starting to work on \arbnull to put back on contextual analysis in case it is broken by TeX commands

3 files changed, 26 insertions, 3 deletions

diff --git a/arabluatex.lua b/arabluatex.lua
index 07968c0..527be3b 100644
--- a/arabluatex.lua
+++ b/arabluatex.lua
@@ -128,6 +128,21 @@ local function indnum(str)
    return str
 end
 
+local function processarbnull(str, scheme)
+   if scheme == "buckwalter" then
+      str = string.gsub(str, "(\\arbnull.?)(%b{})", function(tag, body)
+                           body = string.sub(body, 2, -2)
+                           return string.format("P%sP", body)
+      end)
+   else
+      str = string.gsub(str, "(\\arbnull.?)(%b{})", function(tag, body)
+                           body = string.sub(body, 2, -2)
+                           return string.format("o%so", body)
+      end)
+   end
+   return str
+end
+
 local function takeoutabjad(str)
    str = string.gsub(str, "(\\abjad.?)(%b{})", function(tag, body)
                         body = string.sub(body, 2, -2)
@@ -543,6 +558,7 @@ end
 
 function processtrans(str, mode, rules, scheme)
    str = "\\arb{".. str.."}"
+   str = processarbnull(str, scheme)
    str = takeoutabjad(str)
    str = protectarb(str)
    str = breakcmd(str)
diff --git a/arabluatex_trans.lua b/arabluatex_trans.lua
index ef55a2b..bc44229 100644
--- a/arabluatex_trans.lua
+++ b/arabluatex_trans.lua
@@ -30,7 +30,8 @@ punctuationtr = {
 }
 
 nulltr = {
-   {a="%|", b=""}
+   {a="%|", b=""},
+   {a="o.-o", b=""}
 }
 
 -- cap
@@ -95,7 +96,11 @@ hamzatrdmg = {
    {a="(A)(')(i)", b="%1ʾ%3"}, -- historic madda
    {a="(A)(')", b="%1ʾ"}, -- historic madda
    -- initial (needs both ^ and %W patterns)
-   -- 'aw: the diphthong is to be resoved into 'awi' (next 4 lines)
+   -- 'aw: the diphthong is to be resolved into 'awi' (next 8 lines)
+   {a="^('aw)(o)(['][%_%^%.]?[l'btjghxdrzs`fqkmnwy]%-)(.-o)", b="%1i"},
+   {a="(%W)('aw)(o)(['][%_%^%.]?[l'btjghxdrzs`fqkmnwy]%-)(.-o)", b="%1%2i"},
+   {a="^('aw)(o)([%(%[%|%<]?\"?[uai])(.-o)", b="%1i"},
+   {a="(%W)('aw)(o)([%(%[%|%<]?\"?[uai])(.-o)", b="%1%2i"},
    {a="^('aw)(%s)(['][%_%^%.]?[l'btjghxdrzs`fqkmnwy]%-)", b="%1i%2%3"},
    {a="(%W)('aw)(%s)(['][%_%^%.]?[l'btjghxdrzs`fqkmnwy]%-)", b="%1%2i%3%4"},
    {a="^('aw)(%s)([%(%[%|%<]?\"?[uai])", b="%1i%2%3"},
@@ -229,7 +234,8 @@ trigraphstrdmg = { -- trigraphs or more
 
 idghamtrdmg = {
    -- assimilations
-   {a="(n)(}?)(%s)([rlmnwy])", b="%4%2%3%4"}
+   {a="(n)(}?)(%s)([rlmnwy])", b="%4%2%3%4"},
+   {a="(n)(o)([rlmnwy])(.-o)", b="%3"}
 } 
 
 digraphstrdmg = {
diff --git a/arabluatex_voc.lua b/arabluatex_voc.lua
index 02c3feb..c856818 100644
--- a/arabluatex_voc.lua
+++ b/arabluatex_voc.lua
@@ -161,6 +161,7 @@ buckwalter = {
    {a="F", b="aN"},
    {a="K", b="iN"},
    {a="o", b="\""},
+   {a="P", b="o"}, -- pass on to \arbnull
    -- hard-coded connective alif
    {a="%[", b="ٱ"},
    {a="L", b="ٱ"} -- BW safe