(r67) Pynini < GRM

GRM Web>Pynini (revision 67) (raw view)~~EditAttach~~
---+ OpenGrm Pynini: Finite-state grammar development in Python
---

%ICON{led-red}% *Version 2.1.4 is now available for [[PyniniDownload][download]].*

%ICON{radioactive}% Pynini is now available on [[https://conda-forge.org/][conda-forge]]. Linux (x86) and MacOS users who already have [[https://docs.conda.io/en/latest/][Conda]] can install Pynini and all dependencies using the following command: =conda install -c conda-forge pynini=

%ICON{megaphone}% Linux (x86) users can now install precompiled [[https://github.com/pypa/manylinux][manylinux]] wheels for Pynini directly from PyPI. Unlike the conda-forge installation method mentioned above, this does not install [[http://www.openfst.org][OpenFst]] command-line tools, does not install the [[https://graphviz.org/][Graphviz]] command-line tools [[https://jupyter.org/][Jupyter]] uses to generate state diagrams, and does not work on MacOS. To install a Pynini wheel use the following command: =pip install pynini=

%ICON{strategy}% When using Pynini from a [[https://jupyter.org/][Jupyter]] or [[https://colab.research.google.com/][Colab]] notebook, logging and error messages sent to standard error are lost. To recover them, install [[https://pypi.org/project/wurlitzer/][ =wurlitzer= ]] and then add the following at the top of your notebook: =%load_ext wurlitzer=
---

_OpenGrm_ _Pynini_, like [[Thrax]], compiles grammars expressed as strings, regular expressions, and context-dependent rewrite rules into weighted finite-state transducers. It uses the [[http://www.openfst.org][OpenFst library]] and its [[http://www.python.openfst.org][Python extension]] to create, access and manipulate compiled grammars. Pynini is embedded in a [[https://www.python.org][Python]] module, allowing users to write Thrax-like grammars using Python's flexible syntax (including imperative programming constructs not available in Thrax) and powerful toolchain, including an [[http://ipython.org][interactive development]] ("REPL") environment.

It is (loosely) named after [[https://en.wikipedia.org/wiki/PÄ\x81á¹‡ini][P&#x101;&#x1E47;ini]] (&#x92A;&#x93E;&#x923;&#x93F;&#x928;&#x93F;), the renowned Sanskrit grammarian.

   * [[PyniniDownload][Download]]
   * [[PyniniDocs][Documentation]]
   * C++ API documentation (for advanced users):
      * [[PyniniOptimizeDoc][Optimization]]
      * [[PyniniPathsDoc][Path iteration]]
      * [[PyniniOperatorsDoc][Specialty operators]]
      * [[PyniniStringDoc][String (de)compilation]]

For more information about Pynini, see:

   * K. Gorman & R. Sproat. 2021. [[http://www.morganclaypoolpublishers.com/catalog_Orig/product_info.php?products_id=1636][ _Finite-State Text Processing_ ]]. Morgan & Claypool.
   * K. Gorman & R. Sproat. 2016. [[https://www.oreilly.com/ideas/how-to-get-superior-text-processing-in-python-with-pynini][How to get superior text processing in Python with Pynini]]. O'Reilly Ideas blog, accessed 11/16/16.
   * K. Gorman. 2016. [[https://www.aclweb.org/anthology/W16-2409][Pynini: A Python library for weighted finite-state grammar compilation]]. In _Proceedings of the ACL Workshop on Statistical NLP and Weighted Automata_, pages 75-80.

Other third-party information:

   * [[https://stackoverflow.com/questions/57898360/how-can-i-install-and-use-pynini-on-windows-10][Installing Pynini on Windows 10]]
Topic revision: r67 - 2021-08-06 - KyleGorman