View source for Code Completion Design

[[Category:Developer Documentation]]
==How to build==
===Get the source code===
When you download the  svn source code of code::blocks,(see here [[Installing_Code::Blocks_from_source_on_Windows#Code::Blocks_sources]] the source code of CodeCompletion plugin was already included. 

Here is the screen shot when I opened the codecompletion source code in C::B's source navigator view.

[[Image:Devcc1.png|300x300px| Code completion source tree opened in code::blocks]]

===Build the code completion plug-in===
If you only want to build the Code Completion plug-in, you can select it in the "build target list" (Note: by default, the Build target option was "ALL", which means all the targets in the current workspace will be build), see the screen shot below:

[[Image:Buildcc.PNG|frame|none| Code completion build target option in code::blocks]]

Don't forget to run the batch script file "update.bat" after you build the target, this procedure will update your output folder package and strip the debug information. See the wiki page [[Installing_Code::Blocks_from_source_on_Windows]] for more information.

===A belief description of every project files===
From now on, we use '''CC''' as an abbreviation for '''code completion plug-in'''.
{|border="1" cellpadding="1" cellspacing="0" style="font-size: 100%; border: gray solid 1px; border-collapse: collapse; text-align: left; width: 100% table-layout: unfixed;" 
|-
| ccdebuginfo.cpp || a dialog for debugging CC, can be opened by double click on the code browser tree entry with shift and ctrl key pressed 
|- 
| ccoptionsdlg.cpp|| code completion options dialog, can be opened by menu->setting->editor->code completion and symbols browser
|-
| ccoptionsprjdlg.cpp|| setting the additional parser search path
|-
| classbrowser.cpp||viewing the symbols tree ctrl(token tree).
|-
| classbrowserbuilderthread.cpp|| a thread to build the class browser tree above
|-
| codecompletion.cpp ||The Main file need by code completion plug-in, maintain all the CC's GUI and native parser
|-
| insertclassmethoddlg.cpp|| a dialog to insert class method, can be open by context menu in editor
|-
| nativeparser.cpp||a class derived from wxEvtHandler, NativeParser class has a member variable "Parser m_Parser";
|-
| selectincludefile.cpp|| select multiply matched token names before any jump to declaration or jump to implementation.
|-
|parser/parser.cpp|| Parser class was also derived from wxEvtHandler, can start batch parse... this class has member variables like :cbThreadPool m_Pool(which will create a thread from thread pool for each file need passed);TokensTree* m_pTokens(contains all the Token database);
|-
|parser/parserthread.cpp||will do the syntax analysis for every file in a project, it has a Tokenizer as member variable
|-
|parser/token.cpp|| definition of the "Token" class, and TokensTree(which means the Token dababase)
|-
|parser/tokenizer.cpp|| tokenizer will return every wxString it regard as a symbol by GetToken(), also do a replacement before return
|-
|parser/searchtree.cpp|| implement the patricia search tree using by TokensTree
|-
|Header files|| no description needed
|}

==Low level parser(Lexical analysis, Tokenize)==
For someone haven't heard what does "Token" and "Tokenize" mean, you should read the wikibooks article [http://en.wikibooks.org/wiki/C%2B%2B_Programming/Compiler#Compilation A brief explain of what does a parser do] and [http://en.wikipedia.org/wiki/Lexical_analysis Tokenize on wikipedia]. Shortly, a Tokenizer regards your C++ or C source code as a large array of characters (sometimes, we call it a string), then this big string can be divided to small atomic strings----a '''token''' (Each token has a unique meanings and can't be divided into sub-strings. A token can be a symbol, an identifier, a keyword, a digital number, etc), meanwhile "white spaces" and "comments" were ignored by the Tokenizer.

for a simple c++ program like below
<source lang = "c">
int main()
{
    std::cout << "hello world" << std::endl;
    return 0;
}
</source>
After tokenizing, it should give these 15 tokens
<source lang = "cpp">
1 = string "int"
2 = string "main"
3 = opening parenthesis
4 = closing parenthesis
5 = opening brace
6 = string "std"
7 = namespace operator
8 = string "cout"
9 = << operator
10 = string ""hello world""
11 = string "endl"
12 = semicolon
13 = string "return"
14 = number 0
15 = closing brace
</source>

===Tokenizer class===
A class named '''Tokenizer''' was implemented in "tokenizer.h" and "tokenizer.cpp". Seen from the Tokenizer's level, a token is just a [http://docs.wxwidgets.org/stable/wx_unicode.html unicode] wxString. There are several steps to run an instance of Tokenizer. 
===preparing the big string===
From the previous section, you know the Tokenizer is just a string cutter. So, the first step is preparing the big string. The string can either be loaded from a source file or a memory buffer. Currently, the Unicode wxString is used.(since we are all using Unicode build of code::blocks, and ANSI mode is outdated and deprecated).
===Get or Peek a token===
'''Tokenizer''' contains a '''file position indicator'''----m_TokenIndex(see [http://en.wikipedia.org/wiki/Fopen File Open In C language]) pointing to the current position of string. So, you can '''Get''' a token or '''Peek''' a token. Here is the function prototype
<source lang = "cpp">
    //Get the current token string started from m_TokenIndex, 
    //After that, increase the m_Tokenindex
    wxString GetToken();

    //Peak the current token string and but do *NOT* increase the m_TokenIndex
    wxString PeekToken();
</source>

For example, if the Tokenizer parses the example code above, you can see how these two methods work.

*After initializing the Tokenizer(preparing the big string), You firstly call the GetToken() function, which will return the first token "int" and increase the m_TokenIndex one step to "int".

*Then, if you call the PeekToken(), which will return a token "main", but the tokenindex was still remaining point to "int". 

*If you call the GetToken() again, it will return a "main" immediately and increase the file pointer to "main". 

Note: Internally, the Tokenizer class use a '''undo and peek cache''' to do the trick. Once a token is peeked, it is saved in '''m_Peek''' member, so, next time you call GetToken(), it just immediately return the cached value without calling the "DoGetToken()" procedure again. 

[[Image:Cb token cache.png]]

===Nested Value of braces===
As you know, braces are exist in pairs, such as "{" and "}", or "[" and "]", Also, these braces can be embeded. So the Tokenizer keep a value '''m_NestLevel''' to indicate how deep you stays. If the Tokenizer meets a '''{''', it will increase the nestValue, and if it meets a '''}''', it will decrease the m_NestLevel. See the pseudo code in Tokenizer.cpp below:
<source lang = "cpp">
        if (CurrentChar() == '{')
            ++m_NestLevel;
        else if (CurrentChar() == '}')
            --m_NestLevel;
</source>

===SkipUnwanted tokens===
A '''bool''' member variable '''m_SkipUnwantedTokens''' determines whether a regular assignments statement will be omitted or not. For example, a piece of code below:
<source lang = "cpp">
a = b + c;
</source>

If m_SkipUnwantedTokens == true, then once SkipUnwanted() meets the "=" symbol, it will go on skipping every characters until it meets ''',''' or ''';''' or '''}''', so this statement will be omitted by the Tokenizer. 

Sometimes, this behavior becomes a nightmare when parsing the statement like default argument in template. Or the default arguments in function.
<source lang = "cpp">
// assignment statement in template arguments
template<class T = int> 
class abc {
 T m_a;
 ......
 ......
}

// assignment statement in function arguments

void foo( int arg1 = 60 )
{
    bla bla;
}
</source>
If the Tokenizer finds that an equation sign '''=''', it will skip any character until it meets a '''}''', so, the class declaration or function body will be totally skipped in the unexpected way. In this case, we should manually '''disable''' this functionality by setting m_SkipUnwantedTokens = false to let the Tokenizer parse these statements correctly.

That's why you will see many situations when you enter a function, you should save the m_SkipUnwantedTokens statues and disabled it, when you leave a function, you should manually restore it.(See function implementation in ParseThread.cpp)

===Return a correct token, Macro replacement===
Special token should be replaced for parsing correctly. For example, in the standard c++ header (mingw), there are a string named "_GLIBCXX_STD", this should be replaced to "std". See the dialog below.

[[Image:Cc std replacement.png|400x400px|]]

The inline function in the Tokenizer class will check whether a token should be replaced before return. 

<source lang = "cpp">
wxString Tokenizer::MacroReplace(const wxString str)
{
    wxStringHashMap::const_iterator it = s_Replacements.find(str);

    if (it != s_Replacements.end())
    {
        // match one!
        wxString key   = it->first;
        wxString value = it->second;
        TRACE(_T("MacroReplace() : Replacing '%s' with '%s' (file='%s')."), key.wx_str(), value.wx_str(), m_Filename.wx_str());
        if (value[0]=='+' && CurrentChar()=='(')
        {
            unsigned int start = m_TokenIndex;
            m_Buffer[start] = ' ';
            bool fillSpace = false;
            while (m_Buffer[start]!=')')
            {
                if (m_Buffer[start]==',')
                    fillSpace = true;

                if (fillSpace==true)
                    m_Buffer[start]=' ';

                start++;
            }
            m_Buffer[start] = '{';
            return value.Remove(0,1);
        }
        else if (value[0] == '-')
        {
            unsigned int lenKey = key.Len();
            value = value.Remove(0,1);
            unsigned int lenValue = value.Len();

            for (unsigned int i=1; i<=lenKey; i++)
            {
                if (i < lenValue)
                    m_Buffer[m_TokenIndex-i] = value[lenValue-i];
                else
                    m_Buffer[m_TokenIndex-i] = ' ';
            }

            int firstSpace = value.First(' ');
            // adjust m_TokenIndex
            m_TokenIndex = m_TokenIndex - lenValue + firstSpace;

            return value.Mid(0,firstSpace);
        }
        else
            return value;
    }

    return str;
}

</source>


[[Image:Cc replace dialog.png|400x400px|none| Code completion build target option in code::blocks]]

Setting the replacement mapping. Note that before return a token, a replacement map was searched to check if it matches any entry in the map, so, the bigger this map goes, the slower it will do parsing.

Note: Code Completion plug-in is not a preprocessor, so it is difficult to deal with the source mixed with many macro, or some strange macros. This is something like Ctags' replacement options "?I identifier?list" in [http://ctags.sourceforge.net/ctags.html#OPERATIONAL%20DETAILS ctags option detial] or [http://codelite.org/LiteEditor/FAQ Code Completion macro FAQ]

For discussion of Ctags, the related thread is [http://forums.codeblocks.org/index.php/topic,10564.msg72424.html#msg72424 Re: namespaces and code completion] and [http://forums.codeblocks.org/index.php/topic,11187.msg77135.html#msg77135 Re: New code completion remarks/issues]

===replacement rules===

For example, when 

<source lang = "cpp">
_GLIBCXX_BEGIN_NESTED_NAMESPACE(std, _GLIBCXX_STD_D)

  /// See bits/stl_deque.h's _Deque_base for an explanation.
  template<typename _Tp, typename _Alloc>
    struct _Vector_base
    {
      typedef typename _Alloc::template rebind<_Tp>::other _Tp_alloc_type;

      struct _Vector_impl
      : public _Tp_alloc_type
      {
    typename _Tp_alloc_type::pointer _M_start;
    typename _Tp_alloc_type::pointer _M_finish;
    typename _Tp_alloc_type::pointer _M_end_of_storage;

    _Vector_impl()
    : _Tp_alloc_type(), _M_start(0), _M_finish(0), _M_end_of_storage(0)
    { }

    _Vector_impl(_Tp_alloc_type const& __a)
    : _Tp_alloc_type(__a), _M_start(0), _M_finish(0), _M_end_of_storage(0)
    { }
      };

    ......

_GLIBCXX_END_NESTED_NAMESPACE
</source>

We should replace the string "_GLIBCXX_BEGIN_NESTED_NAMESPACE(std, _GLIBCXX_STD_D)" 
to "namespace std {", here, we introduce a kind of replacement rule:

_GLIBCXX_BEGIN_NESTED_NAMESPACE   +namespace

==High level parser(Syntax Analysis)==
This is the main flow of a parser.

[[Image:Parser_Flow.gif]]
===parser thread===
Basically, we can say, the low level parser(Tokenizer) moves its pointer character by character, and return a wxString(token) to feed the high level parser(Syntax analyzer).All the syntax analysis was done in ParserThread. A thread must be created to parse a source file. see parserthread.cpp and parserthread.h, a thread will be allocated from thread pool. For example, a file contains these statement:
<source lang = "cpp">
    void  f1();
    int   f2(char c);
    float f3(void * p);
    int   f1;
    double f2;
</source>
After the ParserThread finished its parsing, it will recognize five tokens, which has the keyword "f1","f2" and "f3", note, tokens can have the same names, but they differ from different types( variables, functions...).

===Token class===
How can a large number of tokens be recorded? A '''Token'''(note: it as a capital means it's class type) class was introduced to recorded every token. For boosting the speed of allocating Tokens, the "new" and "delete" operator were overloaded in its base class BlockAllocated. See the [http://en.wikipedia.org/wiki/Memory_pool memory pool] page on wikipedia as a reference.
<source lang = "cpp">
class Token  : public BlockAllocated<Token, 10000>
{
        ......
        wxString m_Type; // this is the return value (if any): e.g. const wxString&
        wxString m_ActualType; // this is what the parser believes is the actual return value: e.g. wxString
        wxString m_Name;
        wxString m_Args;
        wxString m_AncestorsString; // all ancestors comma-separated list
        unsigned int m_File;
        unsigned int m_Line;
        unsigned int m_ImplFile;
        unsigned int m_ImplLine; // where the token was met
        unsigned int m_ImplLineStart; // if token is impl, opening brace line
        unsigned int m_ImplLineEnd; // if token is impl, closing brace line
        TokenScope m_Scope;
        TokenKind m_TokenKind;
        bool m_IsOperator;
        bool m_IsLocal; // found in a local file?
        bool m_IsTemp; // if true, the tree deletes it in FreeTemporaries()
        bool m_IsConst;    // the member method is const (yes/no)

        int m_ParentIndex;
        TokenIdxSet m_Children;
        TokenIdxSet m_Ancestors;
        TokenIdxSet m_DirectAncestors;
        TokenIdxSet m_Descendants;
        ......

   
};
</source>
You can see the Token class contains all the information needed for recording its locating, its type or class derived hierarchy... 

For example, in the source code in[Low level parser(Lexical analysis)]. A Token for "main" should contains it's name (obviously , m_Name="main" ), then m_File will record which file dose this Token exist. m_Line will give the line number of "main" in this source file, and so on.

===Memory Pool--BlockAllocated class===
In BlockAllocated class, there is only a static member say "static BlockAllocator<T, pool_size, debug> allocator;" to keep all the pre-allocated memory for all derived class.

'''10000''' means a pool of 10000 Tokens were allocated in the memory pool, so, dynamically allocate a Token object will be fast and efficient. 

[[Image:Cb blockalloc.png|frame|none|  Operator new overloading for fast allocate in the heap]]

===ParserThread===
The function Parse() will do the most job of syntax analysis. See the pseudo code below.
<source lang = "cpp">
ParserThread::Parse()
{
   ......
   do
    {
        ......
        DoParse();
        ......
 
        
    }while(false);

    return result;
}

</source>

In the DoParse(), it checks the token from Tokenizer. For example, if the token words = "enum", then, the ParserThread::HandleEnum() will do the job to parse this enum block.


===A simple look ahead parser===
We can explain a little about it parser, the member variable '''m_Str''' of class ParserThread will be considered as a '''type stack''', for example, we want to parse the statement below:

<source lang = "cpp">
symbolA symbolB symbolC symbolD;
</source>

Only symbolD can be recognized as a variable, and it has a type of "symbolA symbolB symbolC". When the parser meets each symbol, it will look ahead to see the next token is whether ";", if not, the current token will '''pushed''' to m_Str. These iteration will be ended when the parser look ahead one step from symbolD and find the next token is a ";".

===DoParse Examples===
====Handling class declaration====
The main routing of handling class was in 
ParserThread::HandleClass function
If the parserThread meet these statement
<source lang = "cpp">
class AAA{
    int m_a;
    int m_b;
} ;
</source>
It will surely add a Token of "AAA", it's type is "class".

What about these statements
<source lang = "cpp">
class AAA{
    int m_a;
    int m_b;
} x,y,z;
</source>
The parserThread should firstly add a Token "AAA", then it should regard "x", "y", "z" are three variables of type "AAA". This was done by "ReadVarNames()" function, it will read every comma separated variables after a class declaration.

====Handling typedef statement====
Sometimes, you can see these code blocks:
<source lang = "cpp">
typedef class AAA{
    int m_a;
    int m_b;
} BBB,CCC;
</source>
If the parser meets the keyword "typedef", firstly it set the "m_ParsingTypedef = ture" to indicate that we are parsing a typedef statement.

Next, it meets a keyword "class", which also indicate it is a class declaration. So, the HandleClass function will do the task. A Token of "class AAA" will be added to the TokensTree. Wait a minute, how can we deal with "BBB" and "CCC", this is not the same case as previous code, we can't regard "BBB","CCC" as variables. In this case, another function "ReadClsName()" will be called. For simplicity of TokensTree, we just regard "BBB" and "CCC" as derived class of "AAA".

How does the parserThread know between these two cases. Here is the code:

<source lang = "cpp">
                if (m_ParsingTypedef)
                {
                    m_Str.Clear();
                    ReadClsNames(newToken->m_Name);
                    // m_ParsingTypedef = false;
                    break;

                }
                else
                {
                    m_Str = newToken->m_Name;
                    ReadVarNames();
                    m_Str.Clear();
                    break;
                }
</source>

That's the whole thing I would like to write about handling class and handling typedef.

==TokensTree&SearchTree==
Maybe, you would ask a question: where do these Tokens store? The answer is that all the Tokens will be recorded in "TokensTree class".

When a certain Token is identified(whether it's a global variable, a class declaration, a class member function, and so on), it will be inserted to a database(TokensTree). 

Furthermore, for fast Token query, Tokens should be sorted by it's wxString m_Name member; 

A '''compact Patricia tree'''(see the wikipedia [http://en.wikipedia.org/wiki/Radix_tree Patricia tree on wikipedia]) is built to hold all their names. 

For example, If you add three item to the TokensTree.
<source lang = "cpp">
    mytree->AddItem(_T("physics"),_T("1 - uno"));
    mytree->AddItem(_T("physiology"),_T("2 - dos"));
    mytree->AddItem(_T("psychic"),_T("3 - tres"));
</source>

The Patricia tree structure will show as below, the edge of a tree contains a "label string" and the number in parentheses refers to a node Id.

<source lang = "cpp">
- "" (0)
      \- "p" (4)
              +- "hysi" (2)
              |          +- "cs" (1)
              |          \- "ology" (3)
              \- "sychic" (5)
</source>
===Animation of Patricia tree===
A web page gives an animation of building a Patricia tree, you can see how the tree node added. See here [http://bootstrapping.wordpress.com/2008/01/24/prefix-searching-with-radix-tree/ Prefix searching with Radix Tree] 
 
===SearchTree Node===
A Node should contains at least several essential element. They are:
# An edge: in the image below, the gray filled area shows a Node, since each edge belong to the next node, for example, the Node has an edge "abcd".
# SearchTreeItemMap: because our searchTree is a compact Patricia tree which means an edge contains many items. Such as, "ab" , "abc", "abcd" all the suffix string belong to the edge "abcd" 
# SearchTreeLinkMap: This is an associated container map<wxChar,NodeId>, which store all "pointers" to its child Node. In the example below, "e" points to a Node of "efg", while "x" points to "xyz".
# Node depth: depth of Search Tree Node is defined by the string length from the "root node". See a depth of each node on the search tree above. For example, the Node of "hysi" has a m_Depth = 5 ("" + "p" + "hysi" = 5).

===Node Lable===
For example, the node "hysi" (2) has two children, they are "cs" (1) and "ology" (3), show below.

[[Image:Cb tree node lable.png]]

For more information, see the forum discussion here. [http://forums.codeblocks.org/index.php/topic,1696.0.html rickg22's SearchTree development] as a reference.

===How to query a Token by a keyword===
The parser collect all the Token information, and stored them in the TokensTree, the GUI function can query keywords from the database to show function tips or to build a Class Browser tree.

For example, if you want to find all the Tokens named "ab". In the picture above from TokenDatabase(TokensTree). we can search on the Patricia tree containing all the Tokens names, finaly, we find a tree node with a edge "abcd". So, "ab" is in it's Node's items list. Then, we can find a TokenIdxSet in a vector<TokenIdxSet>, this TokenIdxSet has all the index named by "ab", so, we can get the result like: There are many Tokens named "ab".Token "ab" may be a member varialbe name in a class, or a global function name...

<source lang = "cpp">
ClassA::ab
ClassB::ab
void ab()
....
</source>

[[Image:CCTokenTree1.png]]

==Flexible Parser structure==
The Parser class( in Parser.cpp and Parser.h ) has re-factored to support every project associate with a Parser object (svn revision > 6268 ). Which means: One Parser object per project, so, every project (xxx.cbp) will hold its own Token macro defines and Token trees. 

Due to the new introduced conditional preprocessor handling mechanism, the same source file may give different Tokens due to the different macro defines in different project. Also, CC support parsing on the files which does not belong to any project. 


#You open a project in CB, then all the files belong to the project( related to the project) will be parsed in CC.
#Now, you open another file( we call it a "separate file" later) which is not belong to the current project nor any other opened project.
#Then the CC will create a "temporary parser" named "NONE“ parser, and add this separate file to the "NONE" parser.
#So, you can still see the class tree when viewing the separate file.
#Once you close this separate file,  the "NONE" parser will automatically be removed.

The main idea is: Now, We can let the CC do parse on some separate files
and support class browsing in Symbol browser.

==Automatic Code Completion==
===Find the search scope===

[[Image:CC SearchScope.png]]

For example, when you are editing, and the caret position was located at "E" as shown in the above image, the first thing doing an automatic code completion is find the '''search scope'''.

Here are some steps to get the initial(first) search scope:

*First, you need to correct all the "using namespace XXXX" directives at the beginning the the current source file.(labeled with "A"), this means all the child tokens of "namespace Scintilla" should be searched.
*Secondly, find the function body where the current caret position locates. eg, the above code, the caret is in the "RunStyles::RunFromPosition()" function, so the function parameter variable "position"(labeled with "C" and the local variable "run" (labeled as "D") is also a matching Tokens.
*Thirdly, look at the "B", this means we are in the class named "RunStyles", so, "RunStyles" is also a search scope.
*Finally, don't forget the global namespace scope, because the tokens of this scope are exposed to everywhere of the source.

===Break up the current statement===
<source lang = "cpp">
 VariableA.m_VariableB.Functi
</source>
If you are entering the above statement, the caret is behind the "Functi". Then the whole line will be break up to several "parsing component".

There are three components: "'''VariableA'''" , "'''m_VariableB'''" and "'''Functi'''". In the next section, I will introduce that the AI matching routing will start from matching the first component with the initial(first) search scope.

===Do an AI match===
Doing an AI match is just like a walking on a tree, and find the final matching result.See the image below:

[[image:CC AI Match.png]]

'''Here is the matching algorithm''':

Now, suppose we have an initial "search scope", and a "component" array. 

The AI match algorithm starts from matching the "sub search scopes" with the first "component", then, the matched "sub search scope"(red rectangle) will be the initial "search scope" again matching with the second "component", this matching routing is running continuously until we matches with the final component. Then we get the whole matching result, that will be show as a auto code completion list.

==Code completion debugging support==
===Debug Log output===
If you want to debug your plug-in, you may need to Logout the debug message to the "Code::Blocks Debug" panel. Here is the sample code:

<source lang = "cpp">
Manager::Get()->GetLogManager()->DebugLog(_("XXXXX "));

wxString name;
wxString args;
wxString m_Str;
//.....
Manager::Get()->GetLogManager()->DebugLog(F(_T("Add token name='")+name+_T("', args='")+args+_T("', return type='") + m_Str+ _T("'")));

</source>

Also, you need start the Code::Blocks with the command line argument. For example in windows.

'''codeblocks.exe --debug-log'''

then a Code::blocks debug panel will be shown to display the log.
[[Image:CbDebugLog.png|frame|none|  Debug Log output panel]]

===Code-Completion debug tool dialog===
When you press '''shift''' and '''ctrl''' key and double click on any entry of the navigator tree, a debug tool dialog will pop up to give a more detail about the selected token. You can query its information such as its member variables, its Ancestors and so on.

[[Image:CcDebugToolDialog.png]]

==Usefull Links==

*A discussion on search tree in the forum [http://forums.codeblocks.org/index.php/topic,1696.0.html] and [http://forums.codeblocks.org/index.php/topic,1581.0.html].

*Another opensource IDE [http://vcfbuilder.org/?q=node/139 VCF builder]or[https://sourceforge.net/projects/vcfbuilder/ vcfbuilder on sf], By the way , VCF use a CodeStore based on [http://www.antlr.org/ antlr parser] to deal with parsing, the whole source code can be check out from 
<source lang = "cpp">
svn co https://vcfbuilder.svn.sourceforge.net/svnroot/vcfbuilder vcfbuilder
</source>


*online book[http://www.macs.hw.ac.uk/~alison/alg/details.html Data Structures and Algorithms II Course ] and [http://www.macs.hw.ac.uk/~alison/alg/lectures.html pdf lectures]

*A forum discussion on why a standard perprocessor and parser works slowly. [http://forums.codeblocks.org/index.php/topic,2494.msg19801.html#msg19801 parsing iostream header takes 5 seconds]

*[http://www.codelite.org CodeLite] is another open source IDE using wxWidgets, use Ctags as a indexing service, and do context expression syntactic analysis generated from Lex and Yacc tools.See the explanation in forum message by it's author eranif [http://forums.codeblocks.org/index.php/topic,10087.msg70875.html#msg70875 How does Codelite's CC work with CTags and Yacc&Lex]

*Hopefully, we can use the patricia tree from the standard library, see here:[http://gcc.gnu.org/onlinedocs/libstdc++/ext/pb_ds/trie_based_containers.html trie]

* a blog from KDE4's developer, it explain the code completion in KDE4, see here:[http://zwabel.wordpress.com/ zwabel' blog]

* QT creator source code from [http://qt.gitorious.org/qt-creator/qt-creator/trees/master QT-creator source]

* [http://websvn.kde.org/trunk/KDE/kdevelop/ Kdevelop source code] and it's [http://websvn.kde.org/trunk/KDE/kdevelop/languages/cpp/parser/ parser]

* [http://stackoverflow.com/questions/1220099/how-does-code-completion-work How does code completion work? - Stack Overflow]

* [http://forums.codeblocks.org/index.php/topic,11724.0.html Just a reminder Codelite IDE now have improved it's parser]

*[http://padator.org/software-yacfe.php Yacfe, a parser with proprocessor enabled]

*[http://forums.codeblocks.org/index.php/topic,1581.msg11701.html#msg11701 A Mini C++ interpreter] from the book "The Art of C++ by Herbert Schildt", source code can be download [http://books.mcgraw-hill.com/downloads/products/0072255129/0072255129_code.zip here]

*[http://www.macs.hw.ac.uk/~alison/alg/lectures/l7.pdf Lecture 7: Parsing (intro) (pdf)] and [http://www.macs.hw.ac.uk/~alison/alg/lectures/l8.pdf Lecture 8: Recursive Descent Parsing (pdf)] from [http://www.macs.hw.ac.uk/~alison/alg/lectures.html Data Structures and Algorithms II]

*There are some online "Compiler design" books listed in [http://www.freetechbooks.com/compiler-design-and-construction-f14.html Free Online Compiler Design and Construction Books :: FreeTechBooks.com], and the best one I think is : [http://www.diku.dk/hjemmesider/ansatte/torbenm/Basics/index.html Basics of Compiler Design], and you can get a basic understanding.