7360 2006-02-19 04:44:41 -0800 Typing style isn't bold in a bold context 2022-07-06 13:33:22 -0700 1 1 1 Unclassified WebKit HTML Editing 420+ Mac OS X 10.4 NEW InRadar P2 Normal --- 1 Graham.Dennis webkit-unassigned ap ayg bfulgham dwood enrica ian justin.garcia rniwa webkit-bug-importer oldest_to_newest 33363 0 Graham.Dennis 2006-02-19 04:44:41 -0800 Let me prefix this by saying that I'm not sure that this actually is a bug. If one adds styling to the text at the start of a contentEditable div, and then deletes all of the content in the div, any new text that is added has the styling that was added to the text at the start of the div. This is the same behaviour as TextEdit, however, for TextEdit, there is no well-defined default style. For a contentEditable div, it is possible that a style has been applied to the div, which has since been overridden by the span. It makes sense that there should be a way to remove the style applied by the span. One way to do this would be to remove a span once it contains no text. I'll attach a testcase. 33364 1 6607 Graham.Dennis 2006-02-19 04:46:31 -0800 Created attachment 6607 testcase The red colour is applied by a <span> element inside the contentEditable div. If all the text in the div is deleted, this style is still applied to any new text typed in the div. 33365 2 joost 2006-02-19 05:30:55 -0800 The behavior i can confirm... Don't know if this is wrong thoug... Justin, what do you think? 33728 3 justin.garcia 2006-02-22 14:03:41 -0800 (In reply to comment #0) > This is the same behaviour as TextEdit, however, for TextEdit, there is no well-defined default > style. In TextEdit, create a few lines of bold text, with an empty line somewhere in the middle. Click in that empty line, unbold with command-b, and start typing. You'll get unbolded text. Now, delete all the text on that line, and begin typing again. You'll still have unbolded text. So, even though you're inside a bold context, the unbold style hangs onto the caret even though you've deleted all unbold text. Also notice that if you click somewhere else and then click back on the empty line and start typing, you'll get bold text. So clicking away and clicking back blows away the unbold style that was hanging on the caret. We mimic this behavior as well. We consider this behavior correct, since our goal in general has been to mimic the behavior of TextEdit. 33854 4 Graham.Dennis 2006-02-24 00:05:38 -0800 (In reply to comment #3) > Also notice that if you click somewhere else and then click back on the empty > line and start typing, you'll get bold text. So clicking away and clicking > back blows away the unbold style that was hanging on the caret. We mimic this > behavior as well. I just tried to do this, but the behaviour is not what you said. If in the testcase, I create three lines of text, with a blank line in the middle, then bold the lot, text on the blank line is typed as bold. So far, so good. I unbold and type text which is unbolded. Then I delete the the text on this line (but not the line itself), then click on the URL entry field, then back on the empty line. New text I type is not bold, but we would expect bold text (and TextEdit does show bold text). I haven't reopened this bug in case I'm not understanding this properly. 33861 5 justin.garcia 2006-02-24 02:07:15 -0800 (In reply to comment #4) > (In reply to comment #3) > > Also notice that if you click somewhere else and then click back on the empty > > line and start typing, you'll get bold text. So clicking away and clicking > > back blows away the unbold style that was hanging on the caret. We mimic this > > behavior as well. > > I just tried to do this, but the behaviour is not what you said. If in the > testcase, I create three lines of text, with a blank line in the middle, then > bold the lot, text on the blank line is typed as bold. So far, so good. I > unbold and type text which is unbolded. Then I delete the the text on this line > (but not the line itself), then click on the URL entry field, then back on the > empty line. New text I type is not bold, but we would expect bold text (and > TextEdit does show bold text). You're right, this is a bug. Reopening. 34976 6 Graham.Dennis 2006-03-03 13:02:14 -0800 I've had a look at this, and the problem seems to be caused when DeleteSelectionCommand::calculateTypingStyleAfterDelete applies the typing style to the placeholder element. However, I can't figure out how to calculate the correct style. Consider the following two cases: 1: Type three lines, bold all three, and delete the text in the middle line. 2: Type three lines, bold just the first and third lines, and delete the text in the middle line. In case 1, after moving the caret away and back again, new text on the second line should be bold. In case 2, after moving the caret away and back again, new text on the second line should not be bold. In HTML, it will be difficult to tell the difference between case 1 and case 2 because all tags must be nested, and because the newlines aren't <br/>'s, but each paragraph is a block element, there must be a boldface tag for each line separately. So how can we tell the difference between these cases in HTML? For the given two cases, it looks like one solution might be to use the current typing style (this is what is done in the code at present), but consider a third case: 3: Type three lines, bold all three, unbold the middle line, now delete the text in the middle line. In this case, after clicking away and back again, the text should be bold, which is not the previous typing style. Ideas anyone? 38601 7 7526 Graham.Dennis 2006-04-05 06:41:49 -0700 Created attachment 7526 work in progress patch Justin, I've been doing a bit of work on this bug, and what I have been trying to do is to apply styles to the block element when a style should be applied to the entire paragraph (i.e. the style should be persistent when moving the caret away and back to the paragraph). This means that some styles aren't using the apple-style-span span's... Is this reasonable behaviour? The patch I have attached is a work in progress that I'd like your feedback on. I haven't marked it r? as it is missing a testcase, changelog entry, etc. I just want to know if you think I'm on the right track. Other than that, this patch does fix this bug. 38605 8 justin.garcia 2006-04-05 09:06:58 -0700 I'm not sure that this bug is titled appropriately anymore, and the testcase that's attached doesn't describe what it's testing. Let's make sure we're on the same page. There is only one bug discussed here that I am familiar with. If there are others, let's file those in separate reports. Here's the tree we produce, and the typing style resulting from the three examples you gave: > 1: Type three lines, bold all three, and delete the text in the middle line. <div><b>line 1</b></div> <div><b><br></b></div> <div><b>line 3</b></div> Current typing style: bold Typing on the second line produces characters with the right style, and clicking away/clicking back and typing produces characters with the right style. > 2: Type three lines, bold just the first and third lines, and delete the text in the middle line. <div><b>line 1</b></div> <div><br></div> <div><b>line 3</b></div> Current typing style: unbold Typing on the second line produces characters with the right style, and clicking away/clicking back and typing produces characters with the right style. > 3: Type three lines, bold all three, unbold the middle line, now delete the text in the middle line. <div><b>line 1</b></div> <div><br></div> <div><b>line 3</b></div> Current typing style: unbold Typing on the second line produces characters with the right style, but clicking away/clicking back and typing produces unbold text. This is the only bug as far as I can tell. I'm going to retitle this bug and add a testcase that demonstrates this bug. > In HTML, it will be difficult to tell the difference between case 1 and case 2 because all tags must be nested What do you mean by "all tags must be nested"? 38607 9 7528 justin.garcia 2006-04-05 09:20:45 -0700 Created attachment 7528 testcase This is an automated testcase that demonstrates this bug: Type three lines of text Bold them all Unbold the middle line Delete the text on the middle line Click away and click back to the middle line Start typing You should see bold text. You instead see unbold text. 40700 10 Graham.Dennis 2006-04-28 18:17:42 -0700 Sorry for the delay in replying... (In reply to comment #8) > I'm not sure that this bug is titled appropriately anymore, and the testcase > that's attached doesn't describe what it's testing. Let's make sure we're on > the same page. There is only one bug discussed here that I am familiar with. > If there are others, let's file those in separate reports. That's the bug I'm talking about. Of the three cases I was talking about, only the third worked incorrectly. The reason I mentioned the other two is because they are similar, and at present the HTML code produced for 2 and 3 is identical, however I wasn't sure how they should differ when the bug is fixed. > > In HTML, it will be difficult to tell the difference between case 1 and case 2 because all tags must be nested > > What do you mean by "all tags must be nested"? What I mean to say is that in Cocoa, this problem is avoided because we can say that the bolding started in one paragraph, continued through one paragraph and into the next. However this is impossible to do with one tag in HTML as all tags must be nested. The bolding region that starts in the first paragraph must also finish in the first paragraph. So a new bolding region must be used for the second and third paragraphs. My solution to this problem is where a bolding region (or styling region in general) would cover an entire paragraph (actually any Block element), the style is applied to the block element itself, instead of creating a styling span within it. This way, when all text in a block is deleted (and any empty styling spans removed), when the cursor moves away and back again, the style that was applied to the block will still apply to the text. Let me know if I haven't explained myself properly... 44181 11 justin.garcia 2006-05-31 15:13:53 -0700 > This way, when all text in a block is deleted (and any empty > styling spans removed), when the cursor moves away and back again, the style > that was applied to the block will still apply to the text. So, would it be better to fix this bug by not removing a span when its contents are removed? Please add some testcases. // Do not check for legacy styles here. Those styles, like <B> and <I>, only apply for // inline content. + // No longer true. Just remove the comment. r- for now until you add some testcases. 44182 12 justin.garcia 2006-05-31 15:20:23 -0700 + // If the node we are in is an empty paragraph, then we have selected the entire paragraph + // so we want to apply the style to the whole paragraph + VisiblePosition startPos(start, VP_UPSTREAM_IF_POSSIBLE); + if (start.offset() == end.offset() && isStartOfParagraph(startPos) && isEndOfParagraph(startPos)) { + HTMLElement *block = static_cast<HTMLElement *>(start.node()->enclosingBlockFlowElement()); + removeCSSStyle(style, block); + addBlockStyleIfNeeded(style, block); Just because startPos is in an empty paragraph, doesn't mean that it's in a block by itself. For example [br2, 0] example in the tree below: <div><br1><br2></div> 47971 13 mrowe 2006-07-02 02:17:04 -0700 Comments indicates this has been confirmed as a bug. 453351 14 ayg 2011-08-18 14:27:08 -0700 (In reply to comment #10) > My solution to this problem is where a bolding region (or styling region in general) would cover an entire paragraph (actually any Block element), the style is applied to the block element itself, instead of creating a styling span within it. This way, when all text in a block is deleted (and any empty styling spans removed), when the cursor moves away and back again, the style that was applied to the block will still apply to the text. This is not a good idea. Gecko does something like this, and it causes problems: 1) It doesn't work at all if styleWithCSS is false. You can't use CSS for styling in that case: it would break non-CSS UAs, like some mail clients. (Or so I've been told.) 2) If you do it only when styleWithCSS is true, it will produce markup that has different effects in some cases. For instance, <p style="text-decoration:underline"><font color=red>foo</font></p> has a different underline color than <p style="color:red; text-decoration:underline">foo</p>. This is bad, because styleWithCSS vs. not should only affect the markup produced, not the visual effect. 3) It breaks hiliteColor, because background-color does something different on block vs. inline elements. <p style="background-color:teal">foo</p> is different from <p><span style="background-color:teal">foo</span></p>. 4) It makes everything more complicated. You'll have to add lots of different code paths everywhere to properly handle the two different types of style. Instead, I would suggest that if the entire contents of the block are deleted, the styling tag should just be left. So instead of <p><b>foo</b></p> -> <p><br></p> with a typing style set, it should become <p><b><br></b></p> with no typing style. This bug also affects the editing spec. I filed a bug against that: http://www.w3.org/Bugs/Public/show_bug.cgi?id=13831 I plan to specify the solution that I just outlined, if no one has any objections to it. 613247 15 rniwa 2012-04-30 22:41:03 -0700 Also see the bug 34608. 1881077 16 bfulgham 2022-07-06 13:33:07 -0700 Chrome and Safari agree on behavior (the middle line is not bolded), while Firefox shows all three lines in bold. 1881078 17 webkit-bug-importer 2022-07-06 13:33:22 -0700 <rdar://problem/96544639> 6607 2006-02-19 04:46:31 -0800 2006-04-05 09:20:45 -0700 testcase SpanStylesInEditableDivs.html text/html 111 Graham.Dennis PGh0bWw+Cjxib2R5Pgo8ZGl2IGNvbnRlbnRFZGl0YWJsZT0idHJ1ZSI+PHNwYW4gc3R5bGU9ImNv bG9yOiByZ2IoMjU1LCAwLCAwKTsiPkZvbzwvc3Bhbj48L2Rpdj4KPC9ib2R5Pgo8L2h0bWw+ 7526 2006-04-05 06:41:49 -0700 2010-06-10 14:22:23 -0700 work in progress patch BlockStyles test.diff text/plain 5843 Graham.Dennis SW5kZXg6IFdlYkNvcmUvZWRpdGluZy9EZWxldGVTZWxlY3Rpb25Db21tYW5kLmNwcAo9PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09Ci0tLSBXZWJDb3JlL2VkaXRpbmcvRGVsZXRlU2VsZWN0aW9uQ29tbWFuZC5jcHAJKHJldmlz aW9uIDEzNTM1KQorKysgV2ViQ29yZS9lZGl0aW5nL0RlbGV0ZVNlbGVjdGlvbkNvbW1hbmQuY3Bw CSh3b3JraW5nIGNvcHkpCkBAIC02MjUsOCArNjI1LDggQEAgdm9pZCBEZWxldGVTZWxlY3Rpb25D b21tYW5kOjpjYWxjdWxhdGVUeQogICAgICAgICAvLyBpcyBub3QgcmV0YWluZWQgdW50aWwgdGhl IG5leHQgdHlwaW5nIGFjdGlvbi4KIAogICAgICAgICBzZXRFbmRpbmdTZWxlY3Rpb24oU2VsZWN0 aW9uKFBvc2l0aW9uKGluc2VydGVkUGxhY2Vob2xkZXIsIDApLCBET1dOU1RSRUFNKSk7Ci0gICAg ICAgIGFwcGx5U3R5bGUobV90eXBpbmdTdHlsZS5nZXQoKSwgRWRpdEFjdGlvblVuc3BlY2lmaWVk KTsKLSAgICAgICAgbV90eXBpbmdTdHlsZSA9IDA7CisvLyAgICAgICAgYXBwbHlTdHlsZShtX3R5 cGluZ1N0eWxlLmdldCgpLCBFZGl0QWN0aW9uVW5zcGVjaWZpZWQpOworLy8gICAgICAgIG1fdHlw aW5nU3R5bGUgPSAwOwogICAgIH0KICAgICAvLyBTZXQgbV90eXBpbmdTdHlsZSBhcyB0aGUgdHlw aW5nIHN0eWxlLgogICAgIC8vIEl0J3MgcGVyZmVjdGx5IE9LIGZvciBtX3R5cGluZ1N0eWxlIHRv IGJlIG51bGwuCkluZGV4OiBXZWJDb3JlL2VkaXRpbmcvQXBwbHlTdHlsZUNvbW1hbmQuY3BwCj09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT0KLS0tIFdlYkNvcmUvZWRpdGluZy9BcHBseVN0eWxlQ29tbWFuZC5jcHAJKHJldmlz aW9uIDEzNTM1KQorKysgV2ViQ29yZS9lZGl0aW5nL0FwcGx5U3R5bGVDb21tYW5kLmNwcAkod29y a2luZyBjb3B5KQpAQCAtNDEsNiArNDEsNyBAQAogI2luY2x1ZGUgImh0bWxlZGl0aW5nLmgiCiAj aW5jbHVkZSAiSFRNTE5hbWVzLmgiCiAjaW5jbHVkZSAiUmVuZGVyT2JqZWN0LmgiCisjaW5jbHVk ZSAidmlzaWJsZV91bml0cy5oIgogI2luY2x1ZGUgPGt4bWxjb3JlL0Fzc2VydGlvbnMuaD4KIAog bmFtZXNwYWNlIFdlYkNvcmUgewpAQCAtNTkxLDEwICs1OTIsNDIgQEAgdm9pZCBBcHBseVN0eWxl Q29tbWFuZDo6YXBwbHlJbmxpbmVTdHlsZQogICAgICAgICBub2RlID0gbm9kZS0+dHJhdmVyc2VO ZXh0Tm9kZSgpOwogICAgIAogICAgIGlmIChzdGFydC5ub2RlKCkgPT0gZW5kLm5vZGUoKSkgewot ICAgICAgICBhZGRJbmxpbmVTdHlsZUlmTmVlZGVkKHN0eWxlLCBub2RlLCBub2RlKTsKKyAgICAK KyAgICAgICAgLy8gSWYgdGhlIG5vZGUgd2UgYXJlIGluIGlzIGFuIGVtcHR5IHBhcmFncmFwaCwg dGhlbiB3ZSBoYXZlIHNlbGVjdGVkIHRoZSBlbnRpcmUgcGFyYWdyYXBoCisgICAgICAgIC8vIHNv IHdlIHdhbnQgdG8gYXBwbHkgdGhlIHN0eWxlIHRvIHRoZSB3aG9sZSBwYXJhZ3JhcGgKKyAgICAg ICAgVmlzaWJsZVBvc2l0aW9uIHN0YXJ0UG9zKHN0YXJ0LCBWUF9VUFNUUkVBTV9JRl9QT1NTSUJM RSk7CisgICAgICAgIGlmIChzdGFydC5vZmZzZXQoKSA9PSBlbmQub2Zmc2V0KCkgJiYgaXNTdGFy dE9mUGFyYWdyYXBoKHN0YXJ0UG9zKSAmJiBpc0VuZE9mUGFyYWdyYXBoKHN0YXJ0UG9zKSkgewor ICAgICAgICAgICAgSFRNTEVsZW1lbnQgKmJsb2NrID0gc3RhdGljX2Nhc3Q8SFRNTEVsZW1lbnQg Kj4oc3RhcnQubm9kZSgpLT5lbmNsb3NpbmdCbG9ja0Zsb3dFbGVtZW50KCkpOworICAgICAgICAg ICAgcmVtb3ZlQ1NTU3R5bGUoc3R5bGUsIGJsb2NrKTsKKyAgICAgICAgICAgIGFkZEJsb2NrU3R5 bGVJZk5lZWRlZChzdHlsZSwgYmxvY2spOworICAgICAgICB9IGVsc2UKKyAgICAgICAgICAgIGFk ZElubGluZVN0eWxlSWZOZWVkZWQoc3R5bGUsIG5vZGUsIG5vZGUpOwogICAgIH0gZWxzZSB7CiAg ICAgICAgIHdoaWxlICgxKSB7Ci0gICAgICAgICAgICBpZiAobm9kZS0+Y2hpbGROb2RlQ291bnQo KSA9PSAwICYmIG5vZGUtPnJlbmRlcmVyKCkgJiYgbm9kZS0+cmVuZGVyZXIoKS0+aXNJbmxpbmUo KSkgeworICAgICAgICAgICAgVmlzaWJsZVBvc2l0aW9uIHBvcyhub2RlLCAwLCBET1dOU1RSRUFN KTsKKyAgICAgICAgICAgIGJvb2wgdHJ5SW5saW5lID0gdHJ1ZTsKKyAgICAgICAgICAgIGlmIChp c1N0YXJ0T2ZQYXJhZ3JhcGgocG9zKSkgeworICAgICAgICAgICAgICAgIEhUTUxFbGVtZW50ICpi bG9jayA9IHN0YXRpY19jYXN0PEhUTUxFbGVtZW50ICo+KG5vZGUtPmVuY2xvc2luZ0Jsb2NrRmxv d0VsZW1lbnQoKSk7CisgICAgICAgICAgICAgICAgCisgICAgICAgICAgICAgICAgTm9kZSAqbmV4 dCA9IGJsb2NrLT50cmF2ZXJzZU5leHRTaWJsaW5nKCk7CisgICAgICAgICAgICAgICAgCisgICAg ICAgICAgICAgICAgLy8gSWYgdGhlIGVuZCBvZiB0aGUgYmxvY2sgZmxvdyBpcyBiZWZvcmUgKG9y IGVxdWFsIHRvKSB0aGUgZW5kIG9mIHRoZSBzZWxlY3Rpb24sCisgICAgICAgICAgICAgICAgLy8g d2UgYXBwbHkgdGhlIHN0eWxlIHRvIHRoZSBibG9jayBhbmQgbm90IHRoZSBydW5zIHRvIGVuc3Vy ZSB0aGF0IHBhcmFncmFwaAorICAgICAgICAgICAgICAgIC8vIHN0eWxlcyByZW1haW4gaW50YWN0 IGFmdGVyIGRlbGV0aW5nIHRoZWlyIGNvbnRlbnRzCisgICAgICAgICAgICAgICAgaWYgKFJhbmdl Ojpjb21wYXJlQm91bmRhcnlQb2ludHMoUG9zaXRpb24obmV4dCwgMCksIGVuZCkgPD0gMCkgewor ICAgICAgICAgICAgICAgICAgICByZW1vdmVDU1NTdHlsZShzdHlsZSwgYmxvY2spOworICAgICAg ICAgICAgICAgICAgICBhZGRCbG9ja1N0eWxlSWZOZWVkZWQoc3R5bGUsIGJsb2NrKTsKKyAgICAg ICAgICAgICAgICAgICAgCisgICAgICAgICAgICAgICAgICAgIC8vIFdlIG5lZWQgdG8gdHJhdmVy c2UgYmFjayB0byB0aGUgcHJldmlvdXMgbm9kZSBhcyB3ZSBoYXZlIG9ubHkgYXBwbGllZCBzdHls ZXMKKyAgICAgICAgICAgICAgICAgICAgLy8gdG8gdGhlIGJsb2NrIGJlZm9yZSAnbmV4dCcsIG5v dCBuZXh0IGl0c2VsZi4KKyAgICAgICAgICAgICAgICAgICAgbm9kZSA9IG5leHQtPnRyYXZlcnNl UHJldmlvdXNOb2RlKCk7CisgICAgICAgICAgICAgICAgICAgIAorICAgICAgICAgICAgICAgICAg ICAvLyBXZSd2ZSBhbHJlYWR5IGFwcGxpZWQgdGhlIHN0eWxlLCBubyBuZWVkIHRvIGRvIGl0IGlu bGluZSBpbnN0ZWFkCisgICAgICAgICAgICAgICAgICAgIHRyeUlubGluZSA9IGZhbHNlOworICAg ICAgICAgICAgICAgIH0KKyAgICAgICAgICAgIH0gCisgICAgICAgICAgICAKKyAgICAgICAgICAg IGlmICh0cnlJbmxpbmUgJiYgbm9kZS0+Y2hpbGROb2RlQ291bnQoKSA9PSAwICYmIG5vZGUtPnJl bmRlcmVyKCkgJiYgbm9kZS0+cmVuZGVyZXIoKS0+aXNJbmxpbmUoKSkgewogICAgICAgICAgICAg ICAgIE5vZGUgKnJ1blN0YXJ0ID0gbm9kZTsKICAgICAgICAgICAgICAgICB3aGlsZSAoMSkgewog ICAgICAgICAgICAgICAgICAgICBOb2RlICpuZXh0ID0gbm9kZS0+dHJhdmVyc2VOZXh0Tm9kZSgp OwpAQCAtOTM5LDggKzk3MiwxMyBAQCBib29sIEFwcGx5U3R5bGVDb21tYW5kOjpub2RlRnVsbHlT ZWxlY3RlCiAgICAgQVNTRVJUKG5vZGUtPmlzRWxlbWVudE5vZGUoKSk7CiAKICAgICBQb3NpdGlv biBwb3MgPSBQb3NpdGlvbihub2RlLCBub2RlLT5jaGlsZE5vZGVDb3VudCgpKS51cHN0cmVhbSgp OwotICAgIHJldHVybiBSYW5nZTo6Y29tcGFyZUJvdW5kYXJ5UG9pbnRzKG5vZGUsIDAsIHN0YXJ0 Lm5vZGUoKSwgc3RhcnQub2Zmc2V0KCkpID49IDAgJiYKLSAgICAgICAgUmFuZ2U6OmNvbXBhcmVC b3VuZGFyeVBvaW50cyhwb3MsIGVuZCkgPD0gMDsKKyAgICAKKyAgICBpZiAobm9kZS0+aXNCbG9j a0Zsb3coKSkgeworICAgICAgICByZXR1cm4gUmFuZ2U6OmNvbXBhcmVCb3VuZGFyeVBvaW50cyhu b2RlLCAwLCBzdGFydC5ub2RlKCksIHN0YXJ0Lm9mZnNldCgpKSA+PSAwICYmCisgICAgICAgICAg ICBSYW5nZTo6Y29tcGFyZUJvdW5kYXJ5UG9pbnRzKHBvcywgZW5kKSA8IDA7CisgICAgfSBlbHNl CisgICAgICAgIHJldHVybiBSYW5nZTo6Y29tcGFyZUJvdW5kYXJ5UG9pbnRzKG5vZGUsIDAsIHN0 YXJ0Lm5vZGUoKSwgc3RhcnQub2Zmc2V0KCkpID49IDAgJiYKKyAgICAgICAgICAgIFJhbmdlOjpj b21wYXJlQm91bmRhcnlQb2ludHMocG9zLCBlbmQpIDw9IDA7CiB9CiAKIGJvb2wgQXBwbHlTdHls ZUNvbW1hbmQ6Om5vZGVGdWxseVVuc2VsZWN0ZWQoTm9kZSAqbm9kZSwgY29uc3QgUG9zaXRpb24g JnN0YXJ0LCBjb25zdCBQb3NpdGlvbiAmZW5kKSBjb25zdApAQCAtMTE5OCw2ICsxMjM2LDcgQEAg dm9pZCBBcHBseVN0eWxlQ29tbWFuZDo6YWRkQmxvY2tTdHlsZUlmTgogewogICAgIC8vIERvIG5v dCBjaGVjayBmb3IgbGVnYWN5IHN0eWxlcyBoZXJlLiBUaG9zZSBzdHlsZXMsIGxpa2UgPEI+IGFu ZCA8ST4sIG9ubHkgYXBwbHkgZm9yCiAgICAgLy8gaW5saW5lIGNvbnRlbnQuCisgICAgLy8gTm8g bG9uZ2VyIHRydWUuCiAgICAgaWYgKCFub2RlKQogICAgICAgICByZXR1cm47CiAgICAgCkBAIC0x MjE1LDYgKzEyNTQsMjkgQEAgdm9pZCBBcHBseVN0eWxlQ29tbWFuZDo6YWRkQmxvY2tTdHlsZUlm TgogICAgICAgICAgICAgY3NzVGV4dCArPSBkZWNsLT5jc3NUZXh0KCk7CiAgICAgICAgIHNldE5v ZGVBdHRyaWJ1dGUoYmxvY2ssIHN0eWxlQXR0ciwgY3NzVGV4dCk7CiAgICAgfQorCisgICAgaWYg KHN0eWxlQ2hhbmdlLmFwcGx5Rm9udENvbG9yKCkgfHwgc3R5bGVDaGFuZ2UuYXBwbHlGb250RmFj ZSgpIHx8IHN0eWxlQ2hhbmdlLmFwcGx5Rm9udFNpemUoKSB8fAorICAgICAgICBzdHlsZUNoYW5n ZS5hcHBseUJvbGQoKSB8fCBzdHlsZUNoYW5nZS5hcHBseUl0YWxpYygpKSB7CisgICAgICAgIGlm ICghYmxvY2stPmlubGluZVN0eWxlRGVjbCgpKQorICAgICAgICAgICAgYmxvY2stPmNyZWF0ZUlu bGluZVN0eWxlRGVjbCgpOworCisgICAgICAgIENTU011dGFibGVTdHlsZURlY2xhcmF0aW9uICpk ZWNsID0gYmxvY2stPmlubGluZVN0eWxlRGVjbCgpOworICAgICAgICAKKyAgICAgICAgaWYgKHN0 eWxlQ2hhbmdlLmFwcGx5Rm9udENvbG9yKCkpCisgICAgICAgICAgICBkZWNsLT5zZXRQcm9wZXJ0 eShDU1NfUFJPUF9DT0xPUiwgc3R5bGVDaGFuZ2UuZm9udENvbG9yKCkpOworICAgICAgICAgICAg CisgICAgICAgIGlmIChzdHlsZUNoYW5nZS5hcHBseUZvbnRGYWNlKCkpCisgICAgICAgICAgICBk ZWNsLT5zZXRQcm9wZXJ0eShDU1NfUFJPUF9GT05UX0ZBTUlMWSwgc3R5bGVDaGFuZ2UuZm9udEZh Y2UoKSk7CisgICAgICAgICAgICAKKyAgICAgICAgaWYgKHN0eWxlQ2hhbmdlLmFwcGx5Rm9udFNp emUoKSkKKyAgICAgICAgICAgIGRlY2wtPnNldFByb3BlcnR5KENTU19QUk9QX0ZPTlRfU0laRSwg c3R5bGVDaGFuZ2UuZm9udFNpemUoKSk7CisgICAgCisgICAgICAgIGlmIChzdHlsZUNoYW5nZS5h cHBseUJvbGQoKSkKKyAgICAgICAgICAgIGRlY2wtPnNldFByb3BlcnR5KENTU19QUk9QX0ZPTlRf V0VJR0hULCAiYm9sZCIpOworICAgICAgICAKKyAgICAgICAgaWYgKHN0eWxlQ2hhbmdlLmFwcGx5 SXRhbGljKCkpCisgICAgICAgICAgICBkZWNsLT5zZXRQcm9wZXJ0eShDU1NfUFJPUF9GT05UX1NU WUxFLCAiaXRhbGljIik7CisgICAgfQogfQogCiB2b2lkIEFwcGx5U3R5bGVDb21tYW5kOjphZGRJ bmxpbmVTdHlsZUlmTmVlZGVkKENTU011dGFibGVTdHlsZURlY2xhcmF0aW9uICpzdHlsZSwgTm9k ZSAqc3RhcnROb2RlLCBOb2RlICplbmROb2RlKQo= 7528 2006-04-05 09:20:45 -0700 2006-04-05 09:20:45 -0700 testcase 7360-testcase.html text/html 1078 justin.garcia PHByZT5UaGlzIGlzIGFuIGF1dG9tYXRlZCB0ZXN0Y2FzZSB0aGF0IGRlbW9uc3RyYXRlcyB0aGlz IGJ1ZzoKVHlwZSB0aHJlZSBsaW5lcyBvZiB0ZXh0CkJvbGQgdGhlbSBhbGwKVW5ib2xkIHRoZSBt aWRkbGUgbGluZQpEZWxldGUgdGhlIHRleHQgb24gdGhlIG1pZGRsZSBsaW5lCkNsaWNrIGF3YXkg YW5kIGNsaWNrIGJhY2sgdG8gdGhlIG1pZGRsZSBsaW5lClN0YXJ0IHR5cGluZwoKWW91IHNob3Vs ZCBzZWUgYm9sZCB0ZXh0LiAgWW91IGluc3RlYWQgc2VlIHVuYm9sZCB0ZXh0LjwvcHJlPgoKPGRp diBpZD0idGVzdCIgY29udGVudGVkaXRhYmxlPSJ0cnVlIj48L2Rpdj4KCjxzY3JpcHQ+CnZhciBz ID0gd2luZG93LmdldFNlbGVjdGlvbigpOwp2YXIgZSA9IGRvY3VtZW50LmdldEVsZW1lbnRCeUlk KCJ0ZXN0Iik7CnMuc2V0UG9zaXRpb24oZSwgMCk7CmRvY3VtZW50LmV4ZWNDb21tYW5kKCJJbnNl cnRUZXh0IiwgZmFsc2UsICJmb28iKTsKZG9jdW1lbnQuZXhlY0NvbW1hbmQoIkluc2VydFBhcmFn cmFwaCIpOwpkb2N1bWVudC5leGVjQ29tbWFuZCgiSW5zZXJ0VGV4dCIsIGZhbHNlLCAiYmFyIik7 CmRvY3VtZW50LmV4ZWNDb21tYW5kKCJJbnNlcnRQYXJhZ3JhcGgiKTsKZG9jdW1lbnQuZXhlY0Nv bW1hbmQoIkluc2VydFRleHQiLCBmYWxzZSwgImJhciIpOwpkb2N1bWVudC5leGVjQ29tbWFuZCgi U2VsZWN0QWxsIik7CmRvY3VtZW50LmV4ZWNDb21tYW5kKCJCb2xkIik7CnMubW9kaWZ5KCJtb3Zl IiwgImJhY2t3YXJkIiwgImNoYXJhY3RlciIpOwpzLm1vZGlmeSgibW92ZSIsICJmb3J3YXJkIiwg InBhcmFncmFwaCIpOwpzLm1vZGlmeSgiZXh0ZW5kIiwgImZvcndhcmQiLCAid29yZCIpOwpkb2N1 bWVudC5leGVjQ29tbWFuZCgiQm9sZCIpOwpkb2N1bWVudC5leGVjQ29tbWFuZCgiRGVsZXRlIik7 CnMubW9kaWZ5KCJtb3ZlIiwgImJhY2t3YXJkIiwgImNoYXJhY3RlciIpOwpzLm1vZGlmeSgibW92 ZSIsICJmb3J3YXJkIiwgImNoYXJhY3RlciIpOwpkb2N1bWVudC5leGVjQ29tbWFuZCgiSW5zZXJ0 VGV4dCIsIGZhbHNlLCAidGhpcyB0ZXh0IHNob3VsZCBiZSBib2xkIik7Cjwvc2NyaXB0Pg==