if s[i+1] == c1 && Equal(s[i:i+n], sep) {
			return i
		}
		i++
		fails++
		if fails >= 4+i>>4 && i < t {
			// Give up on IndexByte, it isn't skipping ahead
			// far enough to be better than Rabin-Karp.
			// Experiments (using IndexPeriodic) suggest
			// the cutover is about 16 byte skips.
			// TODO: if large prefixes of sep are matching
			// we should cutover at even larger average skips,

	{"foofyfoobarfoobar", "y", 4},
	{"oooooooooooooooooooooo", "r", -1},
	{"oxoxoxoxoxoxoxoxoxoxoxoy", "oy", 22},
	{"oxoxoxoxoxoxoxoxoxoxoxox", "oy", -1},
	// test fallback to Rabin-Karp.
	{"000000000000000000000000000000000000000000000000000000000000000000000001", "0000000000000000000000000000000000000000000000000000000000000000001", 5},
}

var lastIndexTests = []BinOpTest{
	{"", "", 0},

  optional bool hostIPC = 13;

  // Specifies the hostname of the Carp
  // If not specified, the carp's hostname will be set to a system-defined value.
  // +optional
  optional string hostname = 16;

  // If specified, the fully qualified Carp hostname will be "<hostname>.<subdomain>.<carp namespace>.svc.<cluster domain>".
  // If not specified, the carp will not have a domainname at all.
  // +optional

// This list is auto-generated, don't edit it manually.
// aaa : 2015-02-26 American Automobile Association, Inc.
aaa

// aarp : 2015-05-21 AARP
aarp

// abarth : 2015-07-30 Fiat Chrysler Automobiles N.V.
abarth

// abb : 2014-10-24 ABB Ltd
abb

// abbott : 2014-07-24 Abbott Laboratories, Inc.
abbott

2ix.ch 2ix.de 3.azurestaticapps.net 3.bg 32-b.it 3utilities.com 4.bg 4lima.at 4lima.ch 4lima.de 4u.com 5.bg 5g.in 6.bg 611.to 64-b.it 6g.in 7.bg 8.bg 9.bg 9guacu.br a.bg a.prod.fastly.net a.run.app a.se a.ssl.fastly.net aa.no aaa aaa.pro aarborte.no aarp ab.ca abarth abashiri.hokkaido.jp abb abbott abbvie abc abc.br abeno.osaka.jp abiko.chiba.jp abira.hokkaido.jp abkhazia.su able abo.pa abogado abr.it abruzzo.it abu.yamaguchi.jp abudhabi ac ac.ae ac.at ac.be ac.ci ac.cn ac.cr ac.cy ac.fj ac.gn ac.gov.br...

dipmac??pmet&-aiblo?aiblo??rdnos?zal??l?m!a&greb?ret??oc?re&f?lap???n!a&dipmac&-oidem?oidem??lim?tsiro?zlob??ecip&-ilocsa?ilocsa??i&bru&-orasep?orasep??lleva?rot?tnert??r&elas?ovil??ulleb??p?r!a&sep&-onibru?onibru??znatac??oun??s!ivert?sabopmac??t!arp?e&nev?ssorg??n&arat?e&girga?rt?veneb????zz&era?urba???p&a?ohsdaerpsym,s?t??qa?r&a!m?s??b!a??c?f?g?k?me?o?p?s?t?v??s&a&b?iselgi&-ainobrac?ainobrac???b?c?elpan?i?m?o&t?x&bi,obdaili,??rahc21,s?t?v??t&a?b?c?l?m?nomdeip?o!psgolb,?p?v??u&de?l?n?p??v&a?og...

    *   Allow `DNNBoostedTreeCombinedEstimator` to work with core versions of
        feature columns and losses.
    *   Add non-linear image warping ops: `tf.contrib.image.sparse_image_warp`,
        `tf.contrib.image.dense_image_warp`, and
        `tf.contrib.image.interpolate_spline`.
    *   Fix bug in `tf.contrib.opt.MultitaskOptimizerWrapper` where types of
        tensors were mismatched.
*   Other: