1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
|
1.12
- now work with Perl 5.10 using an eval block
- GPL v2 and v3 now
1.11
- test suite compares sorted results of find to fix false alarms on some systems
- test suite now uses tar instead of cp for copying.
- fix a y/n mixup
- at startup we check if Perl::Encode is not too broken (bugs 37757 and 49830 for now). Recent Perl releases are terribly broken. If this won't improve, I'll have to rewrite convmv in a language with more stable core functionality :-|. If you are distributor of convmv, please run make test before packaging to make sure you don't ship convmv with a broken Perl release.
1.10
- now use from_to instead of utf8::decode (which is experimental and might vanish in the future) to check for valid UTF-8
- updates and cleanups of testsuite and include "0" filename
- fix "0" filename bugs
- take into account that charset name "utf-8" resolves to "utf-8-strict" in recent Perl versions
1.09
- work around broken decode_utf8() in Perl 5.8.7 resp. Encode 2.10 (use utf8::decode instead)
- adopt test suite to catch cases like this
- warn about dry test run at start, not just at end of a test run
1.08
- do not complain about wrong from_enc if to_enc is utf8 and files are, too; this happens with from_enc being a multibyte encoding. Thanks to Kuang-che Wu
- fixed bug, where filename "0" was considered to be wrong utf-8 if from_enc was utf8
- some documentations updates
1.07
- add%20option%20to%20undo%20ugly%20http-%20and%20ftp-like%20escape%20sequences
- add option to upper/lowercase virtually any charset
- for upper/lower conversion add a "dotless i" option for Lithuanian, Turkish and Azeri
- check for characters not allowed in POSIX filesystems (just in case someone wants to try to convert to UTF-16)
1.06
- fix some typos and cosmetics
- add missing from_print
- rewrote checkenc(), no more false alarms now with "encodings" like MIME-Q, which can look quite different. As a side effect error messages became more detailed
1.05
- --qfrom/to also mask cntrl characters now
- symlink printout splitted to get correct output
- missing $arg in symlink fixed (bug introduced in 1.02)
- add NFS4 issues to documentation
- add test suite
- add file length check
1.04
- add --exec option to allow arbitrary operations
1.03
- add --lowmem option to keep memory footprint low when millions of files are being converted. However this will deactivate symlink target checks.
- add Samba issues to documentation
- suppress normalization error messages in cases of invalid UTF-8 in checkenc()
1.02
- make default behaviour to not work recursively so we can convert a single directory, use -r to get the old (recursive) mode
- work around find() oddity, which doesn't process directories from @ARGV
- die on unknown option
- check if files exist and die otherwise
1.01
- documentation updates
- check if symlink target is -f or -d, we run into problems in /proc otherwise.
- implement interactive (-i) mode
1.0
- UTF-8 smartness added
0.3
- first stable release
|
