Merge 8dd519c into 21c0a33

documentation/c07-settings.sil

@@ -185,6 +185,10 @@ true value) and also \em{unset} the setting \autodoc:setting{document.spaceskip}
 unset it, just call \autodoc:command{\set} with no \autodoc:parameter{value} parameter:
 \autodoc:command{\set[parameter=document.spaceskip]}.
 
+Note that non-breaking spaces (U+00A0), following the guidelines of Unicode Annex 14 (UAX 14), are treated by default as stretchable or shrinkable akin to regular inter-word spaces, contributing to text justification and alignment for consistent layout.
+
+If you want to disable this behavior, the \autodoc:setting{languages.fixedNbsp} setting may be set to \code{true} to enforce fixed-width non-breaking spaces.
+
 \subsection{Letter spacing settings}
 
 You can also put spaces in between \em{letters} with the \autodoc:setting{document.letterspaceglue} setting.

languages/fr.lua

@@ -172,7 +172,7 @@ function SILE.nodeMakers.fr:mustRemove (i, items)
   -- Clear "manual" spaces we do not want, so that later we only have to
   -- insert the relevant kerns.
   local curr = items[i].text
-  if self:isSpace(curr) then
+  if self:isSpace(curr) or self:isNonBreakingSpace(curr) then
     if i < #items then
       local next = items[i+1].text
       if self:isSpace(next)

languages/unicode.lua

@@ -2,6 +2,13 @@ local icu = require("justenoughicu")
 
 local chardata = require("char-def")
 
+SILE.settings:declare({
+  parameter = "languages.fixedNbsp",
+  type = "boolean",
+  default = false,
+  help = "Whether to treat U+00A0 (NO-BREAK SPACE) as a fixed-width space"
+})
+
 SILE.nodeMakers.base = pl.class({
 
     _init = function (self, options)
@@ -43,6 +50,14 @@ SILE.nodeMakers.base = pl.class({
       self.lastnode = "penalty"
     end,
 
+    makeNonBreakingSpace = function (self)
+      -- Unicode Line Breaking Algorithm (UAX 14) specifies that U+00A0
+      -- (NO-BREAK SPACE) is expanded or compressed like a normal space.
+      coroutine.yield(SILE.nodefactory.kern(SILE.shaper:measureSpace(self.options)))
+      self.lastnode = "glue"
+      self.lasttype = "sp"
+    end,
+
     iterator = function (_, _)
       SU.error("Abstract function nodemaker:iterator called", true)
     end,
@@ -61,6 +76,15 @@ SILE.nodeMakers.base = pl.class({
       return self.isSpaceType[self:charData(char).linebreak]
     end,
 
+    isNonBreakingSpace = function (self, char)
+      local c = self:charData(char)
+      return c.contextname and c.contextname == "nobreakspace"
+    end,
+
+    isActiveNonBreakingSpace = function (self, char)
+      return self:isNonBreakingSpace(char) and not SILE.settings:get("languages.fixedNbsp")
+    end,
+
     isBreaking = function (self, char)
       return self.isBreakingType[self:charData(char).linebreak]
     end,
@@ -85,6 +109,9 @@ function SILE.nodeMakers.unicode:dealWith (item)
   if self:isSpace(item.text) then
     self:makeToken()
     self:makeGlue(item)
+  elseif self:isActiveNonBreakingSpace(item.text) then
+    self:makeToken()
+    self:makeNonBreakingSpace()
   elseif self:isBreaking(item.text) then
     self:addToken(char, item)
     self:makeToken()
@@ -152,7 +179,11 @@ function SILE.nodeMakers.unicode:handleWordBreak (item)
   if self:isSpace(item.text) then
     -- Spacing word break
     self:makeGlue(item)
-  else -- a word break which isn't a space
+  elseif self:isActiveNonBreakingSpace(item.text) then
+    -- Non-breaking space word break
+    self:makeNonBreakingSpace()
+  else
+     -- a word break which isn't a space
     self:addToken(item.text, item)
   end
 end
@@ -161,7 +192,7 @@ function SILE.nodeMakers.unicode:handleLineBreak (item, subtype)
   -- Because we are in charge of paragraphing, we
   -- will override space-type line breaks, and treat
   -- them just as ordinary word spaces.
-  if self:isSpace(item.text) then
+  if self:isSpace(item.text) or self:isActiveNonBreakingSpace(item.text) then
     self:handleWordBreak(item)
     return
   end

tests/feat-unicode-nbsp.expected

@@ -0,0 +1,176 @@
+Set paper size 	297.6377985	419.5275636
+Begin page
+Mx 	14.8819
+My 	28.5447
+Set font 	Gentium Plus;10;400;;normal;;;LTR
+T	47 82 85 72 80 w=26.4014	(Lorem)
+Mx 	45.5228
+T	76 83 86 88 80 w=25.2002	(ipsum)
+Mx 	74.9626
+T	71 82 79 82 85 w=21.9287	(dolor)
+Mx 	101.1308
+T	86 76 87 w=10.0146	(sit)
+Mx 	115.3850
+T	68 80 72 87 w=20.6836	(amet)
+Mx 	136.0686
+T	15 w=2.2900	(,)
+Mx 	142.5982
+T	86 76 87 w=10.0146	(sit)
+Mx 	156.8524
+T	68 80 72 87 w=20.6836	(amet)
+Mx 	181.7755
+T	68 71 76 83 76 86 70 76 81 74 w=41.8457	(adipiscing)
+Mx 	227.8608
+T	72 79 76 87 w=13.4814	(elit)
+Mx 	241.3422
+T	17 w=2.2900	(.)
+Mx 	247.8718
+T	54 72 71 w=14.5801	(Sed)
+Mx 	266.6915
+T	81 82 81 w=16.0645	(non)
+Mx 	14.8819
+My 	40.5447
+T	85 76 86 88 86 w=19.6924	(risus)
+Mx 	34.5743
+T	17 w=2.2900	(.)
+Mx 	39.5220
+T	54 88 86 83 72 81 71 76 86 86 72 w=49.6094	(Suspendisse)
+Mx 	91.7890
+T	79 72 70 87 88 86 w=24.3018	(lectus)
+Mx 	118.7484
+T	87 82 85 87 82 85 w=24.8633	(tortor)
+Mx 	143.6117
+T	17 w=2.2900	(.)
+Mx 	14.8819
+My 	58.5447
+T	47 82 85 72 80 w=26.4014	(Lorem)
+Mx 	49.1701
+T	76 83 86 88 80 w=25.2002	(ipsum)
+Mx 	82.2572
+T	71 82 79 82 85 w=21.9287	(dolor)
+Mx 	112.0727
+T	86 76 87 w=10.0146	(sit)
+Mx 	129.9742
+T	68 80 72 87 w=20.6836	(amet)
+Mx 	150.6578
+T	15 w=2.2900	(,)
+Mx 	160.8347
+T	70 82 81 86 72 70 87 72 87 88 85 w=48.5303	(consectetur)
+Mx 	217.2519
+T	68 71 76 83 76 86 70 76 81 74 w=41.8457	(adipiscing)
+Mx 	266.9844
+T	72 79 76 87 w=13.4814	(elit)
+Mx 	280.4659
+T	17 w=2.2900	(.)
+Mx 	14.8819
+My 	70.5447
+T	54 72 71 w=14.5801	(Sed)
+Mx 	32.1157
+T	81 82 81 w=16.0645	(non)
+Mx 	50.8338
+T	85 76 86 88 86 w=19.6924	(risus)
+Mx 	70.5262
+T	17 w=2.2900	(.)
+Mx 	75.4699
+T	54 88 86 83 72 81 71 76 86 86 72 w=49.6094	(Suspendisse)
+Mx 	127.7330
+T	79 72 70 87 88 86 w=24.3018	(lectus)
+Mx 	154.6884
+T	87 82 85 87 82 85 w=24.8633	(tortor)
+Mx 	179.5517
+T	17 w=2.2900	(.)
+Mx 	14.8819
+My 	82.5447
+T	47 82 85 72 80 w=26.4014	(Lorem)
+Mx 	43.1918
+T	76 83 86 88 80 w=25.2002	(ipsum)
+Mx 	70.3005
+T	71 82 79 82 85 w=21.9287	(dolor)
+Mx 	94.1377
+T	86 76 87 w=10.0146	(sit)
+Mx 	106.0609
+T	68 80 72 87 w=20.6836	(amet)
+Mx 	126.7445
+T	15 w=2.2900	(,)
+Mx 	130.9431
+T	86 76 87 w=10.0146	(sit)
+Mx 	142.8663
+T	68 80 72 87 w=20.6836	(amet)
+Mx 	165.4584
+T	68 71 76 83 76 86 70 76 81 74 w=41.8457	(adipiscing)
+Mx 	209.2126
+T	72 79 76 87 w=13.4814	(elit)
+Mx 	222.6941
+T	17 w=2.2900	(.)
+Mx 	226.8926
+T	54 72 71 w=14.5801	(Sed)
+Mx 	243.3812
+T	81 82 81 w=16.0645	(non)
+Mx 	261.3542
+T	85 76 86 88 86 w=19.6924	(risus)
+Mx 	281.0466
+T	17 w=2.2900	(.)
+Mx 	14.8819
+My 	94.5447
+T	54 88 86 83 72 81 71 76 86 86 72 w=49.6094	(Suspendisse)
+Mx 	67.1516
+T	79 72 70 87 88 86 w=24.3018	(lectus)
+Mx 	94.1137
+T	87 82 85 87 82 85 w=24.8633	(tortor)
+Mx 	118.9770
+T	17 w=2.2900	(.)
+Mx 	14.8819
+My 	106.5447
+T	47 82 85 72 80 w=26.4014	(Lorem)
+Mx 	53.6435
+T	76 83 86 88 80 w=25.2002	(ipsum)
+Mx 	91.2039
+T	71 82 79 82 85 w=21.9287	(dolor)
+Mx 	125.4928
+T	86 76 87 w=10.0146	(sit)
+Mx 	147.8677
+T	68 80 72 87 w=20.6836	(amet)
+Mx 	168.5513
+T	15 w=2.2900	(,)
+Mx 	183.2015
+T	86 76 87 w=10.0146	(sit)
+Mx 	205.5764
+T	68 80 72 87 w=20.6836	(amet)
+Mx 	238.6202
+T	68 71 76 83 76 86 70 76 81 74 w=41.8457	(adipiscing)
+Mx 	280.4659
+T	17 w=2.2900	(.)
+Mx 	14.8819
+My 	118.5447
+T	54 88 86 83 72 81 71 76 86 86 72 w=49.6094	(Suspendisse)
+Mx 	67.1516
+T	79 72 70 87 88 86 w=24.3018	(lectus)
+Mx 	94.1137
+T	87 82 85 87 82 85 w=24.8633	(tortor)
+Mx 	118.9770
+T	17 w=2.2900	(.)
+Mx 	14.8819
+My 	136.5447
+T	41 85 68 81 111 68 76 86 w=34.3799	(Français)
+Mx 	50.5831
+T	34 w=4.3311	(?)
+Mx 	57.5711
+T	41 85 68 81 111 68 76 w=30.5176	(Françai)
+Mx 	88.0887
+T	86 w=3.8623	(s)
+Mx 	94.6080
+T	29 w=2.2900	(:)
+Mx 	99.5550
+T	169 w=4.8535	(«)
+Mx 	106.5269
+T	41 85 68 81 111 w=23.2178	(Franç)
+Mx 	129.7447
+T	68 w=4.5898	(a)
+Mx 	134.3345
+T	76 86 w=6.5723	(is)
+Mx 	142.2281
+T	4 w=2.7197	(!)
+Mx 	147.0662
+T	170 w=4.8535	(»)
+End page
+Finish

tests/feat-unicode-nbsp.sil

@@ -0,0 +1,33 @@
+\begin[papersize=a6]{document}
+\neverindent
+\nofolios
+\language[main=und]
+% Reference text
+Lorem ipsum dolor sit amet, sit amet adipiscing elit.
+Sed non risus. Suspendisse lectus tortor.
+\medskip
+% Four first spaces are U+00A0.
+% Expect proper stretchability as regular spaces.
+Lorem ipsum dolor sit amet, consectetur adipiscing elit.\break
+Sed non risus. Suspendisse lectus tortor.
+
+% All spaces are U+00A0 but two:
+% Expectation:
+% Expect non-breakable spaces, shrinking as regular spaces
+% when necessary.
+Lorem ipsum dolor sit amet, sit amet adipiscing elit.
+Sed non risus. Suspendisse lectus tortor.
+
+\set[parameter=linebreak.emergencyStretch, value=20em]
+Lorem ipsum dolor sit amet, sit amet adipiscing.
+Suspendisse lectus tortor.
+
+% Punctiation spaces are U+00A0 in the following case.
+% Expectation:
+% French punctuation spaces are managed automatically,
+% also replacing U+00A0.
+\medskip
+\language[main=fr]
+Français ? Français : « Français ! »
+
+\end{document}