Python model.Dataset类代码示例

def import_upload(dataset_name, id, account_id,
                  entity_col, alias_col):
    dataset = Dataset.find(dataset_name)
    account = Account.by_id(account_id)
    metadata, row_set = parse_upload(dataset, id)
    headers = detect_headers(row_set)
    for row in row_set:
        data = dict([(c.column, c.value) for c in row])
        entity = data.pop(entity_col) if entity_col else None
        alias = data.pop(alias_col) if alias_col else None
        if alias_col and alias is not None and len(alias) and alias != entity:
            d = {'name': alias, 'data': data}
            alias_obj = Alias.lookup(dataset, d, account,
                                     match_entity=False)
            data = {}
        if entity_col and entity is not None and len(entity):
            d = {'name': entity, 'data': data}
            entity_obj = Entity.by_name(dataset, entity)
            if entity_obj is None:
                entity_obj = Entity.create(dataset, d, account)
            entity_obj.data = data
        if alias_col and entity_col:
            alias_obj.match(dataset, {'choice': entity_obj.id}, account)
    db.session.commit()
    flush_cache(dataset)

def view(dataset, link):
    dataset = Dataset.find(dataset)
    link = Link.find(dataset, link)
    format = response_format()
    if format == 'json':
        return jsonify(link)
    return "Not implemented!"

def match(dataset, link, random=False):
    dataset = Dataset.find(dataset)
    authz.require(authz.dataset_edit(dataset))
    link = Link.find(dataset, link)
    random = random or request.args.get('random')=='True'
    choices = match_op(link.key, dataset,
            query=request.args.get('query'))
    pager = Pager(choices, '.match',
        dataset=dataset.name, link=link.id,
        limit=10)

    # HACK: Fetch only the values on the selected page.
    value_objs = Value.id_map(dataset, map(lambda (c,v,s): v,
        pager.query[pager.offset:pager.offset+pager.limit]))
    for i, (c,v,s) in enumerate(pager.query):
        if v in value_objs:
            pager.query[i] = (c, value_objs.get(v), s)

    html = render_template('link/match.html',
            dataset=dataset, link=link, choices=pager,
            random=random)
    choice = 'INVALID' if link.is_invalid else link.value_id
    if len(choices) and choice is None:
        c, v, s = choices[0]
        choice = 'INVALID' if s <= 50 else v
    return htmlfill.render(html, force_defaults=False,
            defaults={'choice': choice,
                      'value': link.key,
                      'query': request.args.get('query', ''),
                      'random': random})

def edit(dataset):
    dataset = Dataset.find(dataset)
    authz.require(authz.dataset_manage(dataset))
    html = render_template('dataset/edit.html',
                           dataset=dataset,
                           algorithms=get_algorithms())
    return htmlfill.render(html, defaults=dataset.as_dict())

def match(dataset, alias, random=False):
    dataset = Dataset.find(dataset)
    authz.require(authz.dataset_edit(dataset))
    alias = Alias.find(dataset, alias)
    random = random or request.args.get('random')=='True'
    choices = match_op(alias.name, dataset,
            query=request.args.get('query'))
    pager = Pager(choices, '.match',
        dataset=dataset.name, alias=alias.id,
        limit=10)

    # HACK: Fetch only the entities on the selected page.
    entities = Entity.id_map(dataset, map(lambda (c,e,s): e,
        pager.query[pager.offset:pager.offset+pager.limit]))
    for i, (c,e,s) in enumerate(pager.query):
        if e in entities:
            pager.query[i] = (c, entities.get(e), s)

    html = render_template('alias/match.html',
            dataset=dataset, alias=alias, choices=pager,
            random=random)
    choice = 'INVALID' if alias.is_invalid else alias.entity_id
    if len(choices) and choice is None:
        c, e, s = choices[0]
        choice = 'INVALID' if s <= 50 else e.id
    return htmlfill.render(html, force_defaults=False,
            defaults={'choice': choice,
                      'name': alias.name,
                      'query': request.args.get('query', ''),
                      'random': random})

def reconcile_index(dataset):
    domain = url_for('index', _external=True).strip('/')
    urlp = domain + '{{id}}'
    meta = {
        'name': 'nomenklatura',
        'identifierSpace': 'http://rdf.freebase.com/ns/type.object.id',
        'schemaSpace': 'http://rdf.freebase.com/ns/type.object.id',
        'view': {'url': urlp},
        'preview': {
            'url': urlp + '?preview=true', 
            'width': 600,
            'height': 300
            }
        }
    if dataset is not None:
        meta['name'] = dataset.label
        meta['suggest'] = {
            'entity': {
                'service_url': domain,
                'service_path': '/' + dataset.name + '/suggest',
                'flyout_service_path': '/flyout'
                }
            }
        meta['defaultTypes'] = [{'name': dataset.label, 'id': '/' + dataset.name}]
    else:
        meta['defaultTypes'] = [{'name': d.label, 'id': '/' + d.name} for d in Dataset.all()]
    return jsonify(meta)

def view(dataset, value):
    dataset = Dataset.find(dataset)
    value = Value.find(dataset, value)
    format = response_format()
    if format == 'json':
        return jsonify(value)
    query = request.args.get('query', '').strip().lower()
    choices = match_op(value.value, dataset)
    choices = filter(lambda (c,v,s): v != value.id, choices)
    if len(query):
        choices = filter(lambda (c,v,s): query in Value.find(dataset,v).value.lower(),
                         choices)
                         # THIS is very inefficient - rather do this
                         # differently
    pager = Pager(choices, '.view', dataset=dataset.name,
                  value=value.id, limit=10)

    # HACK: Fetch only the values on the selected page.
    value_objs = Value.id_map(dataset, map(lambda (c,v,s): v,
        pager.query[pager.offset:pager.offset+pager.limit]))
    for i, (c,v,s) in enumerate(pager.query):
        if v in value_objs:
            pager.query[i] = (c, value_objs.get(v), s)

    return render_template('value/view.html', dataset=dataset,
                           value=value, values=pager, query=query)

def review(dataset):
    entities = Entity.all()
    dataset = Dataset.find(dataset)
    entities = entities.filter_by(dataset=dataset)
    entities = entities.filter(Entity.reviewed==False)
    entities = entities.offset(randint(0, entities.count()-1))
    return jsonify(entities.first())

def lookup(dataset):
    dataset = Dataset.find(dataset)
    readonly = validators.StringBool(if_empty=False, if_missing=False)\
            .to_python(request.args.get('readonly'))
    readonly = readonly if authz.logged_in() else True
    data = request_content()
    if response_format() != 'json':
        return Response("Not implemented!", status=400)

    try:
        alias = Alias.lookup(dataset, data, request.account,
                             readonly=readonly)
        if alias is None:
            return jsonify({
                'is_matched': False,
                'entity': None,
                'name': data.get('name'),
                'dataset': dataset.name
                }, status=404)

        if isinstance(alias, Entity):
            return jsonify({
                'is_matched': True,
                'entity': alias,
                'name': data.get('name'),
                'dataset': dataset.name
                }, status=200)

        db.session.commit()
        status = 200 if alias.is_matched else 404
        status = 418 if alias.is_invalid else status
        return jsonify(alias, status=status)
    except Invalid, inv:
        return handle_invalid(inv, index, data=data,
                              args=[dataset.name])

def suggest(dataset):
    """ 
    Suggest API, emulates Google Refine API. See:
    http://code.google.com/p/google-refine/wiki/SuggestApi
    """
    dataset = Dataset.by_name(dataset)
    entities = Entity.all().filter(Entity.invalid!=True)
    query = request.args.get('prefix', '').strip()
    entities = entities.filter(Entity.name.ilike('%s%%' % query))
    entities = entities.offset(get_offset(field='start'))
    entities = entities.limit(get_limit(default=20))

    matches = []
    for entity in entities:
        matches.append({
            'name': entity.name,
            'n:type': {
                'id': '/' + dataset.name,
                'name': dataset.label
                },
            'id': entity.id
            })
    return jsonify({
        "code" : "/api/status/ok",
        "status" : "200 OK",
        "prefix" : query,
        "result" : matches
        })

def reconcile(dataset=None):
    """
    Reconciliation API, emulates Google Refine API. See: 
    http://code.google.com/p/google-refine/wiki/ReconciliationServiceApi
    """
    if dataset is not None:
        dataset = Dataset.by_name(dataset)

    # TODO: Add proper support for types and namespacing.
    data = request.args.copy()
    data.update(request.form.copy())
    if 'query' in data:
        # single 
        q = data.get('query')
        if q.startswith('{'):
            try:
                q = json.loads(q)
            except ValueError:
                raise BadRequest()
        else:
            q = data
        return jsonify(reconcile_op(dataset, q))
    elif 'queries' in data:
        # multiple requests in one query
        qs = data.get('queries')
        try:
            qs = json.loads(qs)
        except ValueError:
            raise BadRequest()
        queries = {}
        for k, q in qs.items():
            queries[k] = reconcile_op(dataset, q)
        return jsonify(queries)
    else:
        return reconcile_index(dataset)

def create():
    data = request_data()
    dataset = Dataset.from_form(data)
    authz.require(authz.dataset_edit(dataset))
    entity = Entity.create(dataset, data, request.account)
    db.session.commit()
    return redirect(url_for('.view', id=entity.id))

def view(dataset, entity):
    dataset = Dataset.find(dataset)
    entity = Entity.find(dataset, entity)
    print entity.data
    format = response_format()
    if format == 'json':
        return jsonify(entity)
    query = request.args.get('query', '').strip().lower()
    choices = match_op(entity.name, dataset)
    choices = filter(lambda (c,e,s): e != entity.id, choices)
    if len(query):
        choices = filter(lambda (c,e,s): query in Entity.find(dataset,e).name.lower(),
                         choices)
                         # THIS is very inefficient - rather do this
                         # differently
    pager = Pager(choices, '.view', dataset=dataset.name,
                  entity=entity.id, limit=10)

    # HACK: Fetch only the entities on the selected page.
    entities = Entity.id_map(dataset, map(lambda (c,v,s): v,
        pager.query[pager.offset:pager.offset+pager.limit]))
    for i, (c,e,s) in enumerate(pager.query):
        if e in entities:
            pager.query[i] = (c, entities.get(e), s)

    return render_template('entity/view.html', dataset=dataset,
                           entity=entity, entities=pager, query=query)

def import_upload(dataset_name, sig, account_id,
        value_col, link_col):
    dataset = Dataset.find(dataset_name)
    account = Account.by_id(account_id)
    metadata, row_set = parse_upload(dataset, sig)
    headers = detect_headers(row_set)
    for row in row_set:
        data = dict([(c.column, c.value) for c in row])
        value = data.pop(value_col) if value_col else None
        link = data.pop(link_col) if link_col else None
        if link_col:
            d = {'key': link, 'data': data}
            link_obj = Link.lookup(dataset, d, account,
                            match_value=False)
            data = {}
        if value_col:
            d = {'value': value, 'data': data}
            value_obj = Value.by_value(dataset, value)
            if value_obj is None:
                value_obj = Value.create(dataset,
                        d, account)
            value_obj.data = data
        if link_col and value_col:
            link_obj.match(dataset, {'choice': value_obj.id},
                    account)
        db.session.commit()

def set_template_globals():
    return {
        "datasets": Dataset.all(),
        "authz": authz,
        "avatar_url": session.get("avatar_url", ""),
        "logged_in": request.account is not None,
        "login": request.account.login if request.account else None,
    }