Apache OpenOffice (AOO) Bugzilla – Issue 43190
FAQ xfer: bad output of Korean/special characters
Last modified: 2017-05-20 09:27:33 UTC
Hi Scott, Your FAQs look nice! :) Korean characters used in the hyperlink in section 3b are changed on output side of script. Source is at OOoAuthors.
something weird with characters may be happening here too, I think. If not, cuss me out and tell me to open a separate issue. :) http://scott.progbits.com/doctest/Writer/faqwriterspellcheck/263.html
More of the same character oddities: http://scott.progbits.com/doctest/Writer/faqwriterprinting/257.html At the Reveal Codes question in this TOC: http://scott.progbits.com/doctest/Writer/faqwriterother/ http://scott.progbits.com/doctest/Writer/faqwriterother/245.html
another: http://scott.progbits.com/doctest/Impress/Printing/287.html
At OOoAuthors, removed quotation marks from the answer: http://oooauthors.org/FAQs/Writer/faqwriterspellcheck/263 and http://oooauthors.org/FAQs/Writer/faqwriterprinting/257 and http://oooauthors.org/FAQs/Impress/Printing/287 At OOoAuthors, removed quotation marks from the question: http://oooauthors.org/FAQs/Writer/faqwriterother/245 This is likely a UTF-8 vs 8859-1 character set issue. In early OOoAuthors days, quotation marks often caused issues, until something in the Plone settings were changed. (I'm not sure that it was ever clear what changed: it may have been a magical fix that occurred during one of the upgrades.) On these FAQs, they may have been fixed if the quotation marks were simply replaced with new quotation marks. I purged the quotation marks instead, just as personal preference and hopefully for less xfer issues. Is the spider set up to output in UTF-8 or ISO 8859-1? I'm guessing that a UTF-8 char set will address the Korean characters issue.
1. Updated URL field (above) to point to new location of faq spider's output file. 2. Issues at all other URLS mentioned within have been resolved. 3. Changed component to user-faq project. Reassigned issue qa contact to the user-faq project.
man perlunicode may be helpful...
Manually replaced the Korean characters in the FAQ file located at the user-faq project: http://user-faq.openoffice.org/new-faq/20040314a.html so that Korean characters show correctly in bluefish and firefox html editor and browser used on my local machine. This did not seem to fix how it diplays on the OOo user-faq site, so perhaps the character mars are occurring during the file xfer from my machine to the OOo CVS? Assigning issue to myself; will follow up on analysis to try to isolate where the error occurs in the faq xfer process.
This is interesting. An XHTML 1.0 Transitional DTD and head tags have been added to the HTML file for this FAQ, to define the char set of the html page as UTF-8. This caused the Korean characters in the FAQ to display correctly. But look here: http://user-faq.openoffice.org/source/browse/user-faq/www/new-faq/20040314a.html?rev=1.7&content-type=text/vnd.viewcvs-markup and see the characters shown in a different format in the code! Neat! This seems to display more insight into how character transfers change via the various stages of the file xfer. Leaving this FAQ as is in CVS with the DTD and head tags in place (which is necessary for proper display) will leave loose ends that may hose us in future runs of the faq_spider. I believe this will help us learn the answers to two questions: 1. How will the process work, when the file in CVS differs from the file output from the faq_spider? We need to learn this to understand how we can allow folks to edit via CVS at OOo or via Plone at OOoAuthors, so this will be an excellent experiment to help us understand more about how the two interfaces will mesh with each other. Note that if we attempt to place the head tags into the answer at OOoAuthors, the question portion of the FAQ may not display with the answer as expected, when viewed through the OOoAuthors site (unless a mistake was made in past experiments). 2. Will the faq_spider overwrite the repaired Korean characters in the repaired HTML file? If it does, because it displays correctly at the OOoAuthors site, we will learn that we have a character issue occuring on that front end of the transfer, as well. I hope we can post further details here, so that we have records about char set issues exposed by this experimental file xfer process.
That NOTE should read, "Note that if we attempt to place the head tags into the answer at OOoAuthors, the question portion of the FAQ may not display with the answer as expected, when viewed through the *User-FAQ* site (unless a mistake was made in past experiments)."
Reset assignee on issues not touched by assignee in more than 1000 days.
obsolete