Permalink
Commits on Oct 4, 2005
Commits on Sep 30, 2005
  1. Use more generic SQL_NUMERIC instead of SQL_INTEGER to bind numeric

    parameters.
    
    This fixes a problem on PostgreSQL >= 7.4 and recent versions of DBD::Pg
    that now use server-side prepared statements and binding instead of
    binding in the DBD itself. The PostgreSQL server was rejecting numbers
    with decimals when binding to SQL_INTEGER, because, well, they're not
    integers.
    
    Tested on MySQL 3.23, PostgreSQL 8.0, and Oracle 8i.
    jonjensen committed Sep 30, 2005
Commits on Sep 28, 2005
Commits on Sep 27, 2005
  1. crontab script added

    racke committed Sep 27, 2005
Commits on Sep 26, 2005
  1. Prevent spammer abuse of [email] tag via header injection in inputs such

    as To, From, Subject. Properly handle multiple-line "folded" headers as
    per RFC, but reject any other newlines attempted injection abuses.
    
    Properly tolerate missing MIME::Lite module, as appears to have been
    originally intended.
    
    Bring back cosmetic changes from version 1.6, which were nuked in 1.7.
    jonjensen committed Sep 26, 2005
Commits on Sep 22, 2005
  1. sarge has been released

    racke committed Sep 22, 2005
  2. merged changes 2.52.2.5 vs 2.52.2.6

    +++ 2.52.2.6 +++
    new upstream release
    racke committed Sep 22, 2005
  3. Remove outdated comment.

    jonjensen committed Sep 22, 2005
  4. * Improve ability of AutoModifier, make consistent in both places where

      it can be set.
    
    * Current variations:
    
    	AutoModifier  category
    	AutoModifier  outboard:category
    	AutoModifier  outboard:category:foreign
    
      all stay the same -- only difference is that :: is now accepted same as :.
    
    * New variation:
    
    	  AutoModifier attribute=table:column
    	  AutoModifier attribute=table:column:foreign
    
      is accepted. This removes the requirement that the attribute be named
      the same thing as the table column.
    
    * Added regression test that checks all of the above.
    
    * Will look for existing documentation to update.
    
    * Your regression tests will require a "rm products/products2.sql" to
      update the data for that table.
    perusionmike committed Sep 22, 2005
  5. * Minor improvement in POD.

    perusionmike committed Sep 22, 2005
  6. Swish setting added

    racke committed Sep 22, 2005
Commits on Sep 21, 2005
Commits on Sep 19, 2005
  1. Remove [br] tag.

    Don't see any reason to replace it with [x] or [xt] or other discussed
    alternatives, because it wasn't being used in the core anyway, and at the
    catalog level people will either use XHTML or not, but not dynamically
    decide about trailing slashes, I think.
    jonjensen committed Sep 19, 2005
Commits on Sep 16, 2005
  1. * Add an order check which always passes -- used to update value of v…

    …ariables
    
      no matter what they are when &update=yes in force.
    
    * Updated xmldocs/glossary/profile to add.
    perusionmike committed Sep 16, 2005
Commits on Sep 15, 2005
  1. * Fix input button typo.

    perusionmike committed Sep 15, 2005
Commits on Sep 14, 2005
  1. * Add SpecialSub "guess_cc_type" to allow custom credit card types on a

      per catalog basis.
    
    * Calling parameters -- is passed the digits of the credit card number
      as its only parameter (no spaces or non-word chars).
    
    * If you know what type the number is, you return it and it is used as
      the result. If not, return a false value to let the normal routine do
      it's thing.
    
    * To activate, do in catalog.cfg:
    
    	## Map a subroutine to guess custom credit card types
    	SpecialSub  guess_cc_type  special_cc
    
    	Sub special_cc <<EOS
    	sub {
    		my $num = shift;
    		return 'interchange_test' if $num =~ /^4111/;
    		return undef;
    	}
    	EOS
    
      That will return "interchange_test" if the number begins with 4111
    
    * Where I should document this I don't know. If someone suggests a home
      and helps me find the actual document -- this is important -- I will
      try. But I don't think finding the right file out of hundreds to do
      the documentation in is not a really viable alternative.
    perusionmike committed Sep 14, 2005
  2. Solve the problem of fetching column names on databases that don't su…

    …pport
    
    LIMIT, such as Oracle, by having a successful query that always returns no
    rows.
    
    This also solves an obscure problem in PostgreSQL (and presumably also
    MySQL with InnoDB tables) where an un-vacuumed table with a large
    number of dead rows at the beginning can take a long time to do the
    previous query SELECT * FROM table LIMIT 1, because no index is used
    and a lengthy sequential scan must first pass all the dead rows before
    retrieving a single live row. (Using LIMIT 0 would also have worked,
    but would've been less portable SQL.)
    
    Tested on PostgreSQL 7.4, MySQL 3.23, and Oracle 8i.
    jonjensen committed Sep 14, 2005
Commits on Sep 11, 2005
Commits on Sep 9, 2005
  1. * Add Set::Crontab to list of modules checked (thanks, Davor).

    * Removed a few duplicate module entries.
    
    * Switch module list from hash to array type, since only use is of keys,
      as a list.
    
    * Clean up module definition formatting.
    
    * Make HTML output XHTML-compliant.
    jonjensen committed Sep 9, 2005
Commits on Sep 8, 2005
  1. * Make [query ...] base be transactions to avoid using Vend::SQL_Parser

      when products is not in same DSN as transactions.
    
    * Use proper SQL <> operator instead of the != I propagated all over the
      catalog originally.
    perusionmike committed Sep 8, 2005
Commits on Sep 7, 2005
  1. help users on imports to locate offending rows

    especially useful with Postgresql which is known to be quite picky
    racke committed Sep 7, 2005
Commits on Sep 5, 2005
  1. * Recognize https:// urls (handled by "http" regexp, not a separate one)

    * Make match case insensitive, that is, also match things like HTTP://
    
    * I also benchmarked a situation where I would precompile regexps,
      but probably Perl figures that by itself so it's not needed.
      (I ran the test on 40,000 URLs, and in both cases results were between
      1.310 and 1.330 seconds).
    docelic committed Sep 5, 2005