{"schema_version":1,"metadata":{"branch":"release","release":"7.9","architecture":"amd64","generated_at":"2026-09-09T04:04:02.361425+00:00","source_url":"https://cdn.openbsd.org/pub/OpenBSD/7.9/packages/amd64/sqlports-7.54.tgz","source_sha256":"2de00144847bf9da3a365be982f101dbb55d591ed7a915ea94984a16ab0e0404","package_count":12059,"source_kind":"sqlports"},"package":{"name":"openfst-1.8.4","path":"math/openfst","url":"/packages/release/math/openfst/","comment":"weighted finite-state transducers library","homepage":"https://www.openfst.org/twiki/bin/view/FST/WebHome","maintainer":"The OpenBSD ports mailing-list <ports@openbsd.org>","description":"OpenFst is a library for constructing, combining, optimizing, and\nsearching weighted finite-state transducers (FSTs). Weighted\nfinite-state transducers are automata where each transition has an input\nlabel, an output label, and a weight. The more familiar finite-state\nacceptor is represented as a transducer with each transition's input and\noutput label equal. Finite-state acceptors are used to represent sets of\nstrings (specifically, regular or rational sets); finite-state\ntransducers are used to represent binary relations between pairs of\nstrings (specifically, rational transductions). The weights can be used\nto represent the cost of taking a particular transition.\n\nFSTs have key applications in speech recognition and synthesis, machine\ntranslation, optical character recognition, pattern matching, string\nprocessing, machine learning, information extraction and retrieval among\nothers. Often a weighted transducer is used to represent a probabilistic\nmodel (e.g., an n-gram model, pronunciation model). FSTs can be\noptimized by determinization and minimization, models can be applied to\nhypothesis sets (also represented as automata) or cascaded by\nfinite-state composition, and the best results can be selected by\nshortest-path algorithms.\n\nThis library was developed by contributors from Google Research and\nNYU's Courant Institute. It is intended to be comprehensive, flexible,\nefficient and scale well to large problems. It has been extensively\ntested. It is an open source project distributed under the Apache\nlicense.\n","package_architecture":"amd64","stem":"openfst","readme":null,"dependencies":[],"reverse_dependencies":{"count":0,"url":null},"categories":["math","devel","textproc"],"flavors":[],"only_for_architectures":["aarch64","amd64","arm","i386","mips64","mips64el","powerpc","powerpc64","riscv64","alpha","hppa","sparc64"],"not_for_architectures":[]}}