lemmieshell

A command line tool for querying the Language Bank corpora

Description

Lemmieshell is a text-based corpus tool manufactured at CSC. The program is used interactively on the corpus.csc.fi server.

Lemmieshell allows interactive queries to the Language Bank databases. The software uses the same query language as WWW-Lemmie's web interface, and provides interactive help.

Usage

[hippu.csc.fi]$ lemmieshell

Help, Manuals and Documentation

Lemmie shell is a simple shell-like user interface built using the
experimental Lemmie API. You can use Lemmie shell to

1) retrieve concordance lines matching a query language expression
   given at the prompt

2) retrieve a list of collocates for the word form given at the prompt

The syntax of the query language is a subset of CQP query
language. Here are some examples:

> [wf='kissa']

The above query retrieves a set of kwic concordance lines where the
word form (wf) 'kissa' occurs. If the set larger than 40 the result is
sent to the Unix pager program less. You can set the arrow keys,
return key and space bar to scroll the buffer. Press 'q' to return to
Lemmie shell.

In addition to surface word forms the user can query other attributes
of word form occurences. These attributes are base form (bf), part of
speech (pos) and morpho-syntactic description (msd). The syntax for
querying these attributes is the same as above.

> [bf='kissa']

The above query is exactly like the first example except that this
time it retrieves concordance lines where all the word forms with the
base form (bf) 'kissa' occurs. Several attributes can be combined in a
single query

> [wf='voi' pos='Verb']

It is also possible to query a sequence of two or more tokens in a
single query.

> [pos='Verb'][wf='pois']

The same queries can be used to retrieve a set of sentences or
paragaphs instead of kwic concordances. This is done by adding +s (for
sentences) or +p (for paragraphs) at the end of the query.

> [wf='tieto'] +s

A list of collocates for a given word form.