Update Lua API doc html.

develop
Petr Mrázek 2012-04-05 02:42:08 +02:00
parent 91c20c0607
commit a6dd17919f
1 changed files with 184 additions and 5 deletions

@ -330,6 +330,11 @@ ul.auto-toc {
</li>
<li><a class="reference internal" href="#named-types" id="id7">Named types</a></li>
<li><a class="reference internal" href="#global-functions" id="id8">Global functions</a></li>
<li><a class="reference internal" href="#recursive-table-assignment" id="id9">Recursive table assignment</a></li>
</ul>
</li>
<li><a class="reference internal" href="#dfhack-utilities" id="id10">DFHack utilities</a><ul>
<li><a class="reference internal" href="#persistent-configuration-storage" id="id11">Persistent configuration storage</a></li>
</ul>
</li>
</ul>
@ -430,17 +435,22 @@ or as a result of calling the <tt class="docutils literal">_field()</tt> method.
<ul>
<li><p class="first"><tt class="docutils literal">ref.field</tt>, <tt class="docutils literal">ref.field = value</tt></p>
<p>Valid fields of the structure may be accessed by subscript.</p>
<p>In case of inheritance, <em>superclass</em> fields have precedence
<p>Primitive typed fields, i.e. numbers &amp; strings, are converted
to/from matching lua values. The value of a pointer is a reference
to the target, or nil/NULL. Complex types are represented by
a reference to the field within the structure; unless recursive
lua table assignment is used, such fields can only be read.</p>
<p><strong>NOTE:</strong> In case of inheritance, <em>superclass</em> fields have precedence
over the subclass, but fields shadowed in this way can still
be accessed as <tt class="docutils literal"><span class="pre">ref['subclasstype.field']</span></tt>.</p>
<p>This shadowing order is necessary because vtable-based classes
be accessed as <tt class="docutils literal"><span class="pre">ref['subclasstype.field']</span></tt>.
This shadowing order is necessary because vtable-based classes
are automatically exposed in their exact type, and the reverse
rule would make access to superclass fields unreliable.</p>
</li>
<li><p class="first"><tt class="docutils literal">ref._field(field)</tt></p>
<p>Returns a reference to a valid field. That is, unlike regular
subscript, it returns a pointer reference even for primitive
typed fields.</p>
subscript, it returns a reference to the field within the structure
even for primitive typed fields and pointers.</p>
</li>
<li><p class="first"><tt class="docutils literal"><span class="pre">ref:vmethod(args...)</span></tt></p>
<p>Named virtual methods are also exposed, subject to the same
@ -574,6 +584,175 @@ lightuserdata (step is mandatory then).</p>
</li>
</ul>
</div>
<div class="section" id="recursive-table-assignment">
<h2><a class="toc-backref" href="#id9">Recursive table assignment</a></h2>
<p>Recursive assignment is invoked when a lua table is assigned
to a C++ object or field, i.e. one of:</p>
<ul class="simple">
<li><tt class="docutils literal"><span class="pre">ref:assign{...}</span></tt></li>
<li><tt class="docutils literal">ref.field = <span class="pre">{...}</span></tt></li>
</ul>
<p>The general mode of operation is that all fields of the table
are assigned to the fields of the target structure, roughly
emulating the following code:</p>
<pre class="literal-block">
function rec_assign(ref,table)
for key,value in pairs(table) do
ref[key] = value
end
end
</pre>
<p>Since assigning a table to a field using = invokes the same
process, it is recursive.</p>
<p>There are however some variations to this process depending
on the type of the field being assigned to:</p>
<ol class="arabic">
<li><p class="first">If the table contains an <tt class="docutils literal">assign</tt> field, it is
applied first, using the <tt class="docutils literal">ref:assign(value)</tt> method.
It is never assigned as a usual field.</p>
</li>
<li><p class="first">When a table is assigned to a non-NULL pointer field
using the <tt class="docutils literal">ref.field = <span class="pre">{...}</span></tt> syntax, it is applied
to the target of the pointer instead.</p>
<p>If the pointer is NULL, the table is checked for a <tt class="docutils literal">new</tt> field:</p>
<ol class="loweralpha simple">
<li>If it is <em>nil</em> or <em>false</em>, assignment fails with an error.</li>
<li>If it is <em>true</em>, the pointer is initialized with a newly
allocated object of the declared target type of the pointer.</li>
<li>Otherwise, <tt class="docutils literal">table.new</tt> must be a named type, or an
object of a type compatible with the pointer. The pointer
is initialized with the result of calling <tt class="docutils literal">table.new:new()</tt>.</li>
</ol>
<p>After this auto-vivification process, assignment proceeds
as if the pointer wasn't NULL.</p>
<p>Obviously, the <tt class="docutils literal">new</tt> field inside the table is always skipped
during the actual per-field assignment processing.</p>
</li>
<li><p class="first">If the target of the assignment is a container, a separate
rule set is used:</p>
<ol class="loweralpha">
<li><p class="first">If the table contains neither <tt class="docutils literal">assign</tt> nor <tt class="docutils literal">resize</tt>
fields, it is interpreted as an ordinary <em>1-based</em> lua
array. The container is resized to the #-size of the
table, and elements are assigned in numeric order:</p>
<pre class="literal-block">
ref:resize(#table);
for i=1,#table do ref[i-1] = table[i] end
</pre>
</li>
<li><p class="first">Otherwise, <tt class="docutils literal">resize</tt> must be <em>true</em>, <em>false</em>, or
an explicit number. If it is not false, the container
is resized. After that the usual struct-like 'pairs'
assignment is performed.</p>
<p>In case <tt class="docutils literal">resize</tt> is <em>true</em>, the size is computed
by scanning the table for the largest numeric key.</p>
</li>
</ol>
<p>This means that in order to reassign only one element of
a container using this system, it is necessary to use:</p>
<pre class="literal-block">
{ resize=false, [idx]=value }
</pre>
</li>
</ol>
<p>Since nil inside a table is indistinguishable from missing key,
it is necessary to use <tt class="docutils literal">df.NULL</tt> as a null pointer value.</p>
<p>This system is intended as a way to define a nested object
tree using pure lua data structures, and then materialize it in
C++ memory in one go. Note that if pointer auto-vivification
is used, an error in the middle of the recursive walk would
not destroy any objects allocated in this way, so the user
should be prepared to catch the error and do the necessary
cleanup.</p>
</div>
</div>
<div class="section" id="dfhack-utilities">
<h1><a class="toc-backref" href="#id10">DFHack utilities</a></h1>
<p>DFHack utility functions are placed in the <tt class="docutils literal">dfhack</tt> global tree.</p>
<p>Currently it defines the following features:</p>
<ul>
<li><p class="first"><tt class="docutils literal"><span class="pre">dfhack.print(args...)</span></tt></p>
<p>Output tab-separated args as standard lua print would do,
but without a newline.</p>
</li>
<li><p class="first"><tt class="docutils literal"><span class="pre">print(args...)</span></tt>, <tt class="docutils literal"><span class="pre">dfhack.println(args...)</span></tt></p>
<p>A replacement of the standard library print function that
works with DFHack output infrastructure.</p>
</li>
<li><p class="first"><tt class="docutils literal"><span class="pre">dfhack.printerr(args...)</span></tt></p>
<p>Same as println; intended for errors. Uses red color and logs to stderr.log.</p>
</li>
<li><p class="first"><tt class="docutils literal"><span class="pre">dfhack.color([color])</span></tt></p>
<p>Sets the current output color. If color is <em>nil</em> or <em>-1</em>, resets to default.</p>
</li>
<li><p class="first"><tt class="docutils literal">dfhack.is_interactive()</tt></p>
<p>Checks if the thread can access the interactive console and returns <em>true</em> or <em>false</em>.</p>
</li>
<li><p class="first"><tt class="docutils literal"><span class="pre">dfhack.lineedit([prompt[,history_filename]])</span></tt></p>
<p>If the thread owns the interactive console, shows a prompt
and returns the entered string. Otherwise returns <em>nil, error</em>.</p>
</li>
<li><p class="first"><tt class="docutils literal"><span class="pre">dfhack.interpreter([prompt[,env[,history_filename]]])</span></tt></p>
<p>Starts an interactive lua interpreter, using the specified prompt
string, global environment and command-line history file.</p>
<p>If the interactive console is not accessible, returns <em>nil, error</em>.</p>
</li>
<li><p class="first"><tt class="docutils literal"><span class="pre">dfhack.pcall(f[,args...])</span></tt></p>
<p>Invokes f via xpcall, using an error function that attaches
a stack trace to the error. The same function is used by SafeCall
in C++, and dfhack.safecall.</p>
<p>The returned error is a table with separate <tt class="docutils literal">message</tt> and
<tt class="docutils literal">stacktrace</tt> string fields; it implements <tt class="docutils literal">__tostring</tt>.</p>
</li>
<li><p class="first"><tt class="docutils literal"><span class="pre">safecall(f[,args...])</span></tt>, <tt class="docutils literal"><span class="pre">dfhack.safecall(f[,args...])</span></tt></p>
<p>Just like pcall, but also prints the error using printerr before
returning. Intended as a convenience function.</p>
</li>
<li><p class="first"><tt class="docutils literal"><span class="pre">dfhack.with_suspend(f[,args...])</span></tt></p>
<p>Calls <tt class="docutils literal">f</tt> with arguments after grabbing the DF core suspend lock.
Suspending is necessary for accessing a consistent state of DF memory.</p>
<p>Returned values and errors are propagated through after releasing
the lock. It is safe to nest suspends.</p>
<p>Every thread is allowed only one suspend per DF frame, so it is best
to group operations together in one big critical section. A plugin
can choose to run all lua code inside a C++-side suspend lock.</p>
</li>
</ul>
<div class="section" id="persistent-configuration-storage">
<h2><a class="toc-backref" href="#id11">Persistent configuration storage</a></h2>
<p>This api is intended for storing configuration options in the world itself.
It probably should be restricted to data that is world-dependent.</p>
<p>Entries are identified by a string <tt class="docutils literal">key</tt>, but it is also possible to manage
multiple entries with the same key; their identity is determined by <tt class="docutils literal">entry_id</tt>.
Every entry has a mutable string <tt class="docutils literal">value</tt>, and an array of 7 mutable <tt class="docutils literal">ints</tt>.</p>
<ul>
<li><p class="first"><tt class="docutils literal">dfhack.persistent.get(key)</tt>, <tt class="docutils literal">entry:get()</tt></p>
<p>Retrieves a persistent config record with the given string key,
or refreshes an already retrieved entry. If there are multiple
entries with the same key, it is undefined which one is retrieved
by the first version of the call.</p>
<p>Returns entry, or <em>nil</em> if not found.</p>
</li>
<li><p class="first"><tt class="docutils literal">dfhack.persistent.delete(key)</tt>, <tt class="docutils literal">entry:delete()</tt></p>
<p>Removes an existing entry. Returns <em>true</em> if succeeded.</p>
</li>
<li><p class="first"><tt class="docutils literal"><span class="pre">dfhack.persistent.get_all(key[,match_prefix])</span></tt></p>
<p>Retrieves all entries with the same key, or starting with key..'/'.
Calling <tt class="docutils literal"><span class="pre">get_all('',true)</span></tt> will match all entries.</p>
<p>If none found, returns nil; otherwise returns an array of entries.</p>
</li>
<li><p class="first"><tt class="docutils literal"><span class="pre">dfhack.persistent.save({key=str1,</span> <span class="pre">...}[,new])</span></tt>, <tt class="docutils literal"><span class="pre">entry:save([new])</span></tt></p>
<p>Saves changes in an entry, or creates a new one. Passing true as
new forces creation of a new entry even if one already exists;
otherwise the existing one is simply updated.
Returns <em>entry, did_create_new</em></p>
</li>
</ul>
<p>Since the data is hidden in data structures owned by the DF world,
and automatically stored in the save game, these save and retrieval
functions can just copy values in memory without doing any actual I/O.
However, currently every entry has a 180+-byte dead-weight overhead.</p>
</div>
</div>
</div>
</body>