[ruby-core:81648] [Ruby trunk Bug#13644] Windows - Setting Time.now

[Greg.mpls@gmail.com]

Issue #13644 has been updated by MSP-Greg (Greg L).


Nobuyoshi,

> It's defined in FakeTime module which is prepended to Time.

Thank you for being so polite and taking the time to respond.  That was a very dumb mistake on my part.  I reached that conclusion when I was fighting a poor, old, MinGW embedded ruby 2.2.4 build in a commercial app. It was nice when I swapped in my build of 2.3 stable, and everything worked...

Anyway, I believe I've found the real issue.  It may be occurring because I'm using Win7, or possibly because I have UAC enabled.

On my system, File.utime only sets atime, not mtime.  I suspect that the OS may not allow setting mtime.  As you probably know, `LogDevice.new` uses mtime to determine whether it's time to create a new log file.

Normally, I can't see how that would be an issue, but re this test trying to mock things, it is.

If you have a minute, could you check the following code?  For a MinGW build, it changes LogDevice to use atime to determine whether to create a new log.

```ruby
require 'logger'

module PLogDevice
  def initialize(log = nil, shift_age: nil, shift_size: nil, shift_period_suffix: nil)
    @dev = @filename = @shift_age = @shift_size = @shift_period_suffix = nil
    mon_initialize
    set_dev(log)
    if @filename
      @shift_age = shift_age || 7
      @shift_size = shift_size || 1048576
      @shift_period_suffix = shift_period_suffix || '%Y%m%d'

      unless @shift_age.is_a?(Integer)
        base_time = @dev.respond_to?(:stat) ? 
          (RUBY_PLATFORM !~ /mingw/ ? @dev.stat.mtime : @dev.stat.atime) :
#          @dev.stat.mtime :
          Time.now
        @next_rotate_time = next_rotate_time(base_time, @shift_age)
      end
    end
  end
end
class Logger::LogDevice ; prepend PLogDevice ; end

module FakeTime ; attr_accessor :now ; end
class << Time   ; prepend FakeTime   ; end

log = "log"
File.open(log, "w") {}
File.utime(*[Time.mktime(2014, 1, 2, 0, 0, 0)]*2, log)

Time.now = Time.mktime(2014, 1, 2, 23, 59, 59, 999000)
dev = Logger::LogDevice.new(log, shift_age: 'daily')
dev.write("#{Time.now} hello-1\n")
dev.close

puts "\nTime.now  #{Time.now}\n" \
       "atime     #{File.atime(log)}\n" \
       "mtime     #{File.mtime(log)}\n"

Time.now = Time.mktime(2014, 1, 3, 1, 1, 1)
dev = Logger::LogDevice.new(log, shift_age: 'daily')
dev.write("#{Time.now} hello-2\n")
dev.close

puts "\nTime.now  #{Time.now}\n" \
       "atime     #{File.atime(log)}\n" \
       "mtime     #{File.mtime(log)}\n"
```

It may show more than needed, but as listed, the logs are written correctly.  If you uncomment

```ruby
#          @dev.stat.mtime :
```

and comment

```ruby
          (RUBY_PLATFORM !~ /mingw/ ? @dev.stat.mtime : @dev.stat.atime) :
```

it behaves as normal (using mtime) and will fail.

Hence, I don't believe there's an issue with logger, it's an issue with trying to mock this.

If adding the prepend to LogDevice looks okay, someone should do a PR or commit.

Again, thanks for all your work, both Ruby in general and windows issues.

----------------------------------------
Bug #13644: Windows - Setting Time.now
https://bugs.ruby-lang.org/issues/13644#change-65342

* Author: MSP-Greg (Greg L)
* Status: Open
* Priority: Normal
* Assignee: 
* Target version: 
* ruby -v: ruby 2.5.0dev (2017-06-04 trunk 59013) [x64-mingw32]
* Backport: 2.2: UNKNOWN, 2.3: UNKNOWN, 2.4: UNKNOWN
----------------------------------------
While reviewing a MinGW build `test-all` failure in [TestLogDevice#test_shifting_midnight_exist_file](https://github.com/ruby/ruby/blob/trunk/test/logger/test_logdevice.rb#L687-L727), I determined the cause.  The test in question (and other tests that seem to have newer/better exception handling) sets `Time.now`.

The thread [StackOverflow - Change system date programmatically](https://stackoverflow.com/questions/650849/change-system-date-programmatically) seems to imply that in some instances/configurations, changing the system time on Windows systems requires elevated permissions.

I tested on both a mswin build (`ruby 2.5.0dev (2017-05-27 trunk 58922) [x64-mswin64_140]`) and a MinGW build (`ruby 2.5.0dev (2017-06-08 trunk 59046) [x64-mingw32]`), and both responded to:

```ruby
Time.now = Time.mktime(2017, 1, 3, 1, 1, 1)
```

with the following eror:

```
undefined method `now=' for Time:Class (NoMethodError)
```

Both also had `false = Time.respond_to?(:now=)`.

So, I'm confused as to how this test passes on mswin.  Regardless, the test requires a skip to bypass it and allow MinGW test-all to pass.

I thought I'd file an issue before doing a PR.  I'm building and testing on Win7.  Finally, if anyone has time, where is the method `now=` defined?



-- 
https://bugs.ruby-lang.org/