Skip to main content

An module written with pure Python C Extensions to open a file and cache the more recent accessed lines

Project description

Fast File Package

An module written with pure Python C Extensions to open a file and cache the more recent accessed lines. See these files for an usage example:

  1. tests/fastfileperformance.py
  2. tests/fastfiletest.py
  3. tests/getline_c_performance.cpp
  4. tests/getline_cpp_performance.cpp

Installation

Requires Python 3 , pip3, distutils, setuptools, wheel and a C++ 11 compiler installed:

sudo apt-get install build-essential g++ python3-dev python3 python3-pip
pip3 install setuptools wheel

Then, you can clone this repository with:

git clone https://github.com/evandrocoan/fastfilepackage
cd fastfilepackage
pip3 install .

Or install it with:

pip3 install fastfilepackage

Alternive file reading

There are available 3 alternative implementations for file reading. Both of them, should have the same performance.

  1. You can define FASTFILE_GETLINE=0 to use the Python builtins.open() implementation (default)
  2. You can define FASTFILE_GETLINE=1 to use the C++ std::getline() implementation
  3. You can define FASTFILE_GETLINE=2 to use the POSIX C getline() implementation

Usage examples:

  1. FASTFILE_GETLINE=1 pip3 install . -v
  2. FASTFILE_GETLINE=1 FASTFILE_DEBUG=1 pip3 install . -v

File reading optimizations

You can enable a file reading optimization with the environment variable FASTFILE_REGEX=1. Do not define this variable or define it as FASTFILE_REGEX=0 to disable the optimization.

  1. FASTFILE_REGEX=1 will use the C language builtin regex library regex.h: https://linux.die.net/man/3/regexec
  2. FASTFILE_REGEX=2 will use PCRE2 regex library pcre2.h: http://pcre.org/current/doc/html/pcre2_match.html, it requires the installation of the libpcre2-dev package with sudo apt-get install libpcre2-dev
  3. FASTFILE_REGEX=3 will use RE2 regex library re2/re2.h: https://github.com/google/re2/wiki/CplusplusAPI, it requires the installation of the re2 library with:
    git clone https://github.com/google/re2 re2 &&
    cd re2 &&
    make &&
    sudo make install &&
    make testinstall
    
  4. FASTFILE_REGEX=4 will use Hyperscan regex library hs.h: https://github.com/intel/hyperscan, it requires the installation of the libhyperscan-dev package with sudo apt-get install libhyperscan-dev

Notes:

  • Defining the variable FASTFILE_REGEX only has effect when FASTFILE_GETLINE=2 is set (as/with value 2). If the variable FASTFILE_GETLINE=2 is not defined (as/with value 2), any definition of FASTFILE_REGEX is ignored.

Debugging

You you use the FASTFILE_DEBUG=1 variable specified on the Enable debug mode section, you do not need to build the program with CFLAGS="-O0..." CXXFLAGS="-O0 ..." /usr/bin/pip3 install . because these flags are already set by FASTFILE_DEBUG=1.

If Python got segmentation fault, you need to install the debug symbol packages, and compile the program into mode debug. These instructions works for both Linux and Cygwin.

sudo apt-get install python3-dbg
apt-cyg install libcrypt-devel python36-devel python36-debuginfo python3-debuginfo python3-cython-debuginfo
CFLAGS="-O0 -g -ggdb -fstack-protector-all" CXXFLAGS="-O0 -g -ggdb -fstack-protector-all" /usr/bin/pip3 install .

cd modulename/source
gdb --args /usr/bin/python3 -m modulename.__init__ -p ../tests/light/
python3 -m pdb -m modulename.__init__ -p ../tests/light/

cd /usr/bin
/usr/bin/python3 -u -m trace -t modulename-script.py -p ../tests/light/
  1. https://github.com/spiside/pdb-tutorial
  2. https://docs.python.org/3.7/library/pdb.html
  3. https://docs.python.org/3.7/library/trace.html
  4. https://stackoverflow.com/questions/1629685/when-and-how-to-use-gccs-stack-protection-feature
  5. https://stackoverflow.com/questions/25678978/how-to-debug-python-script-that-is-crashing-python
  6. https://stackoverflow.com/questions/46265835/how-to-debug-a-python-module-run-with-python-m-from-the-command-line

In case you program enter on a deadlock (or livelock), you can use gdb to discover or what is happening. For this, first you need get a core dump file of the system somehow. Once you get the core dump file, you can call gdb program_name coredump. Now, you can use the gdb commands to navigate between the existing threads and to discover which of them are waiting for the other, i.e.,e it is causing the deadlock or livelock.

  1. p / s 0x6ffffdf8ed0 print or content on the address with the format "%s"
  2. x / 100w 0x6ffffdf8ed0 "examines" the specified memory address
  3. frame 0 shows the corresponding stack frame 0 as bt f (backtrace full)
  4. bt f 5 shows the last 5 stack frames with all debugging symbols data
  5. f 0 and p varname print the varname on frame 0 context
  6. https://stackoverflow.com/questions/14659147/how-to-print-pointer-content-in-gdb
  7. https://stackoverflow.com/questions/14502236/how-to-view-a-pointer-like-an-array-in-gdb

Note: Instead of creating a core dump file, you can run your program directly with gdb --args program_name and once you are on the gdb command line, you can use the run command and once your program enters/starts a dead or live lock, you can press Ctrl+C to stop the program execution. Then, gdb will already had "captured" the core dum file.

Enable debug mode

The default debug level is 0 where no debugging message code is generated into the final binary. Guaranteeing maximum performance. If you would like to compile a binary with debug messages, define the environment variable FASTFILE_DEBUG before running the installer. You can see the debugging level available on the file source/debugger.h:

FASTFILE_DEBUG=1 pip install .
FASTFILE_DEBUG=1 pip install -e .

FASTFILE_DEBUG=1 python setup.py install
FASTFILE_DEBUG=1 python setup.py develop

If you are on Windows, run it like this:

set "FASTFILE_DEBUG=1" && pip install .
set "FASTFILE_DEBUG=1" && pip install -e .
set "FASTFILE_DEBUG=1" && python setup.py install
set "FASTFILE_DEBUG=1" && python setup.py develop

To debug refcounts leaks:

  1. CPython compiled with ./configure --with-pydebug
  2. Runtime compiled with C assertion: crash (kill itself with SIGABRT signal) if a C assertion fails (assert(...);).
  3. Use the debug hooks on memory allocators by default, as PYTHONDEBUG=debug environment variable: detect memory under and overflow and misuse of memory allocators.
  4. Compiled without compiler optimizations (-Og or even -O0) to be usable with a debugger like gdb: python-gdb.py should work perfectly. However, the regular runtime is unusable with gdb since most variables and function arguments are stored in registers, and so gdb fails with the <optimized out> message.
  5. https://pythoncapi.readthedocs.io/runtimes.html#debug-build
  6. https://stackoverflow.com/questions/40121749/why-gcc-og-doesnt-generate-source-code-line-mapping
  7. https://stackoverflow.com/questions/8650972/how-do-i-debug-refcounts-in-a-python-c-extension-the-easiest-way

To generate core dumps instead of stack traces on Cygwin

export CYGWIN="$CYGWIN error_start=dumper -d %1 %2"
  1. https://stackoverflow.com/questions/320001/using-a-stackdump-from-cygwin-executable

If you see this when running gdb:

[New Thread 8980.0x626c]
[New Thread 8980.0x1ba0]
[New Thread 8980.0x2454]
[New Thread 8980.0x3fbc]
warning: the debug information found in "/usr/lib/debug//usr/bin/cygwin1.dbg" does not match "/usr/bin/cygwin1.dll" (CRC mismatch).

[New Thread 8980.0x21c4]

Run the command rm /usr/lib/debug//usr/bin/cygwin1.dbg

  1. http://cygwin.1069669.n5.nabble.com/Debugging-with-GDB-td94421.html

License

See the file LICENSE.txt

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Files for fastfilepackage, version 1.0.24
Filename, size File type Python version Upload date Hashes
Filename, size fastfilepackage-1.0.24.tar.gz (35.3 kB) File type Source Python version None Upload date Hashes View

Supported by

AWS AWS Cloud computing Datadog Datadog Monitoring DigiCert DigiCert EV certificate Facebook / Instagram Facebook / Instagram PSF Sponsor Fastly Fastly CDN Google Google Object Storage and Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Salesforce Salesforce PSF Sponsor Sentry Sentry Error logging StatusPage StatusPage Status page