about summary refs log tree commit diff
path: root/Src
Commit message (Collapse)AuthorAgeFilesLines
* 34623: free history more often if "remetafying"Peter Stephenson2015-02-231-2/+6
|
* 34615 + 34619: Remeta one frame earlierMikael Magnusson2015-02-231-35/+30
|
* 34606: fix up nested arithmetic substitutionPeter Stephenson2015-02-221-2/+18
| | | | | | | | Arithmetic within a parameter substitution is a special case that needs fixing with the introduction of the new Inparmath token. Add test.
* 34604: Work around problem with changes in Meta characters.Peter Stephenson2015-02-221-3/+37
| | | | | | If reading in a file that contains characters that should be metafied but are not, fix up on the fly. Only need when using HIST_LEX_WORDS.
* 34590: queue_signals() around more scopes that manipulate global stateBarton E. Schaefer2015-02-202-1/+18
|
* 34587: ensure multibyte characters don't overflow.Peter Stephenson2015-02-201-0/+8
| | | | | | They could start incorporating tokens, with bad karma. Add test.
* 34568: use META_HEAPDUP when passing dlerror() string to metafy()Barton E. Schaefer2015-02-191-1/+1
|
* 34573: Safer failure to handle command substitutionPeter Stephenson2015-02-191-3/+7
|
* 34570: Another nasty command / math substituion thing.Peter Stephenson2015-02-193-51/+85
| | | | | | Mark arithmetic substitutions with tokens to make sure the substitution go knows what to do. Before it was guessing by counting the parentheses at the end.
* Fix up memory allocation for previous patchPeter Stephenson2015-02-171-4/+5
|
* 34560: Fix $(( that's actually a multiline cmd subst.Peter Stephenson2015-02-163-2/+36
|
* 34530: PRINT_EXIT_VALUE with anonymous functions.Peter Stephenson2015-02-161-2/+12
| | | | Simpler fix, avoiding structural changes.
* 34551: Avoid adding an extra "/" to the target path in cd_try_chdir() when ↵Barton E. Schaefer2015-02-151-2/+4
| | | | the current directory is "/"
* 34543: Prevent crash on garbage bytes inside $(...)Barton E. Schaefer2015-02-142-4/+8
| | | | | | Garbage input (nul bytes, etc.) can cause the $(...) parser to become confused during look-ahead and attempt to back up the input too far. This commit catches the error but does not fix the underlying cause.
* 34546: further $_ with anon function fix.Peter Stephenson2015-02-131-4/+13
| | | | Also add tests.
* 34519: $_ for arguments of anonymous functionPeter Stephenson2015-02-121-1/+4
|
* 34514: Back out 34485, an alternate solution needs to be worked out.Barton E. Schaefer2015-02-122-27/+30
| | | | (Tweaked to keep the unrelated hunk of the E01 test.)
* users/19850: add pattern support to watch variablePeter Stephenson2015-02-111-3/+24
|
* 34490: Fix leak on print -zf/-sf errorMikael Magnusson2015-02-101-1/+17
|
* 34488: Fix use-after-free for print -zf and print -sfMikael Magnusson2015-02-101-6/+12
|
* 34485: More rationalisation for anonymous functions.Peter Stephenson2015-02-092-30/+27
| | | | | | | | Don't attempt to treat as "simple" case as there are too many hidden problems. Pull out some post-execution functions to a common case in execcmd().
* 34466: Fix double unmeta in rm verificationMikael Magnusson2015-02-091-1/+1
|
* 34451: use mmap() also on Mac OS XJun-ichi Takimoto2015-02-031-0/+4
| | | | | Let MAP_ANONYMOUS be an alias to MAP_ANON if the former is not defined.
* 34455: Further gdbm parameter fixes.Peter Stephenson2015-02-031-12/+16
| | | | | | | Fix memory leak with name of parameter. Unset old parameter before opening db as it might close an existing db.
* 34447: fix assignment of key-value array to ztie'd parameter.Barton E. Schaefer2015-02-021-5/+56
| | | | | Add gdbmhashsetfn() for this purpose. Erases and reorganizes the database before bulk assign.
* 34446: add "ztie -r" and "zuntie -u", update documentationBarton E. Schaefer2015-02-011-12/+32
|
* 34439: unset variable about to be tiedPeter Stephenson2015-01-301-2/+15
|
* 34430: parameter fixes for gdbm tied hash.Peter Stephenson2015-01-292-8/+36
| | | | | | | | Probably fix the issue with correct parameter hiding or not hiding. A little extra safety checking. Possibly fixed a memory leak with untying.
* 34403: refine 34399 to avoid settyinfo() when the input descriptor is not a TTYBarton E. Schaefer2015-01-261-2/+2
|
* 34387: Avoid loading the main zsh binary as a moduleMikael Magnusson2015-01-271-1/+3
|
* 34382: Handle NULL return from dlerror() more gracefullyMikael Magnusson2015-01-271-1/+2
|
* Unsetting gdbm tied variable basically works.Peter Stephenson2015-01-261-1/+25
| | | | | | | Allows variables in nested scope. However, explicitly untying a variable doesn't properly expose the scope above.
* 34399: fix polltty thinko from 34365Barton E. Schaefer2015-01-251-1/+1
| | | | Also add missing ChangeLog entry for 34365.
* 34389: fix parsing of ">!" when read from histfile with HIST_LEX_WORDSBarton E. Schaefer2015-01-251-1/+2
|
* 34365: History lockfile backoff: randomised time.Peter Stephenson2015-01-252-14/+123
| | | | | | | Time doubles on each lock failure. zsleep() provides microsecond resolution for sleep; uses nanosleep() if available, else select via means of existing tty poll function.
* 34350, 34353: document zsh/db/gdbm module, clean up a few things in the codeBarton E. Schaefer2015-01-251-20/+29
| | | | Still core dumps at this revision if the tied parameter is forced to be a local and is not untied before end of scope.
* 34369: document error and warning codesDaniel Shahaf2015-01-251-1/+15
|
* 34383: ztrdup() should be dupstring() in new cd codePeter Stephenson2015-01-251-1/+1
|
* 34331: better handling of NULL in cd.Peter Stephenson2015-01-222-7/+12
| | | | | Problem was return from symbolic link expander in weird cases where there file system isn't behaving itself properly.
* 34329: add -S option to which and wherePeter Stephenson2015-01-201-2/+2
|
* users/19751: remove error on failure to close file descriptor by number.Peter Stephenson2015-01-201-1/+6
| | | | | Keep it when closing file descriptor stored in a variable, i.e. explicitly opened by the user.
* 34322: bug with interface to parsestr() etc.Peter Stephenson2015-01-189-26/+46
| | | | | | | Was showing up in places like ${(e)...} where command substitution could reallocate the token string, but actually there was never any guarantee that the lexer wouldn't do that, so this was always a bit iffy.
* 34319: fix alias expansion in history for command substitutionPeter Stephenson2015-01-182-7/+21
|
* unposted: another dubious comment in hist.c fixedPeter Stephenson2015-01-161-2/+2
|
* unposted: improve comment on hist_in_word().Peter Stephenson2015-01-161-2/+6
| | | | | Commit f2a2f28f7 was a bit less clear about how hist_in_word() was used than it could be.
* 32413: turn off history word marking in cmd substPeter Stephenson2015-01-162-2/+22
|
* 34304: improve use of new cmd subst in completionPeter Stephenson2015-01-161-10/+103
|
* 34280: more widespread use of FORCE_FLOAT.Peter Stephenson2015-01-151-1/+12
| | | | Add the case of variables read for use in arithmetic expressions.
* 34287: preprocessor for signal name generation.Peter Stephenson2015-01-151-1/+10
| | | | | | Some gcc header files are difficult for the signames2.awk script to process, so if the preprocessor is gcc give the option -P to strip out the unwanted additions.
* Propagate float/integer type in arithmetic assignment.Peter Stephenson2015-01-121-1/+28
| | | | | | Add test. Mention this and also floating point mod change in README.