Ok, more or less found a partial solution. (but anybody with experience in a multilingual world map and labels, please let me know)
Using some python from https://unix.stackexchange.com/questions/247108/how-to-find-out-which-unicode-codepoints-are-defined-in-a-ttf-file I'm now able to find in which ttf font a missing glyph can be found... Using the following in my fonts.lst: sc fonts/DejaVuSansCondensed.ttf scb fonts/DejaVuSansCondensed-Bold.ttf noto fonts/NotoSans-Condensed.ttf notob fonts/NotoSans-CondensedBold.ttf uni fonts/unifont_sample.ttf unib fonts/unifont_sample.ttf tw fonts/UKIJCJK.ttf And using FONT "sc,noto,uni,tw" # does order matter? I have a start to show labels in my small test setup with combined data from Pakistan, Taiwan and Afghanistan... I still have other log messages now: AH01215: Unable to load glyph 62314 for font "uni". Using ? as fallback.: /usr/lib/cgi-bin/mapserv [Fri Apr 14 10:12:20.644638 2023] [cgi:error] [pid 141599] [client ::1:46140] AH01215: msGetGlyphByIndex(): General error message. unable to load glyph 62314 for font "uni": /usr/lib/cgi-bin/mapserv while my script tells me 'glyph 62314' is in 'fonts/unifont_sample.ttf'... Any info or insights are still being appreciated :-) Regards, Richard Duivenvoorde On 4/14/23 09:18, Richard Duivenvoorde via MapServer-users wrote:
Greetings List, Some years ago we used https://github.com/MapServer/basemaps/ to create a raw world map based on OSM data (on hindsight, that map already showed 'Tofu'[0] characters. Trying to refresh this setup, we want multilingual/multiline labels (so the English label/name below the local Locale label/name, IF available).. Using another OSM setup, I used Google Noto-fonts to have all glyphs/fonts/charactersets needed... that worked great But now using basemaps: if I load Pakistan, Taiwanese and Afghanistan data, and point to NotoSans-CondensedBold.ttf NotoSans-Condensed.ttf in my fonts.lst, I still get the three question marks in my Taiwan map... and mapserver telling me: [Thu Apr 13 19:52:52.364863 2023] [cgi:error] [pid 52200] [client ::1:37906] AH01215: Unable to find glyph for codepoint 37129. Using ? as fallback.: /usr/lib/cgi-bin/mapserv [Thu Apr 13 19:52:52.364878 2023] [cgi:error] [pid 52200] [client ::1:37906] AH01215: Unable to find glyph for codepoint 29509. Using ? as fallback.: /usr/lib/cgi-bin/mapserv [Thu Apr 13 19:52:52.364893 2023] [cgi:error] [pid 52200] [client ::1:37906] AH01215: Unable to find glyph for codepoint 28525. Using ? as fallback.: /usr/lib/cgi-bin/mapserv [Thu Apr 13 19:52:52.364913 2023] [cgi:error] [pid 52200] [client ::1:37906] AH01215: Unable to find glyph for codepoint 37129. Using ? as fallback.: /usr/lib/cgi-bin/mapserv etc etc... So no I wonder: SHOULD it be possible to have a world map with all kind of different language names in one mapfile? OR (as in https://github.com/MapServer/basemaps/blob/main/fonts.lst is there just one font used for a label)? OR do I have to create a long list of needed fonts (so mapserver can choose from those and pick one which has that 'codepoint'. OR is defining a Noto ttf not enough? OR do I miss some point... Anybody here has some experience with such an issue: create a map with both Khmer, Arabic, Japanese, Ukranian, Chinese and Dutch place name labels :-) Any pointers appreciated, Regards, Richard Duivenvoorde PS reading about https://mapserver.org/development/rfc/ms-rfc-98.html#rfc98, there is some font-cache I can refresh... ??? [0] https://fonts.google.com/knowledge/glossary/tofu _______________________________________________ MapServer-users mailing list [email protected] https://lists.osgeo.org/mailman/listinfo/mapserver-users
_______________________________________________ MapServer-users mailing list [email protected] https://lists.osgeo.org/mailman/listinfo/mapserver-users
