Permalink
Browse files

initial commit

  • Loading branch information...
flori committed Aug 21, 2009
0 parents commit fd0e8984737188c208b8fb155125ffa3086c6327
Showing with 1,373 additions and 0 deletions.
  1. +3 −0 .gitignore
  2. +69 −0 CHANGES
  3. +340 −0 COPYING
  4. +38 −0 README
  5. +96 −0 Rakefile
  6. +1 −0 VERSION
  7. +54 −0 doc-main.txt
  8. +17 −0 examples/pager.rb
  9. +10 −0 examples/tail.rb
  10. +21 −0 file-tail.gemspec
  11. +21 −0 install.rb
  12. +387 −0 lib/file/tail.rb
  13. +10 −0 lib/file/tail/version.rb
  14. +5 −0 make_doc.rb
  15. +301 −0 tests/test_file-tail.rb
@@ -0,0 +1,3 @@
+.*.sw[pon]
+coverage
+pkg
69 CHANGES
@@ -0,0 +1,69 @@
+2009-08-21 * 1.0.4 * Fixed the threaded tests for Ruby 1.9.
+ * Create a gem spec file.
+ * Some cleanup.
+2008-04-07 * 1.0.3 * Danny Colligan <danny.colligan@sendori.com> reported a
+ memory leak in long running scripts using file-tail. I
+ think this might be a ruby related problem, which is
+ caused/aggravated by using yield after having &block
+ parameter in a method. I changed file-tail to only use
+ block.call, which seems to improve the memory behaviour. I
+ am still not sure, where the problem actually stems
+ from, though.
+2007-04-19 * 1.0.2 * make_doc.rb was missing from the source archive. Thanks to
+ Rick Ohnemus <rick.ohnemus@systemware.com> for reporting it.
+2007-04-19 * 1.0.1 * Bugfix: File::Tail::Logfile#open with block, now closes
+ the file like File#open does. Found by Alex Doan
+ <alex.doan@wachovia.com>, ruby-talk:248383.
+2007-03-30 * 1.0.0 * Bugfix: David.Barzilay@swisscom.com reported, that file
+ tails may skip some log file lines, after rotating it. I
+ think, that I fixed that problem.
+ I added a after_reopen callback as well, that is called
+ after reopening of the tailed file has occured.
+ * Removed rewind/wind methods even earlier than planned: I
+ placed the deprecation warning for rewind method in File
+ instead of File::Tail, which caused rewind to stop working
+ completely after loading file/tail. Duh! I blame vim's
+ matchit, because it jumped to the wrong end keyword.
+2007-02-08 * 0.1.4 * Renamed rewind method to backward, and wind method to
+ forward, because someone already had the good idea to name
+ a method IO#rewind, which was overwritten by the mixed in
+ File::Tail methods. The old methods are now deprecated and
+ will be removed in a new 0.2.x version of the library.
+ * Added a bit more of documentation.
+2005-08-20 * 0.1.3 * Applied LOAD_PATH patch by Daniel Berger, binary mode
+ changes were already in the CVS. Seemed to be like cheating
+ to me, though. ;)
+ * Skipping one windows test for the moment, too. Sigh!
+2004-09-30 * 0.1.2 * First Rubyforge release
+ * Added Rakefile
+ * Supports gem build now.
+2004-09-01 * 0.1.1 * Josh Endries <josh@endries.org> found a bug
+ that caused File::Tail to malfunction on FreeBSD.
+ Hotfix: Use a side effect of seek to clearerr the tailed
+ file handle after EOFError has been raised.
+2004-04-13 * 0.1.0 * API documentation with rdoc.
+ * return_if_eof attribute added.
+ * Added array return mode for finite tail call without block
+ given.
+ * install.rb now uses ruby version site_dir.
+ * Some code and directory structure cleanup.
+2002-08-02 * 0.0.2 * Heavy refactoring, more and smaller methods
+ and expception handling
+ * Added check for inode and device equality of files
+ as suggested by
+ James F.Hranicky <jfh@cise.ufl.edu> and
+ Curt Sampson <cjs@cynic.net> to cover remove
+ rotation
+ * If filesize shrinks suddenly, File::Tail assumes that
+ copy and truncate rotation has happend: The file
+ is reopened and every new line is handled.
+ * NFS-Fix: Errno::ESTALE is caught.
+ * wind added to skip the first n lines, as
+ James F.Hranicky's suggested and changed
+ name of last-method to rewind, because I liked
+ his method names better than mine ;)
+ * Renamed next to tail either.
+ * The API has changed - but I think very few people
+ care at the moment.
+ * Lots of tests added.
+2002-07-30 * 0.0.1 * Initial Release
Oops, something went wrong.

0 comments on commit fd0e898

Please sign in to comment.