]> Repositorios git - scryer-prolog.git/log
scryer-prolog.git
3 years agouse newly available ... //0 from library(dcgs)
Markus Triska [Sat, 13 Aug 2022 19:42:44 +0000 (21:42 +0200)]
use newly available ... //0 from library(dcgs)

3 years agoadd meta_predicate/1 declaration for phrase_to_file/3
Markus Triska [Sat, 13 Aug 2022 19:39:08 +0000 (21:39 +0200)]
add meta_predicate/1 declaration for phrase_to_file/3

3 years agostrip modules from GRBody in phrase/3 (#1565)
Mark Thom [Sun, 14 Aug 2022 02:46:52 +0000 (20:46 -0600)]
strip modules from GRBody in phrase/3 (#1565)

3 years agofail on non-list solutions in setof/3 (#1553)
Mark Thom [Sat, 13 Aug 2022 16:14:37 +0000 (10:14 -0600)]
fail on non-list solutions in setof/3 (#1553)

3 years agoHTTP Server 2.0
Adrián Arroyo Calle [Thu, 11 Aug 2022 18:25:19 +0000 (20:25 +0200)]
HTTP Server 2.0

3 years agoprefix inlined goals with module names in expand_subgoal/5 (#1551)
Mark Thom [Sat, 6 Aug 2022 22:15:25 +0000 (16:15 -0600)]
prefix inlined goals with module names in expand_subgoal/5 (#1551)

3 years agotry to invoke '$call_inline' a second time in call/N (#1543)
Mark Thom [Mon, 25 Jul 2022 19:52:24 +0000 (13:52 -0600)]
try to invoke '$call_inline' a second time in call/N (#1543)

3 years agoENHANCED: CLP(ℤ): Reduce redundant propagator invocations during all_distinct/1 filtering
Markus Triska [Mon, 25 Jul 2022 18:08:54 +0000 (20:08 +0200)]
ENHANCED: CLP(ℤ): Reduce redundant propagator invocations during all_distinct/1 filtering

First, the current propagator is now logged and not re-triggered
during filtering. Second, and more significantly, all neq_num/2
constraints are scheduled and processed before more global constraints
are invoked. In this way, all the distilled information can be taken
into account by subsequently invoked global constraints.

These changes yield a 3-fold improvement in several Sudoku instances,
and a significant runtime reduction in social golfer instance 8-4-9.

3 years agoadd meta_predicate/1 declarations
Markus Triska [Sun, 24 Jul 2022 22:10:32 +0000 (00:10 +0200)]
add meta_predicate/1 declarations

3 years agoresolve phrase modules internally (#1541)
Mark Thom [Sun, 24 Jul 2022 21:20:53 +0000 (15:20 -0600)]
resolve phrase modules internally (#1541)

3 years agofail when exception caught from subgoal expansion (#1535)
Mark Thom [Sun, 24 Jul 2022 19:36:34 +0000 (13:36 -0600)]
fail when exception caught from subgoal expansion (#1535)

3 years agofix module resolution in dcgs, call/N (#1539)
Mark Thom [Sun, 24 Jul 2022 06:59:42 +0000 (00:59 -0600)]
fix module resolution in dcgs, call/N (#1539)

3 years agodeduplicate index ptr inlining for 0-arity atoms (#1538)
Mark Thom [Sat, 23 Jul 2022 19:16:02 +0000 (13:16 -0600)]
deduplicate index ptr inlining for 0-arity atoms (#1538)

3 years agotrim get_structure and put_structure arities when last arg is an index ptr (#1536)
Mark Thom [Sat, 23 Jul 2022 03:34:38 +0000 (21:34 -0600)]
trim get_structure and put_structure arities when last arg is an index ptr (#1536)

3 years agoprint index_ptr offset (#1534)
Mark Thom [Fri, 22 Jul 2022 19:10:12 +0000 (13:10 -0600)]
print index_ptr offset (#1534)

3 years agouse separate predicate for findall cleanup
Mark Thom [Thu, 21 Jul 2022 03:16:17 +0000 (21:16 -0600)]
use separate predicate for findall cleanup

3 years agoinference count call_inline, fail on undefined index
Mark Thom [Wed, 20 Jul 2022 19:31:44 +0000 (13:31 -0600)]
inference count call_inline, fail on undefined index

3 years agostreamline assertz/1, asserta/1
Mark Thom [Wed, 20 Jul 2022 02:29:22 +0000 (20:29 -0600)]
streamline assertz/1, asserta/1

3 years agoinline metacalls
Mark Thom [Wed, 13 Jul 2022 04:39:50 +0000 (22:39 -0600)]
inline metacalls

3 years agoMerge pull request #1630 from epilys/fix-1625
Mark Thom [Thu, 27 Oct 2022 05:37:36 +0000 (23:37 -0600)]
Merge pull request #1630 from epilys/fix-1625

fix type_error with instantiated EOF -1 byte literal in get_byte/2

3 years agofix type_error with instantiated EOF -1 byte literal in get_byte/2
Manos Pitsidianakis [Wed, 26 Oct 2022 10:49:38 +0000 (13:49 +0300)]
fix type_error with instantiated EOF -1 byte literal in get_byte/2

According to ISO Prolog, get_byte/2 predicate can receive an
instantiated input byte:

http://www.gprolog.org/manual/html_node/gprolog037.html#sec156

    get_byte(+stream_or_alias, ?in_byte)

Since in_byte can be -1 if EOF is reached, instantiating it with -1
should work but does not because the implementation is trying to convert
it to a u8 which is unsigned:

?- open("/dev/null", read, S, [type(binary)]), get_byte(S, -1).
   error(type_error(in_byte,-1),get_byte/2).

This commit adds an extra check for -1 before checking for a valid u8
instantiated value if in_byte is an input:

?- open("/dev/null", read, S, [type(binary)]), get_byte(S, -1).
   S = '$stream'(0x55601e65c998).

Closes #1625

3 years agoMerge pull request #1628 from epilys/fix-1626
Mark Thom [Wed, 26 Oct 2022 02:46:49 +0000 (20:46 -0600)]
Merge pull request #1628 from epilys/fix-1626

Fix xor/2 type error reporting wrong argument

3 years agoFix xor/2 type error reporting wrong argument
Manos Pitsidianakis [Mon, 24 Oct 2022 16:26:43 +0000 (19:26 +0300)]
Fix xor/2 type error reporting wrong argument

If first argument n1 in xor/2 is of wrong numerical type, the match
patterns fall through to a catch all case that reports the second
argument in the type error.

Fixes #1626

`xor/2 function reports the wrong argument in type error #1626`

https://github.com/mthom/scryer-prolog/issues/1626

3 years agoMerge pull request #1530 from RossSmyth/FixImports
Mark Thom [Tue, 19 Jul 2022 03:36:46 +0000 (21:36 -0600)]
Merge pull request #1530 from RossSmyth/FixImports

Fixed imports for num-rug-adapter

3 years agoFixed imports num-rug-adapter
Ross Smyth [Sun, 17 Jul 2022 06:06:37 +0000 (02:06 -0400)]
Fixed imports num-rug-adapter

3 years agocapture and propagate (=..)/2 errors in phrase/{2,3} (#1460)
Mark Thom [Wed, 22 Jun 2022 04:47:55 +0000 (22:47 -0600)]
capture and propagate (=..)/2 errors in phrase/{2,3} (#1460)

3 years agopropagate inference_limit_exceeded errors through the loader (#1298)
Mark Thom [Sat, 18 Jun 2022 17:26:31 +0000 (11:26 -0600)]
propagate inference_limit_exceeded errors through the loader (#1298)

3 years agouse setup_cal_n_init_goal_info in dynamic_module_resolution
Mark Thom [Thu, 16 Jun 2022 04:32:05 +0000 (22:32 -0600)]
use setup_cal_n_init_goal_info in dynamic_module_resolution

4 years agoimprove efficiency of call/N, replace '$call_with_default_policy' with
Mark Thom [Mon, 13 Jun 2022 04:17:44 +0000 (22:17 -0600)]
improve efficiency of call/N, replace '$call_with_default_policy' with
'$call_with_inference_counting'

4 years agosimplify detecting undefined goal_ and term_expansion
Mark Thom [Wed, 8 Jun 2022 03:45:26 +0000 (21:45 -0600)]
simplify detecting undefined goal_ and term_expansion

4 years agoMerge pull request #1510 from Skgland/add-gitattributes
Mark Thom [Sun, 12 Jun 2022 21:44:14 +0000 (15:44 -0600)]
Merge pull request #1510 from Skgland/add-gitattributes

add a .gitattributes file

4 years agoadd a .gitattributes file
Skgland [Sun, 12 Jun 2022 13:39:55 +0000 (15:39 +0200)]
add a .gitattributes file

- this should stop problems such as mthom/scryer-prolog#1491
  by specifying the line ending
  as this should take precedence over the global config

4 years agoreplace call with $call in more meta-predicates
Mark Thom [Mon, 6 Jun 2022 02:53:07 +0000 (20:53 -0600)]
replace call with $call in more meta-predicates

4 years agoemit indexing instructions in single clauses of dynamic predicates
Mark Thom [Sun, 5 Jun 2022 05:43:48 +0000 (23:43 -0600)]
emit indexing instructions in single clauses of dynamic predicates

4 years agodo not emit indexing instructions for single clause subsequences (#1503)
Mark Thom [Sat, 4 Jun 2022 21:29:00 +0000 (15:29 -0600)]
do not emit indexing instructions for single clause subsequences (#1503)

4 years agoMerge pull request #1505 from Skgland/optional_rustfmt
Mark Thom [Sat, 4 Jun 2022 15:56:51 +0000 (09:56 -0600)]
Merge pull request #1505 from Skgland/optional_rustfmt

make rustfmt optional

4 years agoadjust error message on failed formatting
Skgland [Fri, 3 Jun 2022 21:50:17 +0000 (23:50 +0200)]
adjust error message on failed formatting

4 years agomake rustfmt optional
Skgland [Fri, 3 Jun 2022 21:35:55 +0000 (23:35 +0200)]
make rustfmt optional

- when rustfmt is not detected (by attempting to run ``rustfmt --version)
  no formatting will be attempted
- this should resolve issue mthom/scryer-prolog#1504

4 years agounmark cells in arith_eval_by_metacall (#1497)
Mark Thom [Wed, 1 Jun 2022 06:02:11 +0000 (00:02 -0600)]
unmark cells in arith_eval_by_metacall (#1497)

4 years agodon't print superfluous round brackets in curly braces (#1416)
Mark Thom [Wed, 1 Jun 2022 05:34:55 +0000 (23:34 -0600)]
don't print superfluous round brackets in curly braces (#1416)

4 years agofix broken quoting in writeq/1 (#1498)
Mark Thom [Wed, 1 Jun 2022 00:12:19 +0000 (18:12 -0600)]
fix broken quoting in writeq/1 (#1498)

4 years agofix off-by-one error reporting in write_term/3 (#1490)
Mark Thom [Tue, 31 May 2022 05:25:44 +0000 (23:25 -0600)]
fix off-by-one error reporting in write_term/3 (#1490)

4 years agodon't quote most characters in strings (#1495, #301)
Mark Thom [Tue, 31 May 2022 05:09:20 +0000 (23:09 -0600)]
don't quote most characters in strings (#1495, #301)

4 years agouse existence_error for dropped streams (#1472)
Mark Thom [Mon, 23 May 2022 17:15:58 +0000 (11:15 -0600)]
use existence_error for dropped streams (#1472)

4 years agoMerge pull request #1485 from triska/in_character
Mark Thom [Mon, 23 May 2022 01:52:19 +0000 (19:52 -0600)]
Merge pull request #1485 from triska/in_character

Implement and use can_be(in_character, ...)

4 years agouse in_character
Markus Triska [Sun, 22 May 2022 20:43:48 +0000 (22:43 +0200)]
use in_character

This addresses one aspect of #1472.

4 years agoADDED: must_be(in_character, ...) and can_be(in_character, ...)
Markus Triska [Sun, 22 May 2022 20:43:01 +0000 (22:43 +0200)]
ADDED: must_be(in_character, ...) and can_be(in_character, ...)

4 years agofix stream position and term comparison bugs (#1472)
Mark Thom [Sun, 22 May 2022 20:15:56 +0000 (14:15 -0600)]
fix stream position and term comparison bugs (#1472)

4 years agoMerge pull request #1484 from aarroyoc/fix-windows-build-2
Mark Thom [Sun, 22 May 2022 16:15:16 +0000 (10:15 -0600)]
Merge pull request #1484 from aarroyoc/fix-windows-build-2

Fix build on Windows 10

4 years agoMSI installer
Adrián Arroyo Calle [Sun, 22 May 2022 12:18:51 +0000 (14:18 +0200)]
MSI installer

4 years agoFix build on Windows 10
Adrián Arroyo Calle [Sun, 22 May 2022 11:18:11 +0000 (13:18 +0200)]
Fix build on Windows 10

4 years agothrow a representation error if max arity is exceeded (#1483)
Mark Thom [Sun, 22 May 2022 06:03:55 +0000 (00:03 -0600)]
throw a representation error if max arity is exceeded (#1483)

4 years agoMerge pull request #1482 from triska/multiple_of_five_answers
Mark Thom [Sat, 21 May 2022 18:35:35 +0000 (12:35 -0600)]
Merge pull request #1482 from triska/multiple_of_five_answers

ENHANCED: "f" on toplevel to report answers up to the next multiple of 5.

4 years agoadd max arity checks at various stages (#1459)
Mark Thom [Sat, 21 May 2022 18:32:48 +0000 (12:32 -0600)]
add max arity checks at various stages (#1459)

4 years agoENHANCED: "f" on toplevel to report answers up to the next multiple of 5.
Markus Triska [Sat, 21 May 2022 16:20:08 +0000 (18:20 +0200)]
ENHANCED: "f" on toplevel to report answers up to the next multiple of 5.

This makes counting solutions easy.

4 years agocheck asserta/2, assertz/2, abolish/2 for uninstantiated Module variables (#1475)
Mark Thom [Tue, 17 May 2022 02:45:45 +0000 (20:45 -0600)]
check asserta/2, assertz/2, abolish/2 for uninstantiated Module variables (#1475)

4 years agodetect character type errors in atom_{chars|codes}/2 (#1472)
Mark Thom [Mon, 16 May 2022 22:32:38 +0000 (16:32 -0600)]
detect character type errors in atom_{chars|codes}/2 (#1472)

4 years agoMerge pull request #1478 from triska/sgml_improvements
Mark Thom [Sun, 15 May 2022 15:44:44 +0000 (09:44 -0600)]
Merge pull request #1478 from triska/sgml_improvements

Various improvements to library(sgml)

4 years agoMerge pull request #1477 from triska/document_tab_completion
Mark Thom [Sun, 15 May 2022 15:44:26 +0000 (09:44 -0600)]
Merge pull request #1477 from triska/document_tab_completion

document atom and predicate completion with TAB

4 years agoindent enumerations
Markus Triska [Sun, 15 May 2022 08:55:56 +0000 (10:55 +0200)]
indent enumerations

4 years agoclarify the representation
Markus Triska [Sun, 15 May 2022 08:55:27 +0000 (10:55 +0200)]
clarify the representation

4 years agostrengthen and improve type checks
Markus Triska [Sun, 15 May 2022 08:25:41 +0000 (10:25 +0200)]
strengthen and improve type checks

4 years agoclarify comment
Markus Triska [Sun, 15 May 2022 08:23:00 +0000 (10:23 +0200)]
clarify comment

4 years agouse phrase_from_file/2 directly on the file name
Markus Triska [Sun, 15 May 2022 08:19:00 +0000 (10:19 +0200)]
use phrase_from_file/2 directly on the file name

4 years agodocument atom and predicate completion with TAB
Markus Triska [Sun, 15 May 2022 08:04:45 +0000 (10:04 +0200)]
document atom and predicate completion with TAB

This feature, added via
https://github.com/mthom/scryer-prolog/pull/1467, makes the toplevel
interaction a lot more convenient in many cases.

Many thanks to @euanlacy!

4 years agoMerge pull request #1474 from triska/correct_can_be_chars
Mark Thom [Sun, 15 May 2022 05:28:04 +0000 (23:28 -0600)]
Merge pull request #1474 from triska/correct_can_be_chars

FIXED: can_be(chars, [a,X]), i.e., if variables occur as elements

4 years agoconsider deallocate a head instruction (#1473)
Mark Thom [Sat, 14 May 2022 17:01:26 +0000 (11:01 -0600)]
consider deallocate a head instruction (#1473)

4 years agoprioritize type errors over instantiation errors
Markus Triska [Sat, 14 May 2022 07:52:34 +0000 (09:52 +0200)]
prioritize type errors over instantiation errors

Example:

    ?- must_be(chars, [a,X,cc]).
       error(type_error(character,cc),can_be/2).

See https://github.com/mthom/scryer-prolog/pull/1474#issuecomment-1126664368

4 years agoFIXED: can_be(chars, [a,X]), i.e., if variables occur as elements
Markus Triska [Sat, 14 May 2022 06:09:01 +0000 (08:09 +0200)]
FIXED: can_be(chars, [a,X]), i.e., if variables occur as elements

This should simplify 2f3de51e554629d2e7b4ea023c9ef181c240692c and
other cases like it.

4 years agoremove non-determinism in number_chars/2 and numbercodes/2 (#1473)
Mark Thom [Sat, 14 May 2022 00:38:02 +0000 (18:38 -0600)]
remove non-determinism in number_chars/2 and numbercodes/2 (#1473)

4 years agouse must_be(chars) in number_chars/2 (#1470)
Mark Thom [Thu, 12 May 2022 06:00:37 +0000 (00:00 -0600)]
use must_be(chars) in number_chars/2 (#1470)

4 years agoMerge pull request #1467 from euanlacy/master
Mark Thom [Thu, 12 May 2022 04:32:09 +0000 (22:32 -0600)]
Merge pull request #1467 from euanlacy/master

Completion of predicates in repl

4 years agowip: add more atoms to completion list
Euan Lacy [Wed, 11 May 2022 20:37:03 +0000 (21:37 +0100)]
wip: add more atoms to completion list

4 years agofix: remove unnecessary string & vec allocations
Euan Lacy [Tue, 10 May 2022 16:54:15 +0000 (17:54 +0100)]
fix: remove unnecessary string & vec allocations

4 years agowip: basic completion of predicates
Euan Lacy [Tue, 10 May 2022 03:40:22 +0000 (04:40 +0100)]
wip: basic completion of predicates

4 years agounify partial string tail later in create_partial_string (#1462, #1465, #1427)
Mark Thom [Sun, 8 May 2022 20:52:22 +0000 (14:52 -0600)]
unify partial string tail later in create_partial_string (#1462, #1465, #1427)

4 years agofix culprit in stream_position permission_error (#1463)
Mark Thom [Fri, 6 May 2022 23:38:25 +0000 (17:38 -0600)]
fix culprit in stream_position permission_error (#1463)

4 years agocorrect cycle detection in unify_* (#1455)
Mark Thom [Fri, 6 May 2022 00:58:22 +0000 (18:58 -0600)]
correct cycle detection in unify_* (#1455)

4 years agoMerge pull request #1461 from triska/clpz_sign
Mark Thom [Thu, 5 May 2022 23:24:54 +0000 (17:24 -0600)]
Merge pull request #1461 from triska/clpz_sign

ADDED: sign/1

4 years agoADDED: sign/1
Markus Triska [Thu, 5 May 2022 20:49:48 +0000 (22:49 +0200)]
ADDED: sign/1

4 years agoremove comment that seems no longer valid
Markus Triska [Thu, 5 May 2022 20:51:21 +0000 (22:51 +0200)]
remove comment that seems no longer valid

4 years agofix rnd_i fixnum casting bug (#1456, #1458)
Mark Thom [Thu, 5 May 2022 04:44:12 +0000 (22:44 -0600)]
fix rnd_i fixnum casting bug (#1456, #1458)

4 years agotrack f64 offsets in Literal (#1190)
Mark Thom [Thu, 5 May 2022 03:54:46 +0000 (21:54 -0600)]
track f64 offsets in Literal (#1190)

4 years agocompact lists of characters to partial strings in read_term_from_heap (#1452, #1453)
Mark Thom [Mon, 2 May 2022 05:13:48 +0000 (23:13 -0600)]
compact lists of characters to partial strings in read_term_from_heap (#1452, #1453)

4 years agofix match_partial_string, compare_pstr_to_string (#1451)
Mark Thom [Sun, 1 May 2022 01:51:19 +0000 (19:51 -0600)]
fix match_partial_string, compare_pstr_to_string (#1451)

4 years agoremove redundant register assignments in numeric comparisons (#1438, #1448)
Mark Thom [Fri, 29 Apr 2022 07:25:36 +0000 (01:25 -0600)]
remove redundant register assignments in numeric comparisons (#1438, #1448)

4 years agoremove redundant skip_stub_try_me_else condition (#1444)
Mark Thom [Fri, 29 Apr 2022 01:25:09 +0000 (19:25 -0600)]
remove redundant skip_stub_try_me_else condition (#1444)

4 years agoMerge pull request #1436 from triska/must_be_term
Mark Thom [Tue, 26 Apr 2022 19:31:43 +0000 (13:31 -0600)]
Merge pull request #1436 from triska/must_be_term

ADDED: can_be(term, ...) and must_be(term, ...)

4 years agoreplace SliceDeque with VecDeque
Mark Thom [Tue, 26 Apr 2022 04:55:28 +0000 (22:55 -0600)]
replace SliceDeque with VecDeque

4 years agocatch existence_error exceptions from top-level calls to attribute_goals/3 (#1432)
Mark Thom [Mon, 25 Apr 2022 02:59:00 +0000 (20:59 -0600)]
catch existence_error exceptions from top-level calls to attribute_goals/3 (#1432)

4 years agounify structures from left to right (#1396)
Mark Thom [Sat, 23 Apr 2022 01:00:42 +0000 (19:00 -0600)]
unify structures from left to right (#1396)

4 years agoupdate to crossterm 0.20.0 (#1439)
Mark Thom [Fri, 22 Apr 2022 00:41:07 +0000 (18:41 -0600)]
update to crossterm 0.20.0 (#1439)

4 years agoADDED: can_be(term, ...) and must_be(term, ...)
Markus Triska [Wed, 20 Apr 2022 18:41:11 +0000 (20:41 +0200)]
ADDED: can_be(term, ...) and must_be(term, ...)

This terminology is a bit harder to understand than that of the other
type checks.

Scryer Prolog comprises two individually consistent but incompatible
notions of terms, namely terms and rational trees (which are also
called "cyclic terms"). This leads to a confusion in terminology,
because cyclic terms are not terms in the former sense even though
they also contain the word "term" in their name.

This terminological confusion is not restricted to terms: We also find
it for example in the notion of lists. Lists are defined inductively:

    - The empty list (written as "" or [], called 'nil') is a list.
    - A term with principal functor '.'/2 is a list if its second argument
      is a list.

Note that a list may contain cyclic terms as elements, and still be
called a list. However, if we post:

    ?- Ls = [_|Ls].

then Ls is not a list, because it does not fit the inductive
definition. Still, it is sometimes referred to as a "cyclic list" or
"circular list", even though it is not a list. Even Prolog
implementors that aim for strict conformance to the Prolog ISO
standard and are deeply familiar with the standard use this
terminology when talking about such terms, see for example:

    https://github.com/ichiban/prolog/pull/198

Consequently, must_be(list, Ls) raises a *type error* in this case.

Terms are distinguished from cyclic terms completely analogously. This
leads to the counterintuitive consequence that *cyclic lists are not
terms*. The reason for this apparent contradiction in terminology is
the mutual incompatibility of different notions of terms. The notion
of "term" precludes rational trees, similar to the notion of "list"
which precludes cyclic lists.

We therefore introduce:

    * must_be(term, Term)
      - raises an instantiation error if Term contains variables (since
        it could still become a rational tree)
      - raises a type error if Term is a rational tree

    * can_be(term, Term)
      - raises a type error if Term is a rational tree

Note that using the notion of "cyclic term" and "acyclic term" in this
terminology would not be a good idea, because this confusion is
precisely what we want to eliminate: must_be(term, Term) should be
usable as a test that reliably determines whether Term can be treated
as a term with the logical properties we expect from terms.

This addresses #1428.

4 years agovery minor optimizations
Mark Thom [Mon, 18 Apr 2022 02:10:13 +0000 (20:10 -0600)]
very minor optimizations

4 years agomark unsafe variables and substitute temporary variables in is/2 when appropriate...
Mark Thom [Sun, 17 Apr 2022 23:33:09 +0000 (17:33 -0600)]
mark unsafe variables and substitute temporary variables in is/2 when appropriate (#1430)

4 years agofix panic caused by jmp offset off by one error (#1429)
Mark Thom [Sun, 17 Apr 2022 18:19:48 +0000 (12:19 -0600)]
fix panic caused by jmp offset off by one error (#1429)

4 years agoavoid writing chars to the atom table in several places
Mark Thom [Sun, 17 Apr 2022 01:57:26 +0000 (19:57 -0600)]
avoid writing chars to the atom table in several places

4 years agoMerge pull request #1421 from gruhn/master
Mark Thom [Sat, 16 Apr 2022 17:11:33 +0000 (11:11 -0600)]
Merge pull request #1421 from gruhn/master

push Docker image on new release

4 years agofix comparisons to partial string tails (#1420)
Mark Thom [Fri, 15 Apr 2022 04:14:23 +0000 (22:14 -0600)]
fix comparisons to partial string tails (#1420)

4 years agopush Docker image on new release
Niklas Gruhn [Thu, 14 Apr 2022 14:33:57 +0000 (16:33 +0200)]
push Docker image on new release

Setup GitHub action which automatically builds and pushes Docker
images, when new release tags are created.