AnyStyle Command Line Interface

anystyle --help

NAME
    anystyle - Finds and parses bibliographic references

SYNOPSIS
    anystyle [global options] command [command options] [arguments...]

VERSION
    1.0.0 (cli 1.0.0, data 1.2.0)

GLOBAL OPTIONS
    --adapter=name       - Set the dictionary adapter (default: ruby)
    -f, --format=name    - Set the output format (default: ["json"])
    --help               - Show this message
    --[no-]stdout        - Print results directly to stdout
    --[no-]verbose       - Print status messages to stderr
    --version            - Display the program version
    -w, --[no-]overwrite - Allow overwriting existing files

COMMANDS
    find    - Find and extract references from text documents
    help    - Shows a list of commands or help for one command
    license - Print license information
    parse   - Parse and convert references

anystyle help find

NAME
    find - Find and extract references from text documents

SYNOPSIS
    anystyle [global options] find [command options] input [output]

DESCRIPTION
    This manual page documents the AnyStyle `find' command. AnyStyle `find'
    analyzes PDF or text documents and extracts all references it finds.

    The input argument can be a single PDF or text document, or a folder
    containing multiple documents. The (optional) output argument specifies
    the folder where the results shall be saved; if no output folder is
    specified, results will be saved in the folder containing the input.

    AnyStyle `find' supports the following formats:
        bib     BibTeX (references only);
        csl     CSL/JSON (references only);
        json    AnyStyle JSON (references only);
        ref     One reference per line, suitable for parser input;
        txt     Plain text document;
        ttx     Tagged document format, used for training the finder model;
        xml     References only, XML, suitable for training the parser model.

    You can specify multiple output formats, separated by commas.

    Anlyzing PDF documents currently depends on `pdftotext' which must be
    installed separately.

EXAMPLES
    anystyle -f csl,xml find thesis.pdf

    Extract references from `thesis.pdf' and save them in `thesis.csl' and
    `thesis.xml'.

    anystyle -f bib find --no-layout thesis.pdf bib

    Extract references from `thesis.pdf' in `no-layout' mode (e.g., use this
    if your document uses a multi-column layout) and save them in BibTeX in
    `./bib/thesis.bib'.

anystyle help parse

COMMAND OPTIONS
    --[no-]layout - Use layout mode for PDF text extraction (default: enabled)
NAME
    parse - Parse and convert references

SYNOPSIS
    anystyle [global options] parse input [output]

DESCRIPTION
    This manual page documents the AnyStyle `parse' command. AnyStyle `parse'
    segments references (one per line) and converts them into structured
    formats.

    The input argument can be a single text document containing one full
    reference per line (blank lines will be ignored), or a folder containing
    multiple documents. The (optional) output argument specifies
    the folder where the results shall be saved; if no output folder is
    specified, results will be saved in the folder containing the input.

    AnyStyle `parse' supports the following formats:
        bib     BibTeX (normalized);
        csl     CSL/JSON (normalized);
        json    AnyStyle JSON (normalized);
        ref     One reference per line, suitable for parser input;
        txt     Same as `ref';
        xml     XML, suitable for training the parser model.

    You can specify multiple output formats, separated by commas.

EXAMPLES
    anystyle -f json,xml parse biblio.txt

    Extract references from `biblio.txt' and save them in `biblio.json' and
    `biblio.xml'.

    anystyle --stdout -f csl parse input.txt

    Extract references from `input.txt' and print them to STDOUT in CSL/JSON.

License

Copyright 2011-2018 Sylvester Keil. All rights reserved.

AnyStyle is distributed under a BSD-style license. See LICENSE for details.