summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* tests for string range length.Kaz Kylheku2024-12-151-0/+12
| | | | * tests/012/seq.tl: New tests.
* lib: fix g++ warning in map_common.Kaz Kylheku2024-12-111-1/+1
| | | | | | * eval.c (map_common): use the all_zero_init macro, defined differently for C and C++ for initializing a struct to all zero.
* quips: jab at Weller.Kaz Kylheku2024-12-081-0/+1
| | | | * stdlib/quips.tl (%quips%): New one.
* switch to .txr-profile and .txr-historyKaz Kylheku2024-11-143-19/+74
| | | | | | | | | | | | | | | | | | | | The profile and history files should have used hyphens from the beginning. Let's switch to that but continue to work with the old files if present, as an obsolescent feature. * parser.c (open_txr_file): Treat files with .txr-profile suffix as Lisp. (load_rcfile): Arguments rearranged. This function now needs the home directory and the existence test function, but does not need the profile file name. It tries .txr-profile first, then .txr-profile. (repl): Call load_rcfile in the new way. Try two history files: first .txr-history and then .txr_history. Remember which one was used so the same one is saved. If neither file existed at startup, then save new history into .txr-history. * txr.1: Documented.
* ffi: bug: flexible structure size calculation.Kaz Kylheku2024-10-071-3/+2
| | | | | | | | | | | * ffi.c (make_ffi_type_struct): We must calculate the size of a flexible structure the way GCC does it. We cannot simply truncate it at the offset of the member. Rather, the size is calculated in the usual way. The alignment of the array is taken into account for the purpose of determining what is the most aligned member of the structure, and then padding is added, if required. Thus, the size may exceed the offset of that member.
* copy: now handles range objects.Kaz Kylheku2024-10-014-22/+20
| | | | | | | | | | | | | | | | Ranges are iterable, denoting abstract sequences. The copy function now copies a range by constructing the array. This is useful when copy is used for the purpose of obtaining a mutable copy. For example, (shuffle 0..100) will now work, returning a shuffled vector of the integers from 0 to 99. * lib.c (copy): Handle RNG case via vec_seq. * tests/012/seq.tl, * tests/012/sort.tl: New test cases. * txr.1: Documented. Documentation for the copy function improved.
* refset, replace: adjust diagnostic for unsupported object.Kaz Kylheku2024-09-301-2/+2
| | | | | | | * lib.c (refset, replace): Word the bad object diagnostic in terms of it not being a modifiable sequence. This covers cases when the object is something abstractly iterable, like a range. We don't want to say that it's not a sequence.
* regex: closure operations don't output epsilon states.Kaz Kylheku2024-09-151-6/+12
| | | | | | | | | | | | * regex.c (nfa_closure, nfa_move_closure): Do not add epsilon states to the output array. We only need to add them to the stack for the spanning traversal. Epsilon states are not real states; they are just a representation of the concept of transitioning to multiple states at the same time. When we add them to the output, they just end up being ignored when nfa_move_closure is called again on that set, since it only cares about states that have real transitions for a character.
* regex: bugfix: not taking full advantage of REGM_MATCH_DONE.Kaz Kylheku2024-09-151-2/+2
| | | | | | | | | | | | * regex.c (nfa_has_transitions): The logical disjunction here is wrong. We would like to test whether a state has transitions if it is not an epsilon state. The code which uses this macro doesn't care about epsilon states, even though they have transitions; those are squeezed out by transitively closure. The wrong condition here makes the code think that a NFA set has transitions when it does not, preventing the result code REGM_MATCH_DONE to be produced which can spare a character from being consumed from a stream.
* regex: misleading comment.Kaz Kylheku2024-09-151-1/+2
| | | | | | * regex.c (mfa_move_closure): Fix misleading wording in comment. The state which matches the character (has a transition for it) is not the one added to the move set.
* read-until-match: fix regression.Kaz Kylheku2024-09-143-2/+21
| | | | | | | | | | | | | | | | | | | | | | | | | Commit 9aa751c8a4f845ef2d2bba091c81ffeded941afd broke things. This fix affects the function read-until-match, scan-until-match and count-until-match which share implementation. * regex.c (scan_until_common): In the REGM_MATCH_DONE and REGM_MATCH cases, we must push the character onto the local stack, before doing the match = stack assignment. Otherwise, it's possible that the stack is empty and so no match is recorded. The REGM_FAIL case will then behave as if no match was found, consuming a character and continuing. * txr.1: Codify an existing behavior: only non-empty matches for the regex are considered by read-until-match. * tests/015/regex.tl: New file. I am amazed to discover that we don't seem to have a test suite for regexes at all. Putting the tests here which confirm this fix and provide coverage for some edge cases in read-until-match.
* quips: new one.Kaz Kylheku2024-09-031-0/+1
| | | | * stdlib/quips.tl (%quips%): New entry.
* mapcar: introduce map as a synonym.Kaz Kylheku2024-08-142-2/+14
| | | | | | | * eval.c (eval_init): Bind new map symbol to the same function as mapcar. * txr.1: Documented.
* loongarch64: fix assembler warning.Kaz Kylheku2024-08-111-2/+2
| | | | | | | | | | This issue was spotted by Alpine Linux people; I'm upstreaming their patch (originally authored by a user named huajingyun01). * jmp.S (jmp_save, jmp_restore): Use recommended register names on loongarch64: a0 and a1 rather than v0 and v1.
* genman: change to new PayPal button code.Kaz Kylheku2024-08-091-6/+4
| | | | * genman.txr: PayPay form replaced.
* Version 296.txr-296Kaz Kylheku2024-08-077-1067/+1149
| | | | | | | | | | | | | | * RELNOTES: Updated. * configure (txr_ver): Bumped version. * stdlib/ver.tl (lib-version): Bumped. * txr.1: Bumped version and date. * txr.vim, tl.vim: Regenerated. * protsym.c: Regenerated.
* keep-if, remove-if, keep-keys-if: mapfun argument.Kaz Kylheku2024-07-305-52/+90
| | | | | | | | | | | | | | | | | | | | | | | | | | | | We introduce a mapfun argument to these functions so that they can additionally transform the accumulated values. The keep-keys-if function is now implemented through the same helper function as keep-if but with the mapfun argument defaulting to a copy of the keyfun argument. * eval.c (eval_init): Update registrations of remove-if, keep-if and keep-keys-if to new arities of C functions. * lib.c (rem_if_impl): Implement new optional mapfun parameter. (remove_if, keep_if): Add mapfun parameter. (keep_keys_if): Implement via rem_if_impl, and add mapfun argument. We do the defaulting of keyfun here, so that we can then use that argument's value to default mapfun. * lib.h (remove_if, keep_if, keep_keys_if): Declarations updated. * tests/012/seq.tl: Couple of test cases exercising mapfun argument of keep-if and remove-if. * txr.1: Documented.
* doc: touch up description of trivial patterns.Kaz Kylheku2024-07-271-5/+9
| | | | | | | * txr.1: patterns are also nontrivial if they contain quasiquotes that contain nontrivial patterns, and quasiliterals that contain operators. Also, we italicize the term trivial pattern as well as nontrivial.
* doc: missing subject in @(hash ...) pattern section.Kaz Kylheku2024-07-271-1/+1
| | | | * txr.1: Fix sentence which needs a subject, "the operator".
* doc: avoid quasiliteral in reference to non-strings.Kaz Kylheku2024-07-271-6/+6
| | | | | | | * txr.1: Fix a number of places that use quasiliteral in reference to a structural quasiquote. In the TXR manual, quasiliteral syntax is the `...` that produces strings, short for quasistring literal. ^(...) is a quasiquote.
* lib: get rid of internal rewindable iter nonsense.Kaz Kylheku2024-07-253-82/+19
| | | | | | | | | | | | | | | | | | | | Iterator rewinding is only used by the three functions isec, isecp and diff, which can easily just re-initialize the iterator. * lib.c (seq_iter_rewind): Static function removed. (seq_iter_init_with_info): Remove support_rewind argument, and adjust all code referencing it on the assumption that it's zero. (seq_iter_init_with_rewind): Static function removed. (seq_iter_init, iter_begin, iter_reset, nullify, find, rfind): Drop last argument from seq_iter_init_with_info. (diff, isec, iescp): Use seq_iter_init rather than seq_iter_init_with_rewind. Instead of seq_iter_rewind, just reinitialize the iterator. * lib.h (seq_iter_init_with_info): Declaration updated. * eval.c (tprint): Drop last argument from seq_iter_init_with_info.
* zip: more permissive implementation.Kaz Kylheku2024-07-245-22/+58
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | zip and transpose should allow non-character data when the leftmost column is a string, falling back on making lists, like seq_build. We can't use seq_build as-is because of the special semantics of transpose/zip with regard to strings. We introduce a "strcat" variant of seq_build for this purpose. * lib.c (seq_build_strcat_add): New static function. (sb_strcat_ops): New static structure like sb_str_ops, but with seq_build_strcat_add as the add operation, which allows string arguments to be appended to the string rather than switching to a list. (seq_build_strcat_init): New function. * lib.h (seq_build_strcat_init): Declared. * eval.c (zip_strcat): New static function; uses seq_build_strcat_init. (zipv): Only recognize strings specially; all else goes through the existing default case. Strings use zip_strcat. * tests/012/seq.tl: New test case. * txr.1: Describe special semantics of zip/tranpose; previously only documented in one example. Clarify that the rows are only sequences of the same kind as the leftmost column if possible, otherwise lists. Remove text which says that it's an error for the other columns to contain non-string, non-character objects if the leftmost column is a string.
* make-like: use seq_build.Kaz Kylheku2024-07-233-50/+31
| | | | | | | | | | | * lib.c (make_like): Simplify implementation using seq_build, which also lets it handle more cases. * tests/012/seq.tl: New tests. Some existing test fixed, including one for tuples*. * txr.1: Documentation updated: mainly that make-like doesn't strictly require a list argument.
* New function: seq-like.Kaz Kylheku2024-07-233-5/+49
| | | | | | | | | | | | * eval.c (zip_fun): Renamed to seq_like. (zipv): Follow rename of zip_fun. (eval_init): Register seq-like intrinsic. * tests/seq.tl: Some tests for make-like and seq-like, revealing a difference: make-like needs to be rewritten to use seq_build. * txr.1: Documented.
* doc: certain JSON-related variables are specials.Kaz Kylheku2024-07-221-3/+3
| | | | | | | * txr.1: The *print-json-format*, *read-bad-json* and *read-json-int* variables are special (i.e. dynamically scoped) variables, and so reaquire the Special Variable heading, not Variable.
* json: new *read-json-int* variable.Kaz Kylheku2024-07-226-18/+74
| | | | | | | | | | | | | | | | | | | | | * parser.h (struct parser): New member, read_json_int. * parser.c (read_json_int_s): New symbol variable for *read-json-int* symbol. (parser_common_init): Look up value of *read-json-int* and store in read_json_int struct member. (parse_init): Initialize read_json_int_s with interned symbol and also register the dynamic variable. * parser.l (grammar): Extend the {JNUM} rule to check the read_json_int flag and produce an integer value if the lexeme does not contain a decimal point, e or E. * tests/010/json.tl: New tests. * txr.1: Documented. * lex.yy.c.shipped: Regenerated.
* parser: remove some wasteful string object allocations.Kaz Kylheku2024-07-224-211/+236
| | | | | | | | | | | | | | | | * lib.c (int_str_wc): New function, made out of int_str. This can be used by the parser to work with a wchar_t * string without having to create a string object. (int_str): Implemented in terms of int_str_wc. * parser.l (grammar): Remove string_own calls from numerous rule bodies that use int_str to return a number. These rules now capture the wchar_t string, pass it to int_str_wc and then immediately free it. Whereas string_own allocates an extra object and leaves it to the garbage collector. * lex.yy.c.shipped: Regenerated.
* vim: #: and : notation in symbols not handled right.Kaz Kylheku2024-07-211-1/+1
| | | | | | | | | * genvim.txr (tl_ident): Move the pattern which matches the # and #: symbols above the more general one. Because, remember, because Vim's syntax match definitions work such that the last definition which matches wins, even if it's not the longest match. Not regenerating the txr.vim and tl.vim files; that will be done at release time.
* doc: fix miscoded backslashes in examples.Kaz Kylheku2024-07-201-5/+5
| | | | | | * txr.1: In examples for append, whereq and tuples*, fix backslashes not encoded as the \e sequence, causing improper rendering.
* oop: special methods to handle missing slots.Kaz Kylheku2024-07-194-8/+158
| | | | | | | | | | | | | | | | | | | | | * struct.h (slotset_s, static_slot_s, static_slot_set_s): New symbol variables declared. (enum special_slot): New enum symbols: slot_m, slotset_m, static_slot_m, static_slot_set_m. * struct.c (slotset_s, static_slot_s, static_slot_set_s): New symbol variables. (special_sym): Associate new symbols with new enums. (struct_init): Intern slotset, static-slot and static-slot-set symbols, initializing the variables. Change the registrations of the same-named functions to use the variables. (slot, maybe_slot, slotset, static_slot, static_slot_set): In the no-such-slot case, check for the special method and call it. * tests/012/oop.tl: New tests. * txr.1: Documented.
* struct: slot warning only for bindable symbols.Kaz Kylheku2024-07-171-1/+2
| | | | | | * stdlib/struct.tl (sys:check-slot): Don't issue the diagnostic "<obj> isn't the name of a struct slot" for slots that are not bindable symbols like obj."abc".
* doc: fix misleading claim about (. pattern).Kaz Kylheku2024-07-161-1/+3
| | | | | | * txr.1: We cannot say that (. pattern) is not a list pattern, since it is indistinguishable from pattern, which could itself be a list pattern.
* doc: misleading info about macro param lists.Kaz Kylheku2024-07-161-5/+4
| | | | | | | | | * txr.1: In the "Comparison to Macro Parameter Lists" section, which compares structural pattern matching to macro parameter lists, we remove the outdated claim that all positions in a macro parameter pattern must bind a variable. This has not been true since t has been supported as a way to support binding variables.
* New functions: find-maxes and find-mins.Kaz Kylheku2024-07-165-2/+86
| | | | | | | | | | | * eval.c (eval_init): New intrinsic functions find-maxes and find-mins. * lib.[ch] (find_maxes, find_mins): New function. * tests/012/seq.tl: New tests. * txr.1: Documented.
* json: new special var *print-json-type*.Kaz Kylheku2024-07-124-11/+45
| | | | | | | | | | | | | | | | | | | | This variable controls whether we emit the "__type" key for structures. * lib.c (out_json_rec): React to the new variable, via the flag in the json_opts structure: include the "__type" key only if it is requested. (out_json, put_json): Initialize the type flag in the josn_opts according to the *print-json-type* dynamic variable. * stream.c (print_json_type_s): New symbol variable. (stream_init): print_json_type_s initialized, and corresponding special variable registered, with intial value t. * stream.h (struct json_opts): New bitfield member, type. (print_json_type_s): Declared. * txr.1: Documented.
* New funtion related to where function.Kaz Kylheku2024-07-115-0/+176
| | | | | | | | | | | | | | * eval.c (eval_init): register intrinsics wheref, whereq, whereql and wherequal. * lib.c (wheref_fun): New static function. (wheref, whereq, whereql, wherequal): New functions. * lib.h (wheref, whereq, whereql, wherequal): Declared. * tests/012/seq.tl: New tests. * txr.1: Documented.
* doc: partition, split, split*: clarifications about indicesKaz Kylheku2024-07-101-2/+12
| | | | | | | | * txr.1: Clarify what repeated values mean in partition, since they are allowed. For split/split*, clarify that indice have to be strictly increasing after negative indicates are displaced by the sequence length, and that the behavior is unspecified otherwise.
* partition, split, split*: bug handling negative indices.Kaz Kylheku2024-07-102-22/+35
| | | | | | | | | | | | | | | | * lib.c (partition_func, split_func, split_star_func): When negative indices occur after the sequence has already been shortened, the conversion to positivce must take into account the base. This must be added so that the positive index produced is relative to the original length of the input sequence. When index_rebased is calculated, the base is subtracted out again. If we based the positive index off the shortened length, it's as if we are subtracting base twice. * tests/012/seq.tl: Dubious test cases for split* are replaced with the new results that make sense. Additional test cases are added which cover this issue, for not only split* but split and partition.
* split, split*: fix poor behavior for beyond-length indices.Kaz Kylheku2024-07-102-30/+32
| | | | | | | | | | * lib.c (split_func, split_Star_func): Ignore indices greater than the length of the sequence, the same as negative indices are ignored which don't become nonnegative after adding the length. * tests/012/seq.tl: Fix questionable test cases, which now confirm the right behavior.
* partition: add negative index tests.Kaz Kylheku2024-07-101-0/+36
| | | | * tests/012/seq.tl: New tests.
* split**: split for far negative indices.Kaz Kylheku2024-07-102-1/+33
| | | | | | | | | * lib.c (split_star_func): In empty index case, convert sequence via sub(seq, zero, t), so that ranges are properly expanded. This was done in split_func and partition_func in recent commits. * tests/012/seq.tl: Test cases added.
* split*: add missing tests.Kaz Kylheku2024-07-101-0/+125
| | | | | | * tests/012/seq.tl: Add tests for split* that were lost in some editing. Corresponding tests exist for split already and for partition.
* split: fix for far negative indices.Kaz Kylheku2024-07-102-1/+29
| | | | | | | | | * lib.c (split_func): In empty index case, convert sequence via sub(seq, zero, t), so that ranges are properly expanded. This was done in partition_func in the previous commit. * tests/012/seq.tl: Test cases added.
* split, split*, partition: tests, fixes.Kaz Kylheku2024-07-102-2/+412
| | | | | | | | | | | | | | * lib.c (partition_func): In empty index list case, run the sequence through sub(seq, zero, t) so that ranges are expanded: e.g. 1..3 becomes (1 2). The corresponding code in split_func and split_star_func also needs this fix, but the current test cases don't reproduce a problem. (partition_split_common): Likewise here. * tests/012/seq.tl: Tests for split, split* and partition. Some tests have questionable results. We accept these as they are for now; will address these.
* partition, split, split*: infinite looping regression.Kaz Kylheku2024-07-081-6/+9
| | | | | | | | | | | | | | This is a bug introduced in 9cfa3435 on 2024-02-24. The underlying cause is lack of test coverage for these functions. * lib.c (partition_func, split_func, split_star_func): The original code iterated through the indies using the pop macro, thus extracting the next index and stepping in one step. The iter_begin rewrite wrongly moved the iter_step into one of the cases. The index iteration must be stepped in the case where the loop is continued vi continue, otherwise an infinite loop results.
* sub: don't produce an iterator.Kaz Kylheku2024-07-072-4/+4
| | | | | | | | | | | | | | | | | | | | | Having the sub function yield an iterator in some cases is a defective requirement, causing problems like this: 1> (partition 1..10 '(2 3)) ((1 2) (3) #<seq-iter: a24c380>) With fix: 1> (partition 1..10 '(2 3)) ((1 2) (3) (4 5 6 7 8 9)) * lib.c (sub_iter): When the interval is open and we are operating on a sequence via iter-begin, do not return an iterator. Convert it to a lazy list. Not subjecting this to -C compat flag; I can't imagine anyone writing code to depend on this, rather than stepping around it as a bugx. * txr.1: Documentation updated.
* json: fix flat-p argument in put-json and put-jsons.Kaz Kylheku2024-07-062-33/+40
| | | | | | | | | | | | | | | | The flat-p flag is not being passed through ther recursion. Some of the formatting code emits newlines unconditionally. Instead of passing down the enum json_fmt, we pass down a new structure which carries the flat-p flag also. * stream.h (struct json_opts): New struct. * lib.c (out_json_rec): Take a struct json_opts argument instead of enum json_fmt. Handle the flat flag to avoid generating line breaks. (out_json, put_json): Prepare json_opt structure and pass to out_json_rec.
* json: support printing structs in JSON format.Kaz Kylheku2024-07-062-0/+78
| | | | | | | * lib.c (out_json_sym): New static function. (out_json_rec): Handle structp. * txr.1: Documented.
* hash: fix: equal hashes being reduced modulo NUM_MAX.Kaz Kylheku2024-07-042-5/+1
| | | | | | | | | | Almost exactly 6 years ago, commit 612f4f57e made hashes use the full width of the ucnum type. However, several reductions of the form hash &= NUM_MAX were forgotten. * hash.c (hash_hash_op): Eliminate reductions modulo NUM_MAX. * struct.c (struct_inst_hash): Likewise.
* regex: eliminate unnecessary stack push.Kaz Kylheku2024-07-031-1/+0
| | | | | | | | * regex.c (scan_until_common): in the REGM_MATCH case, there is no need to push the current character into the unget stack. That character is not supposed to be pushed back, and it won't be because it's below the match point; the stack node just ends up recycled at the end of the function.
generated by cgit v1.2.3 (git 2.25.1) at 2025-03-31 09:14:23 +0000