| Subject: | splits method shows only last digit of Post value |
ActivePerl 5.8.7, build 815 on Win XP SP 2
When you run this command line:
perl -MFinance::QuoteHist -e "print qq(\nSymbol\tDate\t\tPost\tPre\n);
while ( ($Symbol, $date) = (shift, shift) and defined $date ) { $q =
Finance::QuoteHist->new(symbols => [$Symbol], start_date => $date,
end_date => $date, lineup => 'Finance::QuoteHist::Yahoo' ); foreach
$ra_row ($q->splits()) { local ($,,$\) = (qq(\t),qq(\n)); print
@$ra_row; }}" NPBC 2006-09-06 TKC 2006-06-26 EPIX 2006-08-17 ID
2005-12-19 RBNC 2005-11-02
You get this output (I added the [s/b] column to show the correct Post
numbers):
Symbol Date Post [s/b] Pre
NPBC 2006/09/06 3 103 100
TKC 2006/06/26 5 85 16
EPIX 2006/08/17 0 10 15
ID 2005/12/19 0 10 25
RBNC 2005/11/02 1 11 10
Suggested patch (and please forgive me if I over-explained--I've no way
of knowing how much or how little to say):
Replace "my($date, $post, $pre) = /^(\S+).*(\d+):(\d+)/;"
With one of the following, from fastest, but least resistant to format
changes to slowest, but most likely to survive small niggling format
changes like changing [ to " or using leading hyphens or designator
letters or other things bored Yahoo programmers may do in the dark of
the night:
my($date, $post, $pre) = /^(\S+)\s*\[(\d+):(\d+)/m; <-- least robust,
fastest
my($date, $post, $pre) = /^(\S+)\D*(\d+):(\d+)/m;
my($date, $post, $pre) = /^(\S+).*?\b(\d+):(\d+)/m;
my($date, $post, $pre) = /^(\S+).*\b(\d+):(\d+)/m;
my($date, $post, $pre) = /^(\S+).*\D(\d+):(\d+)/m; <-- most robust,
most processing
All of these test okay with the symbol/date samples above. The second
regex is the best, as long as Yahoo never sticks any digits between the
date and the split Post value. The third and fourth ones would survive
that and just about any format changes that don't put word characters on
front of the Post value. The last one would survive even word characters
prepended to the Post value.
Good luck, mate, and thanks for the great module. This thing is saving
my life.