openmedialibrary/oml/fulltext.py

import logging
import os
import subprocess
import sys
from urllib.parse import quote, unquote

from sqlalchemy.sql import operators

logger = logging.getLogger(__name__)


def get_prefix():
    import settings
    return os.path.join(os.path.expanduser(settings.preferences['libraryPath']), 'Books/')

def get_ids(books):
    from item.models import File
    if not books:
        return []
    ids = [b[0] for b in File.query.filter(operators.in_op(File.path, books)).values('sha1')]
    return ids

def find_fulltext_macos(query):
    prefix = get_prefix()
    cmd = ["mdfind", "-onlyin", prefix, query]
    books = subprocess.check_output(cmd).decode().strip().split('\n')
    books = [path[len(prefix):] for path in books]
    return get_ids(books)

def find_fulltext_windows(query):
    prefix = get_prefix()
    prefix = os.path.normpath(prefix)
    cmd = ['findstr', '/i', '/c:' + query, '/d:' + prefix]
    books = subprocess.check_output(cmd).decode().strip().split('\n')
    books = [b.split(':')[0] for b in books]
    return get_ids(books, prefix)

def find_fulltext_linux(query):
    prefix = get_prefix()
    prefix_url = quote(prefix)
    cmd = [
        'tracker',
        'sparql',
        '-q',
        "SELECT nie:url(?f) WHERE { ?f fts:match '%s' FILTER (tracker:uri-is-descendant ('file://%s', nie:url (?f))) }" % (query, prefix_url)
    ]
    books = subprocess.check_output(cmd).decode().strip().split('\n')
    books = [
        unquote(r.strip()).replace('file://', '')[len(prefix):]
        for r in books if r.strip().startswith('file://')
    ]
    return get_ids(books)

def find_fulltext(query):
    ids = []
    if sys.platform == 'darwin':
        ids = find_fulltext_macos(query)
    elif sys.platform == 'linux':
        ids = find_fulltext_linux(query)
    else:
        logger.debug('missing fulltext search implementation for %s', sys.platform)
    return ids

def platform_supported():
    return sys.platform in ('darwin', 'linux')