aboutsummaryrefslogtreecommitdiff
path: root/coq
AgeCommit message (Collapse)Author
2017-05-05Change (eval-when (compile) ...) to (eval-when-compile ...)Clément Pit--Claudel
This fixes a bunch of compilation warnings
2017-05-05Merge pull request #157 from ProofGeneral/elpaClément Pit-Claudel
[WIP] ELPA/MELPA support
2017-04-25Typo from commit 758e679e.Pierre Courtieu
2017-04-24Preparing new warning tags (no more special chars).Pierre Courtieu
2017-04-19Fix #176.Pierre Courtieu
The code uses several token searcher and the wrong one was called somewhere.
2017-03-31Fixing #173.Pierre Courtieu
2017-03-22Added support for future new options (trunk).Pierre Courtieu
2017-03-08Add a FIXME in coq.elClément Pit--Claudel
2017-03-08Remove uses of defpgdefault in coq-abbrevClément Pit--Claudel
This file is `require'-d when compiling coq.el, and at that point the proof assistant isn't set to coq yet, so it would define variables prefixed by `nil-' instead of `coq'.
2017-03-08Remove uses of defpacustom in coq-compile-commonClément Pit--Claudel
This file is `require'-d from other places, at compile time. This doesn't work nicely with compilation by package.el (PG complains and says "defpacustom: Proof assistant setting compile-before-require re-defined!")
2017-03-08Remove compile-time calls to proof-ready-for-assistantClément Pit--Claudel
Compilation used to run in a separate Emacs process for each file, but that's not what happens when installing PG with package.el.
2017-03-08Remove unnecessary calls to 'eval-and-compile'Clément Pit--Claudel
2017-03-08Fix incorrect uses of defvarClément Pit--Claudel
It didn't really matter that these variables were defined and set to nil during compilation, since we ran compilation in a clean Emacs in --batch mode; it does matter now, however, since package.el compiles PG in the user's currently running Emacs instance.
2017-03-08Fix incorrect assumption that noninteractive == byte-compilingClément Pit--Claudel
The PG Makefile does ensure (using --batch) that noninteractive is non-nil while compiling, but package.el doesn't.
2017-03-08Merge commit '06fd76163b857a056ac44e7437efa17656f06e5b'Paul Steckler
2017-03-08Fixing unicode tokens in generic code and in coq.Pierre Courtieu
2017-03-06get threeb frames only when neededPaul Steckler
2017-03-06one more redundant call removedPaul Steckler
2017-03-06remove redundant calls, simplify codePaul Steckler
2017-03-03Merge pull request #163 from ProofGeneral/fix_indentationPierre Courtieu
Fix indentation
2017-03-03Refreshing goal when Set Printing xxx. (#162)Pierre Courtieu
* Close ProofGeneral/PG#161. Issue a "Show" each time a (Uns|S)et Printing is detected (and a proof is open).
2017-03-02use Utf8 from Coq libraryPaul Steckler
2017-02-27serveral coqtags fixes and improvementsHendrik Tews
- precise tags for definitions - fix bug with nonempty white space lines - add several keywords (Proposition, Record, ...) - generate tags for constructors of inductive definitions and for record labels
2017-02-23Fixing #154.Pierre Courtieu
2017-01-26Fixing #147 and #91 + others indentation bugs.Pierre Courtieu
While fixing these I discovered several flaws in indentation (what a suprise). The probem is the following: since we don't have a precise grammar of tactics, smie often decides that the parent of a sub-part of a tactic is the previous command instead of the current tacic. Typical example (fixed now but in general): foo. apply bar with bar'. Since "apply ... bar'" is considered as one leaf of the grammar, it is considered to be a child of the previous dot. . /\ / \ foo apply...bar' Therefore indentation of "with" wants to align with parent "." and hence with "foo". Basically we should try to define a much more precise grammar of complex tactics (with with, as, using etc) in order to fix the problem. Of course this has the drawback of making impossible to support user tactic notations.
2017-01-19save settings not defined with defpacustom (fixes #142)Hendrik Tews
- infrastructure for saving/resetting customizations not defined with defpacustom - improve Coq -> Auto Compilation menu - polish documentation and manual
2017-01-18split emergency-cleanup to handle interrupts properly (fixes #143)Hendrik Tews
Split coq-par-emergency-cleanup into two functions, one for reacting on user interrupts and one for cleaning up after compilation errors.
2017-01-17fix coqtagsHendrik Tews
@psteckler I believe you have this patch already in your branch.
2017-01-17Add {Local ,}Axioms like {Local ,}Axiom, fix Parameters (#106)Jason Gross
2017-01-17Merge pull request #107 from JasonGross/patch-3hendriktews
Add Context to coq-syntax.el
2017-01-14Fix prooftree for Coq 8.6Hendrik Tews
In Coq 8.6 evar status printing is off by default, causing prooftree to crash. This patch inserts invisible commands to switch evar status printing on and off. This is done via the urgent-action-hook.
2016-12-31add second argument to looking-back, required in emacs25Hendrik Tews
2016-12-28properly reset the vio2vo delay timerHendrik Tews
cancel-timer does of course not set the variable that holds the timer to nil
2016-12-16Merge pull request #133 from hendriktews/file-errorhendriktews
die gracefully when visiting files in nonexisting directories
2016-12-15add quote to fix commit e3cc66dc2e60683531d75c12256d059ccbc64576Hendrik Tews
2016-12-15fix :get for coq-search-blacklistHendrik Tews
2016-12-15Merge pull request #101 from tchajed/print-universes-optionhendriktews
Add Set Printing Universes to options menu
2016-12-15die gracefully when visiting files in nonexisting directoriesHendrik Tews
2016-12-14fix generic interrupt procedure to interrupt parallel background compilationHendrik Tews
2016-12-14fix race in vio2vo compilation startHendrik Tews
2016-12-14Merge pull request #129 from hendriktews/keep-goinghendriktews
Keep going
2016-12-14Merge pull request #132 from Matafou/masterPierre Courtieu
Remove default absolute name from coq-prog-name (Fixes #76), but keep displaying…
2016-12-13Same name guessing for coqc/coqdep then for coqtop.Pierre Courtieu
2016-12-12remove default absolute name from coq-prog-name, but keep dipsplaying it ↵Pierre Courtieu
when asking for it.
2016-12-08documentation and CHANGES for coq-compile-keep-goingHendrik Tews
2016-12-08option coq-compile-keep-going for parallel compilationHendrik Tews
With this option set, compilation continues after the first error to compile as much as possible and to potentially report more than one error.
2016-12-02remove ancestor hash in Coq parallel background compilationHendrik Tews
2016-11-30fix customize-group coqHendrik Tews
2016-11-30style change: use when for if coq-debug-auto-compilationHendrik Tews
2016-11-30use coq-- for internal compilation variablesHendrik Tews