Coroutines in Python

Python 3.5 added native support for coroutines. Actually, there were several steps towards the current implementation. See Wikipedia, and it seems a bit messy to me:

  • Python 2.5 implements better support for coroutine-like functionality, based on extended generators (PEP 342).
  • Python 3.3 improves this ability, by supporting delegating to a subgenerator (PEP 380).
  • Python 3.4 introduces a comprehensive asynchronous I/O framework as standardized in PEP 3156, which includes coroutines that leverage subgenerator delegation.
  • Python 3.5 introduces explicit support for coroutines with async/await syntax (PEP 0492).

Before Python 2.5, there were only generators.

In Python 2.5, yield was refined to be an expression rather than a statement, which gave the possibility to implement a simple coroutine. But still a lot of work left for programmers to use it. For instance, a simple conroutine scheduler was required.

In Python 3.3, yield from was added to support subgenerators. Nothing to do with coroutines.

In Python 3.4, the Father of Python (Guido van Rossum) wrote a PEP himself to add an asyncio module to simplify coroutine usage in Python. An official scheduler was added. We can use @asyncio.coroutine to decorate a function. We can use yield from expressions to yield to a specific coroutine.

In Python 3.5, async/await syntax was added, borrowed from C#. The newest PEP made coroutines a native Python language feature, and clearly separated them from generators. A native coroutine now declares with async def syntax, and yield from is replaced with await expression. This removes generator/coroutine ambiguity. So in Python 3.5, coroutines used with asyncio may be implemented using the async def statement, or by using generators. Generator-based coroutines should be decorated with @asyncio.coroutine, although this is not strictly enforced. The decorator enables compatibility with async def coroutines, and also serves as documentation. See Python documents here.

The implementation can be found in this commit.

I wrote a echo server/client sample to try corutines. Server code first:

Client code here, or you can simply use telnet command:

Server output:

Client output:

With Python 3.5 on Ubuntu 16.04, we can also use async/await:

Basic Usage of Boost MultiIndex Containers

Just take a simple note here.
The Boost Multi-index Containers Library provides a class template named multi_index_container which enables the construction of containers maintaining one or more indices with different sorting and access semantics.


To use with pointer values, only limited change needed as highlighted:

Enabing Pretty Permalinks

Well.. long time no see. Just have some time to optimize the site for better analysis.

According to the official tutorial:
1. Enable mod_rewrite in apache2.

2. Enable FollowSymLinks option, which is default.
3. Enable FileInfo directives. Edit /etc/apache2/sites-available/, add:

4. Restart apache:

Updated Dec 31, 2015:
Enabling mod_rewrite rewrites all requests including the one used by mod_status. To disable this, add a rule to the .htaccess file.

Then, change its user & group attributes to prevent overwriting from apache.

3 New Plugins Installed

Just added another personal site for one of my friends. So another chance to speed up loading for my own site, from mainland china >.<

1. Disable google fonts: Remove Google Fonts References

2. Use local avatar: WP User Avatar

3. Use local emoji: WP Local Emoji (installed, but not activated.)

Database Access Layer in C++

We have JDBC in Java… and SOCI in C++… Well, it’s not so easy as it should be. To build with cmake:

The documents seem outdated, many options do not work. Just managed to figure out from the *.cmake source files. You can also download the oracle instant client SDK, and re-arrange the directory structure for build.

Code snippet I extracted from its unit tests:

Updated Apr 20, 2015:

1. Under RHEL5/CentOS5, I got errors like:

It’s due to SELinux security feature. Simply workaround it with:

2. Oracle uses oraociei11.dll or for client data. They are both large files(110+MB), since they support multiple languages. Instead, you can use oraociicus11.dll(30+MB) or These files contain only English support.