history

来自「UUDeview是一个编码解码器」· 代码 · 共 539 行 · 第 1/2 页

  because much less data is gathered, thus many invalid files may get
  processed.
- added e(x)tract command; but (d)ecode stays default.

 0.4.16 (21.01.1996)
--------
- a tiny typo, which was introduced for debugging purposes but not removed,
  broke 0.4.15. This was especially problematic for the Windows version.

 0.4.17 (09.02.1996)
--------
- once again a stupid 16 bit problem introduced in 0.4.15. In uunconc.c:
  uunconc(), the variable pppos, used to remember the file position, was
  declared as int. Because of this, on 16 bit systems, it looped over and
  over, but the file pointer never reached maxpos :-(
  This patchlevel was only released as uudv[dw]04h but not as Unix version
  because 32 or 64 bit systems weren't affected.

 0.4.18 (not publicly released)
--------
- this is an intermediate beta test version. With all the changes going
  on, I expect everything to settle into a 0.5 release sooner or later.
- got rid of some bugs
- Base64-encoded lines can now have arbitrary length (not limited to
  multiples of four)
- support for BinHex decoding added. No encoder yet.
- complete re-layout of the code. All the encoding/decoding stuff has
  been moved into a seperate, system-independent library, while the
  enveloping standalone programs uuencode and uudecode are just front-
  ends to this library.
- another front-end has been added, uutcl, which makes the library
  functions available to Tcl/Tk programs
- A Tcl/Tk script, xdeview, uses these functions and provides a graphical
  frontend
- support for decompressing and tar files has been removed. They didn't
  work reliably, and aren't the job of a decoder anyway.
- BinHex and Base64 lines may have space characters at the end of the
  line

 0.4.19 (not publicly released)
--------
- More changes everywhere. Fixed another problem with Base64 files.
- Improvements everywhere :)
- Added progress meter to the decoder

 0.5.0 (07.06.1996)
-------
- Well, thought I have to release the stuff sooner or later.
- Fed the code into CVS for better revision control.

 0.5.1 (10.06.1996)
-------
- Fixed problem with multiple tiny Base64 files in a single message.

 0.5.2 (15.06.1996)
-------
- Cleaned up some code; now using size_t parameters for fread() and
  fwrite() and such.
- Fixed problem with short BinHex files.
- Use the initial and trailing colon `:' of BinHex files as
  `begin' and `end' markers.
- Only encode to message/partial embedded messages if we really
  have more than one part; otherwise encode directly to toplevel
  MIME body.

 0.5.3 (23.06.1996)
-------
- Complete rewrite of the scanning routines, which have been moved
  to uuscan.c and grown into the largest chunks of code.
- Mime files are properly treated with a pretty complete Mime par-
  ser. Outside of Mime messages, the scanner falls back to the old
  behaviour.
- Offers to `decode' plain text (and quoted-printable text) files
  with -t option.
- Intermediate preview release.

 0.5.4 (21.07.1996)
-------
- Rewrote encoding functions. The system-dependent stuff has moved
  to the outside of the library.
- Minor changes not affecting the frontend.

 0.5.5 (21.08.1996)
-------
- Bugfixes:
- Sometimes, the file pointer wasn't properly rewound to the beginning
  of the headers after bumping into them.
- The Binhex RLE decompressor produced garbage with repetition counts
  greater than 127 (which were expanded into negative integers)
- Stripping the last CRLF of a plain-text message was fatal if the
  piece was really an unhandled message/multipart part.

 0.5.6 (28.08.1996)
-------
- The Mime-Id buffer in uuencode.c:UUE_PrepPartial() was too small,
  causing it to be partially overwritten
- Added code to read part number and file name from UUEnview-Style
  headers. Especially important for filename detection with pure
  Base64. I consider this an ugly hack, but users call for it.

 0.5.7 (04.09.1996)
-------
- Added "attach" option to uuenview: read a msg from stdin and attach
  one or more files using proper MIME multipart/mixed style.
- Fixed configure script: we might need $x_includes for <tk.h>
- Fixed rules for "make install"

 0.5.8 (13.09.1996)
-------
- Fixed output filename when encoding dotfiles
- Fixed uudecode compatibilty mode (would crash when reading stdin)
- Fixed bug with long header lines (was actually a bug in _FP_strncpy)
- Fixed bug when mailing directly from uuenview. The recipient address
  length must be added when composing the command.
- Added configure options --enable-sendmail and --enable-inews

 0.5.9 (29.09.1996)
-------
- Fixed percentage display for >20Meg files on 32bit systems, where
  100*fsize becomes > MAXINT ...
- Fixed problem with "--- cut here ---" lines that were misinterpreted
  as MIME boundaries.
- A user tried to decode a file with an early truncated MIME message
  and another message later with the same boundary. Consequently, all
  data inbetween was ignored. Added check for a Content-Type with the
  same boundary when scanning a subpart.
- Allow to specify more dumbness on UUDeview's command line to ignore
  subjects as well as part numbers. Fixed a problem with dumb mode as
  well.

 0.5.10 (01.10.1996)
--------
- Plugged memory leak in uucheck.c:PreProcessPart()

 0.5.11 (03.10.1996)
--------
- Added workaround for broken uuencoded files sent from MS Exchange.

 0.5.12 (23.12.1996)
--------
- Fixed bug with aborted decoding runs that the tempfile was deleted,
  but UUFILE_TMPFILE was still set.
- Make uudeview return a more or less sensible return value: 0 if all
  files were decoded successfully, 99 for miscellaneous errors, 2 if
  all files failed to decode, or 3 if some files were decoded okay
  and others failed.
- Added -n option to uudeview to not print progress bars.
- You can now set global options for UUDeview in the UUDEVIEW environ-
  ment variable. Options set there may be overrridden on the command
  line.
- Some tweaking to make some troublesome input files (for example with
  interleaved blank lines) decode.
- Fixed bug that caused erroneously decoded data when UUencoded data
  looked like Base64 (because of the overlapping alphabet)

 0.5.13 (09.01.1997)
--------
- Fixed bug in uuscan.c.
- Fixed another bug in uuscan.c which happened when a filename
  on the Content-Type line happened to contain a header keyword.
- Fixed infinite loop when trying to descramble Netscape data;
  in one place we were looking for <a href=, but we would only
  be unscrambling this if a double-quote was following.

 0.5.14 (30.04.1997)
--------
- Fixed problem with the uuencode compatibility of uuenview.
- Do not throw away plain text or quoted-printable parts which
  have explicit filenames.
- Fish out another one of those stupid "--- this is the end ---"
  lines that looked like uuencoding.

 0.5.15 (08.06.2001)
--------
- Fixed problem with quoted-printable MIME messages, whose decoding
  included their header.
- Added handling for AOL folder files which resemble a MIME multipart
  message without a header -- they start off with a boundary line.
- Fixed a buffer overrun problem by adjusting output line size
- Always "decode" MIME bodies which have a filename set in their
  Content-Disposition field
- Fixed an end-of-line problem with MIME bodies
- Hopefully fix some Base64 false detections by ignoring seemingly
  Base64-encoded lines that are shorter than 40 characters
- Added -a option to uudeview to auto-rename a file if the target file
  exists (i.e. foo.gif would become foo.gif.1).
- Make xdeview compile and run with Tcl&Tk > 8.0
- Do not use "dangerous" gets in minews, replace by fgets
- If the file looks like MIME, be more strict when scanning "text/*"
  parts for encoded data.
- Add +o option to never overwrite files (useful in non-interactive
  mode)
- Add -c option to autoclear input files from which anything was decoded.
  Use with care - UUDeview does not care about anything else in that file.
- Add -z option for improved MIME compliance on MIME messages.
- Improved encoding detection, it should now be less sensitive to data
  that looks both like uu and Base 64.
- Add -r (Reply-To:) and -f (From:) options to uuenview
- Support encoding as quoted-printable
- Support "encoding" as plain text
- Add -t and -q options to uuenview for that
- Make the encoding to be used depend on the last -ubxtq option, so that
  you can send plain text and binary attachments on the same command line.
- Make uuenview read the $UUENVIEW command line option
- Tag attachments as "text/plain" if plain text or quoted printable is
  used and no other mime type is appropriate.
- Recognize a boundary when scanning a file in "freestyle" mode. This is
  similar to the AOL handling above.

 0.5.16 (26.08.2001)
--------
- fix handling of "x-uue" encoded MIME attachments
- bug fix for file renaming

 0.5.17 (06.03.2002)
--------
- fix handling of quoted parameters
- some support for yEnc encoding

 0.5.18 (02.04.2002)
--------
- fix for yEnc files in UUInfoFile
- patch by Matthew Mueller:
  - add CRC generation and checks for yEnc
  - accept yEnc files where the tab character is not escaped
  - improved error checking for yEnc files

 0.5.19 (07.10.2003)
--------
- UUDeview: check validity of save path upon 'p' (in interactive mode)
  (Chris Hanson)
- patches by Matthew Mueller:
  - check return code of fclose() after writing target file to detect
    e.g. disk full errors
  - check for errors reading the input file
  - Get suspicious about potentially uuencoded data if the line length
    varies
  - write incomplete files, hoping for some block-level recovery
  - Add a UULoadFileWithPartNo function, in case a user of this library
    is certain of the part number and doesn't need to trust UUDeview's
    heuristics.
- fix a bug in the historic Netscape fixing code, which probably should
  be removed altogether sometime anyway (see log for 0.4.5 above)
  (Michael Newcomb)
- UUDeview: Add an alternative autorenamer that adds a unique sequence
  number before the first dot, keeping the extension intact
- UUDeview: If the filename is "." or "..", replace the first dot with
  an underscore to not conflict with the directory names
- finally implemented RFC 1522 header decoding. The charset is ignored,
  though
- fix some problems with decoding message/partial messages
- In violation of RFC 1521, allow the [;] characters in a token. The
  Klez virus does that, and people who feed their virus scanners with
  the output from UUDeview would like to catch it.
- UUDeview: replace some questionable punctuation characters in
  filenames with underscores. The current list is {<|">}
- Stop tinkering with CFLAGS in ./configure; remove --disable-optimize.
  Use 'CFLAGS=-O ./configure' for non-debug builds.

 0.5.20 (01.03.2004)
--------
- Stop tinkering with CFLAGS in ./uulib/configure
- fix bug in inews/inews.c: fgets doesn't discard the final newline
  (Yuri D'Elia)
- fix bug in parsing of header lines
- fix fgets to accept lines that are exactly of the maximum length
- fix two buffer overflows