Magnus Hagander <magnus(at)hagander(dot)net> wrote:
> On Fri, Apr 9, 2010 at 16:02, Kevin Grittner
>> I assume we reject anything where what precedes the colon doesn't
>> match the current drive's designation?
> Define reject?
I guess I made that comment thinking about the example of usage
> We're just answering the question "is absolute path?". It's then
> up to the caller. For example, in the genfiles function, we will
> take the absolute path and compare it to the path specified for
> the data directory, to make sure we can't go outside it.
I would say that a function which tells you whether a path is
absolute should, under Windows, return false if there isn't a
leading slash or backslash after any drive specification. Whether
lack of a drive specification should cause it to return false or
whether that should be a separate test doesn't seem like it makes a
big difference, as long as it's clear and documented.
In response to
pgsql-hackers by date
|Next:||From: Łukasz Dejneka||Date: 2010-04-09 17:35:11|
|Subject: Re: How to modify default Type (TSQuery) behaviour?|
|Previous:||From: Robert Haas||Date: 2010-04-09 17:19:11|
|Subject: Re: extended operator classes vs. type interfaces|