scc%netscape.com
|
e6aaa0f8f5
|
making string conversions explicit; this checkin stops the back-sliding before we turn on NEW_STRING_APIS.
|
2000-04-27 20:37:54 +00:00 |
scc%netscape.com
|
0ef80f2c28
|
making promises work
|
2000-04-27 20:01:28 +00:00 |
mjudge%netscape.com
|
47511cf02b
|
scc string fixes
|
2000-04-26 00:55:37 +00:00 |
pavlov%netscape.com
|
29d75b355b
|
add typename to make linux compiler happy r=scc
|
2000-04-23 05:54:36 +00:00 |
scc%netscape.com
|
f1766992d8
|
Added an additional permitting and denying signatures for |NS_ConvertASCIItoUCS2|
|
2000-04-23 01:03:28 +00:00 |
tbogard%aol.net
|
76e8db8e09
|
Moved the implementation of AssignWithConversion to the cpp file since Linux and Windows didn't like it inline in the h file. r=scc
|
2000-04-22 09:17:20 +00:00 |
scc%netscape.com
|
08376156e3
|
Added a signature previously covered by |Assign(const nsStr&)|, now |Assign( const nsString& )|.
|
2000-04-22 08:36:28 +00:00 |
pinkerton%netscape.com
|
b138f71e20
|
Fix build bustage r=scc.
|
2000-04-21 07:34:06 +00:00 |
scc%netscape.com
|
c9db033ed7
|
there were a bunch of routines that returned references to this and were abused as much as they were used correctly. now they return |void|
|
2000-04-21 06:59:14 +00:00 |
scc%netscape.com
|
2eb9fa2d3c
|
Fixed |nsPromiseSubstring::GetReadableFragment|, which had been walking off the end. Re-implemented |copy_string| in terms of the new char-source/char-sink stuff that beard and I came up with. Factored some code into |SameFragment|. Did some general shuffling and cleanup.
|
2000-04-21 01:25:19 +00:00 |
scc%netscape.com
|
dd7694898d
|
Added |nsCharSourceTraits| and |nsCharSinkTraits| to better facilitate copying from arbitrary char sources into arbitrary char consumers
|
2000-04-21 01:20:55 +00:00 |
scc%netscape.com
|
d03cfe6e83
|
added |nsWritingIterator<T>::write| to satisfy the char-source/char-sink stuff beard and I put in.
|
2000-04-21 01:19:40 +00:00 |
scc%netscape.com
|
5c3b3872c2
|
needed to make the implementation of |EqualsIgnoreCase(nsIAtom*, ...)| available.
|
2000-04-17 14:50:07 +00:00 |
scc%netscape.com
|
9e78c81481
|
fixing build bustage
|
2000-04-17 12:22:37 +00:00 |
scc%netscape.com
|
45681631d8
|
attempting to fix build bustage
|
2000-04-17 11:49:23 +00:00 |
scc%netscape.com
|
4753a77d13
|
Added private unimplemented signatures of |...WithConversion| to prevent accidents where I should have used the non-converting form. Removed |AppendWithConversion|s that stood in for |AppendInt| and |AppendFloat|.
|
2000-04-17 11:06:41 +00:00 |
scc%netscape.com
|
c26c8ee879
|
no, the |Equals(PRUnichar*, PRUnichar*, ...)| form is not what you think it is. It has to go. Make comparing with |nsIAtoms| remain avalailable even in |NEW_STRINGS_APIS|
|
2000-04-16 11:09:16 +00:00 |
scc%netscape.com
|
52ad71ebb3
|
new ideas about string conversion
|
2000-04-15 00:05:17 +00:00 |
scc%netscape.com
|
4a987edad4
|
Added signatures to prevent automatic incorrect promotion of |char| to |PRUnichar|
|
2000-04-14 23:20:31 +00:00 |
scc%netscape.com
|
11dac06cbd
|
Added a new type, |incompatible_char_type| to help build signatures that prevent automatic incorrect promotion from |char| to |PRUnichar|. Kind of a hack, but the right way (specializing |nsAWritable| for |PRUnichar| can't be made to compile everywhere)
|
2000-04-14 23:19:14 +00:00 |
waterson%netscape.com
|
8d9071615b
|
Ok, the |typename| stuff wasn't really necessary, and caused problems on VC++. r=scc
|
2000-04-04 09:21:57 +00:00 |
waterson%netscape.com
|
443f6f72f6
|
Add |operator!=()| implementation, conditioned on HAVE_CPP_UNAMBIGUOUS_STD_NOTEQUAL. r=scc
|
2000-04-04 06:42:59 +00:00 |
waterson%netscape.com
|
3f84e427ec
|
Explicit conversion required for NEW_STRING_APIS. r=scc
|
2000-04-04 06:03:20 +00:00 |
waterson%netscape.com
|
489d8de9d4
|
Hackery to get NEW_STRING_APIS compiling on gcc-2.7.2.3. r=scc
|
2000-04-04 06:00:01 +00:00 |
waterson%netscape.com
|
b2a20ccd6d
|
Start trying to compile on Unix with new changes. r=scc
|
2000-04-04 05:45:57 +00:00 |
scc%netscape.com
|
b0c94dca55
|
make appending a single character faster for rickg
|
2000-04-04 03:06:15 +00:00 |
scc%netscape.com
|
0cfbea19f8
|
Added a new converting signature |NS_ConvertToString| to substitute for this common practice |nsAutoString("a c string")| where used in parameter lists
|
2000-04-03 04:52:47 +00:00 |
scc%netscape.com
|
af62f49483
|
Unified some signatures of |AppendWithConversion|
|
2000-04-01 00:24:39 +00:00 |
scc%netscape.com
|
62b5f82b89
|
Had to add another signature of |AppendWithConversion|
|
2000-04-01 00:24:32 +00:00 |
alecf%netscape.com
|
6594bf537d
|
fix build bustage
|
2000-03-31 07:55:08 +00:00 |
scc%netscape.com
|
1421ca7839
|
incremental changes. these files are not yet part of the build
|
2000-03-31 07:28:38 +00:00 |
scc%netscape.com
|
ac3e2d0d46
|
more work bringing |ns[C]String| into line with the new implementations. Again, this is no change when |NEW_STRING_APIS| is not defined.
|
2000-03-31 07:26:13 +00:00 |
scc%netscape.com
|
74ccce056f
|
incremental changes. We're getting closer. These files are not yet part of the build.
|
2000-03-30 22:53:41 +00:00 |
scc%netscape.com
|
42d9401c3e
|
adjustments to the string interfaces that should have no effect in the non-|NEW_STRING_APIS| case (yes, I built all changes to verify)
|
2000-03-30 22:51:51 +00:00 |
scc%netscape.com
|
62f74bfd17
|
First Checked In.
|
2000-03-29 08:11:59 +00:00 |
warren%netscape.com
|
d60b12b153
|
Necko API changes: primarily nsIChannel, changing initialization parameters to accessors. Got javascript: evaluation to happen at the right time (when AsyncRead is called) as well as on the right thread.
|
2000-03-29 03:58:50 +00:00 |
scc%netscape.com
|
7379e92bad
|
rename |string_copy| and |string_copy_backward| to |copy_string| and |copy_string_backward|
|
2000-03-29 00:54:12 +00:00 |
scc%netscape.com
|
4f1ac4a894
|
finer grain control of which current string assignments are still available after |NEW_STRING_APIS|. Added |AssignWithConversion| signatures. Made appropriate current |Assign|s be synonyms.
|
2000-03-29 00:06:31 +00:00 |
scc%netscape.com
|
e34e046ae4
|
Adding some macros that allow me to start using literal strings
|
2000-03-28 22:37:51 +00:00 |
waterson%netscape.com
|
e2976c60e9
|
Beef up autoconf tests to differentiate 'ambiguity resolving' and 'access changing' support of the 'using' keyword. Add a test to see if the 'std' namespace is used on the platform. r=scc
|
2000-03-28 07:01:37 +00:00 |
rickg%netscape.com
|
9a6dbab6c9
|
2x perf. improvement StripChar(); r=rods
|
2000-03-28 00:02:41 +00:00 |
scc%netscape.com
|
e3cac710d4
|
fixing an off-by-one bug, and the associated problem when |mCapacity| and |mLength| are both |0|. r=waterson
|
2000-03-27 22:13:41 +00:00 |
jst%netscape.com
|
96a8f00fc0
|
Backing out scc's last checking to this file, it causes linux builds to dump core on startup. a=leaf
|
2000-03-27 15:10:12 +00:00 |
scc%netscape.com
|
92b5b65386
|
traded in |copy_chunky_backward| for |string_copy_backward| and moved it to "nsAReadableString.h" along with the |string_copy|s. Added tests for |SetLength|, which would have caught the problem in |nsString| I fixed earlier, and one for |Insert| that exposed the trouble in |copy_chunky_backward|.
|
2000-03-27 09:06:37 +00:00 |
scc%netscape.com
|
bf02af838a
|
fixing an off-by-one bug
|
2000-03-27 07:46:02 +00:00 |
scc%netscape.com
|
683bbe9fd6
|
Trying to fix for Solaris
|
2000-03-25 07:55:07 +00:00 |
pavlov%netscape.com
|
65f297564c
|
add some #if[n]def STASTANDALONE_STRING_TESTS r=scc
|
2000-03-25 02:07:34 +00:00 |
scc%netscape.com
|
57bf24f8f9
|
Making the tests run on all platforms. Removing some confusion about iostreams.
|
2000-03-25 00:59:29 +00:00 |
scc%netscape.com
|
92964e17ff
|
Incremental changes. These files are not yet part of the build.
|
2000-03-25 00:47:02 +00:00 |
scc%netscape.com
|
ea8e807dda
|
Lots of changes to get this stuff building on Solaris, et al. None of these changes effect the mainline build yet, and won't until |NEW_STRING_APIS| is defined for everyone (coming soon)
|
2000-03-24 22:06:57 +00:00 |