Skip to content

README.md

Latest commit

 

History

History
137 lines (94 loc) · 5.9 KB

README.md

File metadata and controls

137 lines (94 loc) · 5.9 KB
Raw

ChatNoir Resiliparse

Build Packages Codecov Documentation Status

A collection of robust and fast processing tools for parsing and analyzing web archive data written in Rust and Cython/C++ with bindings for Python.

Usage Instructions

For detailed information about the build process, dependencies, APIs, or usage instructions, please read the Resiliparse Documentation

Resiliparse Module Summary

The Resiliparse collection encompasses the following two modules at the moment:

1. Resiliparse

The Resiliparse main module with the following subcomponents:

Parsing Utilities

The Resiliparse Parsing Utilities are highly optimized tools for dealing with encodings, detecting content types of raw protocol payloads, parsing HTML web pages, performing language detection, and more.

Main documentation: Resiliparse Parsing Utilities

Extraction Utilities

The Resiliparse Extraction Utilities are a set of performance-optimized and highly efficient tools for extracting structural or semantic information from noisy raw web data for further processing, such as (main) content extraction / boilerplate removal, schema extraction, general web data cleansing, and more.

Main documentation: Resiliparse Extraction Utilities

Process Guards

The Resiliparse Process Guard module is a set of decorators and context managers for guarding a processing context to stay within pre-defined limits for execution time and memory usage. Process Guards help to ensure the (partially) successful completion of batch processing jobs in which individual tasks may time out or use abnormal amounts of memory, but in which the success of the whole job is not threatened by (a few) individual failures. A guarded processing context will be interrupted upon exceeding its resource limits so that the task can be skipped or rescheduled.

Main documentation: Resiliparse Process Guards

Itertools

Resiliparse Itertools are a collection of convenient and robust helper functions for iterating over data from unreliable sources using other tools from the Resiliparse toolkit.

Main documentation: Resiliparse Itertools

2. FastWARC

FastWARC is a high-performance WARC parsing library for Python written in Rust with bindings for Python. compressed and uncompressed WARC/1.0 and WARC/1.1 streams. Supported compression algorithms are GZip, Zstd, and LZ4.

Main documentation: FastWARC and FastWARC CLI

Installation (Rust)

To use FastWARC as a library in your Rust project, add the following to your Cargo.toml:

[dependencies]
fastwarc = "*" # Specify the correct version here

Then build your project like normal with cargo build.

Resiliparse is not yet available as a stable crate.

Installation (Python)

Pre-built wheels are available for Linux, macOS, and Windows on PyPi, which can be installed via pip:

Resiliparse:

pip install resiliparse

FastWARC:

pip install fastwarc

Building Bindings From Source

You can also build the Resiliparse and FastWARC Python bindings yourself from source. FastWARC has been fully ported to Rust. Resiliparse is still written in Cython.

For detailed build instructions, please refer to the build documentation or the individual READMEs:

Cite Us

Resiliparse is part of the ChatNoir web analytics toolkit. If you use ChatNoir or any of its tools for a publication, you can make us happy by citing our ECIR 2018 demo paper:

@InProceedings{bevendorff:2018,
  address =             {Berlin Heidelberg New York},
  author =              {Janek Bevendorff and Benno Stein and Matthias Hagen and Martin Potthast},
  booktitle =           {Advances in Information Retrieval. 40th European Conference on IR Research (ECIR 2018)},
  editor =              {Leif Azzopardi and Allan Hanbury and Gabriella Pasi and Benjamin Piwowarski},
  month =               mar,
  publisher =           {Springer},
  series =              {Lecture Notes in Computer Science},
  site =                {Grenoble, France},
  title =               {{Elastic ChatNoir: Search Engine for the ClueWeb and the Common Crawl}},
  year =                2018
}

If you use FastWARC, you can also cite our OSSYM 2021 abstract paper:

@InProceedings{bevendorff:2021,
  author =                {Janek Bevendorff and Martin Potthast and Benno Stein},
  booktitle =             {3rd International Symposium on Open Search Technology (OSSYM 2021)},
  editor =                {Andreas Wagner and Christian Guetl and Michael Granitzer and Stefan Voigt},
  month =                 oct,
  publisher =             {International Open Search Symposium},
  site =                  {CERN, Geneva, Switzerland},
  title =                 {{FastWARC: Optimizing Large-Scale Web Archive Analytics}},
  year =                  2021
}