about summary refs log tree commit diff
path: root/docs/tool/Modules/NaturalDocs/Builder/Base.pm
diff options
context:
space:
mode:
authorMagnus Auvinen <magnus.auvinen@gmail.com>2008-08-02 08:21:29 +0000
committerMagnus Auvinen <magnus.auvinen@gmail.com>2008-08-02 08:21:29 +0000
commit61bfe2d70cae6be8c4086a210a5451135ccca9ea (patch)
tree62bf7808b1b2bfe5f56fe1e329871fb0991d0687 /docs/tool/Modules/NaturalDocs/Builder/Base.pm
parenta13b94f9e0bca8ea892311d9d9e0c0bc48616ea7 (diff)
downloadzcatch-61bfe2d70cae6be8c4086a210a5451135ccca9ea.tar.gz
zcatch-61bfe2d70cae6be8c4086a210a5451135ccca9ea.zip
added doc tool
Diffstat (limited to 'docs/tool/Modules/NaturalDocs/Builder/Base.pm')
-rw-r--r--docs/tool/Modules/NaturalDocs/Builder/Base.pm348
1 files changed, 348 insertions, 0 deletions
diff --git a/docs/tool/Modules/NaturalDocs/Builder/Base.pm b/docs/tool/Modules/NaturalDocs/Builder/Base.pm
new file mode 100644
index 00000000..0298264d
--- /dev/null
+++ b/docs/tool/Modules/NaturalDocs/Builder/Base.pm
@@ -0,0 +1,348 @@
+###############################################################################
+#
+#   Class: NaturalDocs::Builder::Base
+#
+###############################################################################
+#
+#   A base class for all Builder output formats.
+#
+###############################################################################
+
+# This file is part of Natural Docs, which is Copyright (C) 2003-2008 Greg Valure
+# Natural Docs is licensed under the GPL
+
+use strict;
+use integer;
+
+package NaturalDocs::Builder::Base;
+
+
+###############################################################################
+# Group: Notes
+
+
+#
+#   Topic: Implementation
+#
+#   Builder packages are implemented as blessed arrayrefs, not hashrefs.  This is done for all objects in Natural Docs for
+#   efficiency reasons.  You create members by defining constants via <NaturalDocs::DefineMembers> and using them as
+#   indexes into the array.
+#
+
+#
+#   Topic: Function Order
+#
+#   The functions in the build process will always be called in the following order.
+#
+#   - <BeginBuild()> will always be called.
+#   - <PurgeFiles()> will be called next only if there's files that need to be purged.
+#   - <PurgeIndexes()> will be called next only if there's indexes that need to be purged.
+#   - <PurgeImages()> will e called next only if there's images that need to be purged.
+#   - <BuildFile()> will be called once for each file that needs to be built, if any.
+#   - <BuildIndex()> will be called once for each index that changed and is part of the menu, if any.
+#   - <UpdateImage()> will be called once for each image that needs to be updated, if any.
+#   - <UpdateMenu()> will be called next only if the menu changed.
+#   - <EndBuild()> will always be called.
+#
+
+#
+#   Topic: How to Approach
+#
+#   Here's an idea of how to approach making packages for different output types.
+#
+#
+#   Multiple Output Files, Embedded Menu:
+#
+#       This example is for when you want to build one output file per source file, each with its own copy of the menu within it.
+#       This is how <NaturalDocs::Builder::HTML> works.
+#
+#       Make sure you create a function that generates just the menu for a particular source file.  We'll need to generate menus for
+#       both building a file from scratch and for updating the menu on an existing output file, so it's better to give it its own function.
+#       You may want to surround it with something that can be easily detected in the output file to make replacing easier.
+#
+#       <BeginBuild()> isn't important.  You don't need to implement it.
+#
+#       Implement <PurgeFiles()> to delete the output files associated with the purged files.
+#
+#       Implement <PurgeIndexes()> to delete the output files associated with the purged indexes.
+#
+#       Implement <BuildFile()> to create an output file for the parsed source file.  Use the menu function described earlier.
+#
+#       Implement <BuildIndex()> to create an output file for each index.  Use the menu function described earlier for each page.
+#
+#       Implement <UpdateMenu()> to go through the list of unbuilt files and update their menus.  You can get the list from
+#       <NaturalDocs::Project->UnbuiltFilesWithContent()>.  You need to open their output files, replace the menu, and save it back
+#       to disk.  Yes, it would be simpler from a programmer's point of view to just rebuild the file completely, but that would be
+#       _very_ inefficient since there could potentially be a _lot_ of files in this group.
+#
+#       Also make sure <UpdateMenu()> goes through the unchanged indexes and updates them as well.
+#
+#       <EndBuild()> isn't important.  You don't need to implement it.
+#
+#
+#   Multiple Output Files, Menu in File:
+#
+#       This example is for when you want to build one output file per source file, but keep the menu in its own separate file.  This
+#       is how <NaturalDocs::Builder::FramedHTML> works.
+#
+#       <BeginBuild()> isn't important.  You don't need to implement it.
+#
+#       Implement <PurgeFiles()> to delete the output files associated with the purged files.
+#
+#       Implement <PurgeIndexes()> to delete the output files associated with the purged indexes.
+#
+#       Implement <BuildFile()> to generate an output file from the parsed source file.
+#
+#       Implement <BuildIndex()> to generate an output file for each index.
+#
+#       Implement <UpdateMenu()> to rebuild the menu file.
+#
+#       <EndBuild()> isn't important.  You don't need to implement it.
+#
+#
+#   Single Output File using Intermediate Files:
+#
+#       This example is for when you want to build one output file, such as a PDF file, but use intermediate files to handle differential
+#       building.  This would be much like how a compiler compiles each source file into a object file, and then a linker stitches them
+#       all together into the final executable file.
+#
+#       <BeginBuild()> isn't important.  You don't need to implement it.
+#
+#       Implement <PurgeFiles()> to delete the intermediate files associated with the purged files.
+#
+#       Implement <PurgeIndexes()> to delete the intermediate files associated with the purged indexes.
+#
+#       Implement <BuildFile()> to generate an intermediate file from the parsed source file.
+#
+#       Implement <BuildIndex()> to generate an intermediate file for the specified index.
+#
+#       Implement <UpdateMenu()> to generate the intermediate file for the menu.
+#
+#       Implement <EndBuild()> so that if the project changed, it stitches the intermediate files together into the final
+#       output file.  Make sure you check the parameter because the function will be called when nothing changes too.
+#
+#
+#   Single Output File using Direct Changes:
+#
+#       This example is for when you want to build one output file, such as a PDF file, but engineering it in such a way that you don't
+#       need to use intermediate files.  In other words, you're able to add, delete, and modify entries directly in the output file.
+#
+#       Implement <BeginBuild()> so that if the project changed, it opens the output file and does anything it needs to do
+#       to get ready for editing.
+#
+#       Implement <PurgeFiles()> to remove the entries associated with the purged files.
+#
+#       Implement <PurgeIndexes()> to remove the entries associated with the purged indexes.
+#
+#       Implement <BuildFile()> to add or replace a section of the output file with a new one generated from the parsed file.
+#
+#       Implement <BuildIndex()> to add or replace an index in the output file with a new one generated from the specified index.
+#
+#       Implement <EndBuild()> so that if the project changed, it saves the output file to disk.
+#
+#       How you handle the menu depends on how the output file references other sections of itself.  If it can do so by name, then
+#       you can implement <UpdateMenu()> to update the menu section of the file and you're done.  If it has to reference itself
+#       by address or offset, it gets trickier.  You should skip <UpdateMenu()> and instead rebuild the menu in <EndBuild()> if
+#       the parameter is true.  This lets you do it whenever anything changes in a file, rather than just when the menu
+#       visibly changes.  How you keep track of the locations and how they change is your problem.
+#
+
+
+###############################################################################
+#
+#   Group: Required Interface Functions
+#
+#   All Builder classes *must* define these functions.
+#
+
+
+#
+#   Function: INIT
+#
+#   Define this function to call <NaturalDocs::Builder->Add()> so that <NaturalDocs::Builder> knows about this package.
+#   Packages are defined this way so that new ones can be added without messing around in other code.
+#
+
+
+#
+#   Function: CommandLineOption
+#
+#   Define this function to return the text that should be put in the command line after -o to use this package.  It cannot have
+#   spaces and is not case sensitive.
+#
+#   For example, <NaturalDocs::Builder::HTML> returns 'html' so someone could use -o html [directory] to use that package.
+#
+sub CommandLineOption
+    {
+    NaturalDocs::Error->SoftDeath($_[0] . " didn't define CommandLineOption().");
+    };
+
+
+#
+#   Function: BuildFile
+#
+#   Define this function to convert a parsed file to this package's output format.  This function will be called once for every source
+#   file that needs to be rebuilt.  However, if a file hasn't changed since the last time Natural Docs was run, it will not be sent to
+#   this function.  All packages must support differential build.
+#
+#   Parameters:
+#
+#       sourceFile  - The name of the source file.
+#       parsedFile  - The parsed source file, as an arrayref of <NaturalDocs::Parser::ParsedTopic> objects.
+#
+sub BuildFile #(sourceFile, parsedFile)
+    {
+    NaturalDocs::Error->SoftDeath($_[0] . " didn't define BuildFile().");
+    };
+
+
+###############################################################################
+#
+#   Group: Optional Interface Functions
+#
+#   These functions can be implemented but packages are not required to do so.
+#
+
+
+#
+#   Function: New
+#
+#   Creates and returns a new object.
+#
+#   Note that this is the only function where the first parameter will be the package name, not the object itself.
+#
+sub New
+    {
+    my $package = shift;
+
+    my $object = [ ];
+    bless $object, $package;
+
+    return $object;
+    };
+
+
+#
+#   Function: BeginBuild
+#
+#   Define this function if the package needs to do anything at the beginning of the build process.  This function will be called
+#   every time Natural Docs is run, even if the project hasn't changed.  This allows you to manage dependencies specific
+#   to the output format that may change independently from the source tree and menu.  For example,
+#   <NaturalDocs::Builder::HTML> needs to keep the CSS files in sync regardless of whether the source tree changed or not.
+#
+#   Parameters:
+#
+#       hasChanged - Whether the project has changed, such as source files or the menu file.  If false, nothing else is going to be
+#                            called except <EndBuild()>.
+#
+sub BeginBuild #(hasChanged)
+    {
+    };
+
+
+#
+#   Function: EndBuild
+#
+#   Define this function if the package needs to do anything at the end of the build process.  This function will be called every time
+#   Natural Docs is run, even if the project hasn't changed.  This allows you to manage dependencies specific to the output
+#   format that may change independently from the source tree.  For example, <NaturalDocs::Builder::HTML> needs to keep the
+#   CSS files in sync regardless of whether the source tree changed or not.
+#
+#   Parameters:
+#
+#       hasChanged - Whether the project has changed, such as source files or the menu file.  If false, the only other function that
+#                            was called was <BeginBuild()>.
+#
+sub EndBuild #(hasChanged)
+    {
+    };
+
+
+#
+#   Function: BuildIndex
+#
+#   Define this function to create an index for the passed topic.  You can get the index from
+#   <NaturalDocs::SymbolTable->Index()>.
+#
+#   The reason it's not passed directly to this function is because indexes may be time-consuming to create.  As such, they're
+#   generated on demand because some output packages may choose not to implement them.
+#
+#   Parameters:
+#
+#       topic  - The <TopicType> to limit the index by.
+#
+sub BuildIndex #(topic)
+    {
+    };
+
+
+#
+#   Function: UpdateImage
+#
+#   Define this function to add or update the passed image in the output.
+#
+#   Parameters:
+#
+#       file - The image <FileName>
+#
+sub UpdateImage #(file)
+    {
+    };
+
+
+#
+#   Function: PurgeFiles
+#
+#   Define this function to make the package remove all output related to the passed files.  These files no longer have Natural Docs
+#   content.
+#
+#   Parameters:
+#
+#       files - An existence hashref of the files to purge.
+#
+sub PurgeFiles #(files)
+    {
+    };
+
+
+#
+#   Function: PurgeIndexes
+#
+#   Define this function to make the package remove all output related to the passed indexes.  These indexes are no longer part
+#   of the menu.
+#
+#   Parameters:
+#
+#       indexes  - An existence hashref of the <TopicTypes> of the indexes to purge.
+#
+sub PurgeIndexes #(indexes)
+    {
+    };
+
+
+#
+#   Function: PurgeImages
+#
+#   Define this function to make the package remove all output related to the passed image files.  These files are no longer used
+#   by the documentation.
+#
+#   Parameters:
+#
+#       files - An existence hashref of the image <FileNames> to purge.
+#
+sub PurgeImages #(files)
+    {
+    };
+
+
+#
+#   Function: UpdateMenu
+#
+#   Define this function to make the package update the menu.  It will only be called if the menu changed.
+#
+sub UpdateMenu
+    {
+    };
+
+
+1;