gnparser

gnparser

gRPC client to gnparser project (Go version)

Overview

gnparser is a Ruby gem that allows to parse scientific names extremely fast using remote gRPC method calls to gnparser written in Go.

Install

gem install gnparser

Usage

Download the latest version of Go gnparser, and start it on your local machine as a gRPC server:

gnparser -g 8778

Now you will be able to use Ruby gnparser with its default settings.

require 'gnparser'
gnp = GNparser::Client.new

# To use different host and port:
gnp = GNparser::Client.new(host = 'parser.example.com', port = 1234 )
gnp = GNparser::Client.new(host = 'parser.example.com', port = '1234' )
# Use default 8778 port
gnp = GNparser::Client.new(host = 'parser.example.com')
# Use default 0.0.0.0 host
gnp = GNparser::Client.new(port = '1234' )

Versions

To see version of the gem:

gnp.version

To see version of the Go gnparser:

ver = gnp.parser_version
puts ver.value
puts ver.build_time.size

Parsing name-strings

Refer to gnparser.proto file to learn available fields.

All parsing methods take the following options:

skip_cleaning: : if true names-strings get stripped from HTML tags, if there are any. For example <i>Homo sapiens</i> <b>L.</b> becomes Homo sapiens L.. It is false by default.

jobs_number: allows to control how many jobs will run in parallel by gnparser. Note that gnparser gRPC server also has a max jobs number option. So if jobs_number is higher than gRPC MaxJobsNumber, it is ignored.

Parse one name

res = gnp.parse('Puma concolor L.')
puts res.canonical.simple
puts res.authorship.value
puts res.name_type
puts res.species

Skipping cleaning from HTML tags:

res = gnp.parse('<i>Puma concolor</i> (Linn.)', skip_cleaning: true)
#parsed will be false
res.parsed
...
res = gnp.parse('<i>Puma concolor</i> (Linn.)')
#parsed will be true
res.parsed
res.normalized

Parse an array of names

There is a limit of 10,000 name strings per batch. Results are always returned in the same order as input.

names = ['Plantago major L.', 'Homo sapiens Linn. 1758', 'Bubo bubo']

# run using 3 jobs
res = gnp.parse_ary(names, jobs_number: 3)
res.each do |r|
  puts r.canonical.full
end


# do not strip HTML tags (in this case there are none in the input)
res = gnp.parse_ary(names, skip_cleaning: true)