Apache OpenOffice (AOO) Bugzilla – Issue 21593
Bad html to e-mail export
Last modified: 2013-08-07 14:38:26 UTC
In case of setting any charset in HTML compatibility (e.g.WIN-1250/WIN latin 2) is every time generated the same HTML Head in "send as e-mail" with 8859-2 charset. But the text are coded in selected charset. The head is following: <META http-equiv=Content-Type content="text/html; charset=iso-8859-2" You may write in Czech, English or German language. Walter
moving to Word Processor
Reassign issue to owner of selected subcomponent
Hi Waltercz, I believe I watched similar problems for website creation, but I was not able to figure out the problem exactly (insufficient knowledge concerning charsets) - so I need your help. Can you please attach some (very!) short documents to show some details? Helpful would be - source document (.sxc), may be with only 1 sentence - document transferred to html with some short explications concerning charset - shown in header 8visible for everyone) - in fact (with some hints where in the text the actual charset can be proved like "'ê' only exists in iso - xxx") Do you see the same problems in CALC - documents saved as HTML ? Do you agree that something like "charset indicated in html document header not in accordance to actual charset in the document) might be a more accurate summary for this issue? Thanks Rainer
Issues that might be related to this one: issue 4453, issue 9128, issue 10373, issue 17037, issue 19514 Rainer
We will duscuss this in [qa-dev] Rainer
In OO.o 680 I have charset as utf-8 (Tools-->Options-->Load/Save-->HTML Compatibility). Using View-->HTML-Source, headers show the correct charset (<META HTTP-EQUIV="CONTENT-TYPE" CONTENT="text/html; charset=utf-8">. Have not yet tested in 1.1.0...
Just repeated the same steps in 1.1.0, with the same results, the header still shows utf-8 as charset... This is all under Linux, perhaps the problem is only under Windows?
I have win98 and live in the UK, on OOo 1.1.0 My defult Character Set for HTML Compatability is Western Europe (Windows-1252/Winlatin 1) And I have no problems with bad coding on this charset, or unicode UTF-8.
When the defult Charset is changed between UTF-8, ISO8859/1 and Windows- 1252/Winlatin 1, the tags in the exported html refer to these. I have no problems with bad tags on my set up.
I was not able to reproduce the problem, too. Reporter, please attach a very detailled step by step instruction and also source- and email document and allso other information that might be useful for further clarification. If I will not see any further action as votes, attachments or confirmations in this issue, I will have to close this issue 2004-01-15 as WFM. CU Rainer
Rainer et al... I suspect the reporter may be Czech (since he uses "cz" as part of his username and refers to teh Czech lnaguage) Could this bug be specific to either locale seetings, or to a specific localiization? I would be very interested to know whether the reporter is using a Czech localization or not. Perhaps this could be referred to the relevant project? I'll send a precis to Pavel Janik and ask if he can reproduce with the cs localization.
waltercz: which version do you use? Do you use 1.1.0cz or original English version? Could you please only try export as HTML instead of e-mail? Your e-mail program can convert the HTML page... I can not reproduce it with original English version of 1.1.0 nor with Czech versions. The resulting HTML page is always encoded as specified in Tools / Options / Loading/Saving / HTML compatibility / Character set.
Created attachment 12147 [details] Source HTML file
Created attachment 12148 [details] O.O setting
Created attachment 12149 [details] Generated e-mail
Dear Reiner, Sorry for late reply, because I was very bussy. You may find the following attachments: Source - source html file Output - an e-mail output seting - O.O. setting (gif file) The e-mail source is following (charset=iso-8859-2, but text are coded in Windows 1250): ----------- <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN"> <HTML><HEAD><TITLE></TITLE> <META http-equiv=Content-Type content="text/html; charset=iso-8859-2"> <META content="MSHTML 5.50.4134.100" name=GENERATOR> <META content="Walter Müller" name=AUTHOR> <META content=20031228;17434215 name=CREATED> <META content="Walter Müller" name=CHANGEDBY> <META content=20031228;18093215 name=CHANGED> <STYLE></STYLE> </HEAD> <BODY lang=cs-CZ dir=ltr bgColor=#ffffff> <DIV> </DIV> <DIV style="FONT: 10pt arial">----- Original Message ----- <DIV style="BACKGROUND: #e4e4e4; font-color: black"><B>From:</B> <A title=muller@waj.cz href="mailto:muller@waj.cz">Walter Müller</A> </DIV> <DIV><B>Sent:</B> Sunday, December 28, 2003 6:10 PM</DIV> <DIV><B>Subject:</B> Output</DIV></DIV> <DIV><BR></DIV> <DIV> </DIV> <P>PÅ™ÃÅ¡ernÄ› žluÅ¥ouÄký kůň pÄ›l Äábelské ódy.</P> <P><BR><BR></P> <P><BR><BR></P></BODY></HTML> ------------------- (System : Windows ME, OO 1.1.0) Regards Walter Müller
JA->ES: reassigned to you. please have a look at it...
es: czech version uses one specific patch that could make a difference here. See ftp://ftp.linux.cz/pub/localization/OpenOffice.org/devel/build/Patches/OOo_1.1.1_source-do-not-export-chars-as-html-entities.diff waltercz: could you please test original English version too?
I checked with 1.1.0 German version WIN98SE: 645m19(Build8693): 0. Start OOo WRITER 1. change settings in accordance to 'Setting.gif'(I only changed charset to 'windows-1250_) 2. open new html-document 3. open attachment 'source.html' 4. Menu: 'File - send document - as email a new email document opened in MOZILLA 1.6b 5. Save email as draft 6. goto email folder "drafts" 7. look for saved email 8. rightclick on attached html-file, 'save' 9. open saved html-file in OOo (from webpage- UI) 10. change to view "html-source" for just opened file 11. goto OOo- window with source file 12. change to 'source view', too 13. compare source expected: same in both windows actual: changed lines: <META NAME="CHANGEDBY" CONTENT="Rainer Bielefeld"> <META NAME="CHANGED" CONTENT="20031229;8133763"> <P>PÅ™íšernÄ› žluÅ¥ouÄký kůň pÄ›l ábelské ódy.</P> These changes are not so extreme like in the attached 'Output e-mail.eml' (and who can know, what MOZILLA, OE and 'eml-export' will do with the files), but it seems that a discussion of these results might be useful. My OOo will always masque some fonts in the body, when I save 'source.html' Reporter, can you please: contribute an email attachment in html created in accordance to my 5 ... 8? I Rainer
Tested this with Mozilla 1.5 and Hebrew OpenOffice 1.1 on Linux, with a simple Hebrew file (one line). Local is set to Hebrew, HTML charset set to windows-1255. Compared both files in a text editor (katoob): the email file had an extra line at the top, with the charset declaration repeated (outside the HEAD element, which is wrong). In OOo Write: When viewing the source, for some reason the Hebrew in the Origianl version displayed incurrectly as high ansi characters, while the version in the email displays properly. Very strange.
Created attachment 12170 [details] original Hebrew HTML
Created attachment 12171 [details] emailed Hebrew HTML
Hi Shoshannah, I saw your funny email file, but I can not reproduce the effect with 1.1.0 German version WIN98SE: 645m19(Build8693) and Mozilla/5.0 (Windows; U; Win98; de-AT; rv:1.6b) Gecko/20031208 My exported email file from 1.html looks quite normal. But it seems that there _are_ several difficulties, even if they are not really reproducible. So: NEW Rainer
I can not reproduce the problem on Win98 OOo1.1 English version. Also, though I think it is correct, is the subcomponent right above?
Can someone reproduce anything anywhere now in a current version? If yes, please detail if the problem can be reproduced with different mailers. I also think it's a framework problem (IF it is an OOo problem). Else I'll close WEM
Hi es! I have the Czech version 1.1.1 - the problems are remaining. If there is any newest version (e.g. beta), please send me link for download, I will check it. Regards Walter
Current version is 1.1.4. Test vesrion is src680m66 Please have a look at it and if you have the same problem, reassign to 'tra'
closed