73502 2011-11-30 17:00:17 -0800 Use const for local variables in KURL::init() and size_t in copyASCII() 2011-12-09 12:49:38 -0800 1 1 1 Unclassified WebKit New Bugs 528+ (Nightly build) Unspecified Unspecified RESOLVED WONTFIX P2 Normal --- 1 ddkilzer ddkilzer darin jamesr jarred koivisto pkasting rniwa webkit.review.bot oldest_to_newest 512251 0 ddkilzer 2011-11-30 17:00:17 -0800 Use const pointers in KURL::init() and size_t in copyASCII() 512252 1 117300 ddkilzer 2011-11-30 17:00:36 -0800 Created attachment 117300 Patch v1 512256 2 webkit.review.bot 2011-11-30 17:04:11 -0800 Attachment 117300 did not pass style-queue: Failed to run "['Tools/Scripts/check-webkit-style', '--diff-files', u'Source/WebCore/ChangeLog', u'Source/WebCor..." exit_code: 1 Source/WebCore/ChangeLog:1: ChangeLog entry has no bug number [changelog/bugnumber] [5] Total errors found: 1 in 2 files If any of these errors are false positives, please file a bug against check-webkit-style. 512258 3 darin 2011-11-30 17:05:15 -0800 Why the emphasis on “pointers” here? The const is just a local variable. The fact that it’s a pointer is not all that interesting. 513049 4 117300 koivisto 2011-12-01 13:30:54 -0800 Comment on attachment 117300 Patch v1 r=me 513051 5 koivisto 2011-12-01 13:31:36 -0800 The title could be better as it is not pointers only. 517656 6 ddkilzer 2011-12-08 12:10:32 -0800 (In reply to comment #3) > Why the emphasis on “pointers” here? The const is just a local variable. The fact that it’s a pointer is not all that interesting. (In reply to comment #5) > The title could be better as it is not pointers only. Fixed title. Will post another patch to incorporate David Levin's feedback on webkit-dev. 517866 7 118482 ddkilzer 2011-12-08 15:56:56 -0800 Created attachment 118482 Patch 517867 8 ddkilzer 2011-12-08 15:58:05 -0800 (In reply to comment #7) > Created an attachment (id=118482) [details] > Patch This addressed the title and ChangeLog wording fixes highlighted by Antti and Darin, and makes baseCharacters a const pointer per email from David Levin on webkit-dev[1]. [1] <message://%3cCACmjMJQzEqCvpODqk--JSmHnX1BEyQDsYMo_=Q1aLddX4crd8A@mail.gmail.com%3e> 517886 9 118482 darin 2011-12-08 16:18:19 -0800 Comment on attachment 118482 Patch View in context: https://bugs.webkit.org/attachment.cgi?id=118482&action=review > Source/WebCore/platform/KURL.cpp:262 > -static void copyASCII(const UChar* src, int length, char* dest) > +static void copyASCII(const UChar* src, const size_t length, char* dest) > { > - for (int i = 0; i < length; i++) > + for (size_t i = 0; i < length; i++) > dest[i] = static_cast<char>(src[i]); > } Why on earth would we use const here? 518074 10 ddkilzer 2011-12-08 20:34:32 -0800 (In reply to comment #9) > (From update of attachment 118482 [details]) > View in context: https://bugs.webkit.org/attachment.cgi?id=118482&action=review > > > Source/WebCore/platform/KURL.cpp:262 > > -static void copyASCII(const UChar* src, int length, char* dest) > > +static void copyASCII(const UChar* src, const size_t length, char* dest) > > { > > - for (int i = 0; i < length; i++) > > + for (size_t i = 0; i < length; i++) > > dest[i] = static_cast<char>(src[i]); > > } > > Why on earth would we use const here? Because the length parameter should never change within the method. Is there a technical reason that it shouldn't be used, or is this more of a style/personal preference issue? 518334 11 darin 2011-12-09 10:33:51 -0800 (In reply to comment #10) > (In reply to comment #9) > > (From update of attachment 118482 [details] [details]) > > View in context: https://bugs.webkit.org/attachment.cgi?id=118482&action=review > > > > > Source/WebCore/platform/KURL.cpp:262 > > > -static void copyASCII(const UChar* src, int length, char* dest) > > > +static void copyASCII(const UChar* src, const size_t length, char* dest) > > > { > > > - for (int i = 0; i < length; i++) > > > + for (size_t i = 0; i < length; i++) > > > dest[i] = static_cast<char>(src[i]); > > > } > > > > Why on earth would we use const here? > > Because the length parameter should never change within the method. > > Is there a technical reason that it shouldn't be used, or is this more of a style/personal preference issue? It is a style issue. Almost all arguments of almost all functions are not changed within the function. For example, this function could be written like this: static void copyASCII(const UChar* const src, const size_t length, char* const dest) This would indicate that src, length, and dest are all unchanged in the implementation of the function. I think that is harder to read than the existing function. And I am not sure why length stands out from the other arguments as worthwhile to call out as unchanged. Does that help people reading the code? It seems unnecessarily cluttered to include const all the time. If const was the default, then I would agree that we should leave all arguments and local variables const except for the ones we plan to modify, but since it’s not the default I prefer that we not put const everywhere. Stating const when there is something unclear and interesting to be clarified or emphasized is a different story. We’ve been discussing cases like that, for example I think you and Antti have been highlighting a case where there is a local variable in a large complex function and it might be useful to emphasize the point that it is never changed. For a two line function, declaring the arguments const seems not to clarify sufficiently to be worth the clutter. I’d like some sort of rule, even if a subjective one, about when we use it. It doesn’t have to be a firm machine-enforceable rule, but I do want consistency and not decisions made differently based on different programmers’ whims. I find it easier to work on WebKit because we use on a consistent basic coding style and adhere to it most of the time. The rule I suggested until now was “don’t use const for the values of arguments or of local variables” -- we’d still use const for members and of for the types the references refer to and pointers point to -- but I believe you wrote this bug at least in part because you are dissatisfied with that rule. So I’d like to make sure we have a new one. 518413 12 ddkilzer 2011-12-09 12:46:19 -0800 (In reply to comment #11) > (In reply to comment #10) > > Is there a technical reason that it shouldn't be used, or is this more of a style/personal preference issue? > > It is a style issue. > > Almost all arguments of almost all functions are not changed within the function. For example, this function could be written like this: > > static void copyASCII(const UChar* const src, const size_t length, char* const dest) Ahh! I agree that would detract the readability of the code. > I’d like some sort of rule, even if a subjective one, about when we use it. It doesn’t have to be a firm machine-enforceable rule, but I do want consistency and not decisions made differently based on different programmers’ whims. I find it easier to work on WebKit because we use on a consistent basic coding style and adhere to it most of the time. > > The rule I suggested until now was “don’t use const for the values of arguments or of local variables” -- we’d still use const for members and of for the types the references refer to and pointers point to -- but I believe you wrote this bug at least in part because you are dissatisfied with that rule. So I’d like to make sure we have a new one. Another way to look at this is that if KURL::init() were split up into smaller functions/methods, the need for making local variables const would greatly decrease since one could read the entire block of code to see how all of the variables are used (and thus which ones should not be changed because they're used for calculating lengths). Or to put it yet another way, my desire to add const to local variables is a really a code smell that probably means the method is too large to easily understand/read. However, refactoring KURL::init() is not at the top of my priority list, so I'm just going to close this bug as RESOLVED/WONTFIX. Thanks for the explanation. 518414 13 118482 ddkilzer 2011-12-09 12:46:37 -0800 Comment on attachment 118482 Patch Clearing flags and marking obsolete. 518419 14 rniwa 2011-12-09 12:49:38 -0800 (In reply to comment #10) > (In reply to comment #9) > > (From update of attachment 118482 [details] [details]) > > View in context: https://bugs.webkit.org/attachment.cgi?id=118482&action=review > > > > > Source/WebCore/platform/KURL.cpp:262 > > > -static void copyASCII(const UChar* src, int length, char* dest) > > > +static void copyASCII(const UChar* src, const size_t length, char* dest) > > > { > > > - for (int i = 0; i < length; i++) > > > + for (size_t i = 0; i < length; i++) > > > dest[i] = static_cast<char>(src[i]); > > > } > > > > Why on earth would we use const here? > > Because the length parameter should never change within the method. > > Is there a technical reason that it shouldn't be used, or is this more of a style/personal preference issue? This is a clear cut case where the constness of a variable is self-evident (the function is only 2 lines!), and there's almost no benefit in explicitly marking it as const. 117300 2011-11-30 17:00:36 -0800 2011-12-08 15:56:51 -0800 Patch v1 bug-73502-20111130170035.patch text/plain 3147 ddkilzer U3VidmVyc2lvbiBSZXZpc2lvbjogMTAxNDQ1CmRpZmYgLS1naXQgYS9Tb3VyY2UvV2ViQ29yZS9D aGFuZ2VMb2cgYi9Tb3VyY2UvV2ViQ29yZS9DaGFuZ2VMb2cKaW5kZXggNGVhYzFiZjMyZTIzN2Vm MDM0ZTVhYWJjYTNmM2ZhZTg2MDAyZjYxNi4uZGQyMDc5NzVmZjk5NzcwNjJlMjBkMGI3ZmVhN2Jh NGY3N2I3MWYwMSAxMDA2NDQKLS0tIGEvU291cmNlL1dlYkNvcmUvQ2hhbmdlTG9nCisrKyBiL1Nv dXJjZS9XZWJDb3JlL0NoYW5nZUxvZwpAQCAtMSwzICsxLDE2IEBACisyMDExLTExLTMwICBEYXZp ZCBLaWx6ZXIgIDxkZGtpbHplckBhcHBsZS5jb20+CisKKyAgICAgICAgVXNlIGNvbnN0IHBvaW50 ZXJzIGluIEtVUkw6OmluaXQoKSBhbmQgc2l6ZV90IGluIGNvcHlBU0NJSSgpCisgICAgICAgIDxo dHRwOi8vd2Via2l0Lm9yZy9iLzAwMDAwPgorCisgICAgICAgIFJldmlld2VkIGJ5IE5PQk9EWSAo T09QUyEpLgorCisgICAgICAgICogcGxhdGZvcm0vS1VSTC5jcHA6CisgICAgICAgIChXZWJDb3Jl Ojpjb3B5QVNDSUkpOiBTd2l0Y2ggZnJvbSBpbnQgdG8gc2l6ZV90IHNpbmNlIHdlJ3JlCisgICAg ICAgIHdvcmtpbmcgd2l0aCBwb2ludGVyIGluZGljZXMuCisgICAgICAgIChXZWJDb3JlOjpLVVJM Ojppbml0KTogTWFrZSBwb2ludGVycyBjb25zdCB0aGF0IGFyZSBub3QgaW50ZW5kZWQKKyAgICAg ICAgdG8gY2hhbmdlLiAgTWFrZSBiYXNlTGVuZ3RoIGNvbnN0LgorCiAyMDExLTExLTI4ICBLZW50 YXJvIEhhcmEgIDxoYXJha2VuQGNocm9taXVtLm9yZz4KIAogICAgICAgICBJbXBsZW1lbnQgW1N1 cHBsZW1lbnRhbF0gSURMIGFuZCBzdXBwb3J0IGl0IGluIHJ1bi1iaW5kaW5ncy10ZXN0cwpkaWZm IC0tZ2l0IGEvU291cmNlL1dlYkNvcmUvcGxhdGZvcm0vS1VSTC5jcHAgYi9Tb3VyY2UvV2ViQ29y ZS9wbGF0Zm9ybS9LVVJMLmNwcAppbmRleCBlNzUyYmI4YTQ2YjUwYWFlNGI4Y2E4NTA2ZmY3NDgw OTJhNjdkN2RmLi5mYjkyMDMzMWYyZWY3OWVkZDg0M2FlYjYyMDNmODZjZDA1YmE0ZmI1IDEwMDY0 NAotLS0gYS9Tb3VyY2UvV2ViQ29yZS9wbGF0Zm9ybS9LVVJMLmNwcAorKysgYi9Tb3VyY2UvV2Vi Q29yZS9wbGF0Zm9ybS9LVVJMLmNwcApAQCAtMjU1LDkgKzI1NSw5IEBAIHN0YXRpYyBpbmxpbmUg Ym9vbCBpc1NjaGVtZUNoYXJhY3Rlck1hdGNoSWdub3JpbmdDYXNlKGNoYXIgY2hhcmFjdGVyLCBj aGFyIHNjaGVtCiAvLyBDb3BpZXMgdGhlIHNvdXJjZSB0byB0aGUgZGVzdGluYXRpb24sIGFzc3Vt aW5nIGFsbCB0aGUgc291cmNlIGNoYXJhY3RlcnMgYXJlCiAvLyBBU0NJSS4gVGhlIGRlc3RpbmF0 aW9uIGJ1ZmZlciBtdXN0IGJlIGxhcmdlIGVub3VnaC4gTnVsbCBjaGFyYWN0ZXJzIGFyZSBhbGxv d2VkCiAvLyBpbiB0aGUgc291cmNlIHN0cmluZywgYW5kIG5vIGF0dGVtcHQgaXMgbWFkZSB0byBu dWxsLXRlcm1pbmF0ZSB0aGUgcmVzdWx0Lgotc3RhdGljIHZvaWQgY29weUFTQ0lJKGNvbnN0IFVD aGFyKiBzcmMsIGludCBsZW5ndGgsIGNoYXIqIGRlc3QpCitzdGF0aWMgdm9pZCBjb3B5QVNDSUko Y29uc3QgVUNoYXIqIHNyYywgY29uc3Qgc2l6ZV90IGxlbmd0aCwgY2hhciogZGVzdCkKIHsKLSAg ICBmb3IgKGludCBpID0gMDsgaSA8IGxlbmd0aDsgaSsrKQorICAgIGZvciAoc2l6ZV90IGkgPSAw OyBpIDwgbGVuZ3RoOyBpKyspCiAgICAgICAgIGRlc3RbaV0gPSBzdGF0aWNfY2FzdDxjaGFyPihz cmNbaV0pOwogfQogCkBAIC00ODYsMTkgKzQ4NiwxOSBAQCB2b2lkIEtVUkw6OmluaXQoY29uc3Qg S1VSTCYgYmFzZSwgY29uc3QgU3RyaW5nJiByZWxhdGl2ZSwgY29uc3QgVGV4dEVuY29kaW5nJiBl bgogICAgICAgICAgICAgICAgIHBhcnNlQnVmZmVyLnJlc2l6ZShidWZmZXJTaXplKTsKIAogICAg ICAgICAgICAgICAgIGNoYXIqIGJ1ZmZlclBvcyA9IHBhcnNlQnVmZmVyLmRhdGEoKTsKLSAgICAg ICAgICAgICAgICBjaGFyKiBidWZmZXJTdGFydCA9IGJ1ZmZlclBvczsKKyAgICAgICAgICAgICAg ICBjaGFyKiBjb25zdCBidWZmZXJTdGFydCA9IGJ1ZmZlclBvczsKIAogICAgICAgICAgICAgICAg IC8vIGZpcnN0IGNvcHkgZXZlcnl0aGluZyBiZWZvcmUgdGhlIHBhdGggZnJvbSB0aGUgYmFzZQot ICAgICAgICAgICAgICAgIHVuc2lnbmVkIGJhc2VMZW5ndGggPSBiYXNlLm1fc3RyaW5nLmxlbmd0 aCgpOworICAgICAgICAgICAgICAgIGNvbnN0IHVuc2lnbmVkIGJhc2VMZW5ndGggPSBiYXNlLm1f c3RyaW5nLmxlbmd0aCgpOwogICAgICAgICAgICAgICAgIGNvbnN0IFVDaGFyKiBiYXNlQ2hhcmFj dGVycyA9IGJhc2UubV9zdHJpbmcuY2hhcmFjdGVycygpOwogICAgICAgICAgICAgICAgIENoYXJC dWZmZXIgYmFzZVN0cmluZ0J1ZmZlcihiYXNlTGVuZ3RoKTsKICAgICAgICAgICAgICAgICBjb3B5 QVNDSUkoYmFzZUNoYXJhY3RlcnMsIGJhc2VMZW5ndGgsIGJhc2VTdHJpbmdCdWZmZXIuZGF0YSgp KTsKLSAgICAgICAgICAgICAgICBjb25zdCBjaGFyKiBiYXNlU3RyaW5nID0gYmFzZVN0cmluZ0J1 ZmZlci5kYXRhKCk7CisgICAgICAgICAgICAgICAgY29uc3QgY2hhciogY29uc3QgYmFzZVN0cmlu ZyA9IGJhc2VTdHJpbmdCdWZmZXIuZGF0YSgpOwogICAgICAgICAgICAgICAgIGNvbnN0IGNoYXIq IGJhc2VTdHJpbmdTdGFydCA9IGJhc2VTdHJpbmc7Ci0gICAgICAgICAgICAgICAgY29uc3QgY2hh ciogcGF0aFN0YXJ0ID0gYmFzZVN0cmluZ1N0YXJ0ICsgYmFzZS5tX3BvcnRFbmQ7CisgICAgICAg ICAgICAgICAgY29uc3QgY2hhciogY29uc3QgcGF0aFN0YXJ0ID0gYmFzZVN0cmluZ1N0YXJ0ICsg YmFzZS5tX3BvcnRFbmQ7CiAgICAgICAgICAgICAgICAgd2hpbGUgKGJhc2VTdHJpbmdTdGFydCA8 IHBhdGhTdGFydCkKICAgICAgICAgICAgICAgICAgICAgKmJ1ZmZlclBvcysrID0gKmJhc2VTdHJp bmdTdGFydCsrOwotICAgICAgICAgICAgICAgIGNoYXIqIGJ1ZmZlclBhdGhTdGFydCA9IGJ1ZmZl clBvczsKKyAgICAgICAgICAgICAgICBjaGFyKiBjb25zdCBidWZmZXJQYXRoU3RhcnQgPSBidWZm ZXJQb3M7CiAKICAgICAgICAgICAgICAgICAvLyBub3cgY29weSB0aGUgYmFzZSBwYXRoCiAgICAg ICAgICAgICAgICAgY29uc3QgY2hhciogYmFzZVN0cmluZ0VuZCA9IGJhc2VTdHJpbmcgKyBiYXNl Lm1fcGF0aEVuZDsK 118482 2011-12-08 15:56:56 -0800 2011-12-09 12:46:37 -0800 Patch bug-73502-20111208155654.patch text/plain 3222 ddkilzer U3VidmVyc2lvbiBSZXZpc2lvbjogMTAxNDQ1CmRpZmYgLS1naXQgYS9Tb3VyY2UvV2ViQ29yZS9D aGFuZ2VMb2cgYi9Tb3VyY2UvV2ViQ29yZS9DaGFuZ2VMb2cKaW5kZXggNGVhYzFiZjMyZTIzN2Vm MDM0ZTVhYWJjYTNmM2ZhZTg2MDAyZjYxNi4uYWEyNTVlYTNjNGZjMjc4NzQwZDkzODgzNTM0OTEz NjRhYjA2OGNhZiAxMDA2NDQKLS0tIGEvU291cmNlL1dlYkNvcmUvQ2hhbmdlTG9nCisrKyBiL1Nv dXJjZS9XZWJDb3JlL0NoYW5nZUxvZwpAQCAtMSwzICsxLDE2IEBACisyMDExLTEyLTA4ICBEYXZp ZCBLaWx6ZXIgIDxkZGtpbHplckBhcHBsZS5jb20+CisKKyAgICAgICAgVXNlIGNvbnN0IGZvciBs b2NhbCB2YXJpYWJsZXMgaW4gS1VSTDo6aW5pdCgpIGFuZCBzaXplX3QgaW4gY29weUFTQ0lJKCkK KyAgICAgICAgPGh0dHA6Ly93ZWJraXQub3JnL2IvNzM1MDI+CisKKyAgICAgICAgUmV2aWV3ZWQg YnkgTk9CT0RZIChPT1BTISkuCisKKyAgICAgICAgKiBwbGF0Zm9ybS9LVVJMLmNwcDoKKyAgICAg ICAgKFdlYkNvcmU6OmNvcHlBU0NJSSk6IFN3aXRjaCBmcm9tIGludCB0byBzaXplX3Qgc2luY2Ug d2UncmUKKyAgICAgICAgd29ya2luZyB3aXRoIHBvaW50ZXIgaW5kaWNlcy4KKyAgICAgICAgKFdl YkNvcmU6OktVUkw6OmluaXQpOiBNYWtlIGxvY2FsIHZhcmlhYmxlcyBjb25zdCB0aGF0IGFyZSBu b3QKKyAgICAgICAgaW50ZW5kZWQgdG8gY2hhbmdlLgorCiAyMDExLTExLTI4ICBLZW50YXJvIEhh cmEgIDxoYXJha2VuQGNocm9taXVtLm9yZz4KIAogICAgICAgICBJbXBsZW1lbnQgW1N1cHBsZW1l bnRhbF0gSURMIGFuZCBzdXBwb3J0IGl0IGluIHJ1bi1iaW5kaW5ncy10ZXN0cwpkaWZmIC0tZ2l0 IGEvU291cmNlL1dlYkNvcmUvcGxhdGZvcm0vS1VSTC5jcHAgYi9Tb3VyY2UvV2ViQ29yZS9wbGF0 Zm9ybS9LVVJMLmNwcAppbmRleCBlNzUyYmI4YTQ2YjUwYWFlNGI4Y2E4NTA2ZmY3NDgwOTJhNjdk N2RmLi5lODM0NGIxNjdkMjFmZWVjOGQxOWQ2OWQ2ODlhMzFiYTVlNTQwNWM0IDEwMDY0NAotLS0g YS9Tb3VyY2UvV2ViQ29yZS9wbGF0Zm9ybS9LVVJMLmNwcAorKysgYi9Tb3VyY2UvV2ViQ29yZS9w bGF0Zm9ybS9LVVJMLmNwcApAQCAtMjU1LDkgKzI1NSw5IEBAIHN0YXRpYyBpbmxpbmUgYm9vbCBp c1NjaGVtZUNoYXJhY3Rlck1hdGNoSWdub3JpbmdDYXNlKGNoYXIgY2hhcmFjdGVyLCBjaGFyIHNj aGVtCiAvLyBDb3BpZXMgdGhlIHNvdXJjZSB0byB0aGUgZGVzdGluYXRpb24sIGFzc3VtaW5nIGFs bCB0aGUgc291cmNlIGNoYXJhY3RlcnMgYXJlCiAvLyBBU0NJSS4gVGhlIGRlc3RpbmF0aW9uIGJ1 ZmZlciBtdXN0IGJlIGxhcmdlIGVub3VnaC4gTnVsbCBjaGFyYWN0ZXJzIGFyZSBhbGxvd2VkCiAv LyBpbiB0aGUgc291cmNlIHN0cmluZywgYW5kIG5vIGF0dGVtcHQgaXMgbWFkZSB0byBudWxsLXRl cm1pbmF0ZSB0aGUgcmVzdWx0Lgotc3RhdGljIHZvaWQgY29weUFTQ0lJKGNvbnN0IFVDaGFyKiBz cmMsIGludCBsZW5ndGgsIGNoYXIqIGRlc3QpCitzdGF0aWMgdm9pZCBjb3B5QVNDSUkoY29uc3Qg VUNoYXIqIHNyYywgY29uc3Qgc2l6ZV90IGxlbmd0aCwgY2hhciogZGVzdCkKIHsKLSAgICBmb3Ig KGludCBpID0gMDsgaSA8IGxlbmd0aDsgaSsrKQorICAgIGZvciAoc2l6ZV90IGkgPSAwOyBpIDwg bGVuZ3RoOyBpKyspCiAgICAgICAgIGRlc3RbaV0gPSBzdGF0aWNfY2FzdDxjaGFyPihzcmNbaV0p OwogfQogCkBAIC00ODYsMTkgKzQ4NiwxOSBAQCB2b2lkIEtVUkw6OmluaXQoY29uc3QgS1VSTCYg YmFzZSwgY29uc3QgU3RyaW5nJiByZWxhdGl2ZSwgY29uc3QgVGV4dEVuY29kaW5nJiBlbgogICAg ICAgICAgICAgICAgIHBhcnNlQnVmZmVyLnJlc2l6ZShidWZmZXJTaXplKTsKIAogICAgICAgICAg ICAgICAgIGNoYXIqIGJ1ZmZlclBvcyA9IHBhcnNlQnVmZmVyLmRhdGEoKTsKLSAgICAgICAgICAg ICAgICBjaGFyKiBidWZmZXJTdGFydCA9IGJ1ZmZlclBvczsKKyAgICAgICAgICAgICAgICBjaGFy KiBjb25zdCBidWZmZXJTdGFydCA9IGJ1ZmZlclBvczsKIAogICAgICAgICAgICAgICAgIC8vIGZp cnN0IGNvcHkgZXZlcnl0aGluZyBiZWZvcmUgdGhlIHBhdGggZnJvbSB0aGUgYmFzZQotICAgICAg ICAgICAgICAgIHVuc2lnbmVkIGJhc2VMZW5ndGggPSBiYXNlLm1fc3RyaW5nLmxlbmd0aCgpOwot ICAgICAgICAgICAgICAgIGNvbnN0IFVDaGFyKiBiYXNlQ2hhcmFjdGVycyA9IGJhc2UubV9zdHJp bmcuY2hhcmFjdGVycygpOworICAgICAgICAgICAgICAgIGNvbnN0IHVuc2lnbmVkIGJhc2VMZW5n dGggPSBiYXNlLm1fc3RyaW5nLmxlbmd0aCgpOworICAgICAgICAgICAgICAgIGNvbnN0IFVDaGFy KiBjb25zdCBiYXNlQ2hhcmFjdGVycyA9IGJhc2UubV9zdHJpbmcuY2hhcmFjdGVycygpOwogICAg ICAgICAgICAgICAgIENoYXJCdWZmZXIgYmFzZVN0cmluZ0J1ZmZlcihiYXNlTGVuZ3RoKTsKICAg ICAgICAgICAgICAgICBjb3B5QVNDSUkoYmFzZUNoYXJhY3RlcnMsIGJhc2VMZW5ndGgsIGJhc2VT dHJpbmdCdWZmZXIuZGF0YSgpKTsKLSAgICAgICAgICAgICAgICBjb25zdCBjaGFyKiBiYXNlU3Ry aW5nID0gYmFzZVN0cmluZ0J1ZmZlci5kYXRhKCk7CisgICAgICAgICAgICAgICAgY29uc3QgY2hh ciogY29uc3QgYmFzZVN0cmluZyA9IGJhc2VTdHJpbmdCdWZmZXIuZGF0YSgpOwogICAgICAgICAg ICAgICAgIGNvbnN0IGNoYXIqIGJhc2VTdHJpbmdTdGFydCA9IGJhc2VTdHJpbmc7Ci0gICAgICAg ICAgICAgICAgY29uc3QgY2hhciogcGF0aFN0YXJ0ID0gYmFzZVN0cmluZ1N0YXJ0ICsgYmFzZS5t X3BvcnRFbmQ7CisgICAgICAgICAgICAgICAgY29uc3QgY2hhciogY29uc3QgcGF0aFN0YXJ0ID0g YmFzZVN0cmluZ1N0YXJ0ICsgYmFzZS5tX3BvcnRFbmQ7CiAgICAgICAgICAgICAgICAgd2hpbGUg KGJhc2VTdHJpbmdTdGFydCA8IHBhdGhTdGFydCkKICAgICAgICAgICAgICAgICAgICAgKmJ1ZmZl clBvcysrID0gKmJhc2VTdHJpbmdTdGFydCsrOwotICAgICAgICAgICAgICAgIGNoYXIqIGJ1ZmZl clBhdGhTdGFydCA9IGJ1ZmZlclBvczsKKyAgICAgICAgICAgICAgICBjaGFyKiBjb25zdCBidWZm ZXJQYXRoU3RhcnQgPSBidWZmZXJQb3M7CiAKICAgICAgICAgICAgICAgICAvLyBub3cgY29weSB0 aGUgYmFzZSBwYXRoCiAgICAgICAgICAgICAgICAgY29uc3QgY2hhciogYmFzZVN0cmluZ0VuZCA9 IGJhc2VTdHJpbmcgKyBiYXNlLm1fcGF0aEVuZDsK