{ "WorkItem": { "AffectedComponent": { "Name": "", "DisplayName": "" }, "ClosedComment": "Fixed in v1.7 prelim. DomZ, please test.", "ClosedDate": "2008-11-02T19:23:15.477-08:00", "CommentCount": 0, "Custom": null, "Description": "Hi Cheeso,\n \nI done some tests will the lastest version it seems that's something is broken with the UseUnicode property.\nI still use the same use cases defined in workitem #6410.\n \nYou describe this property like that :\nIndicates whether to encode entry filenames and entry comments using Unicode (UTF-8) according to the PKWare specification, for those filenames and comments that cannot be encoded in the ASCII character set. \n \nI think there is an error here, you should use UTF-8 only when the entry name contain a char which is not exist in IBM437 charset (IBM437 and ASCII are different charsets).\nIBM437 contains more chars than ASCII.\n \nThanks", "LastUpdatedDate": "2013-05-16T05:32:36.217-07:00", "PlannedForRelease": "", "ReleaseVisibleToPublic": false, "Priority": { "Name": "Low", "Severity": 50, "Id": 1 }, "ProjectName": "DotNetZip", "ReportedDate": "2008-10-28T13:34:48.367-07:00", "Status": { "Name": "Closed", "Id": 4 }, "ReasonClosed": { "Name": "Unassigned" }, "Summary": "UseUnicode not Working as expected", "Type": { "Name": "Issue", "Id": 3 }, "VoteCount": 1, "Id": 6513 }, "FileAttachments": [ { "FileId": 1516, "FileName": "WinFormsApp.zip", "DownloadUrl": ".\\1516" } ], "Comments": [ { "Message": "I have modified the code of WinForms example to handle UseUnicode property, so you can make some tests easily.", "PostedDate": "2008-10-29T03:52:30.887-07:00", "Id": -2147483648 }, { "Message": "", "PostedDate": "2008-10-29T03:52:34.247-07:00", "Id": -2147483648 }, { "Message": "ok, I will have a look.", "PostedDate": "2008-10-29T15:14:30.147-07:00", "Id": -2147483648 }, { "Message": "When you say \"Something is broken\" - what specifically do you see? I still have the \"test cases\" from 6410, but when I ran the tests (zipping and unzipping those \"original files\" directories), everything seemed to work. What problem are you experiencing, specifically? \r\n\r\nI agree with your comment about ASCII versus IBM437. That is a doc error. I have now fixed that in the documentation. Do you observe a runtime behavior that is unexpected? ", "PostedDate": "2008-10-30T14:10:12.973-07:00", "Id": -2147483648 }, { "Message": "I don't understand your comment about the WinForms app. The WinForms app does not use the UseUnicode property. Instead it uses the Encoding property. \r\n\r\nCan you supply a brief test case that reproduces the problem you have observed? ", "PostedDate": "2008-10-30T14:17:19.42-07:00", "Id": -2147483648 }, { "Message": "Hi Cheeso,\r\n\r\nThe description of the Encoding property is not clear, I'm not sure if when using this property every entries are encoded using the choosen Encoding or only entries with chars that are not in IBM437 charset ...\r\n\r\nUseUnicode is well explained and this is exactly what I want (if I understand well, the use of this property produce fully compliant zip files).\r\n\r\nSo if the property Encoding with UTF-8 should produce exactly the same result as UseUnicode, my WinForms modifications are useless :)\r\n\r\nThanks", "PostedDate": "2008-10-31T02:17:43.663-07:00", "Id": -2147483648 }, { "Message": "In reply to :\r\n\"When you say \"Something is broken\" - what specifically do you see? I still have the \"test cases\" from 6410, but when I ran the tests (zipping and unzipping those \"original files\" directories), everything seemed to work. What problem are you experiencing, specifically? \r\n\r\nI agree with your comment about ASCII versus IBM437. That is a doc error. I have now fixed that in the documentation. Do you observe a runtime behavior that is unexpected?\"\r\n\r\nI observe that entries that contains accents (like é, à, ...) are encoded in UTF-8 and should not because these accents are handled in the IBM437 charset.\r\nFor testing I use the Windows integrated Zip archiver (which is not compatible with UTF-8).\r\n\r\nIn the test cases 6410, if you zip the ansi original you should produce a zip that Windows XP should be extract with correct output (keeping accentuation).\r\n\r\nThanks", "PostedDate": "2008-10-31T02:24:45.207-07:00", "Id": -2147483648 }, { "Message": "Ok, I see what you mean. I reproduced the behavior you described UTF8 was being used even when unnecessary. I think this is not a disaster, but it is also not optimal. \r\n\r\nThere is an ambiguity in the documented (Defined) behavior, and also there is a sloppiness in the code, for example around the names of the properties. I have changed the code to use IBM437 when possible, and then use the alternate encoding (either UTF8 or some other code page) when the IBM437 encoding does not work. This requires some mods to the property names, so that the names reflect the true meaning of the property: UseUnicode becomes UseUnicodeAsNecessary. Encoding becomes ProvisionalAlternateEncoding. But changes in behavior and changes in property names will break existing apps - so in both cases I need to put this into a new version of the library. \r\n\r\n", "PostedDate": "2008-11-02T00:17:28.497-07:00", "Id": -2147483648 }, { "Message": "", "PostedDate": "2008-11-02T19:23:15.477-08:00", "Id": -2147483648 }, { "Message": "Thanks Cheeso,\r\n\r\nGreat news !\r\n\r\nCan you provide me a beta test please ?\r\n\r\nThanks", "PostedDate": "2008-11-03T00:55:59.267-08:00", "Id": -2147483648 }, { "Message": "This is perfect Cheeso !\r\n", "PostedDate": "2008-11-03T08:14:31.397-08:00", "Id": -2147483648 }, { "Message": "", "PostedDate": "2013-02-21T18:44:37.57-08:00", "Id": -2147483648 }, { "Message": "", "PostedDate": "2013-05-16T05:32:36.217-07:00", "Id": -2147483648 } ] }