From b2eeeff0498b3ed9453fe73628945ede12f25c53 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Mattias=20Andr=C3=A9e?= Date: Sun, 18 Aug 2013 09:51:21 +0200 Subject: [PATCH] update info manual MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: Mattias Andrée --- manuals/ponysay.texinfo | 43 +++++++++++++++++++++++++++-------------- 1 file changed, 29 insertions(+), 14 deletions(-) diff --git a/manuals/ponysay.texinfo b/manuals/ponysay.texinfo index f5430b09..ee3618b8 100644 --- a/manuals/ponysay.texinfo +++ b/manuals/ponysay.texinfo @@ -2586,22 +2586,37 @@ comment field. Leading line breaks in the comment field is ignored. @cindex pony quote infrastructure @cindex quote infrastructure -When compiling, pony quotes are built to @file{quotes/}, the file names are -lists of ponies joined with plus signs (@code{+}) -- the pony names are the -same as the pony files, except they do not end with @file{.pony} -- with a -index at the end, and a full stop (@code{.}) before the index. +When compiling, two indices for the pony quotes are built. These indices +are dbm databases; if you have gdbm, which you probably have, python will +using gdbm. If you later uninstall gdbm, you will have to reinstall ponysay +so it is built without, gdbm. -The source files are located in @file{ponyquotes/}, where their is a file named -@file{ponies}. This file is called the pony map, and is the basis for how the -compiled files are named. In the ponymap ponies with the same quotes are on the -same line join together with plus signs (@code{+}), if the lines because too -long for file names the line is split into multiple lines with the first pony -in common. +Prior to version 3.1 of ponysay, there was a @file{ponies} in @file{ponyquotes/} +where all ponies are group to gather with plus signs (@code{+}), to identify +which quotes belong to which pony files, because the quote files just hade +the first ponys' name before compilation. -In @file{ponyquotes/} there are also quote files, each contain just one quote, -just as when compiled to @file{quotes/}. The source quote files are identical -to the compiled quote files, except that their name contains just the first -pony. +Since version 3.1 of ponysay, the group is instead saved in index databases, +and built from the pony file metadata tag @code{MASTER}. + +The previous solution has some disadvantages: + +@itemize @bullet +@item +It was not as fast a pony name was needed to be searched for in a all +quote file names. +@item +Was not as flexible for selection another algorithm for randomly selecting +a pony and a quote. +@item +@file{ponyquotes/ponies} was manually edited when a new pony as added, +altough this could have been solved without changing anything in the +installed files. This just happened to be solved at the same time. +@item +Pinkie Pie had so many images that the resolution quote files' names +were to long and quotes had to be duplicated. And Pinkie Pie became +even more likely to be randomly selected. +@end itemize @node Balloon style files