On 2013-10-03, Scott Kostyshak wrote:
> On Thu, Oct 3, 2013 at 4:28 PM, Guenter Milde <[email protected]> wrote:
>> On 2013-10-03, Scott Kostyshak wrote:
>>> On Thu, Oct 3, 2013 at 5:02 AM, Guenter Milde <[email protected]> wrote:
>>>> On 2013-10-03, Scott Kostyshak wrote:
>>>>> Thanks for the confirmation. I started a thread on lyx-devel.
>>>>> Hopefully there's an easy way to have pdfTeX and XeTeX working. Note
>>>>> that if I remove the following lines from the preamble, export with
>>>>> XeTeX
>>>>> works fine (but pdfTeX is broken):
>>>>> \DeclareInputText{176}{\ifmmode^\circ\else\textdegree\fi}
>>>>> \DeclareInputText{183}{\ifmmode\cdot\else\textperiodcentered\fi}
>>>> There is a simple fix: move the offending lines into the
>>>> \ifpdf
>>>> \fi
>>>> block.
>>>> However, the right way is to remove the lines as well as the section
>>>> in the guide, because it is no longer true. Since several versions,
>>>> LYXDIR/unicodesymbols contains the required definitions for text and
>>>> math mode.
>>> I prefer the right way. So you suggest to remove the two
>>> DeclareInputText commands and to remove all of the section
>>> "Miscellaneous special Characters" ?
>>> After I do that, I get errors because of the € symbol. After I remove
>>> those occurrences, XeTeX and pdfTeX both export fine. There is an
>>> entry in unicodesymbols for the euro. Is this a different euro?
>> The \texteuro macro is provided by the "textcomp" package.
>> A quick and dirty fix would be to set the "force" flag so that the textcomp
>> package is auto-loaded. However, this is an overkill as this would also
>> replace the Euro (€) character with the \texteuro macro which is not
>> required for the TeX input encodings
>> utf8, ansinew, cp1250, cp1252, cp1257, cp858, latin10, and latin9
>> A "force-preamble" flag that only forces loading the preamble code without
>> forcing the replacement could solve the issue. For the "textcomp" feature
>> this regards the characters
>> # recognized with 'utf8', if textcomp is loaded
>> textcomp = {
>> # Latin-1 Supplement
>> 0x00a2: ur'\textcent{}', # ¢ CENT SIGN
>> 0x00a4: ur'\textcurrency{}', # ¤ CURRENCY SYMBOL
>> 0x00a5: ur'\textyen{}', # ¥ YEN SIGN
>> 0x00a6: ur'\textbrokenbar{}', # ¦ BROKEN BAR
>> 0x00a7: ur'\textsection{}', # § SECTION SIGN
>> 0x00a8: ur'\textasciidieresis{}', # ¨ DIAERESIS
>> 0x00a9: ur'\textcopyright{}', # © COPYRIGHT SIGN
>> 0x00aa: ur'\textordfeminine{}', # ª FEMININE ORDINAL INDICATOR
>> 0x00ac: ur'\textlnot{}', # ¬ NOT SIGN
>> 0x00ae: ur'\textregistered{}', # ® REGISTERED SIGN
>> 0x00af: ur'\textasciimacron{}', # ¯ MACRON
>> 0x00b0: ur'\textdegree{}', # ° DEGREE SIGN
>> 0x00b1: ur'\textpm{}', # ± PLUS-MINUS SIGN
>> 0x00b2: ur'\texttwosuperior{}', # ² SUPERSCRIPT TWO
>> 0x00b3: ur'\textthreesuperior{}', # ³ SUPERSCRIPT THREE
>> 0x00b4: ur'\textasciiacute{}', # ´ ACUTE ACCENT
>> 0x00b5: ur'\textmu{}', # µ MICRO SIGN
>> 0x00b6: ur'\textparagraph{}', # ¶ PILCROW SIGN # != \textpilcrow
>> 0x00b9: ur'\textonesuperior{}', # ¹ SUPERSCRIPT ONE
>> 0x00ba: ur'\textordmasculine{}', # º MASCULINE ORDINAL INDICATOR
>> 0x00bc: ur'\textonequarter{}', # 1/4 FRACTION
>> 0x00bd: ur'\textonehalf{}', # 1/2 FRACTION
>> 0x00be: ur'\textthreequarters{}', # 3/4 FRACTION
>> 0x00d7: ur'\texttimes{}', # × MULTIPLICATION SIGN
>> 0x00f7: ur'\textdiv{}', # ÷ DIVISION SIGN
>> # others
>> 0x0192: ur'\textflorin{}', # LATIN SMALL LETTER F WITH HOOK
>> 0x02b9: ur'\textasciiacute{}', # MODIFIER LETTER PRIME
>> 0x02ba: ur'\textacutedbl{}', # MODIFIER LETTER DOUBLE PRIME
>> 0x2016: ur'\textbardbl{}', # DOUBLE VERTICAL LINE
>> 0x2022: ur'\textbullet{}', # BULLET
>> 0x2032: ur'\textasciiacute{}', # PRIME
>> 0x2033: ur'\textacutedbl{}', # DOUBLE PRIME
>> 0x2035: ur'\textasciigrave{}', # REVERSED PRIME
>> 0x2036: ur'\textgravedbl{}', # REVERSED DOUBLE PRIME
>> 0x203b: ur'\textreferencemark{}', # REFERENCE MARK
>> 0x203d: ur'\textinterrobang{}', # INTERROBANG
>> 0x2044: ur'\textfractionsolidus{}', # FRACTION SLASH
>> 0x2045: ur'\textlquill{}', # LEFT SQUARE BRACKET WITH QUILL
>> 0x2046: ur'\textrquill{}', # RIGHT SQUARE BRACKET WITH QUILL
>> 0x2052: ur'\textdiscount{}', # COMMERCIAL MINUS SIGN
>> 0x20a1: ur'\textcolonmonetary{}', # COLON SIGN
>> 0x20a3: ur'\textfrenchfranc{}', # FRENCH FRANC SIGN
>> 0x20a4: ur'\textlira{}', # LIRA SIGN
>> 0x20a6: ur'\textnaira{}', # NAIRA SIGN
>> 0x20a9: ur'\textwon{}', # WON SIGN
>> 0x20ab: ur'\textdong{}', # DONG SIGN
>> 0x20ac: ur'\texteuro{}', # EURO SIGN
>> 0x20b1: ur'\textpeso{}', # PESO SIGN
>> 0x20b2: ur'\textguarani{}', # GUARANI SIGN
>> 0x2103: ur'\textcelsius{}', # DEGREE CELSIUS
>> 0x2116: ur'\textnumero{}', # NUMERO SIGN
>> 0x2117: ur'\textcircledP{}', # SOUND RECORDING COYRIGHT
>> 0x211e: ur'\textrecipe{}', # PRESCRIPTION TAKE
>> 0x2120: ur'\textservicemark{}', # SERVICE MARK
>> 0x2122: ur'\texttrademark{}', # TRADE MARK SIGN
>> 0x2126: ur'\textohm{}', # OHM SIGN
>> 0x2127: ur'\textmho{}', # INVERTED OHM SIGN
>> 0x212e: ur'\textestimated{}', # ESTIMATED SYMBOL
>> 0x2190: ur'\textleftarrow{}', # LEFTWARDS ARROW
>> 0x2191: ur'\textuparrow{}', # UPWARDS ARROW
>> 0x2192: ur'\textrightarrow{}', # RIGHTWARDS ARROW
>> 0x2193: ur'\textdownarrow{}', # DOWNWARDS ARROW
>> 0x2212: ur'\textminus{}', # MINUS SIGN
>> 0x2217: ur'\textasteriskcentered{}', # ASTERISK OPERATOR
>> 0x221a: ur'\textsurd{}', # SQUARE ROOT
>> 0x2422: ur'\textblank{}', # BLANK SYMBOL
>> 0x25e6: ur'\textopenbullet{}', # WHITE BULLET
>> 0x25ef: ur'\textbigcircle{}', # LARGE CIRCLE
>> 0x266a: ur'\textmusicalnote{}', # EIGHTH NOTE
>> 0x26ad: ur'\textmarried{}', # MARRIAGE SYMBOL
>> 0x26ae: ur'\textdivorced{}', # DIVORCE SYMBOL
>> 0x27e8: ur'\textlangle{}', # MATHEMATICAL LEFT ANGLE BRACKET
>> 0x27e9: ur'\textrangle{}', # MATHEMATICAL RIGHT ANGLE BRACKET
>> }
>> Many of these already carry the "force" flag in unicodesymbols.
> Thanks Günter
> Attached is a patch that implements your previous suggestions. Can you
> check that it does what you had in mind? If so, I will make the same
> changes for Math.lyx in the other languages.
I thought again about it: we should mention in a suitable way that the »^« and
»_« keys (as well as »{«) are "active" which means that you cannot use them
to insert the ^_{ characters (although "unicodesymbols" will do the right
conversion if the characters are pasted into a math box, say).
> This is an area I know little about. Are you able and willing to
> provide a patch to fix the euro symbol?
The "quick and dirty fix" for the euro would be just
--- /home/milde/.lyx/unicodesymbols~ 2012-05-14 11:56:24.000000000 +0200
+++ /tmp/unicodesymbols10333.3565 2013-10-06 17:52:33.000000000 +0200
@@ -1640,7 +1640,7 @@
0x20a9 "\\textwon" "textcomp" "" # WON SIGN
#0x20aa "" "" "" # NEW SHEQEL SIGN
0x20ab "\\textdong" "textcomp" "" # DONG SIGN
-0x20ac "\\texteuro" "textcomp" "" # EURO SIGN
+0x20ac "\\texteuro" "textcomp" "force" # EURO SIGN
0x20b1 "\\textpeso" "textcomp" "" # PESO SIGN
0x20b2 "\\textguarani" "textcomp" "" # GUARANI SIGN
0x20b9 "\\rupee" "tfrupee" "" # RUPEE SIGN
A better fix would look for all the abovelisted characters and a real fix
would implement a new flag to force the "textpreamble" without conversing
the literal character for all teh abovelisted characters.
I am not able to do this one.
> These workarounds are no longer necessary because of unicodesymbols
> and further they break compilation with XeTeX.
... and pdfTeX with TeX input encoding set to ascii or utf8.
Thanks,
Günter
