summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* @{var1 var2}: bugfix: broken when var2 is regex.Kaz Kylheku2019-07-021-1/+1
| | | | | | | | | | * match.c (h_var): When processing variable bound over a regex or text match, if that datum came from a modifier which was a variable, we must process the substituted modifier, not the original variable symbol. This is a regression that was introduced on November 16, 2011, in commit b86a599bbfcd591f64f31ddfc9ab1a659d39a7c0: "Variable matches can span over function calls.". Reported by user vapnik spaknik.
* @(define): bugfix: incorrect parameter list walk.Kaz Kylheku2019-07-021-2/+2
| | | | | | | | | * parser.y (elem, define_clause): The match_reg_params function is supposed to walk the pattern function parameters, not the argument of the define directive (of which the second one is those arguments). This bug causes spurious unbound variable warnings when function bodies refer to their second or subsequent arguments. Issue reported by user vapnik spaknik.
* Version 219.txr-219Kaz Kylheku2019-07-016-889/+922
| | | | | | | | | | * RELNOTES: Updated. * configure, txr.1: Bumped version and date. * share/txr/stdlib/ver.tl: Likewise. * txr.vim, tl.vim: Regenerated.
* empty: handle buffers.Kaz Kylheku2019-06-301-0/+2
| | | | * lib.c (empty): Handle BUF in switch.
* sha256: C++ fix: void *conversion.Kaz Kylheku2019-06-302-3/+2
| | | | | | | | * chksums/sha256.c (SHA256_update): Change in parameter to const unsigned char *, avoiding the need for a local variable in that type. * chksums/sha256.h (SHA256_update): Declaration updated.
* stdlib: pass env to constantpKaz Kylheku2019-06-292-8/+8
| | | | | | | * share/txr/stdlib/except.tl (catch**): Obtain macro environment and pass to constantp. * share/txr/stdlib/place.tl (last, butlast): Likewise.
* keyparams: bugfix: key list order scrambled.Kaz Kylheku2019-06-291-17/+11
| | | | | | | | | | | | When the keyword parameter list contains a mixture of constant and non-constant default value expressions, the order of the extracted keys is scrambled, so values go to the wrong variables. * share/txr/stdlib/keyparams.tl (sys:build-key-list): Renamed to build-key-list-expr and rewritten to preserve the key order. (:key): Follow rename of build-key-list.
* keyparams: use environment for constantp.Kaz Kylheku2019-06-291-3/+3
| | | | | | * share/txr/stdlib/keyparams.tl (sys:build-key-list): Take macro environment parameter, and pass to constantp. (:key): Pass macro environment to sys:build-key-list.
* defset: small code improvement.Kaz Kylheku2019-06-291-15/+14
| | | | | | * share/txr/stdlib/defset.tl (defset-expander): Move with-gensyms to outermost scope so we can include nvsym in it, avoiding the explicit (nvnsm (gensym)) binding.
* defset: bind new-val-sym to temporary variable.Kaz Kylheku2019-06-283-23/+23
| | | | | | | | | | | | | | | | | | | | | | Users of defset no longer have to ensure that in the store form, the symbol which gives the new value to be stored is inserted only once. * share/txr/stdlib/defset.tl (defset-expander): Transform the store form by inserting a temporary variable using alet. (sub-list, sub-vec, sub-str): These place forms no longer require a local gensym. * txr.1: Updated doc. * tests/012/defset.tl: The expected output for the inc case now incorporates a gensym that comes from the compiled defset macro. Since we can't control that by means of the gensym counter, we resort to extracting it from the expansion itself, then check the test case against a template which incorporates that gensym. We check that the extracted item really is a gensym: it's a symbol with no home package whose name starts with "g".
* seq_info: nullify bugfix.Kaz Kylheku2019-06-281-13/+19
| | | | | | | | | | | | | | | | | | | | | A change in the nullify function to support hash tables has broken various functions which classify an object using seq_info, obtainig a SEQ_HASHLIKE kind, and then work with si.obj using hash functions. But si.obj has been nullified. An example of a broken function is find-max. Basically, this can be attributed to a careless use of nullify in seq_info. The purpose of nullify is to support code which treats any sequence as if it were a list. But seq_info doesn't do that; it classifies sequences and treats them according to their kind. Under seq_info, the only non-list objects that get treated as lists are list-like structures. For these it makes sense to call nullify, in case they have a nullify method. * lib.c (seq_info): Don't unconditionally call nullify on all COBJ objects. Only call nullify on struct objects. If that returns nil, then treat the object as SEQ_NIL; and if it returns an object different from the original, then recurse.
* seq-begin: bugfix: non-lists don't work.Kaz Kylheku2019-06-282-1/+8
| | | | | | | | | | | | | * lib.c (seq_begin): Do not null out si->inf.obj; it's needed for accessing hashes and vector-like objects. This bug means that seq-begin iteration has only worked correctly for lists. The original motivation was not to have spurious retention of the head of a lazy list, which is hereby reintroduced. But iterators can be rewound. Let's just document this away and leave it as a to-do item. * txr.1: Document the limitation of seq-begin w.r.t. lazy lists.
* doc: put-buf/fill-buf clarification of pos.Kaz Kylheku2019-06-271-0/+2
| | | | | * txr.1: The pos argument of put-buff and fill-buf is a position within the buffer (not the stream).
* doc: update syntactic places.Kaz Kylheku2019-06-271-3/+10
| | | | | * txr.1: Update the list of syntactic places to list all the new places that have been added, plus a few additional edits.
* in: allow hash with keyfun and testfun.Kaz Kylheku2019-06-252-4/+31
| | | | | | | * lib.c (in): A simple check and fallthrough lets this function process hash tables more generally in this function. * txr.1: Documented.
* in: use seq_infoKaz Kylheku2019-06-251-26/+28
| | | | | * lib.c (in): Keep the existing specialized cases, but use seq_info in the fallback.
* Factor function name into self variable.Kaz Kylheku2019-06-251-10/+15
| | | | | * lib.c (take, take_while, take_until, drop_while, drop_until): Move repeated function name into self variable.
* drop-{while,until}: convert to seq_info.Kaz Kylheku2019-06-251-16/+16
| | | | | | * lib.c (drop_while, drop_until): Use seq_info, so these functions work with all sequences. Thus now for instance [drop-while zerop #b'0000f00d'] yields #b'f00d'.
* empty: handle carray.Kaz Kylheku2019-06-251-0/+2
| | | | * lib.c (empty): Add carray sub case to COBJ case.
* nullify: handle carray and hashes.Kaz Kylheku2019-06-251-0/+4
| | | | | * lib.c (nullify): Add carray and hash subcases into the COBJ case.
* doc: clarify nullify.Kaz Kylheku2019-06-251-5/+26
| | | | | | * txr.1: Clarify that nullify also just returns its argument if it is not a sequence at all. Clarify that nullify works on objects via the nullify method only.
* Handle buffers in list collector functions.Kaz Kylheku2019-06-251-0/+19
| | | | | | * lib.c (nullify, list_collect, list_collect_nconc, list_collect_append, list_collect_nreconc, list_collect_revappend): Handle buffer type.
* list_collect: handle objects.Kaz Kylheku2019-06-252-1/+11
| | | | | | | | | * lib.c (list_collect): Handle sequence-like COBJ objects. We can add an item to using their respective replace functions. (replace_obj): Change to external linkage. * lib.h (replace_obj): Declared.
* Code clean-up in list collector functions.Kaz Kylheku2019-06-251-19/+23
| | | | | | | * lib.c (list_collect, list_collect_append, list_collect_revappend): Use local variables to avoid repeated expressions. (list_collect_nconc): Only call nullify in necessary cases.
* seqp: expand definition of sequences.Kaz Kylheku2019-06-252-13/+18
| | | | | | | * lib.c (seqp): Use seq_info to classify the object as a sequence. * txr.1: Update description of seqp.
* replace: fix strange diagnostic from bad fallthrough.Kaz Kylheku2019-06-241-2/+2
| | | | | | | | | * lib.c (replace): If a COBJ is passed to replace which doesn't support the operation, we wrongly pass it to replace_buf because the BUF case was added into the fallthrough pass. The end result is that length_buf blows up on the object, resulting in a strange diagnostic. The BUF case must be moved above COBJ.
* New: CRC-32 functions.Kaz Kylheku2019-06-245-1/+243
| | | | | | | | | | | | * Makefile (OBJS): Add chksums/crc32.o. * chksum.c (crc32_stream, crc32): New functions. (chksum_init): Register crc32-stream and crc32 intrinsic functions. * chksums/crc32.c, chksums/crc32.h: New files. * txr.1: Documented.
* lnew: two regressions fixed.Kaz Kylheku2019-06-241-2/+2
| | | | | | | | | * share/txr/stdlib/struct.tl (sys:new-expander): The syntax (lnew (type boa args ...)) is not being handled at all due a typo in the case. Moreover, the syntax (lnew type slot0 val0 slot1 val1 ...) is broken due to a missing unquote on the type gensym in the output template.
* * Makefile (OBJS): New objects chksum.o and chksums/sha256.o.Kaz Kylheku2019-06-238-1/+585
| | | | | | | | | | * chksum.c, chksum.h, chksums/sha256.c, chksums/sha256.h: New files. * lib.c (init): Call chksum_init. * txr.1: Documented. * LICENSE: Add SHA-256 copyright notice.
* doc: move buffers out of dynamic lib section.Kaz Kylheku2019-06-231-1341/+1343
| | | | | | | | * txr.1: Buffers are now documented after strings and vectors. Buffer streams are in the I/O section. Note: the diff between this and the previous commit needs to be generated with git diff --minimal. Otherwise it looks like over 30,000 lines of changes.
* base64: integrate with buffers.Kaz Kylheku2019-06-215-8/+54
| | | | | | | | | | | | | | | * filter.c (base64_encode): Handle an object of either string or buffer type by using make_byte_input_stream. (base64_decode_buf): New function. (filter_init): Register base64-decode-buf intrinsic. * filter.h (base64_decode_buf): Declared. * stream.c (make_byte_input_stream): New function. * stream.h (make_byte_input_stream): Declared. * txr.1: Updated.
* streams: use ~s and ~a properly in errors.Kaz Kylheku2019-06-211-27/+27
| | | | | | | | | | | | | Objects should almost always be printed as ~s. If a bad object is printed as ~a, junk could result on the terminal if it's a buffer, for instance. * stream.c (stdio_maybe_read_error, stdio_maybe_error, stdio_close, pipe_close, parse_mode, normalize_mode, normalize_mode_no_bin, cat_unget_byte, cat_unget_char, unget_byte, open_directory, open_file, open_tail, open_command, open_process, run, remove_path, rename_path): Revise conversion specifiers in error message formatting.
* doc: minor grammar problems in Lisp CompilationKaz Kylheku2019-06-211-2/+2
| | | | | * txr.1: Two issues fixed while proof-reading: missing "a" article and a missing plural.
* doc: cure that-is-itis.Kaz Kylheku2019-06-211-3/+3
| | | | | * txr.1: In two places, fix an awkward "; that is, has ..." and "; that is, is ..." grammatical curiosity.
* doc: dump-compiled-objects: wrong number.Kaz Kylheku2019-06-201-1/+1
| | | | * txr.1: "following three expressions" should be "two".
* dump-to-tlo: move to sys package.Kaz Kylheku2019-06-201-1/+1
| | | | | | * share/txr/stdlib/compiler.tl (usr:dump-to-tlo): This function has no reason to be named by a usr package symbol. Renaming to sys:dump-to-tlo.
* at-exit-do-not-call: wrong return value.Kaz Kylheku2019-06-201-1/+1
| | | | | * sysif.c (at_exit_do_not_call): Reverse Boolean polarity of return value to match documentation.
* Version 218.txr-218Kaz Kylheku2019-06-197-947/+1000
| | | | | | | | | | | | * RELNOTES: Updated. * configure, txr.1: Bumped version and date. * share/txr/stdlib/ver.tl: Likewise. * txr.vim, tl.vim: Regenerated. * protsym.c: Likewise.
* ash: refactor in light of gc bug.Kaz Kylheku2019-06-191-24/+38
| | | | | | | | | * arith.c (ash): Introduce smaller scopes and move variable initializations close to their use. The NUM case of argument a no longer falls through to the BGNUM case, and doesn't cons up a bignum object. Rather, a temporary mp_int is allocated, used and freed. The gc_hint added in the previous commit is now gone.
* ash: gc problem.Kaz Kylheku2019-06-191-0/+1
| | | | | | | | | | | | | | | On 32 bit x86 Solaris 10, with gcc 4.9.0, this issue caused a miscompilation of the pset macro, due to ash abruptly returning zero, causing the op-code field of an instruction to be NOOP rather than GCALL. I'm committing this fix just for reference; I will immediately replace it with a refactoring of the function. * arith.c (ash): Add a gc_hint to prevent the a bignum from being reclaimed if the b = make_bignum() triggers gc. This happens in the case when the previous case computes a = make_bignum() and falls through.
* internal debug support: reduce noise from sweep_one.Kaz Kylheku2019-06-191-1/+1
| | | | | | | | * gc.c (sweep_one): If the swept object matches the break_obj, only log it if the object isn't free. Otherwise if break_obj is sitting on the free list, we will uselessly report on it on every sweep operation, which is very often under --gc-debug torture operation.
* packages: generational gc bug.Kaz Kylheku2019-06-191-4/+4
| | | | | | | | | | | | * lib.c (make_package_common): The way the two hashes are assigned into the new package here is not correct. The problem is that the first make_hash can trigger gc. Then it is possible that the package object will move into the mature generation, after which the assignment of the second package is a wrong-way assignment requiring the set macro. Instead of bringing in that macro, the obvious way to solve it is to just allocate the hashes first, and then the package: exactly the way we build a cons cell from existing values.
* make-struct-type: fix far-fetched gc/stability issue.Kaz Kylheku2019-06-191-2/+2
| | | | | | | * struct.c (make_struct_type): The length and plus operations can cons if the number of static slots doesn't fit into a fixnum. Let's move that operation out of the sensitive object construction region.
* structs: gc bug due to derived hook.Kaz Kylheku2019-06-191-3/+5
| | | | | | | | | | | * struct.c (make_struct_type): The static slot lookup for for a derived method in the supertype can potentially allocate a cons cell when there is a cache miss, which can trigger gc. So we can't be doing this while a the new type object object is partially initialized, because then the garbage collector will visit uninitialized memory. We move the derived method lookup down farther to where that value is needed to do the derived call, just before we return the new type.
* buffers: bug: wrong object in format arguments.Kaz Kylheku2019-06-181-1/+1
| | | | | | | | | * buf.c (buf_strm_seek): index is a BSD function; we want to refer to the local variable npos here. This was fortuitiously caught by GNU C++ which is obtains multiple overloaded declarations of index from somewhere, and so in this variable argument list situation thinks that there isn't enough context to resolve the overload.
* logcount: crash in 64 bit build.Kaz Kylheku2019-06-181-1/+1
| | | | | | | * arith.c (logcount): NUM case doesn't have a return statement in 64 bit build, statement causing fall-through to BGNUM case where the integer object is treated as a bignum. This bug has existed in the function ab initio.
* New function: bitset.Kaz Kylheku2019-06-183-1/+122
| | | | | | | | | | | * arith.c (bitset_s): New symbol variable. (bitset): New function. (arith_init): bitset_s initialized, bitset intrinsic registered. * lib.h (bitset): Declared. * txr.1: Documented bitset function and method.
* carray-replace: use seq_iter.Kaz Kylheku2019-06-181-15/+5
| | | | | * ffi.c (carray_replace): Use generic sequence iteration for range replacement instead of separate list/vector code.
* sub-list, sub-vec, sub-str: becoming accessors.Kaz Kylheku2019-06-173-5/+107
| | | | | | | | | | | * lisplib.c (defset_instantiate): Add sub-list, sub-vec and sub-str symbols as autoload triggers for defset.tl. * share/txr/stdlib/defset.tl (sub-list, sub-vec, sub-str): New syntactic places defined with defset. * txr.1: sub-list, sub-vec and sub-str documented as accessors.
* expander: relax sys:setq and lisp-1 stringency.Kaz Kylheku2019-06-171-23/+21
| | | | | | | | | | | | | | | | | | | | | | | | The motivation here is that there are behaviors in the expander which hinder symbol-macro-based renaming techniques. For instance (expand '(symacrolet ((x y)) (sys:setq x x))) throws. The right hand side of the setq is fine, but the left hand one is a forbidden symbol macro. Yet, we would just like this to expand to (sys:setq y y). The original idea was that sys:setq occurs as a result of macro-expansion. Therefore, if its argument is a symbol macro, something must be wrong; it didn't get expanded. That reasoning is wrong in the face of explicit expansion techniques that make multiple expansion passes. For instance (set a b) can become something like (sys:setq #:g0005 #:g0007) when the intent that in another round of renaming these gensyms will be defined as symbol macros which perform one more renaming. * eval.c (expand_lisp1_value, expand_lisp1_setq): Macro-expand the symbol and work with the expanded one. We still keep the check for a symbol macro; these situations can arise if a symbol macro cannot expand due to circularity. (do_expand): When checking sys:setq for a bad symbol or symbol macro, work with the expanded argument.
generated by cgit v1.2.3 (git 2.25.1) at 2025-03-21 21:16:53 +0000