Development reference: Difference between revisions

From Bitpost wiki
No edit summary
No edit summary
Line 499: Line 499:
** windows: Qt Creator (OR Visual Studio OR eclipse?)
** windows: Qt Creator (OR Visual Studio OR eclipse?)
|}
|}
|}
<!--
===========================================================================================================================================================================================================================================================================================
-->
{| class="mw-collapsible mw-collapsed wikitable"
! C++
|-
|
{| class="mw-collapsible mw-collapsed wikitable"
! Deleting multiple elements in a collection
|-
|
* For sequences (vector, list...), best practice is to use the [https://en.wikipedia.org/wiki/Erase%E2%80%93remove_idiom erase-remove] STL pattern
* For associative containers (set, map...): just recurse and remove iterators, they won't muck up the sequence.  [http://stackoverflow.com/questions/800955/remove-if-equivalent-for-stdmap example]
See MemoryModel::saveDirtyObjectsAsNeeded() for example code.
|}
{| class="mw-collapsible mw-collapsed wikitable"
! c++ Create a portable C/C++ project
|-
| Use QtCreator + CMake, both available everywhere!
    set up:
        create a Qt project as: New Project-> Non-Qt/Plain C Project (CMake Build)
            pick the PARENT and use the name of the existing folder
            set up a build folder in the existing folder, called (base)/qtcreator-release-build
            it creates CMakeLists.txt (THIS IS BASICALLY THE PROJECT FILE!)
            also creates main.c or something
            build it!  make sure it works
            NOW we can edit CMakeLists.txt - change . to ./src to get it to scan the src folder for code!
            and easily add "known" libs!  this was all i needed:
                TARGET_LINK_LIBRARIES(${PROJECT_NAME} pthread websockets)
            make sure it builds
            now we want a DEBUG build as well!
            Projects->Build&Run->pick Build in weird Build/Run pillbox->Edit build configs:
                Rename: release
                CMake build dir: Make sure build dir is set to release ie qtcreator-debug-build
                Add: Clone: release, name: debug
                    for debug, add a custom build step BEFORE make step:
                        command: /usr/bin/cmake
                        args: -DCMAKE_BUILD_TYPE=Debug .
                        working dir: %{buildDir}
                        (add it, and move it up above make)
                        (build em both)
            Projects->Build&Run->pick Run in weird Build/Run pillbox
                Run Configuration: rename to "{...} release"
                clone it into "{...} debug"
                change working dir to match Build path
                change any needed params
    Now you should have build+run debug+release configurations, selectable in the weird project icon in bottom of left-side toolbar ("mode selector")
|}
{| class="mw-collapsible mw-collapsed wikitable"
! boost release and debug build for linux
|-
|
    # download latest boost, eg: boost_1_59_0
    m@wallee:~/development$ 7z x boost_1_59_0.7z
    cd boost_1_##_0
    build_boost_release_and_debug.sh
    # then patch .bashrc as instructed
    # eclipse and server/nix/bootstrap.sh are customized to match
To upgrade a project:
    cd nix
    make distclean  # removes nasty .deps folders that link to old boost if you let them
    make clean      # removes .o files etc
    cd ../build-Release && make distclean && make clean
    cd ../build-Debug && make distclean && make clean
    cd ..
    ./bootstrap force release
    ./bootstrap force debug
|}
{| class="mw-collapsible mw-collapsed wikitable"
! boost release and debug build for Windows
|-
| Open a VS2015 x64 Native Tools Command Prompt. 
EITHER: for new installs, you have to run bootstrap.bat first, it will build b2;
OR: for reruns, remove boost dirs: [bin.v2, stage].
Then build 64-bit:
cd "....\boost_1_59_0"
b2 toolset=msvc variant=release,debug link=static address-model=64
rem trying to avoid excessive options, assuming I don't need these: threading=multi
(old stuff)
      --toolset=msvc-14.0 address-model=64 --build-type=complete --stagedir=windows_lib\x64 stage
      Now open VS2013 x86 Native Tools Command Prompt and build 32-bit:
      cd "C:\Michael's Data\development\sixth_column\boost_1_55_0"
      bjam --toolset=msvc-12.0 address-model=32 --build-type=complete --stagedir=windows_lib\x86 stage
|}
{| class="mw-collapsible mw-collapsed wikitable"
! c++11 include base constructor (etc) in derived class
|-
| The awesome C++11 way to steal the constructor from the base class, oh man I've been waiting for this one...
  class HttpsServer : public Server<HTTPS>
  {
    // DOES THIS FOR YOU!
    /*
    HttpsServer(unsigned short port, size_t num_threads, const std::string& cert_file, const std::string& private_key_file)
    :
        // Call base class
        Server<HTTPS>::Server(port, num_threads, cert_file, private_key_file)
    {}
    */
    using Server<HTTPS>::Server;
    ....
|}
{| class="mw-collapsible mw-collapsed wikitable"
! c++11 containers
|-
| sorted_vector || use when doing lots of unsorted insertions and maintaining constant sort would be expensive; vector is good for a big pile of things that only occasionally needs a sorted lookup
|-
| map || sorted binary search tree; always sorted by key; you can walk through in sorted order (choose unordered if not needed!)
|-
| multimap || same as map but allows dupe keys (not as common)
|-
| unordered_map || hashmap; always sorted by key; additional bucket required for hash collisions; no defined order when walking through
|-
| unordered_multimap || same as map but allows dupe keys; dupes are obviously in the same bucket, and you can walk just the dupes if needed
|-
| set<br>multiset<br>unordered_set<br>unordered_multiset || sets are just like maps, except the key is embedded in the object, nice for encapsulation.<br />
Items must be const (!) since they are the key - sounds bad, but this is mitigated by the mutable keyword.<br />
You can use mutable on the variables that are not part of the key to remove the const.<br />
This changes the constness of the object from binary (completely const) to logical (constness is defined by the developer).<br />
So... set is a good way to achieve both encapsulation and logical const - make const work for you, not against!  :-)<br />
|-
| set (etc.) of pointers || sets of pointers are the pinnacle of object stores<br />
The entire object can be dereferenced and accessed then without const issues.<br />
A pointer functor can be provided that does a sort by dereferencing the pointer to the object.<br />
Two requirements: you must make sure yourself that you do not change the key values - you can mark them const, provided in constructor;<br />
you must create sort/equal/hash functors that dereference the pointers to use object contents<br />
(the default will be by pointer address).<br />
The arguably biggest advantage, as a result, is that you can create multiple sets<br />
to reference the same group of objects with different sort funtors to create multiple indices.<br />
You just have to manage the keys carefully, so that they don't change (which would invalidate the sorting).<br />
The primary container can manage object allocation; using a heap-based unique_ptr allocation<br />
    map vs key redux
               
        use a key in the set, derive a class from it with the contents
            + small key
            + encapsulation
            - requires mutable to solve the const problem
        use a key in the set, key includes a mutable object
            + encapsulation
            - weird bc everything uses a const object but we have const functions like save() that change the mutable subobject
        use a map
            + small key
            - no encapsulation, have to deal with a pair instead of an object
                can we just put a ref to key in the value?  sure why not - err, bc we don't have access to it
            + solves const problem bc value is totally mutable by design
            + we can have multiple keys - and the value can have multiple refs to them
            + simpler equal and hash functions
        map:
            create an object with internal key(s)
            create map index(es) with duplicate key values outside the object - dupe data is the downside
        use set(s) with one static key for find():
            create an object with internal key(s)
            create set index(es) with specific hash/equals functor(s)
            when finding, use one static key object (even across indexes!) so there isn't a big construction each time; just set the necessary key values
                that proves difficult when dealing with member vars that are references
                but to solve it, just set up a structure of dummy static key objects that use each other; then provide a function to setKey(Object& keyref) { keyref_ = keyref; }
                nope, can't reassign refs
                the solution: use pointers not references
                yes that's right
                just do it
                apparently there was a reason i was anti-reference for all those years
                two reasons to use pointers:
                    dynamically allocated
                    reassignment required
                there ya go.  simple.  get it done.
            when accessing find results from the set, use a const_cast on the object!
            WARNING: a separate base class with the key sounds good... but fails when you have more than one index on the object.  just use a static key object for them all!
|}
{| class="mw-collapsible mw-collapsed wikitable"
! c++11 example for large groups of objects with frequent crud AND search
|-
| Best solution is an unordered set of pointers:
typedef boost::unordered_set<MajorObject*> MajorObjects;
|}
{| class="mw-collapsible mw-collapsed wikitable"
! c++11 example for large groups of objects with infrequent crud and frequent search
|-
| Best solution is a vector of pointers sorted on demand (sorted_vector):
TODO
|}
{| class="mw-collapsible mw-collapsed wikitable"
! c++11 example to associate two complex objects (one the map key, one the map value)
|-
| Use unordered_map with a custom object as key.  You must add hash and equals functions.  Boost makes it easy:
static bool operator==(MyKeyObject const& m1, MyKeyObject const& m2)
{
    return
            m1.id_0 == m2.id_0
        &&  m1.id_1 == m2.id_1;
}
static std::size_t hash_value(MyKeyObject const& mko)
{
    std::size_t seed = 0;
    boost::hash_combine(seed, mko.id_0);
    boost::hash_combine(seed, mko.id_1);
    return seed;
}
typedef boost::unordered_map<MyKeyObject, MyValueObject*> MyMap;
Note that you can extend this to use a pointer to a key object, whoop.
|}
{| class="mw-collapsible mw-collapsed wikitable"
! c++11 example for multiple unordered_set indexes into one group of objects
|-
| Objects will be dynamically created.  One set should include them all and be responsible for memory allocation cleanup:
TODO
|}
{| class="mw-collapsible mw-collapsed wikitable"
! c++11 example for set with specific sorting
|-
| Use set with a specific sort functor.  You can create as many of these indexes as you want!
struct customers_set_sort_functor
{
    bool operator()(const MyObject* l, const MyObject* r) const
    {
        // the id is the key
        return l->id_ < r->id_;
    }
};
typedef set<MyObject*,myobject_sort_by_id_functor> MyObjectsById;
|}
{| class="mw-collapsible mw-collapsed wikitable"
! c++11 loop through vector to erase some items
|-
| Note that other containers' iterators may not be invalidated so you can just erase() as needed...
For vectors, you have to play with iterators to get it right - watch for proper ++ pre/postfix!
for (it = numbers.begin(); it != numbers.end(); )  // NOTE we increment below, only if we don't erase
{
    if (*it.no_good())
    {
        numbers.erase(it++);  // NOTE that we ERASE THEN INCREMENT here.
    }
    else
    {
        ++it;
    }
}
I thought I had always looped backwards to do this, I *think* that's ok too, but I don't see it used in my code, I think I'll avoid.  :-)
|}
{| class="mw-collapsible mw-collapsed wikitable"
! c++11 range based for loop, jacked with boost index if needed
|-
| No iterator usage at all.  Nice at times, not enough at others.  Make SURE to always use a reference or you will be working on a COPY.  Make it const if you aren't changing the object.
for (auto& mc : my_container)
    mc.setString("default");
for (const auto& cmc : my_container)
    cout << cmc.asString();
boost index can give you the index if you need it, sweet:
#include <boost/range/adaptor/indexed.hpp>
...
for (const auto &element: boost::adaptors::index(mah_container))
    cout << element.value() << element.index();
|}
{| class="mw-collapsible mw-collapsed wikitable"
! c++11 for loop using lambda
|-
| This C++11 for loop is clean and elegant and a perfect way to check if your compiler is ready for c++11:
vector<int> v;
for_each( v.begin(), v.end(), [] (int val)
{
    cout << val;
} );
This is using a lambda function, we should switch from iterators and functors to those - but not quite yet, since we're writing cross-platform code.  Do not touch this until we can be sure that all platforms provide compatible C++11 handling.
|}
{| class="mw-collapsible mw-collapsed wikitable"
! c++11 integer types
|-
| I really like the "fast" C++11 types, that give best performance for a guaranteed minimum bit width. 
Use them when you know a variable will not exceed the maximum value of that bit width, but does not have to be a precise bit width in memory or elsewhere.
Pick specific-width fields whenever data is shared with other processes and components and you want a guarantee of its bit width.
And when using pointer size and array indices you should use types defined for those specific situations.
FAST types:
    int_fast8_t
    int_fast16_t                fastest signed integer type with width of
    int_fast32_t                at least 8, 16, 32 and 64 bits respectively
    int_fast64_t
    uint_fast8_t
    uint_fast16_t              fastest unsigned integer type with width of
    uint_fast32_t              at least 8, 16, 32 and 64 bits respectively
    uint_fast64_t
SMALL types:
    int_least8_t
    int_least16_t              smallest signed integer type with width of
    int_least32_t              at least 8, 16, 32 and 64 bits respectively
    int_least64_t
    uint_least8_t
    uint_least16_t smallest unsigned integer type with width of
    uint_least32_t at least 8, 16, 32 and 64 bits respectively
    uint_least64_t
EXACT types:
    int8_t                      signed integer type with width of
    int16_t                    exactly 8, 16, 32 and 64 bits respectively
    int32_t                    with no padding bits and using 2's complement for negative values
    int64_t                    (provided only if the implementation directly supports the type)
    uint8_t                    unsigned integer type with width of
    uint16_t                    exactly 8, 16, 32 and 64 bits respectively
    uint32_t                    (provided only if the implementation directly supports the type)
    uint64_t
SPECIFIC-USE types:
    intptr_t                    integer type capable of holding a pointer
    uintptr_t                  unsigned integer type capable of holding a pointer
    size_t                      unsigned integer type capable of holding an array index (same size as uintptr_t)
|}
{| class="mw-collapsible mw-collapsed wikitable"
! C++11 scoped enumeration
|-
| C++11 has scoped enumeration, which lets you specify the SPECIFIC VARIABLE TYPE for the enum.  Perfect, let's use uint_fast32_t.
enum class STRING_PREF_INDEX int_fast32_t: { ... };
Unfortunately, gcc gives me a scary warning, and stuff fails.  For some reason, it does not know about the provided type, although it is definitely defined.  Revisit this later if you have time.
warning: elaborated-type-specifier for a scoped enum must not use the ‘class’ keyword
Old skool is still cool:
typedef enum
{
    // assert( SP_COUNT == 2 );
    SP_FIRST = 0                ,
    SP_SOME_PREF = SP_FIRST    ,
    SP_ANOTHA                  ,
    SP_COUNT
} STRING_PREF_INDEX;
|}
{| class="mw-collapsible mw-collapsed wikitable"
! c++ in-memory storage of "major" objects
|-
|
    OBSERVATION ONE
    Consider An Important Qt Design: QObjects cannot normally be copied
        their copy constructors and assignment operators are private
        why?  A Qt Object...
            might have a unique QObject::objectName(). If we copy a Qt Object, what name should we give the copy?
            has a location in an object hierarchy. If we copy a Qt Object, where should the copy be located?
            can be connected to other Qt Objects to emit signals to them or to receive signals emitted by them. If we copy a Qt Object, how should we transfer these connections to the copy?
            can have new properties added to it at runtime that are not declared in the C++ class. If we copy a Qt Object, should the copy include the properties that were added to the original?
    in other words, a QObject is a pretty serious object that has the ability to be tied to other objects and resources in ways that make copying dangerous
    isn't this true of all serious objects?  pretty much
    OBSERVATION TWO
    if you have a vector of objects, you often want to track them individually outside the vector
    if you use a vector of pointers, you can move the object around much more cheaply, and not worry about costly large vector reallocations
    a vector of objects (not pointers) only makes sense if the number of objects is initially known and does not change over time
    OBSERVATION THREE
    STL vectors can store your pointers, iterate thru them, etc.
    for a vector of any substantial size, you want to keep objects sorted so you can find them quickly
    that's what my sorted_vector class is for; it simply bolts vector together with sort calls and a b_sorted status
    following STL practices, to get sorting, you have to provide operator< for whatever is in your vector
    BUT... you are not allowed to do operator<(const MyObjectPtr* right) because it would require a reference to a pointer which is not allowed
    BUT... you can provide a FUNCTOR to do the job, then provide it when sorting/searching
    a functor is basically a structure with a bool operator()(const MyObjectPtr* left, const MyObjectPtr* right)
    OBSERVATION FOUR
    unordered_set works even better when combining frequent CRUD with frequent lookups
    SUMMARY
    Dealing with tons of objects is par for the course in any significant app.
    Finding a needle in the haystack of those objects is also standard fare.
    Having multiple indices into those objects is also essential.
    Using unordered_set with object pointers and is very powerful.
|}
{| class="mw-collapsible mw-collapsed wikitable"
! c++ stl reverse iterator skeleton
|-
| From [http://www.sgi.com/tech/stl/ReverseIterator.html SGI]...
reverse_iterator rfirst(V.end());
reverse_iterator rlast(V.begin());
while (rfirst != rlast)
{
    cout << *rfirst << endl;
    ...
    rfirst++;
}
|}
{| class="mw-collapsible mw-collapsed wikitable"
! c++ stl reading a binary file into a string
|-
|
    std::ifstream in("my.zip",std::ios::binary);
    if (!in)
    {
      std::cout << "problem with file open" << std::endl;
      return 0;
    }
    in.seekg(0,std::ios::end);
    unsigned long length = in.tellg();
    in.seekg(0,std::ios::beg);
 
    string str(length,0);
    std::copy(
        std::istreambuf_iterator< char >(in) ,
        std::istreambuf_iterator< char >() ,
        str.begin()
    );
For more, see [[Reading a binary file|c++ stl reading a binary file]]
|}
{| class="mw-collapsible mw-collapsed wikitable"
! C/C++ best-in-class tool selection
|-
| I need to have easy setup of debug-level tool support for portable C++11 code.  And I need to decide and stick to it to be efficient.
* '''Compiler selection'''
** linux and mac: gcc
** windows: Visual Studio
* '''IDE selection'''
** linux and mac: Qt Creator
** windows: Qt Creator (OR Visual Studio OR eclipse?)
* '''Debugger selection'''
** linux and mac: Qt Creator
** windows: Qt Creator (OR Visual Studio OR eclipse?)
|}
|}
<!--
===========================================================================================================================================================================================================================================================================================
-->
{| class="mw-collapsible mw-collapsed wikitable"
! C++
|-
|
{| class="mw-collapsible mw-collapsed wikitable"
! Deleting multiple elements in a collection
|-
|
* For sequences (vector, list...), best practice is to use the [https://en.wikipedia.org/wiki/Erase%E2%80%93remove_idiom erase-remove] STL pattern
* For associative containers (set, map...): just recurse and remove iterators, they won't muck up the sequence.  [http://stackoverflow.com/questions/800955/remove-if-equivalent-for-stdmap example]
See MemoryModel::saveDirtyObjectsAsNeeded() for example code.
|}
{| class="mw-collapsible mw-collapsed wikitable"
! c++ Create a portable C/C++ project
|-
| Use QtCreator + CMake, both available everywhere!
    set up:
        create a Qt project as: New Project-> Non-Qt/Plain C Project (CMake Build)
            pick the PARENT and use the name of the existing folder
            set up a build folder in the existing folder, called (base)/qtcreator-release-build
            it creates CMakeLists.txt (THIS IS BASICALLY THE PROJECT FILE!)
            also creates main.c or something
            build it!  make sure it works
            NOW we can edit CMakeLists.txt - change . to ./src to get it to scan the src folder for code!
            and easily add "known" libs!  this was all i needed:
                TARGET_LINK_LIBRARIES(${PROJECT_NAME} pthread websockets)
            make sure it builds
            now we want a DEBUG build as well!
            Projects->Build&Run->pick Build in weird Build/Run pillbox->Edit build configs:
                Rename: release
                CMake build dir: Make sure build dir is set to release ie qtcreator-debug-build
                Add: Clone: release, name: debug
                    for debug, add a custom build step BEFORE make step:
                        command: /usr/bin/cmake
                        args: -DCMAKE_BUILD_TYPE=Debug .
                        working dir: %{buildDir}
                        (add it, and move it up above make)
                        (build em both)
            Projects->Build&Run->pick Run in weird Build/Run pillbox
                Run Configuration: rename to "{...} release"
                clone it into "{...} debug"
                change working dir to match Build path
                change any needed params
    Now you should have build+run debug+release configurations, selectable in the weird project icon in bottom of left-side toolbar ("mode selector")
|}
{| class="mw-collapsible mw-collapsed wikitable"
! boost release and debug build for linux
|-
|
    # download latest boost, eg: boost_1_59_0
    m@wallee:~/development$ 7z x boost_1_59_0.7z
    cd boost_1_##_0
    build_boost_release_and_debug.sh
    # then patch .bashrc as instructed
    # eclipse and server/nix/bootstrap.sh are customized to match
To upgrade a project:
    cd nix
    make distclean  # removes nasty .deps folders that link to old boost if you let them
    make clean      # removes .o files etc
    cd ../build-Release && make distclean && make clean
    cd ../build-Debug && make distclean && make clean
    cd ..
    ./bootstrap force release
    ./bootstrap force debug
|}
{| class="mw-collapsible mw-collapsed wikitable"
! boost release and debug build for Windows
|-
| Open a VS2015 x64 Native Tools Command Prompt. 
EITHER: for new installs, you have to run bootstrap.bat first, it will build b2;
OR: for reruns, remove boost dirs: [bin.v2, stage].
Then build 64-bit:
cd "....\boost_1_59_0"
b2 toolset=msvc variant=release,debug link=static address-model=64
rem trying to avoid excessive options, assuming I don't need these: threading=multi
(old stuff)
      --toolset=msvc-14.0 address-model=64 --build-type=complete --stagedir=windows_lib\x64 stage
      Now open VS2013 x86 Native Tools Command Prompt and build 32-bit:
      cd "C:\Michael's Data\development\sixth_column\boost_1_55_0"
      bjam --toolset=msvc-12.0 address-model=32 --build-type=complete --stagedir=windows_lib\x86 stage
|}
{| class="mw-collapsible mw-collapsed wikitable"
! c++11 include base constructor (etc) in derived class
|-
| The awesome C++11 way to steal the constructor from the base class, oh man I've been waiting for this one...
  class HttpsServer : public Server<HTTPS>
  {
    // DOES THIS FOR YOU!
    /*
    HttpsServer(unsigned short port, size_t num_threads, const std::string& cert_file, const std::string& private_key_file)
    :
        // Call base class
        Server<HTTPS>::Server(port, num_threads, cert_file, private_key_file)
    {}
    */
    using Server<HTTPS>::Server;
    ....
|}
{| class="mw-collapsible mw-collapsed wikitable"
! c++11 containers
|-
| sorted_vector || use when doing lots of unsorted insertions and maintaining constant sort would be expensive; vector is good for a big pile of things that only occasionally needs a sorted lookup
|-
| map || sorted binary search tree; always sorted by key; you can walk through in sorted order (choose unordered if not needed!)
|-
| multimap || same as map but allows dupe keys (not as common)
|-
| unordered_map || hashmap; always sorted by key; additional bucket required for hash collisions; no defined order when walking through
|-
| unordered_multimap || same as map but allows dupe keys; dupes are obviously in the same bucket, and you can walk just the dupes if needed
|-
| set<br>multiset<br>unordered_set<br>unordered_multiset || sets are just like maps, except the key is embedded in the object, nice for encapsulation.<br />
Items must be const (!) since they are the key - sounds bad, but this is mitigated by the mutable keyword.<br />
You can use mutable on the variables that are not part of the key to remove the const.<br />
This changes the constness of the object from binary (completely const) to logical (constness is defined by the developer).<br />
So... set is a good way to achieve both encapsulation and logical const - make const work for you, not against!  :-)<br />
|-
| set (etc.) of pointers || sets of pointers are the pinnacle of object stores<br />
The entire object can be dereferenced and accessed then without const issues.<br />
A pointer functor can be provided that does a sort by dereferencing the pointer to the object.<br />
Two requirements: you must make sure yourself that you do not change the key values - you can mark them const, provided in constructor;<br />
you must create sort/equal/hash functors that dereference the pointers to use object contents<br />
(the default will be by pointer address).<br />
The arguably biggest advantage, as a result, is that you can create multiple sets<br />
to reference the same group of objects with different sort funtors to create multiple indices.<br />
You just have to manage the keys carefully, so that they don't change (which would invalidate the sorting).<br />
The primary container can manage object allocation; using a heap-based unique_ptr allocation<br />
    map vs key redux
               
        use a key in the set, derive a class from it with the contents
            + small key
            + encapsulation
            - requires mutable to solve the const problem
        use a key in the set, key includes a mutable object
            + encapsulation
            - weird bc everything uses a const object but we have const functions like save() that change the mutable subobject
        use a map
            + small key
            - no encapsulation, have to deal with a pair instead of an object
                can we just put a ref to key in the value?  sure why not - err, bc we don't have access to it
            + solves const problem bc value is totally mutable by design
            + we can have multiple keys - and the value can have multiple refs to them
            + simpler equal and hash functions
        map:
            create an object with internal key(s)
            create map index(es) with duplicate key values outside the object - dupe data is the downside
        use set(s) with one static key for find():
            create an object with internal key(s)
            create set index(es) with specific hash/equals functor(s)
            when finding, use one static key object (even across indexes!) so there isn't a big construction each time; just set the necessary key values
                that proves difficult when dealing with member vars that are references
                but to solve it, just set up a structure of dummy static key objects that use each other; then provide a function to setKey(Object& keyref) { keyref_ = keyref; }
                nope, can't reassign refs
                the solution: use pointers not references
                yes that's right
                just do it
                apparently there was a reason i was anti-reference for all those years
                two reasons to use pointers:
                    dynamically allocated
                    reassignment required
                there ya go.  simple.  get it done.
            when accessing find results from the set, use a const_cast on the object!
            WARNING: a separate base class with the key sounds good... but fails when you have more than one index on the object.  just use a static key object for them all!
|}
{| class="mw-collapsible mw-collapsed wikitable"
! c++11 example for large groups of objects with frequent crud AND search
|-
| Best solution is an unordered set of pointers:
typedef boost::unordered_set<MajorObject*> MajorObjects;
|}
{| class="mw-collapsible mw-collapsed wikitable"
! c++11 example for large groups of objects with infrequent crud and frequent search
|-
| Best solution is a vector of pointers sorted on demand (sorted_vector):
TODO
|}
{| class="mw-collapsible mw-collapsed wikitable"
! c++11 example to associate two complex objects (one the map key, one the map value)
|-
| Use unordered_map with a custom object as key.  You must add hash and equals functions.  Boost makes it easy:
static bool operator==(MyKeyObject const& m1, MyKeyObject const& m2)
{
    return
            m1.id_0 == m2.id_0
        &&  m1.id_1 == m2.id_1;
}
static std::size_t hash_value(MyKeyObject const& mko)
{
    std::size_t seed = 0;
    boost::hash_combine(seed, mko.id_0);
    boost::hash_combine(seed, mko.id_1);
    return seed;
}
typedef boost::unordered_map<MyKeyObject, MyValueObject*> MyMap;
Note that you can extend this to use a pointer to a key object, whoop.
|}
{| class="mw-collapsible mw-collapsed wikitable"
! c++11 example for multiple unordered_set indexes into one group of objects
|-
| Objects will be dynamically created.  One set should include them all and be responsible for memory allocation cleanup:
TODO
|}
{| class="mw-collapsible mw-collapsed wikitable"
! c++11 example for set with specific sorting
|-
| Use set with a specific sort functor.  You can create as many of these indexes as you want!
struct customers_set_sort_functor
{
    bool operator()(const MyObject* l, const MyObject* r) const
    {
        // the id is the key
        return l->id_ < r->id_;
    }
};
typedef set<MyObject*,myobject_sort_by_id_functor> MyObjectsById;
|}
{| class="mw-collapsible mw-collapsed wikitable"
! c++11 loop through vector to erase some items
|-
| Note that other containers' iterators may not be invalidated so you can just erase() as needed...
For vectors, you have to play with iterators to get it right - watch for proper ++ pre/postfix!
for (it = numbers.begin(); it != numbers.end(); )  // NOTE we increment below, only if we don't erase
{
    if (*it.no_good())
    {
        numbers.erase(it++);  // NOTE that we ERASE THEN INCREMENT here.
    }
    else
    {
        ++it;
    }
}
I thought I had always looped backwards to do this, I *think* that's ok too, but I don't see it used in my code, I think I'll avoid.  :-)
|}
{| class="mw-collapsible mw-collapsed wikitable"
! c++11 range based for loop, jacked with boost index if needed
|-
| No iterator usage at all.  Nice at times, not enough at others.  Make SURE to always use a reference or you will be working on a COPY.  Make it const if you aren't changing the object.
for (auto& mc : my_container)
    mc.setString("default");
for (const auto& cmc : my_container)
    cout << cmc.asString();
boost index can give you the index if you need it, sweet:
#include <boost/range/adaptor/indexed.hpp>
...
for (const auto &element: boost::adaptors::index(mah_container))
    cout << element.value() << element.index();
|}
{| class="mw-collapsible mw-collapsed wikitable"
! c++11 for loop using lambda
|-
| This C++11 for loop is clean and elegant and a perfect way to check if your compiler is ready for c++11:
vector<int> v;
for_each( v.begin(), v.end(), [] (int val)
{
    cout << val;
} );
This is using a lambda function, we should switch from iterators and functors to those - but not quite yet, since we're writing cross-platform code.  Do not touch this until we can be sure that all platforms provide compatible C++11 handling.
|}
{| class="mw-collapsible mw-collapsed wikitable"
{| class="mw-collapsible mw-collapsed wikitable"
! c/c++ gdb debugging
! c++11 integer types
|-
| I really like the "fast" C++11 types, that give best performance for a guaranteed minimum bit width. 
Use them when you know a variable will not exceed the maximum value of that bit width, but does not have to be a precise bit width in memory or elsewhere.
 
Pick specific-width fields whenever data is shared with other processes and components and you want a guarantee of its bit width.
 
And when using pointer size and array indices you should use types defined for those specific situations.
 
FAST types:
    int_fast8_t
    int_fast16_t                fastest signed integer type with width of
    int_fast32_t                at least 8, 16, 32 and 64 bits respectively
    int_fast64_t
 
    uint_fast8_t
    uint_fast16_t              fastest unsigned integer type with width of
    uint_fast32_t              at least 8, 16, 32 and 64 bits respectively
    uint_fast64_t
SMALL types:
    int_least8_t
    int_least16_t              smallest signed integer type with width of
    int_least32_t              at least 8, 16, 32 and 64 bits respectively
    int_least64_t
 
    uint_least8_t
    uint_least16_t smallest unsigned integer type with width of
    uint_least32_t at least 8, 16, 32 and 64 bits respectively
    uint_least64_t
EXACT types:
    int8_t                      signed integer type with width of
    int16_t                    exactly 8, 16, 32 and 64 bits respectively
    int32_t                    with no padding bits and using 2's complement for negative values
    int64_t                    (provided only if the implementation directly supports the type)
 
    uint8_t                    unsigned integer type with width of
    uint16_t                    exactly 8, 16, 32 and 64 bits respectively
    uint32_t                    (provided only if the implementation directly supports the type)
    uint64_t
SPECIFIC-USE types:
    intptr_t                    integer type capable of holding a pointer
    uintptr_t                  unsigned integer type capable of holding a pointer
    size_t                      unsigned integer type capable of holding an array index (same size as uintptr_t)
|}
 
{| class="mw-collapsible mw-collapsed wikitable"
! C++11 scoped enumeration
|-  
|-  
| C++11 has scoped enumeration, which lets you specify the SPECIFIC VARIABLE TYPE for the enum.  Perfect, let's use uint_fast32_t.
enum class STRING_PREF_INDEX int_fast32_t: { ... };
Unfortunately, gcc gives me a scary warning, and stuff fails.  For some reason, it does not know about the provided type, although it is definitely defined.  Revisit this later if you have time.
warning: elaborated-type-specifier for a scoped enum must not use the ‘class’ keyword
Old skool is still cool:
typedef enum
{
    // assert( SP_COUNT == 2 );
    SP_FIRST = 0                ,
    SP_SOME_PREF = SP_FIRST    ,
    SP_ANOTHA                  ,
    SP_COUNT
} STRING_PREF_INDEX;
|}
{| class="mw-collapsible mw-collapsed wikitable"
! c++ in-memory storage of "major" objects
|-
|  
|  
  (gdb) help break
    OBSERVATION ONE
  Set breakpoint at specified line or function.
   
Argument may be line number, function name, or "*" and an address.
    Consider An Important Qt Design: QObjects cannot normally be copied
If line number is specified, break at start of code for that line.
        their copy constructors and assignment operators are private
If function is specified, break at start of code for that function.
        why? A Qt Object...
If an address is specified, break at that exact address.
            might have a unique QObject::objectName(). If we copy a Qt Object, what name should we give the copy?
  With no arg, uses current execution address of selected stack frame.
            has a location in an object hierarchy. If we copy a Qt Object, where should the copy be located?
This is useful for breaking on return to a stack frame.
            can be connected to other Qt Objects to emit signals to them or to receive signals emitted by them. If we copy a Qt Object, how should we transfer these connections to the copy?
            can have new properties added to it at runtime that are not declared in the C++ class. If we copy a Qt Object, should the copy include the properties that were added to the original?
    in other words, a QObject is a pretty serious object that has the ability to be tied to other objects and resources in ways that make copying dangerous
    isn't this true of all serious objects?  pretty much
 
    OBSERVATION TWO
   
    if you have a vector of objects, you often want to track them individually outside the vector
    if you use a vector of pointers, you can move the object around much more cheaply, and not worry about costly large vector reallocations
    a vector of objects (not pointers) only makes sense if the number of objects is initially known and does not change over time
 
    OBSERVATION THREE
    STL vectors can store your pointers, iterate thru them, etc.
    for a vector of any substantial size, you want to keep objects sorted so you can find them quickly
    that's what my sorted_vector class is for; it simply bolts vector together with sort calls and a b_sorted status
    following STL practices, to get sorting, you have to provide operator< for whatever is in your vector
    BUT... you are not allowed to do operator<(const MyObjectPtr* right) because it would require a reference to a pointer which is not allowed
    BUT... you can provide a FUNCTOR to do the job, then provide it when sorting/searching
    a functor is basically a structure with a bool operator()(const MyObjectPtr* left, const MyObjectPtr* right)
 
    OBSERVATION FOUR
   
   
  Multiple breakpoints at one place are permitted, and useful if conditional.  
    unordered_set works even better when combining frequent CRUD with frequent lookups
 
    SUMMARY
    Dealing with tons of objects is par for the course in any significant app.
    Finding a needle in the haystack of those objects is also standard fare.
    Having multiple indices into those objects is also essential.
    Using unordered_set with object pointers and is very powerful.
|}
 
{| class="mw-collapsible mw-collapsed wikitable"
! c++ stl reverse iterator skeleton
|-
| From [http://www.sgi.com/tech/stl/ReverseIterator.html SGI]...
  reverse_iterator rfirst(V.end());
reverse_iterator rlast(V.begin());
   
   
  Do "help breakpoints" for info on other commands dealing with breakpoints.
  while (rfirst != rlast)
|-
{
| ddd gives you a front end.  I need to use it more, compare to other options
    cout << *rfirst << endl;
    ...
    rfirst++;
}
|}
 
{| class="mw-collapsible mw-collapsed wikitable"
! c++ stl reading a binary file into a string
|-
|
    std::ifstream in("my.zip",std::ios::binary);
    if (!in)
    {
      std::cout << "problem with file open" << std::endl;
      return 0;
    }
    in.seekg(0,std::ios::end);
    unsigned long length = in.tellg();
    in.seekg(0,std::ios::beg);
 
    string str(length,0);
    std::copy(
        std::istreambuf_iterator< char >(in) ,
        std::istreambuf_iterator< char >() ,
        str.begin()
    );
 
For more, see [[Reading a binary file|c++ stl reading a binary file]]
|}
 
{| class="mw-collapsible mw-collapsed wikitable"
! C/C++ best-in-class tool selection
|-  
| I need to have easy setup of debug-level tool support for portable C++11 codeAnd I need to decide and stick to it to be efficient.
* '''Compiler selection'''
** linux and mac: gcc
** windows: Visual Studio
* '''IDE selection'''
** linux and mac: Qt Creator
** windows: Qt Creator (OR Visual Studio OR eclipse?)
* '''Debugger selection'''
** linux and mac: Qt Creator
** windows: Qt Creator (OR Visual Studio OR eclipse?)
|}
|}
|}
|}
Line 626: Line 1,485:
     at first i thought this would work: /usr/local/lib
     at first i thought this would work: /usr/local/lib
     but i did a static build so i needed teh actual exe!  COOL
     but i did a static build so i needed teh actual exe!  COOL
|}
{| class="mw-collapsible mw-collapsed wikitable"
! c/c++ gdb debugging
|-
|
(gdb) help break
Set breakpoint at specified line or function.
Argument may be line number, function name, or "*" and an address.
If line number is specified, break at start of code for that line.
If function is specified, break at start of code for that function.
If an address is specified, break at that exact address.
With no arg, uses current execution address of selected stack frame.
This is useful for breaking on return to a stack frame.
Multiple breakpoints at one place are permitted, and useful if conditional.   
Do "help breakpoints" for info on other commands dealing with breakpoints.
|-
| ddd gives you a front end.  I need to use it more, compare to other options
|}
|}
|}
|}

Revision as of 01:41, 8 June 2017

Design, programming and version control.

Patterns
Major Objects
Quick-http
Continuous Integration
Architecture of major open source apps
Model View Controller
String escape formatting across different languages and systems
  • c++ to JSON: always use nlohmann::json j.dump() to encode, to ensure strings are properly escaped
  • JSON to c++: always use nlohmann::json j.parse() "
  • c++ to Javascript: use raw_to_Javascript() to properly escape
  • c++ to sqlite: use SqliteLocalModel::safestr(), which uses double_doublequotes(str)
Web design mantra
  1. DESIGN INPUT TO THE INPUT DEVICE
    1. respect the input device
    2. detect it and use a reasonable default guess
    3. allow manual override
    4. [mouse/pen]...[finger]
    5. sm-md-lg ... sm-md-lg
  2. DESIGN VISUALS TO THE SCREEN
    1. high-res = SHOW LOTS OF DETAIL
    2. responsive, zoomable
git central shared repository
Use bare repos for any central shared repositories. Use the [####.git] suffix on bare repo names.

Bare repositories are designed to be shared. Maintenance on the central server is easier because you don't have local files to manage permissions or constant flux. Plus, you can always have a second repo on the central server where you check out a specific branch (e.g. to serve up with apache). If you want a dynamically updated central repo, clone the ###.git repo to ###, and add a post-receive hook (see bitpost quick-http.git for a good example).

To configure the repo as shared:

git config core.sharedRepository true

To set it on a new git repo during initial setup, make sure devs are in the same group, and use:

git init --shared=group
Occam's razor
Among competing hypotheses, the one with the fewest assumptions should be selected.
C++
Deleting multiple elements in a collection
  • For sequences (vector, list...), best practice is to use the erase-remove STL pattern
  • For associative containers (set, map...): just recurse and remove iterators, they won't muck up the sequence. example

See MemoryModel::saveDirtyObjectsAsNeeded() for example code.

c++ Create a portable C/C++ project
Use QtCreator + CMake, both available everywhere!
   set up:
       create a Qt project as: New Project-> Non-Qt/Plain C Project (CMake Build)
           pick the PARENT and use the name of the existing folder
           set up a build folder in the existing folder, called (base)/qtcreator-release-build
           it creates CMakeLists.txt (THIS IS BASICALLY THE PROJECT FILE!)
           also creates main.c or something
           build it!  make sure it works
           NOW we can edit CMakeLists.txt - change . to ./src to get it to scan the src folder for code!
           and easily add "known" libs!  this was all i needed:
               TARGET_LINK_LIBRARIES(${PROJECT_NAME} pthread websockets)
           make sure it builds
           now we want a DEBUG build as well!
           Projects->Build&Run->pick Build in weird Build/Run pillbox->Edit build configs:
               Rename: release
               CMake build dir: Make sure build dir is set to release ie qtcreator-debug-build
               Add: Clone: release, name: debug
                   for debug, add a custom build step BEFORE make step:
                       command: /usr/bin/cmake
                       args: -DCMAKE_BUILD_TYPE=Debug .
                       working dir: %{buildDir}
                       (add it, and move it up above make)
                       (build em both)
           Projects->Build&Run->pick Run in weird Build/Run pillbox
               Run Configuration: rename to "{...} release"
               clone it into "{...} debug"
               change working dir to match Build path
               change any needed params
   Now you should have build+run debug+release configurations, selectable in the weird project icon in bottom of left-side toolbar ("mode selector")
boost release and debug build for linux
   # download latest boost, eg: boost_1_59_0 
   m@wallee:~/development$ 7z x boost_1_59_0.7z
   cd boost_1_##_0
   build_boost_release_and_debug.sh
   # then patch .bashrc as instructed
   # eclipse and server/nix/bootstrap.sh are customized to match

To upgrade a project:

   cd nix
   make distclean  # removes nasty .deps folders that link to old boost if you let them
   make clean      # removes .o files etc
   cd ../build-Release && make distclean && make clean
   cd ../build-Debug && make distclean && make clean
   cd ..
   ./bootstrap force release
   ./bootstrap force debug
boost release and debug build for Windows
Open a VS2015 x64 Native Tools Command Prompt.

EITHER: for new installs, you have to run bootstrap.bat first, it will build b2; OR: for reruns, remove boost dirs: [bin.v2, stage]. Then build 64-bit:

cd "....\boost_1_59_0"
b2 toolset=msvc variant=release,debug link=static address-model=64
rem trying to avoid excessive options, assuming I don't need these: threading=multi
(old stuff)
      --toolset=msvc-14.0 address-model=64 --build-type=complete --stagedir=windows_lib\x64 stage
      Now open VS2013 x86 Native Tools Command Prompt and build 32-bit:
      cd "C:\Michael's Data\development\sixth_column\boost_1_55_0"
      bjam --toolset=msvc-12.0 address-model=32 --build-type=complete --stagedir=windows_lib\x86 stage
c++11 include base constructor (etc) in derived class
The awesome C++11 way to steal the constructor from the base class, oh man I've been waiting for this one...
 class HttpsServer : public Server<HTTPS>
 {
   // DOES THIS FOR YOU!
   /*
   HttpsServer(unsigned short port, size_t num_threads, const std::string& cert_file, const std::string& private_key_file)
   :
       // Call base class
       Server<HTTPS>::Server(port, num_threads, cert_file, private_key_file)
   {}
   */
   using Server<HTTPS>::Server;
   ....
c++11 containers
sorted_vector use when doing lots of unsorted insertions and maintaining constant sort would be expensive; vector is good for a big pile of things that only occasionally needs a sorted lookup
map sorted binary search tree; always sorted by key; you can walk through in sorted order (choose unordered if not needed!)
multimap same as map but allows dupe keys (not as common)
unordered_map hashmap; always sorted by key; additional bucket required for hash collisions; no defined order when walking through
unordered_multimap same as map but allows dupe keys; dupes are obviously in the same bucket, and you can walk just the dupes if needed
set
multiset
unordered_set
unordered_multiset
sets are just like maps, except the key is embedded in the object, nice for encapsulation.

Items must be const (!) since they are the key - sounds bad, but this is mitigated by the mutable keyword.
You can use mutable on the variables that are not part of the key to remove the const.
This changes the constness of the object from binary (completely const) to logical (constness is defined by the developer).
So... set is a good way to achieve both encapsulation and logical const - make const work for you, not against!  :-)

set (etc.) of pointers sets of pointers are the pinnacle of object stores

The entire object can be dereferenced and accessed then without const issues.
A pointer functor can be provided that does a sort by dereferencing the pointer to the object.
Two requirements: you must make sure yourself that you do not change the key values - you can mark them const, provided in constructor;
you must create sort/equal/hash functors that dereference the pointers to use object contents
(the default will be by pointer address).
The arguably biggest advantage, as a result, is that you can create multiple sets
to reference the same group of objects with different sort funtors to create multiple indices.
You just have to manage the keys carefully, so that they don't change (which would invalidate the sorting).
The primary container can manage object allocation; using a heap-based unique_ptr allocation

   map vs key redux
               
       use a key in the set, derive a class from it with the contents
           + small key
           + encapsulation
           - requires mutable to solve the const problem
       use a key in the set, key includes a mutable object
           + encapsulation
           - weird bc everything uses a const object but we have const functions like save() that change the mutable subobject
       use a map
           + small key
           - no encapsulation, have to deal with a pair instead of an object
               can we just put a ref to key in the value?  sure why not - err, bc we don't have access to it
           + solves const problem bc value is totally mutable by design
           + we can have multiple keys - and the value can have multiple refs to them
           + simpler equal and hash functions
       map:
           create an object with internal key(s)
           create map index(es) with duplicate key values outside the object - dupe data is the downside
       use set(s) with one static key for find(): 
           create an object with internal key(s)
           create set index(es) with specific hash/equals functor(s)
           when finding, use one static key object (even across indexes!) so there isn't a big construction each time; just set the necessary key values
               that proves difficult when dealing with member vars that are references
               but to solve it, just set up a structure of dummy static key objects that use each other; then provide a function to setKey(Object& keyref) { keyref_ = keyref; }
               nope, can't reassign refs
               the solution: use pointers not references
               yes that's right
               just do it
               apparently there was a reason i was anti-reference for all those years
               two reasons to use pointers:
                   dynamically allocated
                   reassignment required
               there ya go.  simple.  get it done. 
           when accessing find results from the set, use a const_cast on the object!
           WARNING: a separate base class with the key sounds good... but fails when you have more than one index on the object.  just use a static key object for them all!
c++11 example for large groups of objects with frequent crud AND search
Best solution is an unordered set of pointers:
typedef boost::unordered_set<MajorObject*> MajorObjects;
c++11 example for large groups of objects with infrequent crud and frequent search
Best solution is a vector of pointers sorted on demand (sorted_vector):
TODO
c++11 example to associate two complex objects (one the map key, one the map value)
Use unordered_map with a custom object as key. You must add hash and equals functions. Boost makes it easy:
static bool operator==(MyKeyObject const& m1, MyKeyObject const& m2)
{
    return 
            m1.id_0 == m2.id_0
        &&  m1.id_1 == m2.id_1;
}
static std::size_t hash_value(MyKeyObject const& mko)
{
    std::size_t seed = 0;
    boost::hash_combine(seed, mko.id_0);
    boost::hash_combine(seed, mko.id_1);
    return seed;
}
typedef boost::unordered_map<MyKeyObject, MyValueObject*> MyMap;

Note that you can extend this to use a pointer to a key object, whoop.

c++11 example for multiple unordered_set indexes into one group of objects
Objects will be dynamically created. One set should include them all and be responsible for memory allocation cleanup:
TODO
c++11 example for set with specific sorting
Use set with a specific sort functor. You can create as many of these indexes as you want!
struct customers_set_sort_functor
{
    bool operator()(const MyObject* l, const MyObject* r) const
    {
        // the id is the key
        return l->id_ < r->id_;
    }
};
typedef set<MyObject*,myobject_sort_by_id_functor> MyObjectsById;
c++11 loop through vector to erase some items
Note that other containers' iterators may not be invalidated so you can just erase() as needed...

For vectors, you have to play with iterators to get it right - watch for proper ++ pre/postfix!

for (it = numbers.begin(); it != numbers.end(); )  // NOTE we increment below, only if we don't erase
{
    if (*it.no_good()) 
    {
        numbers.erase(it++);  // NOTE that we ERASE THEN INCREMENT here.
    }
    else 
    {
        ++it;
    }
}

I thought I had always looped backwards to do this, I *think* that's ok too, but I don't see it used in my code, I think I'll avoid.  :-)

c++11 range based for loop, jacked with boost index if needed
No iterator usage at all. Nice at times, not enough at others. Make SURE to always use a reference or you will be working on a COPY. Make it const if you aren't changing the object.
for (auto& mc : my_container)
    mc.setString("default");
for (const auto& cmc : my_container)
    cout << cmc.asString();

boost index can give you the index if you need it, sweet:

#include <boost/range/adaptor/indexed.hpp>
...
for (const auto &element: boost::adaptors::index(mah_container))
    cout << element.value() << element.index();
c++11 for loop using lambda
This C++11 for loop is clean and elegant and a perfect way to check if your compiler is ready for c++11:
vector<int> v;
for_each( v.begin(), v.end(), [] (int val)
{
   cout << val;
} );

This is using a lambda function, we should switch from iterators and functors to those - but not quite yet, since we're writing cross-platform code. Do not touch this until we can be sure that all platforms provide compatible C++11 handling.

c++11 integer types
I really like the "fast" C++11 types, that give best performance for a guaranteed minimum bit width.

Use them when you know a variable will not exceed the maximum value of that bit width, but does not have to be a precise bit width in memory or elsewhere.

Pick specific-width fields whenever data is shared with other processes and components and you want a guarantee of its bit width.

And when using pointer size and array indices you should use types defined for those specific situations.

FAST types:

   int_fast8_t
   int_fast16_t                fastest signed integer type with width of
   int_fast32_t                at least 8, 16, 32 and 64 bits respectively
   int_fast64_t
   uint_fast8_t
   uint_fast16_t               fastest unsigned integer type with width of
   uint_fast32_t               at least 8, 16, 32 and 64 bits respectively
   uint_fast64_t

SMALL types:

   int_least8_t
   int_least16_t               smallest signed integer type with width of
   int_least32_t               at least 8, 16, 32 and 64 bits respectively
   int_least64_t
   uint_least8_t
   uint_least16_t		smallest unsigned integer type with width of
   uint_least32_t		at least 8, 16, 32 and 64 bits respectively
   uint_least64_t

EXACT types:

   int8_t                      signed integer type with width of
   int16_t                     exactly 8, 16, 32 and 64 bits respectively
   int32_t                     with no padding bits and using 2's complement for negative values
   int64_t                     (provided only if the implementation directly supports the type)
   uint8_t                     unsigned integer type with width of
   uint16_t                    exactly 8, 16, 32 and 64 bits respectively
   uint32_t                    (provided only if the implementation directly supports the type)
   uint64_t

SPECIFIC-USE types:

   intptr_t                    integer type capable of holding a pointer
   uintptr_t                   unsigned integer type capable of holding a pointer 
   size_t                      unsigned integer type capable of holding an array index (same size as uintptr_t)
C++11 scoped enumeration
C++11 has scoped enumeration, which lets you specify the SPECIFIC VARIABLE TYPE for the enum. Perfect, let's use uint_fast32_t.
enum class STRING_PREF_INDEX int_fast32_t: { ... };

Unfortunately, gcc gives me a scary warning, and stuff fails. For some reason, it does not know about the provided type, although it is definitely defined. Revisit this later if you have time.

warning: elaborated-type-specifier for a scoped enum must not use the ‘class’ keyword

Old skool is still cool:

typedef enum
{
    // assert( SP_COUNT == 2 );
    SP_FIRST = 0                ,
    SP_SOME_PREF = SP_FIRST     ,
    SP_ANOTHA                   ,

    SP_COUNT
} STRING_PREF_INDEX;
c++ in-memory storage of "major" objects
   OBSERVATION ONE

   Consider An Important Qt Design: QObjects cannot normally be copied
       their copy constructors and assignment operators are private
       why?  A Qt Object...
           might have a unique QObject::objectName(). If we copy a Qt Object, what name should we give the copy?
           has a location in an object hierarchy. If we copy a Qt Object, where should the copy be located?
           can be connected to other Qt Objects to emit signals to them or to receive signals emitted by them. If we copy a Qt Object, how should we transfer these connections to the copy?
           can have new properties added to it at runtime that are not declared in the C++ class. If we copy a Qt Object, should the copy include the properties that were added to the original?
   in other words, a QObject is a pretty serious object that has the ability to be tied to other objects and resources in ways that make copying dangerous
   isn't this true of all serious objects?  pretty much
   OBSERVATION TWO

   if you have a vector of objects, you often want to track them individually outside the vector
   if you use a vector of pointers, you can move the object around much more cheaply, and not worry about costly large vector reallocations
   a vector of objects (not pointers) only makes sense if the number of objects is initially known and does not change over time
   OBSERVATION THREE

   STL vectors can store your pointers, iterate thru them, etc.
   for a vector of any substantial size, you want to keep objects sorted so you can find them quickly
   that's what my sorted_vector class is for; it simply bolts vector together with sort calls and a b_sorted status
   following STL practices, to get sorting, you have to provide operator< for whatever is in your vector
   BUT... you are not allowed to do operator<(const MyObjectPtr* right) because it would require a reference to a pointer which is not allowed
   BUT... you can provide a FUNCTOR to do the job, then provide it when sorting/searching
   a functor is basically a structure with a bool operator()(const MyObjectPtr* left, const MyObjectPtr* right)
   OBSERVATION FOUR

   unordered_set works even better when combining frequent CRUD with frequent lookups
   SUMMARY
   Dealing with tons of objects is par for the course in any significant app.
   Finding a needle in the haystack of those objects is also standard fare.
   Having multiple indices into those objects is also essential.
   Using unordered_set with object pointers and is very powerful.
c++ stl reverse iterator skeleton
From SGI...
reverse_iterator rfirst(V.end());
reverse_iterator rlast(V.begin());

while (rfirst != rlast) 
{
    cout << *rfirst << endl;
    ...
    rfirst++;
}
c++ stl reading a binary file into a string
   std::ifstream in("my.zip",std::ios::binary);
   if (!in)
   {
      std::cout << "problem with file open" << std::endl;
      return 0;
   }
   in.seekg(0,std::ios::end);
   unsigned long length = in.tellg();
   in.seekg(0,std::ios::beg);
 
   string str(length,0);
   std::copy( 
       std::istreambuf_iterator< char >(in) ,
       std::istreambuf_iterator< char >() ,
       str.begin() 
   );

For more, see c++ stl reading a binary file

C/C++ best-in-class tool selection
I need to have easy setup of debug-level tool support for portable C++11 code. And I need to decide and stick to it to be efficient.
  • Compiler selection
    • linux and mac: gcc
    • windows: Visual Studio
  • IDE selection
    • linux and mac: Qt Creator
    • windows: Qt Creator (OR Visual Studio OR eclipse?)
  • Debugger selection
    • linux and mac: Qt Creator
    • windows: Qt Creator (OR Visual Studio OR eclipse?)
C++
Deleting multiple elements in a collection
  • For sequences (vector, list...), best practice is to use the erase-remove STL pattern
  • For associative containers (set, map...): just recurse and remove iterators, they won't muck up the sequence. example

See MemoryModel::saveDirtyObjectsAsNeeded() for example code.

c++ Create a portable C/C++ project
Use QtCreator + CMake, both available everywhere!
   set up:
       create a Qt project as: New Project-> Non-Qt/Plain C Project (CMake Build)
           pick the PARENT and use the name of the existing folder
           set up a build folder in the existing folder, called (base)/qtcreator-release-build
           it creates CMakeLists.txt (THIS IS BASICALLY THE PROJECT FILE!)
           also creates main.c or something
           build it!  make sure it works
           NOW we can edit CMakeLists.txt - change . to ./src to get it to scan the src folder for code!
           and easily add "known" libs!  this was all i needed:
               TARGET_LINK_LIBRARIES(${PROJECT_NAME} pthread websockets)
           make sure it builds
           now we want a DEBUG build as well!
           Projects->Build&Run->pick Build in weird Build/Run pillbox->Edit build configs:
               Rename: release
               CMake build dir: Make sure build dir is set to release ie qtcreator-debug-build
               Add: Clone: release, name: debug
                   for debug, add a custom build step BEFORE make step:
                       command: /usr/bin/cmake
                       args: -DCMAKE_BUILD_TYPE=Debug .
                       working dir: %{buildDir}
                       (add it, and move it up above make)
                       (build em both)
           Projects->Build&Run->pick Run in weird Build/Run pillbox
               Run Configuration: rename to "{...} release"
               clone it into "{...} debug"
               change working dir to match Build path
               change any needed params
   Now you should have build+run debug+release configurations, selectable in the weird project icon in bottom of left-side toolbar ("mode selector")
boost release and debug build for linux
   # download latest boost, eg: boost_1_59_0 
   m@wallee:~/development$ 7z x boost_1_59_0.7z
   cd boost_1_##_0
   build_boost_release_and_debug.sh
   # then patch .bashrc as instructed
   # eclipse and server/nix/bootstrap.sh are customized to match

To upgrade a project:

   cd nix
   make distclean  # removes nasty .deps folders that link to old boost if you let them
   make clean      # removes .o files etc
   cd ../build-Release && make distclean && make clean
   cd ../build-Debug && make distclean && make clean
   cd ..
   ./bootstrap force release
   ./bootstrap force debug
boost release and debug build for Windows
Open a VS2015 x64 Native Tools Command Prompt.

EITHER: for new installs, you have to run bootstrap.bat first, it will build b2; OR: for reruns, remove boost dirs: [bin.v2, stage]. Then build 64-bit:

cd "....\boost_1_59_0"
b2 toolset=msvc variant=release,debug link=static address-model=64
rem trying to avoid excessive options, assuming I don't need these: threading=multi
(old stuff)
      --toolset=msvc-14.0 address-model=64 --build-type=complete --stagedir=windows_lib\x64 stage
      Now open VS2013 x86 Native Tools Command Prompt and build 32-bit:
      cd "C:\Michael's Data\development\sixth_column\boost_1_55_0"
      bjam --toolset=msvc-12.0 address-model=32 --build-type=complete --stagedir=windows_lib\x86 stage
c++11 include base constructor (etc) in derived class
The awesome C++11 way to steal the constructor from the base class, oh man I've been waiting for this one...
 class HttpsServer : public Server<HTTPS>
 {
   // DOES THIS FOR YOU!
   /*
   HttpsServer(unsigned short port, size_t num_threads, const std::string& cert_file, const std::string& private_key_file)
   :
       // Call base class
       Server<HTTPS>::Server(port, num_threads, cert_file, private_key_file)
   {}
   */
   using Server<HTTPS>::Server;
   ....
c++11 containers
sorted_vector use when doing lots of unsorted insertions and maintaining constant sort would be expensive; vector is good for a big pile of things that only occasionally needs a sorted lookup
map sorted binary search tree; always sorted by key; you can walk through in sorted order (choose unordered if not needed!)
multimap same as map but allows dupe keys (not as common)
unordered_map hashmap; always sorted by key; additional bucket required for hash collisions; no defined order when walking through
unordered_multimap same as map but allows dupe keys; dupes are obviously in the same bucket, and you can walk just the dupes if needed
set
multiset
unordered_set
unordered_multiset
sets are just like maps, except the key is embedded in the object, nice for encapsulation.

Items must be const (!) since they are the key - sounds bad, but this is mitigated by the mutable keyword.
You can use mutable on the variables that are not part of the key to remove the const.
This changes the constness of the object from binary (completely const) to logical (constness is defined by the developer).
So... set is a good way to achieve both encapsulation and logical const - make const work for you, not against!  :-)

set (etc.) of pointers sets of pointers are the pinnacle of object stores

The entire object can be dereferenced and accessed then without const issues.
A pointer functor can be provided that does a sort by dereferencing the pointer to the object.
Two requirements: you must make sure yourself that you do not change the key values - you can mark them const, provided in constructor;
you must create sort/equal/hash functors that dereference the pointers to use object contents
(the default will be by pointer address).
The arguably biggest advantage, as a result, is that you can create multiple sets
to reference the same group of objects with different sort funtors to create multiple indices.
You just have to manage the keys carefully, so that they don't change (which would invalidate the sorting).
The primary container can manage object allocation; using a heap-based unique_ptr allocation

   map vs key redux
               
       use a key in the set, derive a class from it with the contents
           + small key
           + encapsulation
           - requires mutable to solve the const problem
       use a key in the set, key includes a mutable object
           + encapsulation
           - weird bc everything uses a const object but we have const functions like save() that change the mutable subobject
       use a map
           + small key
           - no encapsulation, have to deal with a pair instead of an object
               can we just put a ref to key in the value?  sure why not - err, bc we don't have access to it
           + solves const problem bc value is totally mutable by design
           + we can have multiple keys - and the value can have multiple refs to them
           + simpler equal and hash functions
       map:
           create an object with internal key(s)
           create map index(es) with duplicate key values outside the object - dupe data is the downside
       use set(s) with one static key for find(): 
           create an object with internal key(s)
           create set index(es) with specific hash/equals functor(s)
           when finding, use one static key object (even across indexes!) so there isn't a big construction each time; just set the necessary key values
               that proves difficult when dealing with member vars that are references
               but to solve it, just set up a structure of dummy static key objects that use each other; then provide a function to setKey(Object& keyref) { keyref_ = keyref; }
               nope, can't reassign refs
               the solution: use pointers not references
               yes that's right
               just do it
               apparently there was a reason i was anti-reference for all those years
               two reasons to use pointers:
                   dynamically allocated
                   reassignment required
               there ya go.  simple.  get it done. 
           when accessing find results from the set, use a const_cast on the object!
           WARNING: a separate base class with the key sounds good... but fails when you have more than one index on the object.  just use a static key object for them all!
c++11 example for large groups of objects with frequent crud AND search
Best solution is an unordered set of pointers:
typedef boost::unordered_set<MajorObject*> MajorObjects;
c++11 example for large groups of objects with infrequent crud and frequent search
Best solution is a vector of pointers sorted on demand (sorted_vector):
TODO
c++11 example to associate two complex objects (one the map key, one the map value)
Use unordered_map with a custom object as key. You must add hash and equals functions. Boost makes it easy:
static bool operator==(MyKeyObject const& m1, MyKeyObject const& m2)
{
    return 
            m1.id_0 == m2.id_0
        &&  m1.id_1 == m2.id_1;
}
static std::size_t hash_value(MyKeyObject const& mko)
{
    std::size_t seed = 0;
    boost::hash_combine(seed, mko.id_0);
    boost::hash_combine(seed, mko.id_1);
    return seed;
}
typedef boost::unordered_map<MyKeyObject, MyValueObject*> MyMap;

Note that you can extend this to use a pointer to a key object, whoop.

c++11 example for multiple unordered_set indexes into one group of objects
Objects will be dynamically created. One set should include them all and be responsible for memory allocation cleanup:
TODO
c++11 example for set with specific sorting
Use set with a specific sort functor. You can create as many of these indexes as you want!
struct customers_set_sort_functor
{
    bool operator()(const MyObject* l, const MyObject* r) const
    {
        // the id is the key
        return l->id_ < r->id_;
    }
};
typedef set<MyObject*,myobject_sort_by_id_functor> MyObjectsById;
c++11 loop through vector to erase some items
Note that other containers' iterators may not be invalidated so you can just erase() as needed...

For vectors, you have to play with iterators to get it right - watch for proper ++ pre/postfix!

for (it = numbers.begin(); it != numbers.end(); )  // NOTE we increment below, only if we don't erase
{
    if (*it.no_good()) 
    {
        numbers.erase(it++);  // NOTE that we ERASE THEN INCREMENT here.
    }
    else 
    {
        ++it;
    }
}

I thought I had always looped backwards to do this, I *think* that's ok too, but I don't see it used in my code, I think I'll avoid.  :-)

c++11 range based for loop, jacked with boost index if needed
No iterator usage at all. Nice at times, not enough at others. Make SURE to always use a reference or you will be working on a COPY. Make it const if you aren't changing the object.
for (auto& mc : my_container)
    mc.setString("default");
for (const auto& cmc : my_container)
    cout << cmc.asString();

boost index can give you the index if you need it, sweet:

#include <boost/range/adaptor/indexed.hpp>
...
for (const auto &element: boost::adaptors::index(mah_container))
    cout << element.value() << element.index();
c++11 for loop using lambda
This C++11 for loop is clean and elegant and a perfect way to check if your compiler is ready for c++11:
vector<int> v;
for_each( v.begin(), v.end(), [] (int val)
{
   cout << val;
} );

This is using a lambda function, we should switch from iterators and functors to those - but not quite yet, since we're writing cross-platform code. Do not touch this until we can be sure that all platforms provide compatible C++11 handling.

c++11 integer types
I really like the "fast" C++11 types, that give best performance for a guaranteed minimum bit width.

Use them when you know a variable will not exceed the maximum value of that bit width, but does not have to be a precise bit width in memory or elsewhere.

Pick specific-width fields whenever data is shared with other processes and components and you want a guarantee of its bit width.

And when using pointer size and array indices you should use types defined for those specific situations.

FAST types:

   int_fast8_t
   int_fast16_t                fastest signed integer type with width of
   int_fast32_t                at least 8, 16, 32 and 64 bits respectively
   int_fast64_t
   uint_fast8_t
   uint_fast16_t               fastest unsigned integer type with width of
   uint_fast32_t               at least 8, 16, 32 and 64 bits respectively
   uint_fast64_t

SMALL types:

   int_least8_t
   int_least16_t               smallest signed integer type with width of
   int_least32_t               at least 8, 16, 32 and 64 bits respectively
   int_least64_t
   uint_least8_t
   uint_least16_t		smallest unsigned integer type with width of
   uint_least32_t		at least 8, 16, 32 and 64 bits respectively
   uint_least64_t

EXACT types:

   int8_t                      signed integer type with width of
   int16_t                     exactly 8, 16, 32 and 64 bits respectively
   int32_t                     with no padding bits and using 2's complement for negative values
   int64_t                     (provided only if the implementation directly supports the type)
   uint8_t                     unsigned integer type with width of
   uint16_t                    exactly 8, 16, 32 and 64 bits respectively
   uint32_t                    (provided only if the implementation directly supports the type)
   uint64_t

SPECIFIC-USE types:

   intptr_t                    integer type capable of holding a pointer
   uintptr_t                   unsigned integer type capable of holding a pointer 
   size_t                      unsigned integer type capable of holding an array index (same size as uintptr_t)
C++11 scoped enumeration
C++11 has scoped enumeration, which lets you specify the SPECIFIC VARIABLE TYPE for the enum. Perfect, let's use uint_fast32_t.
enum class STRING_PREF_INDEX int_fast32_t: { ... };

Unfortunately, gcc gives me a scary warning, and stuff fails. For some reason, it does not know about the provided type, although it is definitely defined. Revisit this later if you have time.

warning: elaborated-type-specifier for a scoped enum must not use the ‘class’ keyword

Old skool is still cool:

typedef enum
{
    // assert( SP_COUNT == 2 );
    SP_FIRST = 0                ,
    SP_SOME_PREF = SP_FIRST     ,
    SP_ANOTHA                   ,

    SP_COUNT
} STRING_PREF_INDEX;
c++ in-memory storage of "major" objects
   OBSERVATION ONE

   Consider An Important Qt Design: QObjects cannot normally be copied
       their copy constructors and assignment operators are private
       why?  A Qt Object...
           might have a unique QObject::objectName(). If we copy a Qt Object, what name should we give the copy?
           has a location in an object hierarchy. If we copy a Qt Object, where should the copy be located?
           can be connected to other Qt Objects to emit signals to them or to receive signals emitted by them. If we copy a Qt Object, how should we transfer these connections to the copy?
           can have new properties added to it at runtime that are not declared in the C++ class. If we copy a Qt Object, should the copy include the properties that were added to the original?
   in other words, a QObject is a pretty serious object that has the ability to be tied to other objects and resources in ways that make copying dangerous
   isn't this true of all serious objects?  pretty much
   OBSERVATION TWO

   if you have a vector of objects, you often want to track them individually outside the vector
   if you use a vector of pointers, you can move the object around much more cheaply, and not worry about costly large vector reallocations
   a vector of objects (not pointers) only makes sense if the number of objects is initially known and does not change over time
   OBSERVATION THREE

   STL vectors can store your pointers, iterate thru them, etc.
   for a vector of any substantial size, you want to keep objects sorted so you can find them quickly
   that's what my sorted_vector class is for; it simply bolts vector together with sort calls and a b_sorted status
   following STL practices, to get sorting, you have to provide operator< for whatever is in your vector
   BUT... you are not allowed to do operator<(const MyObjectPtr* right) because it would require a reference to a pointer which is not allowed
   BUT... you can provide a FUNCTOR to do the job, then provide it when sorting/searching
   a functor is basically a structure with a bool operator()(const MyObjectPtr* left, const MyObjectPtr* right)
   OBSERVATION FOUR

   unordered_set works even better when combining frequent CRUD with frequent lookups
   SUMMARY
   Dealing with tons of objects is par for the course in any significant app.
   Finding a needle in the haystack of those objects is also standard fare.
   Having multiple indices into those objects is also essential.
   Using unordered_set with object pointers and is very powerful.
c++ stl reverse iterator skeleton
From SGI...
reverse_iterator rfirst(V.end());
reverse_iterator rlast(V.begin());

while (rfirst != rlast) 
{
    cout << *rfirst << endl;
    ...
    rfirst++;
}
c++ stl reading a binary file into a string
   std::ifstream in("my.zip",std::ios::binary);
   if (!in)
   {
      std::cout << "problem with file open" << std::endl;
      return 0;
   }
   in.seekg(0,std::ios::end);
   unsigned long length = in.tellg();
   in.seekg(0,std::ios::beg);
 
   string str(length,0);
   std::copy( 
       std::istreambuf_iterator< char >(in) ,
       std::istreambuf_iterator< char >() ,
       str.begin() 
   );

For more, see c++ stl reading a binary file

C/C++ best-in-class tool selection
I need to have easy setup of debug-level tool support for portable C++11 code. And I need to decide and stick to it to be efficient.
  • Compiler selection
    • linux and mac: gcc
    • windows: Visual Studio
  • IDE selection
    • linux and mac: Qt Creator
    • windows: Qt Creator (OR Visual Studio OR eclipse?)
  • Debugger selection
    • linux and mac: Qt Creator
    • windows: Qt Creator (OR Visual Studio OR eclipse?)
C++
Deleting multiple elements in a collection
  • For sequences (vector, list...), best practice is to use the erase-remove STL pattern
  • For associative containers (set, map...): just recurse and remove iterators, they won't muck up the sequence. example

See MemoryModel::saveDirtyObjectsAsNeeded() for example code.

c++ Create a portable C/C++ project
Use QtCreator + CMake, both available everywhere!
   set up:
       create a Qt project as: New Project-> Non-Qt/Plain C Project (CMake Build)
           pick the PARENT and use the name of the existing folder
           set up a build folder in the existing folder, called (base)/qtcreator-release-build
           it creates CMakeLists.txt (THIS IS BASICALLY THE PROJECT FILE!)
           also creates main.c or something
           build it!  make sure it works
           NOW we can edit CMakeLists.txt - change . to ./src to get it to scan the src folder for code!
           and easily add "known" libs!  this was all i needed:
               TARGET_LINK_LIBRARIES(${PROJECT_NAME} pthread websockets)
           make sure it builds
           now we want a DEBUG build as well!
           Projects->Build&Run->pick Build in weird Build/Run pillbox->Edit build configs:
               Rename: release
               CMake build dir: Make sure build dir is set to release ie qtcreator-debug-build
               Add: Clone: release, name: debug
                   for debug, add a custom build step BEFORE make step:
                       command: /usr/bin/cmake
                       args: -DCMAKE_BUILD_TYPE=Debug .
                       working dir: %{buildDir}
                       (add it, and move it up above make)
                       (build em both)
           Projects->Build&Run->pick Run in weird Build/Run pillbox
               Run Configuration: rename to "{...} release"
               clone it into "{...} debug"
               change working dir to match Build path
               change any needed params
   Now you should have build+run debug+release configurations, selectable in the weird project icon in bottom of left-side toolbar ("mode selector")
boost release and debug build for linux
   # download latest boost, eg: boost_1_59_0 
   m@wallee:~/development$ 7z x boost_1_59_0.7z
   cd boost_1_##_0
   build_boost_release_and_debug.sh
   # then patch .bashrc as instructed
   # eclipse and server/nix/bootstrap.sh are customized to match

To upgrade a project:

   cd nix
   make distclean  # removes nasty .deps folders that link to old boost if you let them
   make clean      # removes .o files etc
   cd ../build-Release && make distclean && make clean
   cd ../build-Debug && make distclean && make clean
   cd ..
   ./bootstrap force release
   ./bootstrap force debug
boost release and debug build for Windows
Open a VS2015 x64 Native Tools Command Prompt.

EITHER: for new installs, you have to run bootstrap.bat first, it will build b2; OR: for reruns, remove boost dirs: [bin.v2, stage]. Then build 64-bit:

cd "....\boost_1_59_0"
b2 toolset=msvc variant=release,debug link=static address-model=64
rem trying to avoid excessive options, assuming I don't need these: threading=multi
(old stuff)
      --toolset=msvc-14.0 address-model=64 --build-type=complete --stagedir=windows_lib\x64 stage
      Now open VS2013 x86 Native Tools Command Prompt and build 32-bit:
      cd "C:\Michael's Data\development\sixth_column\boost_1_55_0"
      bjam --toolset=msvc-12.0 address-model=32 --build-type=complete --stagedir=windows_lib\x86 stage
c++11 include base constructor (etc) in derived class
The awesome C++11 way to steal the constructor from the base class, oh man I've been waiting for this one...
 class HttpsServer : public Server<HTTPS>
 {
   // DOES THIS FOR YOU!
   /*
   HttpsServer(unsigned short port, size_t num_threads, const std::string& cert_file, const std::string& private_key_file)
   :
       // Call base class
       Server<HTTPS>::Server(port, num_threads, cert_file, private_key_file)
   {}
   */
   using Server<HTTPS>::Server;
   ....
c++11 containers
sorted_vector use when doing lots of unsorted insertions and maintaining constant sort would be expensive; vector is good for a big pile of things that only occasionally needs a sorted lookup
map sorted binary search tree; always sorted by key; you can walk through in sorted order (choose unordered if not needed!)
multimap same as map but allows dupe keys (not as common)
unordered_map hashmap; always sorted by key; additional bucket required for hash collisions; no defined order when walking through
unordered_multimap same as map but allows dupe keys; dupes are obviously in the same bucket, and you can walk just the dupes if needed
set
multiset
unordered_set
unordered_multiset
sets are just like maps, except the key is embedded in the object, nice for encapsulation.

Items must be const (!) since they are the key - sounds bad, but this is mitigated by the mutable keyword.
You can use mutable on the variables that are not part of the key to remove the const.
This changes the constness of the object from binary (completely const) to logical (constness is defined by the developer).
So... set is a good way to achieve both encapsulation and logical const - make const work for you, not against!  :-)

set (etc.) of pointers sets of pointers are the pinnacle of object stores

The entire object can be dereferenced and accessed then without const issues.
A pointer functor can be provided that does a sort by dereferencing the pointer to the object.
Two requirements: you must make sure yourself that you do not change the key values - you can mark them const, provided in constructor;
you must create sort/equal/hash functors that dereference the pointers to use object contents
(the default will be by pointer address).
The arguably biggest advantage, as a result, is that you can create multiple sets
to reference the same group of objects with different sort funtors to create multiple indices.
You just have to manage the keys carefully, so that they don't change (which would invalidate the sorting).
The primary container can manage object allocation; using a heap-based unique_ptr allocation

   map vs key redux
               
       use a key in the set, derive a class from it with the contents
           + small key
           + encapsulation
           - requires mutable to solve the const problem
       use a key in the set, key includes a mutable object
           + encapsulation
           - weird bc everything uses a const object but we have const functions like save() that change the mutable subobject
       use a map
           + small key
           - no encapsulation, have to deal with a pair instead of an object
               can we just put a ref to key in the value?  sure why not - err, bc we don't have access to it
           + solves const problem bc value is totally mutable by design
           + we can have multiple keys - and the value can have multiple refs to them
           + simpler equal and hash functions
       map:
           create an object with internal key(s)
           create map index(es) with duplicate key values outside the object - dupe data is the downside
       use set(s) with one static key for find(): 
           create an object with internal key(s)
           create set index(es) with specific hash/equals functor(s)
           when finding, use one static key object (even across indexes!) so there isn't a big construction each time; just set the necessary key values
               that proves difficult when dealing with member vars that are references
               but to solve it, just set up a structure of dummy static key objects that use each other; then provide a function to setKey(Object& keyref) { keyref_ = keyref; }
               nope, can't reassign refs
               the solution: use pointers not references
               yes that's right
               just do it
               apparently there was a reason i was anti-reference for all those years
               two reasons to use pointers:
                   dynamically allocated
                   reassignment required
               there ya go.  simple.  get it done. 
           when accessing find results from the set, use a const_cast on the object!
           WARNING: a separate base class with the key sounds good... but fails when you have more than one index on the object.  just use a static key object for them all!
c++11 example for large groups of objects with frequent crud AND search
Best solution is an unordered set of pointers:
typedef boost::unordered_set<MajorObject*> MajorObjects;
c++11 example for large groups of objects with infrequent crud and frequent search
Best solution is a vector of pointers sorted on demand (sorted_vector):
TODO
c++11 example to associate two complex objects (one the map key, one the map value)
Use unordered_map with a custom object as key. You must add hash and equals functions. Boost makes it easy:
static bool operator==(MyKeyObject const& m1, MyKeyObject const& m2)
{
    return 
            m1.id_0 == m2.id_0
        &&  m1.id_1 == m2.id_1;
}
static std::size_t hash_value(MyKeyObject const& mko)
{
    std::size_t seed = 0;
    boost::hash_combine(seed, mko.id_0);
    boost::hash_combine(seed, mko.id_1);
    return seed;
}
typedef boost::unordered_map<MyKeyObject, MyValueObject*> MyMap;

Note that you can extend this to use a pointer to a key object, whoop.

c++11 example for multiple unordered_set indexes into one group of objects
Objects will be dynamically created. One set should include them all and be responsible for memory allocation cleanup:
TODO
c++11 example for set with specific sorting
Use set with a specific sort functor. You can create as many of these indexes as you want!
struct customers_set_sort_functor
{
    bool operator()(const MyObject* l, const MyObject* r) const
    {
        // the id is the key
        return l->id_ < r->id_;
    }
};
typedef set<MyObject*,myobject_sort_by_id_functor> MyObjectsById;
c++11 loop through vector to erase some items
Note that other containers' iterators may not be invalidated so you can just erase() as needed...

For vectors, you have to play with iterators to get it right - watch for proper ++ pre/postfix!

for (it = numbers.begin(); it != numbers.end(); )  // NOTE we increment below, only if we don't erase
{
    if (*it.no_good()) 
    {
        numbers.erase(it++);  // NOTE that we ERASE THEN INCREMENT here.
    }
    else 
    {
        ++it;
    }
}

I thought I had always looped backwards to do this, I *think* that's ok too, but I don't see it used in my code, I think I'll avoid.  :-)

c++11 range based for loop, jacked with boost index if needed
No iterator usage at all. Nice at times, not enough at others. Make SURE to always use a reference or you will be working on a COPY. Make it const if you aren't changing the object.
for (auto& mc : my_container)
    mc.setString("default");
for (const auto& cmc : my_container)
    cout << cmc.asString();

boost index can give you the index if you need it, sweet:

#include <boost/range/adaptor/indexed.hpp>
...
for (const auto &element: boost::adaptors::index(mah_container))
    cout << element.value() << element.index();
c++11 for loop using lambda
This C++11 for loop is clean and elegant and a perfect way to check if your compiler is ready for c++11:
vector<int> v;
for_each( v.begin(), v.end(), [] (int val)
{
   cout << val;
} );

This is using a lambda function, we should switch from iterators and functors to those - but not quite yet, since we're writing cross-platform code. Do not touch this until we can be sure that all platforms provide compatible C++11 handling.

c++11 integer types
I really like the "fast" C++11 types, that give best performance for a guaranteed minimum bit width.

Use them when you know a variable will not exceed the maximum value of that bit width, but does not have to be a precise bit width in memory or elsewhere.

Pick specific-width fields whenever data is shared with other processes and components and you want a guarantee of its bit width.

And when using pointer size and array indices you should use types defined for those specific situations.

FAST types:

   int_fast8_t
   int_fast16_t                fastest signed integer type with width of
   int_fast32_t                at least 8, 16, 32 and 64 bits respectively
   int_fast64_t
   uint_fast8_t
   uint_fast16_t               fastest unsigned integer type with width of
   uint_fast32_t               at least 8, 16, 32 and 64 bits respectively
   uint_fast64_t

SMALL types:

   int_least8_t
   int_least16_t               smallest signed integer type with width of
   int_least32_t               at least 8, 16, 32 and 64 bits respectively
   int_least64_t
   uint_least8_t
   uint_least16_t		smallest unsigned integer type with width of
   uint_least32_t		at least 8, 16, 32 and 64 bits respectively
   uint_least64_t

EXACT types:

   int8_t                      signed integer type with width of
   int16_t                     exactly 8, 16, 32 and 64 bits respectively
   int32_t                     with no padding bits and using 2's complement for negative values
   int64_t                     (provided only if the implementation directly supports the type)
   uint8_t                     unsigned integer type with width of
   uint16_t                    exactly 8, 16, 32 and 64 bits respectively
   uint32_t                    (provided only if the implementation directly supports the type)
   uint64_t

SPECIFIC-USE types:

   intptr_t                    integer type capable of holding a pointer
   uintptr_t                   unsigned integer type capable of holding a pointer 
   size_t                      unsigned integer type capable of holding an array index (same size as uintptr_t)
C++11 scoped enumeration
C++11 has scoped enumeration, which lets you specify the SPECIFIC VARIABLE TYPE for the enum. Perfect, let's use uint_fast32_t.
enum class STRING_PREF_INDEX int_fast32_t: { ... };

Unfortunately, gcc gives me a scary warning, and stuff fails. For some reason, it does not know about the provided type, although it is definitely defined. Revisit this later if you have time.

warning: elaborated-type-specifier for a scoped enum must not use the ‘class’ keyword

Old skool is still cool:

typedef enum
{
    // assert( SP_COUNT == 2 );
    SP_FIRST = 0                ,
    SP_SOME_PREF = SP_FIRST     ,
    SP_ANOTHA                   ,

    SP_COUNT
} STRING_PREF_INDEX;
c++ in-memory storage of "major" objects
   OBSERVATION ONE

   Consider An Important Qt Design: QObjects cannot normally be copied
       their copy constructors and assignment operators are private
       why?  A Qt Object...
           might have a unique QObject::objectName(). If we copy a Qt Object, what name should we give the copy?
           has a location in an object hierarchy. If we copy a Qt Object, where should the copy be located?
           can be connected to other Qt Objects to emit signals to them or to receive signals emitted by them. If we copy a Qt Object, how should we transfer these connections to the copy?
           can have new properties added to it at runtime that are not declared in the C++ class. If we copy a Qt Object, should the copy include the properties that were added to the original?
   in other words, a QObject is a pretty serious object that has the ability to be tied to other objects and resources in ways that make copying dangerous
   isn't this true of all serious objects?  pretty much
   OBSERVATION TWO

   if you have a vector of objects, you often want to track them individually outside the vector
   if you use a vector of pointers, you can move the object around much more cheaply, and not worry about costly large vector reallocations
   a vector of objects (not pointers) only makes sense if the number of objects is initially known and does not change over time
   OBSERVATION THREE

   STL vectors can store your pointers, iterate thru them, etc.
   for a vector of any substantial size, you want to keep objects sorted so you can find them quickly
   that's what my sorted_vector class is for; it simply bolts vector together with sort calls and a b_sorted status
   following STL practices, to get sorting, you have to provide operator< for whatever is in your vector
   BUT... you are not allowed to do operator<(const MyObjectPtr* right) because it would require a reference to a pointer which is not allowed
   BUT... you can provide a FUNCTOR to do the job, then provide it when sorting/searching
   a functor is basically a structure with a bool operator()(const MyObjectPtr* left, const MyObjectPtr* right)
   OBSERVATION FOUR

   unordered_set works even better when combining frequent CRUD with frequent lookups
   SUMMARY
   Dealing with tons of objects is par for the course in any significant app.
   Finding a needle in the haystack of those objects is also standard fare.
   Having multiple indices into those objects is also essential.
   Using unordered_set with object pointers and is very powerful.
c++ stl reverse iterator skeleton
From SGI...
reverse_iterator rfirst(V.end());
reverse_iterator rlast(V.begin());

while (rfirst != rlast) 
{
    cout << *rfirst << endl;
    ...
    rfirst++;
}
c++ stl reading a binary file into a string
   std::ifstream in("my.zip",std::ios::binary);
   if (!in)
   {
      std::cout << "problem with file open" << std::endl;
      return 0;
   }
   in.seekg(0,std::ios::end);
   unsigned long length = in.tellg();
   in.seekg(0,std::ios::beg);
 
   string str(length,0);
   std::copy( 
       std::istreambuf_iterator< char >(in) ,
       std::istreambuf_iterator< char >() ,
       str.begin() 
   );

For more, see c++ stl reading a binary file

C/C++ best-in-class tool selection
I need to have easy setup of debug-level tool support for portable C++11 code. And I need to decide and stick to it to be efficient.
  • Compiler selection
    • linux and mac: gcc
    • windows: Visual Studio
  • IDE selection
    • linux and mac: Qt Creator
    • windows: Qt Creator (OR Visual Studio OR eclipse?)
  • Debugger selection
    • linux and mac: Qt Creator
    • windows: Qt Creator (OR Visual Studio OR eclipse?)
C++ libraries

Simple-Web-Server

C++ https libraries

Configure Qt development on Windows + Mac + linux

Build the TagLib library with Visual Studio 2013

C/C++ building/linking
gcc makefile pain
  • I went through a LOT of pain to determine that gcc requires libraries to be listed AFTER the source and output parameters
  • set LD_LIBRARY_PATH to point to your libs if they are not in system location
  • Make sure Makefile uses TABS NOT SPACES. it's a FACT OF LIFE. hate it if you want. plenty more things to hate about linux/C as well.
gcc install multiple versions in ubuntu (4 and 5 in wily, eg)
My code will not compile with gcc 5, the version provided with Ubuntu wily.

It gives warnings like this:

/home/m/development/boost_1_59_0/boost/smart_ptr/shared_ptr.hpp:547:34: warning: ‘template<class> class std::auto_ptr’ is deprecated [-Wdeprecated-declarations]

and outright errors like this:

depbase=`echo AtServer.o | sed 's|[^/]*$|.deps/&|;s|\.o$||'`;\
g++ -DPACKAGE_NAME=\"at_server\" -DPACKAGE_TARNAME=\"at_server\" -DPACKAGE_VERSION=\"1.0\" -DPACKAGE_STRING=\"at_server\ 1.0\" -DPACKAGE_BUGREPORT=\"m@abettersoftware.com\" -DPACKAGE_URL=\"\" -DPACKAGE=\"at_server\" -DVERSION=\"1.0\" -I. -I../../src  -I/home/m/development/Reusable/c++ -I/home/m/development/Reusable/c++/sqlite -std=c++11 -I/home/m/development/boost_1_59_0  -ggdb3 -O0 -std=c++11 -MT AtServer.o -MD -MP -MF $depbase.Tpo -c -o AtServer.o ../../src/AtServer.cpp &&\
mv -f $depbase.Tpo $depbase.Po
In file included from /usr/include/c++/5/bits/stl_algo.h:60:0,
                from /usr/include/c++/5/algorithm:62,
                from /usr/include/c++/5/ext/slist:47,
                from /home/m/development/boost_1_59_0/boost/algorithm/string/std/slist_traits.hpp:16,
                from /home/m/development/boost_1_59_0/boost/algorithm/string/std_containers_traits.hpp:23,
                from /home/m/development/boost_1_59_0/boost/algorithm/string.hpp:18,
                from /home/m/development/Reusable/c++/utilities.hpp:4,
                from ../../src/MemoryModel.hpp:11,
                from ../../src/SqliteLocalModel.hpp:13,
                from ../../src/AtServer.cpp:70:
/usr/include/c++/5/bits/algorithmfwd.h:573:13: error: initializer provided for function
    noexcept(__and_<is_nothrow_move_constructible<_Tp>,
            ^
/usr/include/c++/5/bits/algorithmfwd.h:582:13: error: initializer provided for function
    noexcept(noexcept(swap(*__a, *__b)))

You can set up the update-alternatives tool to switch out the symlinks:

sudo update-alternatives --install /usr/bin/gcc gcc /usr/bin/gcc-4.9 60 --slave /usr/bin/g++ g++ /usr/bin/g++-4.9
sudo update-alternatives --install /usr/bin/gcc gcc /usr/bin/gcc-5 20 --slave /usr/bin/g++ g++ /usr/bin/g++-5

BUT that seems BAD to me to switch out the compiler used by the SYSTEM. Instead, we should specify the compiler required by the PROJECT. This is supposed to do it, but it still uses /usr/include/c++/5, which seems wrong, and gives me errors:

CC=gcc-4.9
C/C++ debugging
Qt Creator
  • I was able to build lws in debug, install in /usr/local, then I could step right into its code
m@case:~/development/causam/git/np/nop-bigress-client-c$ ./build_lws_debug
 Install the project...
 -- Install configuration: "DEBUG"
 -- Installing: /usr/local/lib/pkgconfig/libwebsockets.pc
 -- Installing: /usr/local/lib/libwebsockets.a
 -- Up-to-date: /usr/local/include/libwebsockets.h
 (etc)
  • valgrind is easy to use. Just find debugger pane top-left dropdown, switch mode from Debugger to Memcheck, and restart debugger.
  • Signals: At some point I needed to force gdb to pass SIGINT, SIGPIPE... signals to the app instead of gdb, after ssl kept causing Qt to pop a message and break in the debugger. To do so:
       Qt Creator->Options->Debugger->
           [ ] Show a message box when receiving a signal
           Debugger Helper Customization: handle SIGPIPE pass nostop noprint
  • You should be able to install debug builds of libraries into /usr/local (as mentioned above). Older notes if you can't get that going: To debug into an included library that is not installed system-wide:
use Tools->Options->General->Source Paths Map
 source: /home/m/development/causam/git/nop-bigress-client-c/cmake-debug
   that's the COMPLETE PATH TO DEBUG EXE
 target: /home/m/development/causam/git/libwebsockets-master/lib
   at first i thought this would work: /usr/local/lib
   but i did a static build so i needed teh actual exe!  COOL
c/c++ gdb debugging
(gdb) help break
Set breakpoint at specified line or function.
Argument may be line number, function name, or "*" and an address.
If line number is specified, break at start of code for that line.
If function is specified, break at start of code for that function.
If an address is specified, break at that exact address.
With no arg, uses current execution address of selected stack frame.
This is useful for breaking on return to a stack frame.

Multiple breakpoints at one place are permitted, and useful if conditional.    

Do "help breakpoints" for info on other commands dealing with breakpoints.
ddd gives you a front end. I need to use it more, compare to other options
C
Library handling in linux
gcc library basics (good to know for autotools and CMake too)
  • The gcc -l command line option specifies a specific library NAME. It will add the rest to give you (eg) lib[NAME].a
  • Specify the library search path with the -L option. Things that you as a user install seem to default to /usr/local/lib
  • This will link myprogram with the static library libfoo.a in the folder /home/me/foo/lib.
       gcc -o myprogram -lfoo -L/home/me/foo/lib myprogram.c

Tools to check library dependencies:

  • ldd [executable]
  • objdump -p /path/to/program | grep NEEDED
  • objdump -x [object.so]
  • readelf -d [exe]
  • sudo pldd <PID>
  • sudo pmap <PID>

Cross Compiling

C - Create a portable command line C project in Visual Studio
   Visual Studio: File -> New -> project
   Visual C++ -> Win32 -> Win32 Console Application
   name: oms_with_emap
   next -> click OFF precompiled header checkbox (even tho it didn't seem to respect it)
   you'll get a _tmain(..., TCHAR*...)
   change it to main(..., char*...)
   change the project to explicitly say "Not using precompiled header"
   remove the f'in stdafx.h
   recompile!  should be clean
   vs will recognize C files and compile accordingly
Bootstrap
Single Page Application
css:
  .mdm-scrollable-div
  {
    height: 100%; 
    overflow: auto;
  }

html:

  <-- header -->

  <div class="frame" id="frame1top">
    <div class="mdm-scrollable-div">
      ...
    </div>
  </div>

  <-- footer -->

js:

  // We need to adjust layout on resize, in ready handler, and programmatically as needed.
  $(window).resize(adjustLayout).resize();

  $( document ).ready(function() {

    // We need to adjust in ready handler, on resize, and programmatically as needed
    adjustLayout();

    /* your other page load code here*/

  });

  function adjustLayout(){

      // ----------------------
      // Fundamental dimensions
      var hh = $('#theheader').outerHeight();
      var fh = $('#thefooter').outerHeight();
      var workspace_height = $(window).height() - hh - fh;
      var workspace_width = $(window).width(); 
      // ----------------------
      
      var cols = 1;
      var col_gap = 16;
      
      // Margin is based on percent of one column width, so it goes to zero before column width does.  :-)
      // Width is based on total width minus margins, then split between columns.
      var margin = ($(window).width() / cols) * .04;
      var w = ($(window).width() - (margin * (cols + 1))) / cols;
      
      var h1 = workspace_height;
      
      $('#frame1top').css({
          display:'block', 
          position:'absolute',
          left: margin * 1 + w * 0,
          top: hh,
          width: w,
          height: h1
      });
  }
Button handlers
Buttons can be represented with labels; in that case, use a click handler:
    jQuery('<label id="'+cycle.apsID+'" class="btn">MahButt</label>')
    .appendTo('#'+chartid+'-apsbar')
    .on('click', function(){
      $(location).attr('href','/v1/parameters/'+this.id+'.html');
    });

You can also get the checked state, and prevent the button from being pressed, among other things:

      jQuery('<label id="'+cycle.run+'" ...
      .on('click', function(e){
        var run = this.id;
        // Look for the (label->input)
        if (this.firstChild.checked)
        ...
        // To prevent checked/pressed state if desired:
        e.stopPropagation();
      }  

Button bars are represented by labels wrapped around inputs:

    <div class="btn-group live-buttons" data-toggle="buttons"><div class="btn-group"><label class="btn"><input type="checkbox">text...

In that case use a change handler on the input:

      var label = jQuery('<label class="btn btn-sm btn-'+color+'"></label>').appendTo(action_button_bar);
      var input = jQuery('<input class="run-'+cmd+'" type="checkbox" autocomplete="off" value="'+cycle.run+'">').appendTo(label)
      .on('change', cycle.owned? function(){
        patchPick(this.value,'{ "action" : "run-hold"       }');
      } : function() {
        patchPick(this.value,'{ "action" : "run-buy"        }');
      });
      var text = jQuery('<span class="glyphicon glyphicon-'+glyph+'"></span> <span class="hidden-xs">'+cmd+'</span>').appendTo(label);

If you just need clicks from the button bar, you do NOT NEED input:

    <div class="btn-group live-buttons" data-toggle="buttons"><div class="btn-group"><label class="btn">text...

Then you can put the change handler right on the label:

    var applybar = jQuery('<div id="applybar-'+cycle.run+'" class="btn-group pull-right" data-toggle="buttons" />');
    var apply_button = jQuery('<label id="apply-'+cycle.run+'" class="btn btn-sm btn-moneygreen"><span class="glyphicon glyphicon-ok"></span><span class="hidden-xs"> Apply</span></input></label>')
    .on('click', function(e) {    
    
        // Do work

        e.stopPropagation();
    })
    .appendTo(applybar)
    ;
Programmatically pressing a button
You MUST DO TWO THINGS:
  • Set the active class on the button/label
  • Set the input to checked
<label class="btn active">
  <input type="checkbox" autocomplete="off" checked>
  Click me
</label>
Collapsible panel
  • Use a unique id on content div, data-target to connect panel header to content, mdm-panel-collapser so javascript can find icon, abt-scrollable-panel for margin.
  • HTML for each panel (starts collapsed)
    <div class="panel panel-default abt-scrollable-panel">
      <div class="panel-heading collapsed" data-toggle="collapse" data-target="#my-content-block-one">Software development skills<span class="mdm-panel-collapser text-muted glyphicon glyphicon-chevron-down pull-right"></span></div>
      <div class="collapse" id="my-content-block-one">
        <div class="mdm-panel-body">
  
         <!-- CONTENT, can include another nested panel, just add .mdm-nested-panel to class; example: -->
  
          <div class="panel panel-default mdm-scrollable-panel mdm-nested-panel">
            <div class="panel-heading collapsed mdm-job-panel-heading" data-toggle="collapse" data-target="#toshiba-job"><p><strong>Senior Developer and Offshore Manager</strong><i> - Toshiba Global Commerce Solutions, Inc.</i><span class="mdm-panel-collapser text-muted glyphicon glyphicon-chevron-up pull-right"></span></p><p><small>April 2014 – August 2015</small></p></div>
            <div class="collapse in" id="toshiba-job">
              <div class="mdm-panel-body">
                <!-- SUBCONT(IN)ENT im so funny -->
              </div>
            </div>
          </div>
          

        </div>
      </div>
    </div>
  • For an expanded panel, simply change icon from down to up, and add "in" to content div:
    ... glyphicon-chevron-up ...
    <div class="collapse in mdm-panel-body" id="collapseOrderItems1">
  • Define Javascript collapser once on load
 $('.collapse').on('hide.bs.collapse show.bs.collapse', 
   toggleCollapser
 );
  • CSS
   .mdm-nested-panel {
     margin-top: 1em;
     margin-left: 1em;
   }
   collapse.mdm-panel-body collapsing.mdm-panel-body {
       margin: 1em;
   }
  • Common reusable function:
 function toggleCollapser(e) {
     $(e.target)
         .prev('.panel-heading')
         .find('.mdm-panel-collapser')
         .toggleClass('glyphicon-chevron-down glyphicon-chevron-up');
 
   // Prevent bubble up to any parent collapsers.
   // This allows nested collapsers, whoop.
   e.stopPropagation();
 }
D3
ALWAYS initialize incoming data
Hard-earned data-import lessons:
  • tsv() will do complete date formatting, so when you remove it to use JSON directly, you HAVE TO convert JSON timestamps of ANY kind to full Javascript date objects:
 var data = [{"date":"2017-04-01T04:00:00.000Z",...
 // MDM WE MUST MASSAGE DATA HERE
 // JSON VALUES ARE ALWAYS STRINGS, we need to change to Javascript DATE!
 // Incoming datasets need a bit of massaging.
 // We do that in a function so we can reuse it on incoming dataset updates.
 function initializeDataset(dataset) {
     // TIME ON X
     // Convert date strings to actual date values.
     // MDM MAKE SURE this matches the incoming format.
     // Adding date so we can run the chart across days without trouble.
     // var parseDate = d3.time.format("%H:%M:%S %m-%d-%Y").parse;
     // var parseDate = d3.timeParse("%Y %b %d");
     var parseDate = d3.utcParse("%Y-%m-%dT%H:%M:%S.%LZ");
     dataset.forEach(function(d) {
         // there is NO SKIPING THIS STEP, you have to get valid Javascript date objects out of JSON strings
         d.date = parseDate(d.date);
         // we WOULD have to divide all data by zero,
         // but we already grabbed post-data that was already converted
         // This WORKS but makes data / 10000 (very small numbers)
         //for (var i = 1, n = dataset.columns.length; i < n; ++i) d[dataset.columns[i]] = d[dataset.columns[i]] / 100;
     });
 }
 initializeDataset(data);

  • tsv() will create an ARRAY but it also jams in a columns PROPERY; it takes two steps to duplicate that:
  var data = [
    {"date":"2015-06-15T04:00:00.000Z","Google Chrome":0.48090000000000005,...},
    {"date":"2015-06-22T04:00:00.000Z","Google Chrome":0.48979999999999996,...),
 ];
 data.columns = ["date","Google Chrome","Internet Explorer",...  ];
ALWAYS set range and domain properly
You have to know your display size (range) and the min and max of your data (domain), on each axis. From Mike's docs:

D3’s scales specify a mapping from data space (domain) to display space (range).

       D3’s scales can also be used to interpolate many other 
       types of display-space values, such as paths, color spaces 
       and geometric transforms.
       
         var x = d3.scale.linear()
             .domain([0, d3.max(data)])
             .range([0, 420]);
 
       Although x here looks like an object, it is also a function 
       that returns the scaled display value in the range 
       for a given data value in the domain. 
       For example, an input value of 4 returns 40, and an input value of 16 
       returns 160. To use the new scale, simply replace the 
       hard-coded multiplication by calling the scale function:
       
       d3.select(".chart")
         .selectAll("div")
           .data(data)
         .enter().append("div")
           .style("width", function(d) { return x(d) + "px"; })
           .text(function(d) { return d; });

Watch out for Mike's examples where he (trickily) doesn't bother to set a domain because his data is in the [0..1] set, and that's the default domain, apparently.

RESTFul http
Verbs
  • get - to make requests that are simple -response may be complex
  • post - to push a form or JSON to the server
Javascript and JQuery
Formatting examples (numbers dates etc.)
 // Javascript sucks at dates.
 // d3 to the rescue!  
 var parseInitDate = d3.utcParse("%Y-%m-%dT%H:%M:%SZ");
 var initDate = parseInitDate(initial_managed_value_date);
 var initDateString = 
     initDate.getUTCFullYear() 
     +"/"+ (initDate.getUTCMonth()+1) 
     +"/"+ initDate.getUTCDate()
     + " " + initDate.getUTCHours() 
     + ":" + initDate.getUTCMinutes() 
     + ":" + initDate.getUTCSeconds();
 var rightNow = new Date;
 var initToNow_DecimalDays = parseFloat(rightNow - initDate) / 864e5;  // 86,400,000 ms/day
 var initToNow_PercentOfYear = initToNow_DecimalDays / 365.0;
 var change_in_value = (parseFloat(total_managed_value) - parseFloat(initial_managed_value))/parseFloat(initial_managed_value);
 $('#initial_managed_value').val(initial_managed_value.toFixed(2)); 
 $('#initial_managed_value_date').val(initDateString);
 $('#change_in_value').val((change_in_value*100.0).toFixed(1)+ "%");
 $('#equivalent_apr').val((change_in_value/initToNow_PercentOfYear*100.0).toFixed(1)+ "%");
 $('#total_managed_value_readonly').val(total_managed_value.toFixed(2));
Array vs Object lifetime
       // If the JSON data were an array, we would need to slice-to-clone to keep it around:
       //     var dataCopy = data.slice();
       //     updateStockData(run,dataCopy);
       // We have an object so we can pass a reference, and it will keep the object around.
Event handlers
Three basic methods:
  • Attach a function to an event of a specific element; charts.js example:
     var label = jQuery('<label class="btn btn-sm btn-'+color+'"></label>').appendTo(action_button_bar);
     var input = jQuery('<input class="run-'+cmd+'" type="checkbox" autocomplete="off" value="'+cycle.run+'">').appendTo(label)
     .on('change', cycle.owned? function(){
       patchPick(this.value,'{ "action" : "run-hold"       }');
     } : function() {
       patchPick(this.value,'{ "action" : "run-buy"        }');
     });
     var text = jQuery(' ').appendTo(label);
  • Add a function handler once, for a class of elements; picks.js example:
 $("input[type='checkbox'][class='run-top'    ]").change(function() { patchPick(this.value,'{ "action" : "run-move-top"   }');  });
  • Event delegation: Javascript takes events and "bubbles them up" through the chain of parents; set up a handler on a parent to listen for these bubbles.
Debug clicks on ANY PAGE
  • Press F12 => Sources => Event Listener Breakpoints righthand pane => Mouse => [x] click

B O O M we have it captured and can step into ANYTHING.

JWT flow
Client is easy!
   Client application                                            API
   --------                                              -----------
        |                                                      |
        |                   GET /api/employees                 |
        |----------------------------------------------------->|
        |                     403 Forbidden                    |
        |<-----------------------------------------------------|
        |                                                      |
        |                                                      |
        |                 POST /api/authenticate               |
        |     { login: "john.doe", password: "password" }      |
        |----------------------------------------------------->|
        |                      200 Success                     |
        |             { token: "my.personal.token" }           |
        |<-----------------------------------------------------|
        |                                                      |
        |                                                      |
        |                 GET /api/employees                   |
        | Header { "Authorization: Token "my.personal.token" } |
        |----------------------------------------------------->|
        |                      200 Success                     |
        |<-----------------------------------------------------|
        |                                                      |
Node.js
Node.js installation
Node create a new module
Register with npm
A one-time registration is required on a new machine if you want to publish from it:
npm adduser
Username: moodboom
Password: (see private)
Email: (this IS public) moodboom@gmail.com
Publish a node module
sudo npm install -g # keep doing this until you are happy with local install
# update version in package.json
# this creates a FULL "annotated" tag, not a "lightweight" tag that doesn't show up for [git describe] - it also removes the need for a separate commit
git tag -a 1.0.5 -m "changes include..."
git push && git push --tags  # NOTE: bitpost has a git hook to push changes all the way up to github
npm publish
Update a node module's dependencies
# make sure dependency in package.json has a carat at the beginning of its version (^x means "at least" version x)
# make sure the dependency has a new version available - completely publish it first if it is your own
# then you can simply reinstall from within the module folder to get all dependencies upgraded
sudo npm install -g
Develop several node modules at once
Convert dependencies to use local packages instead of published versions, eg:
cd ~/development/mah-haus
npm install -S /home/m/development/thedigitalage/rad-scripts

Then reinstall everything (local dependent modules, then parent modules, pita - consider links if doing longer-term dev)

sudo npm install -g

Then convert back to published versions as they become available (it's up to me to stabilize and publish new module versions):

cd ~/development/mah-haus
npm install -S rad-scripts
auto AWS
  • npm install -g aws-sdk
  • Add credentials here: C:\Users\Administrator\.aws
  • see existing scripts, anything is possible
install bootstrap
  • npm install -g grunt-cli
  • mkdir mysite && cd mysite
  • npm install bootstrap
  • cd node_modules/bootstrap
  • npm install # to actually pull down dependencies
  • grunt dist # builds and minifies so you're good to go!
css
 /* class="first second" */
 .first.second {}
 /* class="first" OR class="second" */
 .first, .second {}
 /* class="first second", or class="second", or class="third second", or class="second third" */
 .second {}
 /* apply to any .child at any depth under .parent */
 .parent .child {}
 /* apply to .child if it is DIRECTLY under .parent */
 .parent > .child {}
SQL
Sqlite timestamp-to-readable-date query
select quote, strftime('%d-%m-%Y %H:%M:%S', datetime(timestamp, 'unixepoch')) from StockQuotes as s where s.symbol="TSLA" order by timestamp;
Java
Scala
PHP
php debugging
Tail these:
tail -f /var/log/apache2/sitelogs/thedigitalage.org/ssl_error_log
tail -f /var/log/ampache-tda/ampache.(today).log
This leads to too much noise, not needed...
emacs /etc/php/apache2-php5.3/php.ini
  display_errors = On
/etc/init.d/apache restart
git
misc
SVN repo move across servers
  • use a tool like VisualSVN to stop SVN server
  • copy the repo's entire svn server directory, eg: c:\svn\Software
  • copy it into the new server under a unique name, eg: c:\svn\NewSoftware
  • use a tool like VisualSVN to restart SVN server

You should now have the new code, accessible as usual from svn. NOTE from Tom: If you move code around, make sure you tag it, then copy the tag, to preserve history (as opposed to directly moving the folder). Weird but true.

SQL Server 2008+ proper upsert using MERGE
       -- We need an "upsert": if record exists, update it, otherwise insert.
       -- There are several options to do that.
       -- Trying to do it correctly means...
       --		1) use a lock or transaction to make the upsert atomic
       --		2) use the best-available operation to maximize performance
       -- SQL Server 2008 has MERGE which may be slightly more efficient than 
       -- separate check && (insert||update) steps.  And we can do it with
       -- a single lock instead of a full transaction (which may be better?).
       -- It's messy to code up though since three blocks of fields must be specified.  
       -- Cest la vie.
       MERGE [dbo].[FACT_DCSR_RemPeriodMonthlyReport] WITH (HOLDLOCK) AS rpmr
       USING (SELECT @ID AS ID) AS new_foo
             ON rpmr.ID = new_foo.ID



               @last_months_year as DCSRYear,
               @last_month as DCSRMonth,
               @last_month_name as MonthName,
               Device_Type_ID,




       WHEN MATCHED THEN
           UPDATE
                   SET f.UpdateSpid = @@SPID, 
                   UpdateTime = SYSDATETIME() 
       WHEN NOT MATCHED THEN
           INSERT
             (
                   ID, 
                   InsertSpid, 
                   InsertTime
             )
           VALUES
             (
                   new_foo.ID, 
                   @@SPID, 
                   SYSDATETIME()
             );
Web Services
Firefox Addon development
c++ Create a portable autotools C++ project in linux
(OLD, USE CMAKE NOW)

For anything serious, it's best to clone an existing project's skeleton.

  • main.cpp
  • MVC code
  • nix/copy_from folder
  • make sure .bashrc is configured for boost
  • nix$ bootstrap_autotools_project.sh force release debug
  • set up eclipse according to screenshots in Eclipse
c++ Create a portable C++ project in Visual Studio
(OLD, USE CMAKE NOW)

If you don't have existing code, it's probably best to create a project_name.cpp file with your main() function.

int main( int argc, char * argv[] )
{ 
    return 0;
}

Then in Visual Studio...

File->New->Project from existing code
C++
(then use mostly defaults on this page, once you provide file location and project name)
Project file location:  <base>\project_name
Project name: project_name
[x] Add files from these folders
   Add subs  
   [x]        <base>\project_name
NEXT
Use Visual Studio
  Console application project
  No ATL, MFC, CLR
NEXT
NEXT
FINISH

Then add Reusable and boost include and lib paths: Example if you built boost according to notes below:

Project->Properties->All Configurations->Configuration Properties->VC++ Directories->Include Directories->
   $(VC_IncludePath);$(WindowsSDK_IncludePath);..\..\..\Reusable\c++
INCLUDE: C:\Software Development\boost_1_53_0
LIB: C:\Software Development\boost_1_53_0\stage\lib