X-Git-Url: https://git.saurik.com/wxWidgets.git/blobdiff_plain/e12be2f7fb5f55553795b7c4168007173511d581..02a3b39120a618998dc690e95c3983606ea97745:/docs/latex/wx/list.tex diff --git a/docs/latex/wx/list.tex b/docs/latex/wx/list.tex index 5e49493589..92259d14ae 100644 --- a/docs/latex/wx/list.tex +++ b/docs/latex/wx/list.tex @@ -9,7 +9,7 @@ when random access is required. While wxList class in the previous versions of wxWindows only could contain elements of type wxObject and had essentially untyped interface (thus allowing you to put apples in the list and read back oranges from it), the new wxList -classes family may contain elements of any type and has much more stricter type +classes family may contain elements of any type and has much more strict type checking. Unfortunately, it also requires an additional line to be inserted in your program for each list class you use (which is the only solution short of using templates which is not done in wxWindows because of portability issues). @@ -19,13 +19,12 @@ data but make all of its dangerous (because untyped) functions protected, so that they can only be used from derived classes which, in turn, expose a type safe interface. With this approach a new wxList-like class must be defined for each list type (i.e. list of ints, of wxStrings or of MyObjects). This is done -with {\it WX\_DECLARE\_LIST} and {\it WX\_IMPLEMENT\_LIST} macros like this +with {\it WX\_DECLARE\_LIST} and {\it WX\_DEFINE\_LIST} macros like this (notice the similarity with WX\_DECLARE\_OBJARRAY and WX\_IMPLEMENT\_OBJARRAY macros): \wxheading{Example} -{\small% \begin{verbatim} // this part might be in a header or source (.cpp) file class MyListElement @@ -51,8 +50,8 @@ macros): // also have MyList::Node type which is the type-safe version of wxNode. MyList list; MyListElement element; - list.Add(element); // ok - list.Add(17); // error: incorrect type + list.Append(element); // ok + list.Append(17); // error: incorrect type // let's iterate over the list for ( MyList::Node *node = list.GetFirst(); node; node = node->GetNext() ) @@ -62,11 +61,12 @@ macros): ...process the current element... } \end{verbatim} -} For compatibility with previous versions wxList and wxStringList classes are still defined, but their usage is deprecated and they will disappear in the -future versions completely. +future versions completely. The use of the latter is especially discouraged as +it is not only unsafe but is also much less efficient than +\helpref{wxArrayString}{wxarraystring} class. In the documentation of the list classes below, you should replace wxNode with wxListName::Node and wxObject with the list element type (i.e. the first @@ -175,7 +175,8 @@ The key string is copied and stored by the list implementation. \func{void}{Clear}{\void} -Clears the list (but does not delete the client data stored with each node). +Clears the list (but does not delete the client data stored with each node +unless you called DeleteContents(TRUE), in which case it deletes data). \membersection{wxList::DeleteContents}\label{wxlistdeletecontents} @@ -227,7 +228,7 @@ Returns the last node in the list (NULL if the list is empty). \func{int}{IndexOf}{\param{wxObject*}{ obj }} -Returns the index of {\it obj} within the list or NOT\_FOUND if {\it obj} +Returns the index of {\it obj} within the list or wxNOT\_FOUND if {\it obj} is not found in the list. \membersection{wxList::Insert}\label{wxlistinsert} @@ -289,8 +290,12 @@ Returns the number of elements in the list. Allows the sorting of arbitrary lists by giving a function to compare two list elements. We use the system {\bf qsort} function -for the actual sorting process. The sort function receives pointers to wxObject pointers (wxObject **), -so be careful to dereference appropriately. +for the actual sorting process. + +If you use untyped wxList the sort function receives pointers to wxObject +pointers (wxObject **), so be careful to dereference appropriately - but, +of course, a better solution is to use list of appropriate type defined with +{\tt WX\_DECLARE\_LIST}. Example: