Commit Graph
100 Commits
Author SHA1 Message Date
Markus Triska 4422ffe39f add meta_predicate/1 declarations 2022-10-26 23:36:07 -06:00
Markus Triska 6b49653754 use in_character
This addresses one aspect of #1472.
2022-05-22 23:04:27 +02:00
Markus Triska d5fa8cc211 ADDED: must_be(in_character, ...) and can_be(in_character, ...) 2022-05-22 23:04:27 +02:00
Markus Triska 47a6b498e3 ENHANCED: "f" on toplevel to report answers up to the next multiple of 5.
This makes counting solutions easy.
2022-05-21 18:20:08 +02:00
Markus Triska 18c52e076f indent enumerations 2022-05-15 10:55:56 +02:00
Markus Triska 1810dabc14 clarify the representation 2022-05-15 10:55:27 +02:00
Markus Triska 609a3a229f strengthen and improve type checks 2022-05-15 10:51:00 +02:00
Markus Triska dff56643f0 clarify comment 2022-05-15 10:23:00 +02:00
Markus Triska 77f8d52271 use phrase_from_file/2 directly on the file name 2022-05-15 10:19:00 +02:00
Markus Triska 1e9821ec0c 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!
2022-05-15 10:05:30 +02:00
Markus Triska a5e72679bc 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
2022-05-14 09:53:21 +02:00
Markus Triska 15cc916333 FIXED: can_be(chars, [a,X]), i.e., if variables occur as elements
This should simplify 2f3de51e55 and
other cases like it.
2022-05-14 08:39:20 +02:00
Markus Triska d58dd7cc91 ADDED: sign/1 2022-05-05 22:51:49 +02:00
Markus Triska 7e06abebd5 remove comment that seems no longer valid 2022-05-05 22:51:49 +02:00
Markus Triska eea83d9474 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.
2022-04-20 22:15:32 +02:00
Markus Triska 1fa8a0a969 ENHANCED: more memory efficient call_nth/2 (#1264)
I hope that this will allow more extensive tests of Scryer Prolog. Enjoy!
2022-04-02 17:42:45 +02:00
Markus Triska f7e31fceac as --> is 2022-03-30 21:33:46 +02:00
Markus Triska 1a7992e524 ENHANCED: use newly available must_be(chars, ...), which is very fast 2022-03-26 19:09:54 +01:00
Markus Triska 69e52d1ed8 ENHANCED: call_nth(Goal, 0) now fails, correct error for N < 0.
This addresses #1307.
2022-03-07 23:38:30 +01:00
Markus Triska c6264cf098 ADDED: library(simplex), for solving linear programming problems. 2022-03-06 21:49:52 +01:00
Markus Triska 8ad4f188f2 DOC: Better explanation of strings and partial strings. 2022-03-06 11:05:31 +01:00
Markus Triska 9f864574de ADDED: "f" to enumerate the next 5 solutions, similar to GUPU. 2022-03-02 21:05:49 +01:00
Markus Triska 102adb3544 ADDED: "a" to enumerate all solutions
Example:

    ?- member(X, "hello").
    %@    X = h   <-- press "a"
    %@ ;  X = e
    %@ ;  X = l
    %@ ;  X = l
    %@ ;  X = o
    %@ ;  false.
2022-03-01 23:52:51 +01:00
Markus Triska 1bec1b7002 FIXED: load_html/3: use str_loc_as_cell! to store attribute structures 2022-03-01 20:21:08 +01:00
Markus Triska 8fb673e93e use newly available get_n_chars/3 from library(charsio) 2022-02-23 00:03:36 +01:00
Markus Triska d1372d9b3b FIXED: library(sgml): load_html/3 and load_xml/3 now both work again.
This addresses #1249.
2022-02-23 00:03:27 +01:00
Markus Triska 2e57789d10 ENHANCED: Make toplevel output a Prolog term that can be read back also for "...".
This addresses #1240.
2022-01-28 18:47:55 +01:00
Markus Triska b703303dd4 omit module prefix, because the goals are already module-qualified 2022-01-20 00:48:30 +01:00
Markus Triska 0da9d1c036 ENHANCED: more specific warning when initialization/1 fails (#1149) 2022-01-20 00:04:09 +01:00
Markus Triska 6a610ac57d FIXED: toplevel to use (newly renamed) read_from_chars/2 2022-01-18 01:12:47 +01:00
Markus Triska 516ed1fd5b ENHANCED: library(crypto): Retain a compact internal representation of the ciphertext throughout decryption.
This allows efficient decryption also of very large files and long
strings.

It is usually highly advisable to stick to common and portable library
predicates such as append/3. However, since append/3 does not yet
recognize this opportunity for improvement, I apply it manually in
this case, so that also very large files can be efficiently decrypted
using the compact internal string representation.

Without this change, decrypting a 1 GB file takes 48 GB of RAM,
whereas with this change, it only takes 2 GB (one for the string, one
for its copy with the appended tag).
2022-01-17 23:02:56 +01:00
Markus Triska 23adfef281 MODIFIED: library(charsio): read_n_chars/3 is now called get_n_chars/3
This is because "read" is associated with general Prolog terms.
2022-01-17 20:37:55 +01:00
Markus Triska 20b6816562 MODIFIED: library(charsio): read_term_from_chars/2 is now called read_from_chars/2
This is for compatibility with SICStus Prolog and GNU Prolog.
2022-01-17 20:21:18 +01:00
Markus Triska 177c98fa95 correct reference to '$working_directory'/2 2022-01-16 23:20:43 +01:00
Markus Triska 11d504e8f4 retain the string "[]" as is, instead of converting it to '[]' (i.e., "")
This addresses #1215.
2022-01-16 17:50:20 +01:00
Markus Triska 91aa0f200e add links to the rebis-dev branch and its announcement 2022-01-13 08:46:24 +01:00
Markus Triska cfbb05fb1b deterministic --> semi-deterministic 2022-01-12 21:58:23 +01:00
Markus Triska 2889a4438b ADDED: format specifier ~NL to limit the number of digits per line
Example:

    %?- format("~65L", [2^1000]).
    %@ 10715086071862673209484250490600018105614048117055336074437503883_
    %@ 70351051124936122493198378815695858127594672917553146825187145285_
    %@ 69231404359845775746985748039345677748242309854210746050623711418_
    %@ 77954182153046474983581941267398767559165543946077062914571196477_
    %@ 686542167660429831652624386837205668069376   true.
2022-01-10 17:52:42 +01:00
Markus Triska f5952088e3 ADDED: format specifier ~NU, using underscores to separate groups of digits
Example:

    ?- format("~2U", [10^12]).
    %@ 10_000_000_000.00   true.
2022-01-10 16:07:05 +01:00
Markus Triska 57d15936cb ENHANCED: Use newly available fast chars test from library(error). 2021-12-12 17:43:17 +01:00
Markus Triska 12c561cee0 report time with "s" instead of "seconds", in analogy to time(1)
This addresses https://github.com/mthom/scryer-prolog/pull/1131#issuecomment-988670049 .
2021-12-08 16:13:54 +01:00
Markus Triska f79b8a1ca5 ENHANCED: support full layout text sequence (also comments) in integers with underscores
Example:

    ?- X = 1_/**/2.
    %@    X = 12.

Following the comment in https://github.com/mthom/scryer-prolog/pull/1112#issuecomment-981540485
2021-12-08 10:42:19 +01:00
Markus Triska 637daa5bda remove redundant test: '\n' is a layout char. 2021-12-08 10:37:37 +01:00
Markus Triska dc5e935ecd use newly available predicates from library(tls) for HTTPS 2021-12-05 18:16:47 +01:00
Markus Triska 538085169a ADDED: library(tls), providing all predicates for TLS connections.
The currently available predicates for TLS-connections can be extended
with predicates to load and reason about certificates etc.
2021-12-05 18:16:47 +01:00
Markus Triska 7e8a635e7e implement tls_client_negotiate/3 for explicit negotiation 2021-12-05 17:39:22 +01:00
Markus Triska 74e76b6f97 MODIFIED: Remove TLS-related predicates from library(sockets).
They will become available in a new library, library(tls).
2021-12-05 17:39:22 +01:00
Markus Triska cb6309d370 walltime --> CPU time 2021-12-05 00:56:16 +01:00
Markus Triska 77c04c3a14 ADDED: Support for creating TLS servers.
The new predicates tls_server_context/2 and tls_server_negotiate/3 can
be used to negotiate TLS connections with clients for encrypted and
authenticated communication.
2021-12-03 16:45:27 +01:00
Markus Triska 48cea6efdf FIXED: retract/1 issues due to module qualification
This addresses #1125.
2021-12-01 21:46:12 +01:00
Markus Triska 89a90d9522 ENHANCED: Better reporting for time/1.
This addresses #378.
2021-12-01 21:23:00 +01:00
Markus Triska 0026f3fdef use anonymous variable 2021-12-01 20:54:06 +01:00
Markus Triska ee054fd99c several corrections to test cases 2021-12-01 19:45:36 +01:00
Markus Triska 79a50de697 ENHANCED: Correct type error for char_type/2.
This addresses #905.
2021-11-29 19:33:22 +01:00
Markus Triska fc8205d375 ENHANCED: indicate failed initialization/1 goal
This addresses #1109.
2021-11-28 14:50:42 +01:00
Markus Triska 5e55732cb0 ENHANCED: Character type checks for get_char/[1,2].
This addresses #906.
2021-11-28 10:49:40 +01:00
Markus Triska 45cfa6c8d8 use seq//1 2021-11-27 16:32:45 +01:00
Markus Triska 25a06b0bea use seq//1 and ... //0 which are now available in library(dcgs) 2021-11-27 16:32:44 +01:00
Markus Triska 91af72e8b1 use seq//1 2021-11-27 16:17:45 +01:00
Markus Triska fd9b354c70 use seq//1 which is now available in library(dcgs) 2021-11-27 16:17:15 +01:00
Markus Triska 1647e67dd4 ENHANCED: Support '_' followed by layout text sequence in integers.
This is handy for writing large integers, for example in cryptographic
applications.

This addresses #994.

Examples:

    ?- X = 1_000_000.
    %@    X = 1000000.

    ?- X = 1_
    000_
    000.
    %@    X = 1000000.
2021-11-27 10:09:39 +01:00
Markus Triska 803e120a51 use must_be(list, ...) earlier, since it is now faster 2021-11-26 17:11:51 +01:00
Markus Triska 6ba374c62b ENHANCED: use '$skip_max_list'/4 for fast list tests in can_be/2 and must_be/2
This also (partially) addresses #1108.
2021-11-26 17:11:35 +01:00
Markus Triska 0d4c38138a use newly available must_be(chars, ...) for fast string tests 2021-11-22 20:00:06 +01:00
Markus Triska fe291e90f0 use newly available must_be(chars, ...) for fast string tests 2021-11-22 19:58:36 +01:00
Markus Triska 220e1e8d83 ADDED: must_be(chars, ...) and can_be(chars, ...) for string tests.
Internal features of Scryer Prolog are used to make these tests fast.
2021-11-22 19:55:34 +01:00
Markus Triska faf74519dc use round brackets for meta_predicate/1 declaration 2021-11-20 22:43:53 +01:00
Markus Triska 66becaf91c FIXED: meta_predicate/1 declaration of cond_t/3. 2021-11-20 22:43:33 +01:00
Markus Triska 6610ba67c4 FIXED: meta_predicate/1 declaration of (',')/3 and (;)/3. 2021-11-20 22:43:07 +01:00
Markus Triska e05dd5ebb5 FIXED: meta_predicate/1 declaration for tfilter/3. 2021-11-20 22:39:38 +01:00
Markus Triska cf77f29988 FIXED: meta_predicate/1 declaration for tmember_t/3. 2021-11-20 22:38:27 +01:00
Markus Triska 2be8ed886b FIXED: meta_predicate/1 declaration for tmember/2. 2021-11-20 22:37:59 +01:00
Markus Triska 21e5b0ab52 FIXED: meta_predicate/1 declaration for tpartition/4.
This addresses #1101.
2021-11-20 22:37:31 +01:00
Markus Triska d72cb74ffa mention phrase_to_file/3 which is being prepared in #1100 2021-11-17 21:33:15 +01:00
Markus Triska 3c988d555f 'f' --> f 2021-11-17 21:20:40 +01:00
Markus Triska 5981a65cc2 better layout 2021-11-17 21:20:40 +01:00
Markus Triska 9900762747 use newly available seq//1 from library(dcgs)
I am using seq(Ls) instead of plain Ls so that DCGs can be debugged
declaratively, by generalizing away nonterminals in rules.
2021-11-17 21:20:40 +01:00
Markus Triska bd75e9c184 more elegant solution for parsing numeric arguments, using a DCG 2021-11-16 18:25:35 +01:00
Markus Triska 23df16ecf9 mention the format("~s", [Ls]) pattern and new library(pio) predicates 2021-11-16 18:20:58 +01:00
Markus Triska 57877107d8 use functional notation for all meta_predicate/1 directives 2021-11-11 19:29:32 +01:00
Markus Triska e9b4a99c8f FIXED: meta_predicate/1 directive for call_nth/2.
This addresses #1021.
2021-11-11 19:26:27 +01:00
Markus Triska 3045f327d5 FIXED: instantiation errors for alias(Var) in open/4 options
This addresses #1030.
2021-11-08 22:41:04 +01:00
Markus Triska 2d1f182c49 FIXED: instantiation errors for type(Var) in open/4 options
This addresses #1030.
2021-11-08 22:37:46 +01:00
Markus Triska 0677f51717 add motivation and comments about library(pio) predicates 2021-11-08 22:18:30 +01:00
Markus Triska a9bfeb0e96 use newly available library predicate read_n_chars/3 2021-11-08 21:51:03 +01:00
Markus Triska c44d27fb7f mention pure output predicates in the README 2021-11-08 20:29:39 +01:00
Markus Triska a0a01b8ba7 ADDED: library(format): DCG nonterminal portray_clause_//1, in analogy to format_//2. 2021-11-07 17:47:05 +01:00
Markus Triska e32215e19a express format/3 and portray_clause/2 in terms of phrase_to_stream/2 2021-11-07 17:47:05 +01:00
Markus Triska 893fb0e3cc ADDED: phrase_to_stream/2, writing a list of characters to a stream. 2021-11-07 17:44:45 +01:00
Markus Triska 52af955463 import also partial_string/3 (used in #1071) 2021-11-04 18:50:40 +01:00
Markus Triska 09e886db0b ADDED: read_n_chars/3, reading N characters from a stream. 2021-11-04 18:36:27 +01:00
Markus Triska f668640e3d move error handling to Prolog by using the new '$first_non_octet'/2 2021-11-04 18:35:22 +01:00
Markus Triska afcd44deaa ENHANCED: open/4 to allow opening a stream by specifying stream(S).
This allows switching standard output to binary, using for example:

  ?- current_output(S0), open(stream(S0), write, S, [type(binary)]).

format/3 can then be used to write binary data to S. This is needed
for example when piping binary data to other programs.

This addresses #614, please read the discussion for more information.

The current implementation is very preliminary: Specifically, it works
by destructively modifiying the parameters of the underlying stream,
making it no longer usable in its original mode. Currently, if the
type of standard output is set to binary, then the toplevel no longer
works. Therefore, after writing binary output to standard output,
the program should either halt, or set the stream type back to text.
2021-11-04 18:33:40 +01:00
Markus Triska 8c8c21c63b use partial_string/1 for much quicker test 2021-11-04 00:12:00 +01:00
Markus Triska 55dabbe16a use '$first_non_octet'/2 for much faster domain check 2021-11-04 00:12:08 +01:00
Markus Triska 494bd7b79c ADDED: '$first_non_octet'/2, for much more efficient domain checks in library(crypto).
'$first_non_octet'(Cs, C) <=> C is the leftmost character in the
string Cs whose character code is not in 0..255.
2021-11-04 00:12:00 +01:00
Markus Triska 320ee072e6 ENHANCED: Much faster Base64 encoding/decoding, using partial_string/1 for a quick check. 2021-11-03 22:07:20 +01:00
Markus Triska ded4a75b9c use round brackets for meta_predicate declarations
This is because meta_predicate is not a standard operator.
2021-09-04 22:02:17 +02:00
Markus Triska 9e75eb35a0 ADDED: Rudimentary version of phrase_to_file/2 in library(pio)
See #691 for more information.
2021-09-04 21:13:51 +02:00
Markus Triska 9e13f18463 ENHANCED: Domain error non_empty_list for empty arguments, addressing #1024.
Example:

    ?- format("~d", []).
    %@ caught: error(domain_error(non_empty_list,[]),format_//2)
2021-08-15 12:29:28 +02:00