a term abbreviation only if it is used more than once

  • why the mapping of ITLIST does not work anymore ?

  • in lp output, replace T by ⊤ ?

  • write progress in an ocaml program, and estimate time wrt size of files

  • use a single term_abbrev file for each theorem, and split it

  • pre-compute once and for all the type and term variables of axioms and definitions

  • compute proof tree sizes in parallel ?

  • really remove useless proof steps to save memory (=> use files are not necessary afterwards)

  • record times to generate proofs and term_abbrevs files of each theorem ?

  • optimize number of let's (use a let only if an abbreviation is used more than once)

  • remove base argument in all commands by using BASE file

  • rename file _term by _sig

  • use subterm_abbrevs even without using --sharing ? keep subterm_abbrevs at all ?

  • Makefile: add targets for dk output

  • simplify dk output of proofs by replacing [c : x:A -> B := x:A => t] by [c (x:A) : B := t].

  • use exact_no_check/vm_cast_no_check/native_cast_no_check in lambdapi export to coq ? can work only if there is no implicit argument and no need for implicit coercions

  • improve hash function on big terms

  • beta-normalize hol-light terms ?

  • remove the need for sed after lambdapi export -o stt_coq

  • remove/comment code for mk? (i.e. use split as default)

  • improve efficiency of code outputing dk/lp (avoid multiple term traversals)

  • extend split command to dedukti output

  • get rid of use file? in pos files, set pos to -1 for unused theorems?

  • simplify/purge proofs in parallel?

  • create doc on readthedocs.io

  • replace terms of the form (@f _ x) by x if f is a function equal to the identity like I, NUMERAL

  • replace type variables like _1718 by better names like A

  • add v file checking in ci

  • make export incremental (1 lp file for each ml file)

  • make or_elim more implicit ?

  • detect invalid lp identifiers

  • minimize and clean xprelude.ml

  • keep in memory theorems used more than n>1 times ?

  • check whether Proof constructor is still useful

  • factorize code between dk and lp? using a functor?

  • uniformize names between no part and part

hol2dk xci.dk generates . xci_types . xci_terms . xci_axioms . xci_theorems . xci_proofs . xci_term_abbrevs . xci_type_abbrevs . xci (concatenation of previous files)

renaming: . xci_proofs -> xci_proofs_1 . xci_term_abbrevs -> xci_proofs_1_term_abbrevs . xci_type_abbrevs -> xci_proofs_1_type_abbrevs

hol2dk dg 7 xci && hol2dk mk 7 xci coq.v && make -f xci.mk dk generates . xci_types . xci_terms . xci_axioms . xci_theorems . xci_part_* . xci_part_*_term_abbrevs . xci_part_*_type_abbrevs . xci (concatenation of previous files)

renaming: . part -> proofs

hol2dk xci.lp generates . xci_types . xci_terms . xci_axioms . xci . xci_proofs . xci_term_abbrevs . xci_type_abbrevs

hol2dk dg 7 xci && hol2dk mk 7 xci coq.v && make -f xci.mk lp generates . xci_types . xci_terms . xci_axioms . xci . xci_part_* . xci_part_*_type_abbrevs . xci_part_*_term_abbrevs