The big thing is locale dependent collation support.

Also added outdigit support and (legacy) YESSTR/NOSTR support.

1 files changed, 185 insertions, 0 deletions

diff --git a/extra/locale/collation/fa_IR b/extra/locale/collation/fa_IR
new file mode 100644
index 000000000..a17159192
--- /dev/null
+++ b/extra/locale/collation/fa_IR
@@ -0,0 +1,185 @@
+copy "iso14651_t1"
+
+% The Persian alphabet order is: ALEF WITH MADDA ABOVE, ALEF, HAMZA (all
+% kinds), BEH, PEH, TEH, JEEM, TCHEH, HAH, KHAH, DAL, THAL, REH, ZAIN, JEH,
+% SEEN, SHEEN, SAD, DAD, TAH, ZAH, AIN, GHAIN, FEH, QAF, KAF, GAF, LAM,
+% MEEM, NOON, WAW, HEH, YEH.
+% The various kind of HAMZA are sorted as ALEF WITH HAMZA ABOVE, ALEF WITH
+% HAMZA BELOW, WAW WITH HAMZA ABOVE, YEH WITH HAMZA ABOVE.
+
+collating-symbol <AHY> % accent hamza over yeh
+collating-symbol <ADL> % dotless
+collating-symbol <ADO> % with dots over
+collating-symbol <AWO> % with wasla over
+
+collating-symbol <alefmadda>
+collating-symbol <yeh>
+
+% Alternate representations displayed the same
+collating-symbol <ALT1> 
+collating-symbol <ALT2>
+
+collating-element <Alef-Madda> from "<U0627><U0653>"
+collating-element <Alef-HamzaBelow> from "<U0627><U0655>"
+collating-element <Waw-Hamza> from "<U0648><U0654>"
+collating-element <AlefMaksura-Hamza> from "<U0649><U0654>"
+collating-element <Yeh-Hamza> from "<U064A><U0654>"
+collating-element <FarsiYeh-Hamza> from "<U06CC><U0654>"
+
+reorder-after <BAS>
+<AHA>
+<AHS>
+<AWO>
+<AHW>
+<AHY>
+<ADL>
+<ADO>
+<AYE>
+<YBA>
+
+reorder-after <LIG>
+<ALT1>
+<ALT2>
+
+reorder-after <th>
+<alefmadda>
+<alef>
+<hamza>
+
+reorder-after <waw>
+<heh>
+<yeh>
+
+% Persian uses fatha, kasra, damma, fathatan, kasratan, dammatan order.
+
+reorder-after <U066D>
+<U064E> IGNORE;IGNORE;IGNORE;<U064E> %<fatha_no>
+<UFE76> IGNORE;IGNORE;IGNORE;<UFE76> %<fatha_is>
+<UFE77> IGNORE;IGNORE;IGNORE;<UFE77> %<fatha_me>
+<U0650> IGNORE;IGNORE;IGNORE;<U0650> %<kasra_no>
+<UFE7A> IGNORE;IGNORE;IGNORE;<UFE7A> %<kasra_is>
+<UFE7B> IGNORE;IGNORE;IGNORE;<UFE7B> %<kasra_me>
+<U064F> IGNORE;IGNORE;IGNORE;<U064F> %<damma_no>
+<UFE78> IGNORE;IGNORE;IGNORE;<UFE78> %<damma_is>
+<UFE79> IGNORE;IGNORE;IGNORE;<UFE79> %<damma_me>
+<U064B> IGNORE;IGNORE;IGNORE;<U064B> %<fathatan_no>
+<UFE70> IGNORE;IGNORE;IGNORE;<UFE70> %<fathatan_is>
+<UFE71> IGNORE;IGNORE;IGNORE;<UFE71> %<fathatan_me>
+<U064D> IGNORE;IGNORE;IGNORE;<U064D> %<kasratan_no>
+<UFE74> IGNORE;IGNORE;IGNORE;<UFE74> %<kasratan_is>
+<U064C> IGNORE;IGNORE;IGNORE;<U064C> %<dammatan_no>
+<UFE72> IGNORE;IGNORE;IGNORE;<UFE72> %<dammatan_is>
+
+reorder-after <UFE7F>
+<U0653> IGNORE;IGNORE;IGNORE;<U0653> %<madda_no>   
+<U0654> IGNORE;IGNORE;IGNORE;<U0654> %<hamzaabove_no>
+<U0655> IGNORE;IGNORE;IGNORE;<U0655> %<hamzabelow_no>
+<U0670> IGNORE;IGNORE;IGNORE;<U0670> %<supalef_no>
+
+% Persian digits are sorted before Arabic ones: they are the basic forms.
+reorder-after <U0660>
+<U06F0> <0>;<BAS>;<MIN>;IGNORE
+<U0660> <0>;<PCL>;<MIN>;IGNORE
+<U06F1> <1>;<BAS>;<MIN>;IGNORE
+<U0661> <1>;<PCL>;<MIN>;IGNORE
+<U06F2> <2>;<BAS>;<MIN>;IGNORE
+<U0662> <2>;<PCL>;<MIN>;IGNORE
+<U06F3> <3>;<BAS>;<MIN>;IGNORE
+<U0663> <3>;<PCL>;<MIN>;IGNORE
+<U06F4> <4>;<BAS>;<MIN>;IGNORE
+<U0664> <4>;<PCL>;<MIN>;IGNORE
+<U06F5> <5>;<BAS>;<MIN>;IGNORE
+<U0665> <5>;<PCL>;<MIN>;IGNORE
+<U06F6> <6>;<BAS>;<MIN>;IGNORE
+<U0666> <6>;<PCL>;<MIN>;IGNORE
+<U06F7> <7>;<BAS>;<MIN>;IGNORE
+<U0667> <7>;<PCL>;<MIN>;IGNORE
+<U06F8> <8>;<BAS>;<MIN>;IGNORE
+<U0668> <8>;<PCL>;<MIN>;IGNORE
+<U06F9> <9>;<BAS>;<MIN>;IGNORE
+<U0669> <9>;<PCL>;<MIN>;IGNORE
+
+% And then the letters:
+<U0622> <alefmadda>;<BAS>;<MIN>;IGNORE % Alef With Madda Above
+<Alef-Madda> <alefmadda>;<BAS>;<MIN>;IGNORE
+<U0627> <alef>;<BAS>;<MIN>;IGNORE       % Alef
+<U0671> <alef>;<AWO>;<MIN>;IGNORE       % Alef Wasla
+<U0621> <hamza>;<BAS>;<MIN>;IGNORE      % Hamza
+<U0623> <hamza>;<AHA>;<MIN>;IGNORE      % Alef With Hamza Above
+<Alef-Hamza> <hamza>;<AHA>;<MIN>;IGNORE
+<U0625> <hamza>;<AHS>;<MIN>;IGNORE      % Alef With Hamza Below
+<Alef-HamzaBelow> <hamza>;<AHS>;<MIN>;IGNORE
+<U0624> <hamza>;<AHW>;<MIN>;IGNORE      % Waw With Hamza Above
+<Waw-Hamza> <hamza>;<AHW>;<MIN>;IGNORE
+<U0626> <hamza>;<AHY>;<MIN>;IGNORE      % Yeh With Hamza Above
+<FarsiYeh-Hamza> <hamza>;<AHY>;<ALT1>;IGNORE
+<AlefMaksura-Hamza> <hamza>;<AHY>;<ALT2>;IGNORE
+<Yeh-Hamza> <hamza>;<AHY>;<MIN>;IGNORE
+reorder-after <U0642>
+<U06A9> <kaf>;<BAS>;<MIN>;IGNORE        % Keheh
+<U0643> <kaf>;<PCL>;<MIN>;IGNORE        % Kaf
+reorder-after <U0648>
+<U0647> <heh>;<BAS>;<MIN>;IGNORE        % Heh
+<U0629> <heh>;<ADO>;<MIN>;IGNORE        % Teh Marbuta
+<U06C0> <heh>;<AHA>;<MIN>;IGNORE        % Heh With Yeh Above
+<U06CC> <yeh>;<BAS>;<MIN>;IGNORE        % Farsi Yeh
+<U0649> <yeh>;<ADL>;<MIN>;IGNORE        % Alef Maksura
+<U064A> <yeh>;<AYE>;<MIN>;IGNORE        % Yeh
+
+% Finally the letters in Presentation Form:
+reorder-after <UFE80>
+<UFE81> <alefmadda>;<BAS>;<AIS>;IGNORE
+<UFE82> <alefmadda>;<BAS>;<AFI>;IGNORE
+<UFE8D> <alef>;<BAS>;<AIS>;IGNORE
+<UFE8E> <alef>;<BAS>;<AFI>;IGNORE
+<UFB50> <alef>;<AWO>;<AIS>;IGNORE
+<UFB51> <alef>;<AWO>;<AFI>;IGNORE
+<UFE80> <hamza>;<BAS>;<AIS>;IGNORE
+<UFE83> <hamza>;<AHA>;<AIS>;IGNORE
+<UFE84> <hamza>;<AHA>;<AFI>;IGNORE
+<UFE87> <hamza>;<AHS>;<AIS>;IGNORE
+<UFE88> <hamza>;<AHS>;<AFI>;IGNORE
+<UFE85> <hamza>;<AHW>;<AIS>;IGNORE
+<UFE86> <hamza>;<AHW>;<AFI>;IGNORE
+<U0689> <hamza>;<AHY>;<AIS>;IGNORE
+<U068A> <hamza>;<AHY>;<AFI>;IGNORE
+reorder-after <UFEAE>
+<UFDFC> "<reh><yeh><alef><lam>";"<LIG><LIG><LIG><LIG>";"<AII><AME><AFI><AIS>";IGNORE % Rial Sign
+reorder-after <UFED8>
+<UFB8E> <kaf>;<BAS>;<AIS>;IGNORE
+<UFB8F> <kaf>;<BAS>;<AFI>;IGNORE
+<UFB90> <kaf>;<BAS>;<AII>;IGNORE
+<UFB91> <kaf>;<BAS>;<AME>;IGNORE
+<UFED9> <kaf>;<PCL>;<AIS>;IGNORE
+<UFEDA> <kaf>;<PCL>;<AFI>;IGNORE
+<UFEDB> <kaf>;<PCL>;<AII>;IGNORE
+<UFEDC> <kaf>;<PCL>;<AME>;IGNORE
+reorder-after <UFEEE>
+<UFEE9> <heh>;<BAS>;<AIS>;IGNORE
+<UFEEA> <heh>;<BAS>;<AFI>;IGNORE
+<UFEEB> <heh>;<BAS>;<AII>;IGNORE
+<UFEEC> <heh>;<BAS>;<AME>;IGNORE
+<UFE93> <heh>;<ADO>;<AIS>;IGNORE
+<UFE94> <heh>;<ADO>;<AFI>;IGNORE
+<UFBA4> <heh>;<AHA>;<AIS>;IGNORE
+<UFBA5> <heh>;<AHA>;<AFI>;IGNORE
+<UFBFC> <yeh>;<BAS>;<AIS>;IGNORE
+<UFBFD> <yeh>;<BAS>;<AFI>;IGNORE
+<UFBFE> <yeh>;<BAS>;<AII>;IGNORE
+<UFBFF> <yeh>;<BAS>;<AME>;IGNORE
+<UFEEF> <yeh>;<ADL>;<AIS>;IGNORE
+<UFEF0> <yeh>;<ADL>;<AFI>;IGNORE
+<UFEF1> <yeh>;<AYE>;<AIS>;IGNORE
+<UFEF2> <yeh>;<AYE>;<AFI>;IGNORE
+<UFEF3> <yeh>;<AYE>;<AII>;IGNORE
+<UFEF4> <yeh>;<AYE>;<AME>;IGNORE
+<UFEF5> "<lam><alefmadda>";"<BAS><BAS>";"<AIS><AFI>";IGNORE
+<UFEF6> "<lam><alefmadda>";"<BAS><BAS>";"<AFI><AFI>";IGNORE
+<UFEF7> "<lam><hamza>";"<BAS><AHA>";"<AIS><AFI>";IGNORE
+<UFEF8> "<lam><hamza>";"<BAS><AHA>";"<AFI><AFI>";IGNORE
+<UFEF9> "<lam><hamza>";"<BAS><AHS>";"<AIS><AFI>";IGNORE
+<UFEFA> "<lam><hamza>";"<BAS><AHS>";"<AFI><AFI>";IGNORE
+<UFEFB> "<lam><alef>";"<BAS><BAS>";"<AIS><AFI>";IGNORE
+<UFEFC> "<lam><alef>";"<BAS><BAS>";"<AFI><AFI>";IGNORE
+reorder-end
+