The R Project SVN R

Rev

Show changed files | Details | Compare with Previous | Blame | RSS feed

Filtering Options

Rev Age Author Path Log message Diff
89928 97 d 6 h luke /trunk/ Out of bounds error conditions for arrays now include the subscript number.  
89130 229 d 9 h maechler /trunk/ x["a"] <- val should not end up calling `EnlargeVector()` and hence losing dim+dimnames  
87901 506 d 9 h ripley /trunk/src/main/ C standard has attributes before return values  
87891 508 d 23 h ripley /trunk/src/ add 2025 BLAS sources  
87184 674 d 21 h maechler /trunk/ -Wall; cosmetic  
87180 675 d 11 h kalibera /trunk/src/ Memcpy, memmove and memset should be only called on valid pointers even when
the size is zero to avoid undefined behavior.		  
86648 788 d 1 h luke /trunk/src/main/ Hide yet more entry points not used and not in installed headers.  
85209 1038 d 22 h luke /trunk/src/main/ Replace two uses of translateChar(PRINTNAME()).  
85146 1047 d 0 h luke /trunk/src/ Add R_typeToChar() ans use to replace type2char(TYPEOF()).
[From OBJSXP branch.]		  
83454 1321 d 6 h ripley /trunk/src/ more moving NORET to the front  
83446 1321 d 10 h ripley /trunk/src/ C23 requires attributes before return values  
82544 1483 d 11 h maechler /trunk/ x[[]] now should fail (w/ classed error) in all cases -- PR#18367  
81380 1685 d 8 h maechler /trunk/ x[<non-integer num>] as documented, i.e. ~= x[as.integer(.)]  
81261 1701 d 2 h maechler /trunk/src/main/ more code consistency between [ and [<-; declaration before use; cosmetic  
81258 1701 d 8 h maechler /trunk/ <no-dimnames-matrix>[ <2-column-char> ] gave "random" garbage}  
81186 1718 d 2 h ripley /trunk/src/main/ add cast to avoid compiler warning  
81144 1724 d 17 h luke /trunk/ Subscript out of bounds errors are now signales as errors of class
subscriptOutOfBoundsError.		  
80533 1863 d 8 h maechler /trunk/src/main/ _() must act on bare strings, not full ternary clauses; from report and patch by Michael Chirico in PR#18125  
79295 2121 d 6 h maechler /trunk/ x[[2^31]] <- v now works, PR#17330  
78184 2300 d 6 h maechler /trunk/ fix PR#17756 and err msg for x[[-Inf]] etc; comment "strange exception"  
78127 2308 d 6 h kalibera /trunk/src/main/ Simplify memory protection.  
75532 2825 d 6 h kalibera /trunk/src/main/ Protect fixes.  
74684 3007 d 9 h ripley /trunk/src/main/ suppress gcc[78] compilation warning  
74660 3012 d 23 h lawrence /trunk/src/main/ avoid quadratic complexity in subassignment by names  
74145 3110 d 1 h maechler /trunk/src/main/ simplification; notably do not recompute xlength(x) in OneIndex()  
74021 3128 d 6 h luke /trunk/src/ Added RAISE_NAMED and used in place of many SET_NAMED calls.  
73729 3177 d 10 h luke /trunk/src/main/ More use of read-only vector data pointers.  
73694 3182 d 8 h luke /trunk/src/main/ Speed improvements in integerSubscript loop.  
73693 3182 d 8 h luke /trunk/src/main/ Remove not needed suplicate calls from makeSubscript.  
73567 3202 d 0 h luke /trunk/src/main/ Reduce calls to DATAPTR via REAL, INTEGER, etc..  
73521 3212 d 3 h luke /trunk/src/main/ OneIndex now uses PRINTNAME in the SYMSXP case instead of STRING_ELT.  
73114 3260 d 10 h maechler /trunk/ fix `x[-n]` segfault (for large n)  
72958 3290 d 10 h maechler /trunk/src/main/ + break: PR#17314  
72506 3393 d 5 h kalibera /trunk/src/main/ Protect fix.  
69588 3922 d 23 h maechler /trunk/ revert large erronous r69587  
69587 3922 d 23 h maechler /trunk/ var() and hence sd() now signal an error for factor arguments  
69576 3926 d 9 h maechler /trunk/src/main/ ECALL3() - for visual simplification and distinguishing *location* of subscripting error  
68947 4005 d 4 h ripley /trunk/src/ use https  
68923 4005 d 10 h ripley /trunk/src/main/ update copyright dates, whitespace cleanup  
68754 4016 d 0 h maechler /trunk/ Patch to logical subscripting from Tomas Kalibera