Re: [PATCH] Improve Malayalam language transliteration

[James Thomas <jimjoe@gmx.net>]

[-- Attachment #1: Type: text/plain, Size: 796 bytes --]

Eli Zaretskii <eliz@gnu.org> writes:

>
> That's okay, but we don't like deleting existing capabilities without
> some period during which it's deprecated.  So maybe say in NEWS that
> ITRANS is deprecated, and add a comment in quail/indian.el that it
> should be removed in some future release.
>
>> Sorry I missed the reference in lisp/language/indian.el. It seems to be
>> just defining the default input method for the language. Can be changed
>> easily if decided.
>
> If you think this new method is much better, it's okay to make it the
> default, I think.  Just mention that in NEWS.
>
> Thanks.
>
> P.S. And please keep the list address on the CC, so that this
> discussion is recorded.

Sorry for the bother, but this newer patch implements the Inscript stuff
in an easier way.

--
Jim

[-- Attachment #2: 0003-Improve-Malayalam-language-transliteration.patch --]
[-- Type: text/x-diff, Size: 9310 bytes --]

From 0b655da7d98d6ff5c6211d1a56e879ac291f9c34 Mon Sep 17 00:00:00 2001
From: James Thomas <jimjoe@gmx.net>
Date: Mon, 27 Apr 2020 08:06:48 +0530
Subject: [PATCH] Improve Malayalam language transliteration

The existing ITRANS scheme does not support some characters and
language quirks like 'chillu's.  The Inscript method has errors.
* lisp/language/ind-util.el
  (indian-mlm-base-table): + archaic chars, Mozhi combos; cleanup.
  (indian-mlm-mozhi-table): For new scheme Mozhi.
* lisp/leim/quail/indian.el
  (inscript-mlm-keytable): Correct errors.
  Add Inscript chillus & zero-width chars, Mozhi scheme.
* etc/NEWS: Mention change
Add a sufficient implementation of the Mozhi scheme.  Complete Inscript
implementation.
---
 etc/NEWS                  |  7 +++
 lisp/language/ind-util.el | 40 +++++++++++++++---
 lisp/leim/quail/indian.el | 89 +++++++++++++++++++++++++++++++++++----
 3 files changed, 122 insertions(+), 14 deletions(-)

diff --git a/etc/NEWS b/etc/NEWS
index 025d5c14a7..aa551177d1 100644
--- a/etc/NEWS
+++ b/etc/NEWS
@@ -288,6 +288,13 @@ prefix on the Subject line in various languages.
 These new navigation commands are bound to 'n' and 'p' in
 'apropos-mode'.

+** Quail
+
+---
+*** Improved Malayalam language transliteration
+Added new Mozhi scheme. The inapplicable ITRANS scheme is now
+deprecated. Errors in Inscript method corrected.
+
 \f
 * New Modes and Packages in Emacs 28.1

diff --git a/lisp/language/ind-util.el b/lisp/language/ind-util.el
index 4319e5537e..62885227f1 100644
--- a/lisp/language/ind-util.el
+++ b/lisp/language/ind-util.el
@@ -232,8 +232,8 @@ indian-mlm-base-table
   '(
     (;; VOWELS
      (?അ nil) (?ആ ?ാ) (?ഇ ?ി) (?ഈ ?ീ) (?ഉ ?ു) (?ഊ ?ൂ)
-     (?ഋ ?ൃ) (?ഌ nil) nil (?ഏ ?േ) (?എ ?െ) (?ഐ ?ൈ)
-     nil (?ഓ ?ോ) (?ഒ ?ൊ) (?ഔ ?ൌ) nil nil)
+     (?ഋ ?ൃ) (?ഌ ?ൢ) (?ൡ ?ൣ) (?ഏ ?േ) (?എ ?െ) (?ഐ ?ൈ)
+     nil (?ഒ ?ൊ) (?ഓ ?ോ) (?ഔ ?ൗ) (?് ?്) (?ൠ ?ൄ))
     (;; CONSONANTS
      ?ക ?ഖ ?ഗ ?ഘ ?ങ                  ;; GUTTRULS
      ?ച ?ഛ ?ജ ?ഝ ?ഞ                  ;; PALATALS
@@ -243,13 +243,16 @@ indian-mlm-base-table
      ?യ ?ര ?റ ?ല ?ള ?ഴ ?വ          ;; SEMIVOWELS
      ?ശ ?ഷ ?സ ?ഹ                    ;; SIBILANTS
      nil nil nil nil nil nil nil nil      ;; NUKTAS
-     "ജ്ഞ" "ക്ഷ")
+     "ജ്ഞ" "ക്ഷ"
+     "റ്റ" "ന്റ" "ത്ത" "ത്ഥ" "ഞ്ഞ" "ങ്ങ" "ന്ന"
+     "ഞ്ച" "ന്ക" "ങ്ക" "ച്ച" "ച്ഛ" "ക്ക"
+     "ബ്ബ" "ക്ക" "ഗ്ഗ" "ജ്ജ" "മ്മ" "പ്പ" "വ്വ" "ക്സ" "ശ്ശ")
     (;; Misc Symbols
      nil ?ം ?ഃ nil ?് nil nil)
     (;; Digits
      ?൦ ?൧ ?൨ ?൩ ?൪ ?൫ ?൬ ?൭ ?൮ ?൯)
-    (;; Inscript-extra (4)  (#, $, ^, *, ])
-     "്ര" "ര്" "ത്ര" "ശ്ര" nil)))
+    (;; Chillus
+     "ണ്" ?ൺ "ന്" ?ൻ "ര്" ?ർ "ല്" ?ൽ "ള്" ?ൾ)))

 (defvar indian-tml-base-table
   '(
@@ -323,6 +326,29 @@ indian-itrans-v5-table-for-tamil
     (;; misc -- 7
      ".N" (".n" "M") "H" ".a" ".h" ("AUM" "OM") "..")))

+(defvar indian-mlm-mozhi-table
+  '(;; for encode/decode
+    (;; vowels -- 18
+     "a" ("aa" "A") "i" ("ii" "I") "u" ("uu" "U")
+     "R" "Ll" "Lll" ("E" "ae") "e" "ai"
+     nil  "o"   "O"   "au"  "~" "RR")
+    (;; consonants -- 40
+     ("k" "c")   "kh"  "g"   "gh"  "ng"
+     "ch" ("Ch" "chh") "j" "jh" "nj"
+     "T"   "Th"  "D"   "Dh"  "N"
+     "th"  "thh" "d"   "dh"  "n"   nil
+     "p"   ("ph" "f")  "b"   "bh"  "m"
+     "y"   "r"   "rr"  "l"  "L" "zh" ("v" "w")
+     ("S" "z") "sh" "s" "h"
+     nil nil nil nil nil nil nil nil
+     nil "X"
+     ;; some of these are extra to Mozhi
+     ("t" "tt") "nt" "tth" "tthh" "nnj" "nng" "nn"
+     "nch" "nc" "nk" "cch" "cchh" "cc"
+     "B" ("C" "K" "q") "G" "J" "M" "P" "V" "x" "Z")
+    (;; misc -- 7
+     nil nil "H")))
+
 (defvar indian-kyoto-harvard-table
   '(;; for encode/decode
     (;; vowel
@@ -524,6 +550,10 @@ indian-mlm-itrans-v5-hash
   (indian-make-hash indian-mlm-base-table
 			  indian-itrans-v5-table))

+(defvar indian-mlm-mozhi-hash
+  (indian-make-hash indian-mlm-base-table
+			  indian-mlm-mozhi-table))
+
 (defvar indian-tml-itrans-v5-hash
   (indian-make-hash indian-tml-base-table
 			  indian-itrans-v5-table-for-tamil))
diff --git a/lisp/leim/quail/indian.el b/lisp/leim/quail/indian.el
index 2681eab0e5..100ae63f6a 100644
--- a/lisp/leim/quail/indian.el
+++ b/lisp/leim/quail/indian.el
@@ -117,6 +117,7 @@ "\\''"
  indian-knd-itrans-v5-hash "kannada-itrans" "Kannada" "KndIT"
  "Kannada transliteration by ITRANS method.")

+;; ITRANS not applicable to Malayalam & could be removed eventually
 (if nil
     (quail-define-package "malayalam-itrans" "Malayalam" "MlmIT" t "Malayalam ITRANS"))
 (quail-define-indian-trans-package
@@ -358,24 +359,23 @@ inscript-mlm-keytable
   '(
     (;; VOWELS  (18)
      (?D nil) (?E ?e) (?F ?f) (?R ?r) (?G ?g) (?T ?t)
-     (?+ ?=) ("F]" "f]") (?! ?@) (?S ?s) (?Z ?z) (?W ?w)
-     (?| ?\\) (?~ ?`) (?A ?a) (?Q ?q) ("+]" "=]") ("R]" "r]"))
+     (?= ?+) nil nil (?S ?s) (?Z ?z) (?W ?w)
+     nil (?~ ?`) (?A ?a) (?Q ?q))
     (;; CONSONANTS (42)
      ?k ?K ?i ?I ?U                ;; GRUTTALS
      ?\; ?: ?p ?P ?}               ;; PALATALS
      ?' ?\" ?\[ ?{ ?C              ;; CEREBRALS
-     ?l ?L ?o ?O ?v ?V             ;; DENTALS
+     ?l ?L ?o ?O ?v nil            ;; DENTALS
      ?h ?H ?y ?Y ?c                ;; LABIALS
-     ?/ ?j ?J ?n ?N "N]" ?b        ;; SEMIVOWELS
+     ?/ ?j ?J ?n ?N ?B ?b          ;; SEMIVOWELS
      ?M ?< ?m ?u                   ;; SIBILANTS
-     "k]" "K]" "i]" "p]" "[]" "{]" "H]" "/]" ;; NUKTAS
-     ?% ?&)
+     nil nil nil nil nil nil nil nil nil) ;; NUKTAS
     (;; Misc Symbols (7)
-     ?X ?x ?_ ">]" ?d "X]" ?>)
+     nil ?x ?_ nil ?d)
     (;; Digits
      ?0 ?1 ?2 ?3 ?4 ?5 ?6 ?7 ?8 ?9)
-    (;; Inscripts
-     ?# ?$ ?^ ?* ?\])))
+    (;; Chillus
+     "Cd" "Cd]" "vd" "vd]" "jd" "jd]" "nd" "nd]" "Nd" "Nd]")))

 (defvar inscript-tml-keytable
   '(
@@ -463,6 +463,9 @@ inscript-tml-keytable
  "malayalam-inscript" "Malayalam" "MlmIS"
  "Malayalam keyboard Inscript.")

+(quail-defrule "\\" ?‌)
+(quail-defrule "X" ?​)
+
 (if nil
     (quail-define-package "tamil-inscript" "Tamil" "TmlIS" t "Tamil keyboard Inscript"))
 (quail-define-inscript-package
@@ -571,4 +574,72 @@ inscript-tml-keytable
   ("?" ?\?)
   ("/" ?্))

+(defun indian-mlm-mozhi-update-translation (control-flag)
+  (let ((len (length quail-current-key)) chillu
+	(vowels '(?a ?e ?i ?o ?u ?A ?E ?I ?O ?U ?R)))
+    (cond ((numberp control-flag)
+	   (progn (if (= control-flag 0)
+		      (setq quail-current-str quail-current-key)
+		    (cond (input-method-exit-on-first-char)
+			  ((and (memq (aref quail-current-key
+					    (1- control-flag))
+				      vowels)
+				(setq chillu (cl-position
+					      (aref quail-current-key
+						    control-flag)
+					      '(?m ?N ?n ?r ?l ?L))))
+			   ;; conditions for putting chillu
+			   (and (or (and (= control-flag (1- len))
+					 (not (setq control-flag nil)))
+				    (and (= control-flag (- len 2))
+					 (let ((temp (aref quail-current-key
+							   (1- len))))
+                                           ;; is it last char of word?
+					   (not
+					    (or (and (>= temp ?a) (<= temp ?z))
+						(and (>= temp ?A) (<= temp ?Z))
+						(eq temp ?~))))
+					 (setq control-flag (1+ control-flag))))
+				(setq quail-current-str     ;; put chillu
+				      (concat (if (not (stringp
+							quail-current-str))
+						  (string quail-current-str)
+						quail-current-str)
+					      (string
+					       (nth chillu '(?ം ?ൺ ?ൻ ?ർ ?ൽ ?ൾ)))))))))
+		  (and (not input-method-exit-on-first-char) control-flag
+		       (while (> len control-flag)
+			 (setq len (1- len))
+			 (setq unread-command-events
+			       (cons (aref quail-current-key len)
+				     unread-command-events))))
+		  ))
+	  ((null control-flag)
+	   (unless quail-current-str
+	     (setq quail-current-str quail-current-key)
+	     ))
+	  ((equal control-flag t)
+	   (if (memq (aref quail-current-key (1- len))  ;; If vowel ending,
+		     vowels)                            ;; may have to put
+	       (setq control-flag nil)))))              ;; chillu. So don't
+  control-flag)                                         ;; end translation
+
+(quail-define-package "malayalam-mozhi" "Malayalam" "MlmMI" t
+                      "Malayalam transliteration by Mozhi method."
+                      nil nil t nil nil nil t nil
+                      'indian-mlm-mozhi-update-translation)
+
+(maphash
+ (lambda (key val)
+   (quail-defrule key (if (= (length val) 1)
+			  (string-to-char val)
+			(vector val))))
+ (cdr indian-mlm-mozhi-hash))
+
+(defun indian-mlm-mozhi-underscore (key len) (throw 'quail-tag nil))
+
+(quail-defrule "_" 'indian-mlm-mozhi-underscore)
+(quail-defrule "|" ?‌)
+(quail-defrule "||" ?​)
+
 ;;; indian.el ends here
--
2.20.1