Add glyphs to unicode reserved plane

[driusan]

I'm trying to use FiraCode w/ ligatures in a text editor I'm writing (where I'm currently using Deja Vu Sans Mono), but there doesn't seem to be any way to access the ligatures from Go, where the standard font drawing methods are based on converting unicode runes to glyphs and have no knowledge of ligatures.

The Unicode private use area planes is designed for this purpose, according to wikipedia, but doesn't seem to be used by Firacode (at least as far as I can tell by printing runes starting at 0x100000 or 0x0f0000 and visually scanning). Would it be possible to map the ligatures to the supplementary plan? If they were accessible as runes in a private use area with a documented mapping, I could manually do the context-sensitive mapping when rendering.

It's possible that I'm missing something, because all the "how to use the font" documentation seems to be user documentation, and I can't find any developer documentation..

[siegebell]

Providing the fancy glyphs via private use area (PUA) codes would resolve #42 via prettify-symbols-mode.

[siegebell]

@mordocai I believe this is what you're looking for.

[siegebell]

@mordocai Here's a workaround: I've extracted the glyphs that Fira Code adds into a separate font, and assigned each a Unicode ID in the private character area. Download the symbols font: FiraCode-Regular-Symbol.zip

This symbols font should be assigned as a fallback font for either Fira Code or Fira Mono because it does not define the regular characters.

Character map:

 www  \ue100     **   \ue101    ***  \ue102    **/  \ue103
  *>  \ue104     */   \ue105     \\  \ue106    \\\  \ue107
  {-  \ue108     []   \ue109     ::  \ue10a    :::  \ue10b
  :=  \ue10c     !!   \ue10d     !=  \ue10e    !==  \ue10f
  -}  \ue110     --   \ue111    ---  \ue112    -->  \ue113
  ->  \ue114    ->>   \ue115     -<  \ue116    -<<  \ue117
  -~  \ue118     #{   \ue119     #[  \ue11a     ##  \ue11b
 ###  \ue11c   ####   \ue11d     #(  \ue11e     #?  \ue11f
  #_  \ue120    #_(   \ue121     .-  \ue122     .=  \ue123
  ..  \ue124    ..<   \ue125    ...  \ue126     ?=  \ue127
  ??  \ue128     ;;   \ue129     /*  \ue12a    /**  \ue12b
  /=  \ue12c    /==   \ue12d     />  \ue12e     //  \ue12f
 ///  \ue130     &&   \ue131     ||  \ue132    ||=  \ue133
  |=  \ue134     |>   \ue135     ^=  \ue136     $>  \ue137
  ++  \ue138    +++   \ue139     +>  \ue13a     +>  \ue13a
 =:=  \ue13b     ==   \ue13c    ===  \ue13d    ==>  \ue13e
  =>  \ue13f    =>>   \ue140     <=  \ue141    =<<  \ue142
 =/=  \ue143     >-   \ue144     >=  \ue145    >=>  \ue146
  >>  \ue147    >>-   \ue148    >>=  \ue149    >>>  \ue14a
  <*  \ue14b    <*>   \ue14c     <|  \ue14d    <|>  \ue14e
  <$  \ue14f    <$>   \ue150   <!--  \ue151     <-  \ue152
 <--  \ue153    <->   \ue154     <+  \ue155    <+>  \ue156
  <=  \ue157    <==   \ue158    <=>  \ue159    <=<  \ue15a
  <>  \ue15b     <<   \ue15c    <<-  \ue15d    <<=  \ue15e
 <<<  \ue15f     <~   \ue160    <~~  \ue161     </  \ue162
 </>  \ue163     ~@   \ue164     ~-  \ue165     ~=  \ue166
  ~>  \ue167     ~~   \ue168    ~~>  \ue169     %%  \ue16a
   x  \ue16b      :   \ue16c      +  \ue16d      *  \ue16f

[siegebell]

And if you're using the prettify-symbols-mode extension for vscode, you can use the following substitution settings. (prettify-symbols-mode for Emacs should look similar-ish, but you'll combine the pre, ugly, and post into the same regex and \ue00a0 maybe can be replaced with a normal space.)

"prettifySymbolsMode.substitutions": [
  {
    "language": "*",
    "substitutions": [
      { "ugly": "\\{2}", "pretty": "\u03bb", "post": "\\s*(?:\\w|_).*?\\s*->" },
      { "ugly": "\\.=", "pretty": " \ue123", "pre": "[^.=:!/<>]|^", "post": "[^.=:!/<>]|$" },
      { "ugly": "\\.-", "pretty": " \ue122", "pre": "[^.=:!/<>]|^", "post": "[^.=:!/<>]|$" },
      { "ugly": ":=", "pretty": " \ue10c", "pre": "[^.=:!/<>]|^", "post": "[^.=:!/<>]|$" },
      { "ugly": "=:=", "pretty": "\u00a0  \ue13b", "pre": "[^.=:!/<>]|^", "post": "[^.=:!/<>]|$" },
      { "ugly": "==", "pretty": " \ue13c", "pre": "[^.=:!/<>]|^", "post": "[^.=:!/<>]|$" },
      { "ugly": "!=", "pretty": " \ue10e", "pre": "[^.=:!/<>]|^", "post": "[^.=:!/<>]|$" },
      { "ugly": "===", "pretty": "\u00a0  \ue13d", "pre": "[^.=:!/<>]|^", "post": "[^.=:!/<>]|$" },
      { "ugly": "!==", "pretty": "\u00a0  \ue10f", "pre": "[^.=:!/<>]|^", "post": "[^.=:!/<>]|$" },
      { "ugly": "=/=", "pretty": "\u00a0  \ue143", "pre": "[^.=:!/<>]|^", "post": "[^.=:!/<>]|$" },
      { "ugly": "<-", "pretty": " \ue152", "pre": "[^<\\->=]|^", "post": "[^<\\-=>]|$" },
      { "ugly": "->", "pretty": " \ue114", "pre": "[^<\\->=]|^", "post": "[^<\\-=>]|$" },
      { "ugly": "<--", "pretty": "\u00a0  \ue153", "pre": "[^<\\->=]|^", "post": "[^<\\->=]|$" },
      { "ugly": "-->", "pretty": "\u00a0  \ue113", "pre": "[^<\\->=]|^", "post": "[^<\\->=]|$" },
      { "ugly": "<->", "pretty": "\u00a0  \ue154", "pre": "[^<\\->=]|^", "post": "[^<\\->=]|$" },
      { "ugly": "<<-", "pretty": "\u00a0  \ue15d", "pre": "[^<\\->=]|^", "post": "[^<\\-=>]|$" },
      { "ugly": "->>", "pretty": "\u00a0  \ue115", "pre": "[^<\\->=]|^", "post": "[^<\\-=>]|$" },
      { "ugly": "=>", "pretty": " \ue13f", "pre": "[^<\\->=]|^", "post": "[^<\\-=>]|$" },
      { "ugly": "<==", "pretty": "\u00a0  \ue158", "pre": "[^<\\->=]|^", "post": "[^<\\->=]|$" },
      { "ugly": "==>", "pretty": "\u00a0  \ue13e", "pre": "[^<\\->=]|^", "post": "[^<\\->=]|$" },
      { "ugly": "<=>", "pretty": "\u00a0  \ue159", "pre": "[^<\\->=]|^", "post": "[^<\\->=]|$" },
      { "ugly": "<=<", "pretty": "\u00a0  \ue15a", "pre": "[^<\\->=]|^", "post": "[^<\\->=]|$" },
      { "ugly": "<<=", "pretty": "\u00a0  \ue15e", "pre": "[^<\\->=]|^", "post": "[^<\\->=]|$" },
      { "ugly": "=>>", "pretty": "\u00a0  \ue140", "pre": "[^<\\->=]|^", "post": "[^<\\->=]|$" },
      { "ugly": ">=>", "pretty": "\u00a0  \ue146", "pre": "[^<\\->=]|^", "post": "[^<\\->=]|$" },
      { "ugly": ">>=", "pretty": "\u00a0  \ue149", "pre": "[^<\\->=]|^", "post": "[^<\\->=]|$" },
      { "ugly": ">>-", "pretty": "\u00a0  \ue148", "pre": "[^<\\->=]|^", "post": "[^<\\->=]|$" },
      { "ugly": ">-", "pretty": " \ue144", "pre": "[^<\\->=]|^", "post": "[^<\\->=]|$" },
      { "ugly": "-<", "pretty": " \ue116", "pre": "[^<\\->=]|^", "post": "[^<\\->=]|$" },
      { "ugly": "-<<", "pretty": "\u00a0  \ue117", "pre": "[^<\\->=]|^", "post": "[^<\\->=]|$" },
      { "ugly": "=<<", "pretty": "\u00a0  \ue142", "pre": "[^<\\->=]|^", "post": "[^<\\->=]|$" },
      { "ugly": "<~~", "pretty": "\u00a0  \ue161", "pre": "[^<\\->=~]|^", "post": "[^<\\->=~]|$" },
      { "ugly": "<~", "pretty": " \ue160", "pre": "[^<\\->=~]|^", "post": "[^<\\->=~]|$" },
      { "ugly": "~~", "pretty": " \ue168", "pre": "[^<\\->=~]|^", "post": "[^<\\->=~]|$" },
      { "ugly": "~>", "pretty": " \ue167", "pre": "[^<\\->=~]|^", "post": "[^<\\->=~]|$" },
      { "ugly": "~~>", "pretty": "\u00a0  \ue169", "pre": "[^<\\->=~]|^", "post": "[^<\\->=~]|$" },
      { "ugly": "<<<", "pretty": "\u00a0  \ue15f", "pre": "[^<\\->=]|^", "post": "[^<\\->=]|$" },
      { "ugly": "<<", "pretty": " \ue15c", "pre": "[^<\\->=]|^", "post": "[^<\\->=]|$" },
      { "ugly": "<=", "pretty": " \ue141", "pre": "[^<\\->=]|^", "post": "[^<\\->=]|$" },
      { "ugly": "<>", "pretty": " \ue15b", "pre": "[^<\\->=]|^", "post": "[^<\\->=]|$" },
      { "ugly": ">=", "pretty": " \ue145", "pre": "[^<\\->=]|^", "post": "[^<\\->=]|$" },
      { "ugly": ">>", "pretty": " \ue147", "pre": "[^<\\->=]|^", "post": "[^<\\->=]|$" },
      { "ugly": ">>>", "pretty": "\u00a0  \ue14a", "pre": "[^<\\->=]|^", "post": "[^<\\->=]|$" },
      { "ugly": "<\\|", "pretty": " \ue14d", "pre": "[^<\\->=]|^", "post": "[^<\\->=]|$" },
      { "ugly": "<\\|>", "pretty": "\u00a0  \ue14e", "pre": "[^<\\->=|]|^", "post": "[^<\\->=|]|$" },
      { "ugly": "\\|>", "pretty": " \ue135", "pre": "[^<\\->=|]|^", "post": "[^<\\->=|]|$" },
      { "ugly": "<\\$", "pretty": " \ue14f", "pre": "[^<\\->=$]|^", "post": "[^<\\->=$]|$" },
      { "ugly": "<\\$>", "pretty": "\u00a0  \ue150", "pre": "[^<\\->=$]|^", "post": "[^<\\->=$]|$" },
      { "ugly": "\\$>", "pretty": " \ue137", "pre": "[^<\\->=$]|^", "post": "[^<\\->=$]|$" },
      { "ugly": "<\\+", "pretty": " \ue155", "pre": "[^<\\->=+]|^", "post": "[^<\\->=+]|$" },
      { "ugly": "<\\+>", "pretty": "\u00a0  \ue156", "pre": "[^<\\->=+]|^", "post": "[^<\\->=+]|$" },
      { "ugly": "\\+>", "pretty": " \ue13a", "pre": "[^<\\->=+]|^", "post": "[^<\\->=+]|$" },
      { "ugly": "<\\*", "pretty": " \ue14b", "pre": "[^<\\->=*]|^", "post": "[^<\\->=*]|$" },
      { "ugly": "<\\*>", "pretty": "\u00a0  \ue14c", "pre": "[^<\\->=*]|^", "post": "[^<\\->=*]|$" },
      { "ugly": "\\*>", "pretty": " \ue104", "pre": "[^<\\->=*]|^", "post": "[^<\\->=*]|$" },
      { "ugly": "\\\\\\\\", "pretty": " \ue106", "pre": "[^<\\\\\\->=]|^", "post": "[^<\\\\\\->=]|$" },
      { "ugly": "\\\\\\\\\\\\", "pretty": "\u00a0  \ue107", "pre": "[^<\\\\\\->=]|^", "post": "[^<\\\\\\->=]|$" },
      { "ugly": "\\{-", "pretty": " \ue108", "pre": "[^<\\\\\\->={}]|^", "post": "[^<\\\\\\->={}]|$" },
      { "ugly": "-}", "pretty": " \ue110", "pre": "[^<\\->={}]|^", "post": "[^<\\->={}]|$" },
      { "ugly": "//", "pretty": " \ue12f", "pre": "[^<\\->=/]|^", "post": "[^<\\->=/]|$" },
      { "ugly": "///", "pretty": "\u00a0  \ue130", "pre": "[^<\\->=/]|^", "post": "[^<\\->=/]|$" },
      { "ugly": "/\\*", "pretty": " \ue12a", "pre": "[^<\\->=/*]|^", "post": "[^<\\->=/*]|$" },
      { "ugly": "/\\*\\*", "pretty": "\u00a0  \ue12b", "pre": "[^<\\->=/*]|^", "post": "[^<\\->=/*]|$" },
      { "ugly": "\\*\\*/", "pretty": "\u00a0  \ue103", "pre": "[^<\\->=/*]|^", "post": "[^<\\->=/*]|$" },
      { "ugly": "\\*/", "pretty": " \ue105", "pre": "[^<\\->=/*]|^", "post": "[^<\\->=/*]|$" },
      { "ugly": "</", "pretty": " \ue162", "pre": "[^<\\->=/]|^", "post": "[^<\\->=/]|$" },
      { "ugly": "<\\!--", "pretty": "\u00a0 \u00a0\ue151", "pre": "[^<\\->=!]|^", "post": "[^<\\->=!]|$" },
      { "ugly": "</>", "pretty": "\u00a0  \ue163", "pre": "[^<\\->=]|^", "post": "[^<\\->=]|$" },
      { "ugly": "/>", "pretty": " \ue12e", "pre": "[^<\\->=/]|^", "post": "[^<\\->=/]|$" },
      { "ugly": "x", "pretty": "\ue16b", "pre": "[0-9a-fA-F]", "post": "[0-9a-fA-F]" },
      { "ugly": ":", "pretty": "\ue16c", "pre": "[0-9]{1,2}", "post": "[0-9]{1,2}" },
      { "ugly": "\\+", "pretty": "\ue16d", "pre": "[0-9a-zA-Z]", "post": "[0-9a-zA-Z]" },
      { "ugly": "-", "pretty": "-", "pre": "[0-9a-zA-Z]", "post": "[0-9a-zA-Z]" },
      { "ugly": "\\*", "pretty": "*", "pre": "", "post": "" },
      { "ugly": "www", "pretty": "\u00a0  \ue100", "pre": "\\b", "post": "\\b" },
      { "ugly": ";;", "pretty": " \ue129", "pre": "[^;]|^", "post": "[^;]|$" },
      { "ugly": "::", "pretty": " \ue10a", "pre": "[^:]|^", "post": "[^:]|$" },
      { "ugly": ":::", "pretty": "\u00a0  \ue10b", "pre": "[^:]|^", "post": "[^:]|$" },
      { "ugly": "!!", "pretty": " \ue10d", "pre": "[^!]|^", "post": "[^!]|$" },
      { "ugly": "\\?{2}", "pretty": " \ue128", "pre": "[^?]|^", "post": "[^?]|$" },
      { "ugly": "%%", "pretty": " \ue16a", "pre": "[^%]|^", "post": "[^%]|$" },
      { "ugly": "&&", "pretty": " \ue131", "pre": "[^&]|^", "post": "[^&]|$" },
      { "ugly": "\\|{2}", "pretty": " \ue132", "pre": "[^|=]|^", "post": "[^|=]|$" },
      { "ugly": "\\.{2}", "pretty": " \ue124", "pre": "[^.<]|^", "post": "[^.<]|$" },
      { "ugly": "\\.{3}", "pretty": "\u00a0  \ue126", "pre": "[^.<]|^", "post": "[^.<]|$" },
      { "ugly": "\\.\\.<", "pretty": "\u00a0  \ue125", "pre": "[^.<]|^", "post": "[^.<]|$" },
      { "ugly": "\\[\\]", "pretty": " \ue109", "pre": "[^[\\]]|^", "post": "[^[\\]]|$" },
      { "ugly": "--", "pretty": " \ue111", "pre": "[^\\-+*=<>~]|^", "post": "[^\\-+*=<>~]|$" },
      { "ugly": "---", "pretty": "\u00a0 \ue112", "pre": "[^\\-+*=<>~]|^", "post": "[^\\-+*=<>~]|$" },
      { "ugly": "\\+{2}", "pretty": " \ue138", "pre": "[^\\-+*=<>~]|^", "post": "[^\\-+*=<>~]|$" },
      { "ugly": "\\+{3}", "pretty": "\u00a0  \ue139", "pre": "[^\\-+*=<>~]|^", "post": "[^\\-+*=<>~]|$" },
      { "ugly": "\\*{2}", "pretty": " \ue101", "pre": "[^\\-+*=<>~]|^", "post": "[^\\-+*=<>~]|$" },
      { "ugly": "\\*{3}", "pretty": "\u00a0  \ue102", "pre": "[^\\-+*=<>~]|^", "post": "[^\\-+*=<>~]|$" },
      { "ugly": "~=", "pretty": " \ue166", "pre": "[^~=\\-]|^", "post": "[^~=\\-]|$" },
      { "ugly": "~-", "pretty": " \ue165", "pre": "[^~=\\-<>]|^", "post": "[^~=\\-<>]|$" },
      { "ugly": "-~", "pretty": " \ue118", "pre": "[^~=\\-<>]|^", "post": "[^~=\\-<>]|$" },
      { "ugly": "~@", "pretty": " \ue164", "pre": "[^~=\\-@~]|^", "post": "[^~=\\-@~]|$" },
      { "ugly": "\\^=", "pretty": " \ue136", "pre": "[^~=\\-^]|^", "post": "[^~=\\-^]|$" },
      { "ugly": "\\?=", "pretty": " \ue127", "pre": "[^~=\\-/?]|^", "post": "[^~=\\-/?]|$" },
      { "ugly": "/=", "pretty": " \ue12c", "pre": "[^~=\\-/]|^", "post": "[^~=\\-/]|$" },
      { "ugly": "/==", "pretty": "\u00a0  \ue12d", "pre": "[^~=\\-/]|^", "post": "[^~=\\-/]|$" },
      { "ugly": "\\|=", "pretty": " \ue134", "pre": "[^~=\\-/|]|^", "post": "[^~=\\-/|]|$" },
      { "ugly": "\\|{2}=", "pretty": "\u00a0  \ue133", "pre": "[^~=\\-/|]|^", "post": "[^~=\\-/|]|$" },
      { "ugly": "##", "pretty": " \ue11b", "pre": "[^#_?({[\\]]|^", "post": "[^#_?({[\\]]|$" },
      { "ugly": "###", "pretty": "\u00a0  \ue11c", "pre": "[^#_?({[\\]]|^", "post": "[^#_?({[\\]]|$" },
      { "ugly": "####", "pretty": "\u00a0\u00a0 \ue11d", "pre": "[^#_?({[\\]]|^", "post": "[^#_?({[\\]]|$" },
      { "ugly": "#{", "pretty": " \ue119", "pre": "[^#_?({[\\]]|^", "post": "[^#_?({[\\]]|$" },
      { "ugly": "#\\[", "pretty": " \ue11a", "pre": "[^#_?({[\\]]|^", "post": "[^#_?({[\\]]|$" },
      { "ugly": "#\\(", "pretty": " \ue11e", "pre": "[^#_?({[\\]]|^", "post": "[^#_?({[\\]]|$" },
      { "ugly": "#\\?", "pretty": " \ue11f", "pre": "[^#_?({[\\]]|^", "post": "[^#_?({[\\]]|$" },
      { "ugly": "#_", "pretty": " \ue120", "pre": "[^#_?({[\\]]|^", "post": "[^#_?({[\\]]|$" },
      { "ugly": "#_\\(", "pretty": "\u00a0  \ue121", "pre": "[^#_?({[\\]]|^", "post": "[^#_?({[\\]]|$" }
    ]
  }
]

@siegebell There's a bug and i'm not sure whether it is fira code or emacs.

So with font set to Fira Code and setting the fontset to use fira code symbol for \ue100 to \ue16f I'm getting some weird behavior.

The symbols appear to the left of the original text a significant amount. So, for instance, if I do <-- as the first text on the line i'll get most of it rendering off the left side of the screen. Including emacs config and screenshot here, not yet sure what the problem is.

Edit: If there is other text to the left this bug will actually cause the symbol to overlap with the text!

Edit2: Loading it with no other emacs config has the same effect so shouldn't be some package interfering. In addition, it looks like it is about two spaces/characters to the left of where it should be.

Edit3: I see now that you are dealing with this via \u00a0 in the vscode config? Not sure if I can do that with emacs or not.

Emacs config:

;;; Prettify symbols
(setq prettify-symbols-unprettify-at-point t)

(set-fontset-font t '(#Xe100 . #Xe16f) "Fira Code Symbol")

(defconst fira-code-font-lock-symbols-alist
  (mapcar (lambda (s)
            (cons (car s) (decode-char 'ucs (car (cdr s)))))
          (list '("www" #Xe100)
                '("<--" #Xe153))))

(add-hook 'prog-mode-hook
          (lambda ()
            (dolist (alias fira-code-font-lock-symbols-alist)
              (push alias prettify-symbols-alist))
            (prettify-symbols-mode)))

@siegebell So this definitely looks like it is because emacs expects to be replacing multiple characters with a character with a width of one where as the fira code symbols are designed to take up multiple character widths. Haven't yet figured out a solution.

Found an uglier way that works.

(add-hook 'prog-mode-hook
          (lambda ()
            (font-lock-add-keywords nil
                                    `(("\\(###\\)"
                                       (0 (prog1 ()
                                            (compose-region (match-beginning 1)
                                                            (match-end 1)
                                                            ;; The first argument to concat is a string containing a literal tab character inserted via C-q <tab>. It is important!
                                                            ,(concat "  " (list (decode-char 'ucs #Xe11c)))))))))))

I'm going to turn it into a helper function then I just need to figure out emacs's awful regex syntax for each thing and we should be good.

[tonsky]

@mordocai FYI in Fira symbols visually take multiple character places but marked as taking just one (rightmost) single place:

So in fact when I do ligature substitution in font code, I replace e.g. hyphen hyphen greater with space space hyphen_hyphen_greater.liga (CR stands for space):

    sub CR CR greater' by hyphen_hyphen_greater.liga;
    sub CR hyphen' greater  by CR;
    sub hyphen' hyphen  greater  by CR;

The reason for that is that editors will still see result of substitution as three single-width characters, and will allow to “step inside” the ligature glyph. Without that, ligature will act as it’s a single character, which changes editing behaviour which we want to avoid. We want to keep feeling that substitution is purely visual and underlying code is there unchanged.

The downside of that is that glyphs span outside their container. This is usually perfectly fine with all known editors, but might surprise anyone who tries to use them directly without knowing about that gotcha.

Hope that helps.

@tonsky Yeah, I ended up figuring out what was going on! Luckily the new way I'm doing it with emacs knows how to handle things like that by using a special string beginning with a tab.

Everyone: I got it working! Here is the emacs code:

https://gist.github.com/mordocai/50783defab3c3d1650e068b4d1c91495

Linking to gist so I can update it later if necessary.

Edit: Screenshot. Also, sorry for spamming this issue so bad. I kept thinking I was going to give up for the night, but then didn't...

[emmanueltouzery]

@mordocai thanks for the work! I tried the gist but I get mostly garbage. See the picture. Maybe I'm using the wrong version of the font or something? I'm pretty sure I'm using 0.200 although I did install older versions in the past.

EDIT OK SORRY IT WORKS. Need to install the Fira code symbol font additionnally. The "x" does look weird, in "Text" for instance. It's replaced by a cross. Also the ligature for =<< doesn't work while fira code appears to support it.

[emmanueltouzery]

to fix the =<< in the gist =>

          ("\\(=<<\\)"                   #Xe142)    (changed from ==<)
          ("[^-=]\\(<<\\)"               #Xe15c)    (excluded the = before the << as well)

this is for the overzealous cross:

          ;; ("\\(x\\)"                     #Xe16b)

would need to require non-word characters before & after but my emacs regex-fu is weak for that, so I just commented that locally.

[siegebell]

For example, this should only match 'x' if the preceding character is a number and the subsequent character is either a number or letter:

          ("[0-9]\\(x\\)[0-9a-fA-F]"                     #Xe16b)

edit: changed a-F to a-f

[emmanueltouzery]

@siegebell i think the ideal would be a "word boundary" regex before & after the x. I was a bit lazy earlier, but this seems to work great =>

          ("\\b\\(x\\)\\b"               #Xe16b)