Averell
Averell, the python library and command line interface that facilitates working with existing repositories of annotated poetry. Averell is able to download an annotated corpus and reconcile different TEI entities to provide a unified JSON output at the desired granularity. That is, for their investigations some researchers might need the entire poem, poems split line by line, or even word by word if that is available. Averell allows to specify the granularity of the final generated dataset, which is a combined JSON with all the entities in the selected corpora. Each corpus in the catalog must specify the parser to produce the expected data format.
- Free software: Apache Software License 2.0
Available corpora (version 1.1.0)
id | name | lang | size | docs | words | granularity | license |
---|---|---|---|---|---|---|---|
1 | Disco V2.1 (disco2_1) | es | 22M | 4088 | 381539 | stanza line | CC-BY |
2 | Disco V3 (disco3) | es | 28M | 4080 | 377978 | stanza line | CC-BY |
3 | Sonetos Siglo de Oro (adso) | es | 6.8M | 5078 | 466012 | stanza line | CC-BY-NC 4.0 |
4 | ADSO 100 poems corpus (adso100) | es | 128K | 100 | 9208 | stanza line | CC-BY-NC 4.0 |
5 | PoesÃa LÃrica Castellana Siglo de Oro (plc) | es | 3.8M | 475 | 299402 | stanza line word syllable | CC-BY-NC 4.0 |
6 | Gongocorpus (gongo) | es | 9.2M | 481 | 99079 | stanza line word syllable | CC-BY-NC-ND 3.0 FR |
7 | Eighteenth Century Poetry Archive (ecpa) | en | 2400M | 3084 | 2063668 | stanza line word | CC BY-SA 4.0 |
8 | For Better For Verse (4b4v) | en | 39.5M | 103 | 41749 | stanza line | Unknown |
9 | Métrique en Ligne (mel) | fr | 183M | 5081 | 1850222 | stanza line | Unknown |
10 | Biblioteca Italiana (bibit) | it | 242M | 25341 | 7121246 | stanza line word | Unknown |
11 | Corpus of Czech Verse (czverse) | cs | 4100M | 66428 | 12636867 | stanza line word | CC-BY-SA |
12 | Stichotheque (stichopt) | pt | 11.8M | 1702 | 168411 | stanza line | Unkwown |
Documentation
https://averell.readthedocs.io/
Installation
To install averell, run this command in your terminal:
pip install averell
This is the preferred method to install averell, as it will always install the most recent stable release.
If you don't have pip installed, this Python installation guide can guide you through the process.
Usage
To show averell help:
averell --help
To list all available corpora:
averell list
Visualization example of one of the available corpora:
id name lang size docs words granularity license
---- ------------------- ------ ------ ------ ------- ------------- -----------
1 Disco V2.1 es 22M 4088 381539 stanza CC-BY
line
download
Download desired corpora into "mycorpora" folder:
averell download 2 3 --corpora-folder my_corpora
Example of poem in TEI format obtained from one of the corpora:
<TEI xmlns="http://www.tei-c.org/ns/1.0">
<teiHeader>
<fileDesc>
<titleStmt>
<title> Spanish Metrical Patterns Bank: Golden Age Sonnets.</title>
<principal>Borja Navarro Colorado</principal>
<respStmt>
<name>MarÃa Ribes Lafoz</name>
<name>Noelia Sánchez López</name>
<name>Borja Navarro Colorado</name>
<resp>Metrical patterns annotation</resp>
</respStmt>
</titleStmt>
<publicationStmt>
<publisher>Natural Language Processing Group. Department of Software and Computing Systems. University of Alicante (Spain)</publisher>
</publicationStmt>
<sourceDesc>
<bibl><title>Sonetos</title> de <author>Garcilaso de La Vega</author>. <publisher>Biblioteca Virtual Miguel de Cervantes</publisher>, edición de <editor role="editor">Ramón GarcÃa González</editor>.</bibl>
</sourceDesc>
</fileDesc>
<encodingDesc>
<metDecl xml:id="bncolorado" type="met" pattern="((\+|\-)+)*">
<metSym value="+">stressed syllable</metSym>
<metSym value="-">unstressed syllable</metSym>
</metDecl>
<metDecl>
<p>All metrical patterns have been manually checked.</p>
</metDecl>
</encodingDesc>
</teiHeader>
<text>
<body>
<head>
<title>-XX-</title>
</head>
<lg type="cuarteto">
<l n="1" met="-++--++--+-">Con tal fuerza y vigor son concertados</l>
<l n="2" met="-----+-+-+-">para mi perdición los duros vientos,</l>
<l n="3" met="--+--+---+-">que cortaron mis tiernos pensamientos</l>
<l n="4" met="+----++--+-">luego que sobre mà fueron mostrados.</l>
</lg>
<lg type="terceto">
<l n="5" met="-++--+---+-">El mal es que me quedan los cuidados</l>
<l n="6" met="---+-----+-">en salvo de estos acontecimientos,</l>
<l n="7" met="-++--+---+-">que son duros, y tienen fundamentos</l>
</lg>
</body>
</text>
</TEI>
Generated example JSON file from input XML/TEI poem into my_corpora/{corpus}/averell/parser/{author_name}/{poem_name}.json
{
"manually_checked": true,
"poem_title": "-XX-",
"author": "Garcilaso de La Vega",
"stanzas": [
{
"stanza_number": "1",
"stanza_type": "cuarteto",
"lines": [
{
"line_number": "1",
"line_text": "Con tal fuerza y vigor son concertados",
"metrical_pattern": "-++--++--+-"
},
{
"line_number": "2",
"line_text": "para mi perdición los duros vientos,",
"metrical_pattern": "-----+-+-+-"
},
{
"line_number": "3",
"line_text": "que cortaron mis tiernos pensamientos",
"metrical_pattern": "--+--+---+-"
},
{
"line_number": "4",
"line_text": "luego que sobre mà fueron mostrados.",
"metrical_pattern": "+----++--+-"
}
],
"stanza_text": "Con tal fuerza y vigor son concertados\npara mi perdición los duros vientos,\nque cortaron mis tiernos pensamientos\nluego que sobre mà fueron mostrados."
},
{
"stanza_number": "2",
"stanza_type": "terceto",
"lines": [
{
"line_number": "5",
"line_text": "El mal es que me quedan los cuidados",
"metrical_pattern": "-++--+---+-"
},
{
"line_number": "6",
"line_text": "en salvo de estos acontecimientos,",
"metrical_pattern": "---+-----+-"
},
{
"line_number": "7",
"line_text": "que son duros, y tienen fundamentos",
"metrical_pattern": "-++--+---+-"
}
],
"stanza_text": "El mal es que me quedan los cuidados\nen salvo de estos acontecimientos,\nque son duros, y tienen fundamentos"
}
]
}
export
Now we can combine and join these corpora through "granularity" selection:
averell export 2 3 --granularity line --corpora-folder my_corpora --filename export_1
It produces an single JSON file with information about all the lines in those corpora. Example of two random lines in the file mycorpora/export_1.json:
{
"line_number": "5",
"line_text": "¿Has visto que en el mismo lugar donde",
"metrical_pattern": "++---+--++-",
"stanza_number": "2",
"manually_checked": false,
"poem_title": " - II - ",
"author": "Mira de Amescua",
"stanza_text": "¿Has visto que en el mismo lugar donde\nbordado estuvo el cristalino velo\nun bordado terliz de escarcha y hielo\nhace que el campo de verdor se monde?",
"stanza_type": "cuarteto"
}
{
"line_number": "10",
"line_text": "el que a lo cierto no a lo incierto mira,",
"metrical_pattern": "---+-+-+-+-",
"stanza_number": "3",
"manually_checked": false,
"poem_title": "- VIII - Considerando un sepulcro y los que están en él ",
"author": "Lope de Zarate",
"stanza_text": "De aquà si que consigue el ser dichoso\nel que a lo cierto no a lo incierto mira,\npues le adorna lo eterno fastuoso;",
"stanza_type": "terceto"
}
By default, export
will download corpora if needed. To avoid this behaviour, the flag --no-download
can be passed in.
Exported corpora can be easily loaded into Pandas
averell export adso100 --filename adso100.json
import pandas as pd
adso100 = pd.read_json(open("adso100.json"))
A note on IDS
IDS can be numeric identifiers in the averell list
output, corpus shortcodes (shown between parenthesis), the speciall literal all
to refer to all corpora, or two-letter ISO language codes to refer to avaliable corpora in a specific language.
For example, the command averell export 1 bibit fr
will export DISCO V2.1, the Biblioteca Italiana poetry corpus, and all corpora tagged with the French languge tag in a single file spliting poems line by line.