01: Inventory of the Sinhala Writing System

Scope: the inventory of the Sinhala writing system (what symbols exist, what they are called, how they are classified, and which ones are live, rare or obsolete) together with how they are encoded in Unicode and SLS 1134. Spelling rules, sandhi, and the design of romanization schemes are out of scope except for the closing "Implications for romanization and transliteration" section.

Compiled October 2026.

Letter IDs (used throughout this repository): consonants ka kha ga gha nga nnga(ඟ) ca cha ja jha nya(ඤ) jnya(ඥ) nyja(ඦ) tta ttha dda ddha nna nndda(ඬ) ta tha da dha na nda(ඳ) pa pha ba bha ma mba(ඹ) ya ra la va sha(ශ) ssa(ෂ) sa ha lla(ළ) fa; vowels a aa ae aee i ii u uu ru ruu ilu iluu e ee ai o oo au; hal = al-lakuna.

0. Summary

The Unicode Sinhala block (U+0D80–U+0DFF) encodes 3 "various signs" (ඁ ං ඃ), 18 independent vowels, 41 consonants, 1 al-lakuna, 19 dependent vowel signs (17 in the main run + 2 "additional"), 10 Lith digits and 1 punctuation mark (෴). Sinhala Archaic Numbers (U+111E1–U+111F4) adds 20 historical numerals. ZWJ (U+200D) is mandatory for every conjunct, yansaya, rakaransaya and repaya; ZWNJ (U+200C) has only marginal, display-oriented roles.
The "60-letter" modern alphabet (NIE 1989: 18 vowels + 42 "consonants", the 42 including ං and ඃ) is the school standard. SLS 1134 counts 61 (18 vowels + 41 consonants + 2 semi-consonants) because it adds ඥ (jnya), which the 1989 NIE list lacks.
Historically: Sidat Sangarā (13th c.) 30 letters → Eḷu / Śuddha 32 (adds ඇ ඈ) → Vadan-kavi 50 → Miśra 54 → NIE 1989 60 (adds ෆ + five sanyaka) → SLS/Unicode 61 (adds ඥ). Counts and what was added at each stage differ between sources (see §11).
In modern use, ඏ ඐ ඎ ෟ(alone) ෳ ඁ ඦ are effectively dead; ඞ ඣ ඪ are Pali/Sanskrit-only rarities; ෆ and ඇ/ඈ do most of the work for English sounds; there is no letter for /z/.

1. Sources (cited by key in the tables)

Key	Source	Notes
U-CH	Unicode 18.0 code chart, Sinhala: https://www.unicode.org/charts/PDF/U0D80.pdf	Normative names, aliases, decompositions. Read directly.
U-AN	Unicode 18.0 code chart, Sinhala Archaic Numbers: https://www.unicode.org/charts/PDF/U111E0.pdf	Read directly.
U-SP	Unicode Core Spec ch. 13 (v16.0), §13.2 Sinhala: https://www.unicode.org/versions/Unicode16.0.0/core-spec/chapter-13/	ZWJ rules, u/uu contextual forms, candrabindu, kunddaliya. Consulted in summary form.
U-NS	Unicode NamedSequences.txt: https://www.unicode.org/Public/UCD/latest/ucd/NamedSequences.txt	Yansaya / rakaransaya / repaya sequences. Read directly.
U-L2	L2/08-105, J. B. Disanayaka for ICTA & SLSI, "Observations on the Encoding of Archaic Sinhala Numerals", Feb 2008: https://www.unicode.org/L2/L2008/08105-sinhala.pdf	Read directly.
CP	codepoints.net (Unicode version of first encoding): https://codepoints.net/U+0D81 , https://codepoints.net/U+0DE6 , https://codepoints.net/U+0DC6	Secondary, derived from UCD DerivedAge.
SLS	SLS 1134 Draft, 2nd revision (SLSI, draft dated 2004-04-26): https://sinhala.sourceforge.net/archive/akuru.org/att-0028/sls1134.pdf	Read directly (pages 1–20). The final published SLS 1134:2004 / :2011 texts were not available; the draft may differ.
WP-EN	Wikipedia, "Sinhala script": https://en.wikipedia.org/wiki/Sinhala_script	Lead only.
WP-SI	සිංහල විකිපීඩියා, "සිංහල හෝඩි": https://si.wikipedia.org/wiki/සිංහල_හෝඩි	Lead only; consulted in summary form.
WP-NUM	Wikipedia, "Sinhala numerals": https://en.wikipedia.org/wiki/Sinhala_numerals	Lead only.
KLN	University of Kelaniya, නෙමඩල, "සිංහල හෝඩියේ උත්පත්තිය", බී. ලක්ෂිකා මදුශානි, 2022-07-11: https://units.kln.ac.lk/nemadala/index.php/visheshanga/gaweshana/295-2022-07-11-03-29-22	University student-research portal; the most detailed Sinhala-language source consulted for alphabet history.
LD	Lankadeepa, "'ෆ' අක්ෂරය සිංහල හෝඩියට ආ හැටි": https://www.lankadeepa.lk/mathaka_ha_mathaka/ෆ-අකෂරය-සහල-හඩයට-ආ-හට/287-656225	Newspaper; history of ෆ.
B-AK	akurusinhala.blogspot.com, "සිංහල අක්‍ෂර වර්ගීකරණය" (2016): http://akurusinhala.blogspot.com/2016/10/blog-post.html	Blog; classification tables (corroborative only).
B-BN	bingunada.blogspot.com, "සිංහල භාෂාවේ පිලි" (2014): http://bingunada.blogspot.com/2014/09/blog-post.html	Blog; pili names (corroborative only).
B-IH	ihhodiya.blogspot.com, "අපේ සිංහල හෝඩිය" (2018): http://ihhodiya.blogspot.com/2018/09/2.html	Blog; alphabet history.
B-S4	sinhala4all.weebly.com, "අක්ෂර මාලාව": https://sinhala4all.weebly.com/34613482353035223515-35123535351735353520.html	Teacher site; alphabet history.
YM	Yamu.lk, "සිංහල හෝඩිය ගැන හැංගුණු කතාවක්" (2019-11-22): https://www.yamu.lk/trending/sinhala-alphabet-doubts/	Popular article; ඥ/z commentary.
ST	Sunday Times (LK), 1998-10-04: https://www.sundaytimes.lk/981004/plus2.html	Disanayaka interview: "from around 37 letters… up to 60".

Not consulted: the NIE / Educational Publications Department Grade 6 textbook lesson "අක්ෂරමාලාව හා පිල්ලම්" (listed at https://govdoc.lk/lesson-view?id=3986&fid=62e8a270be286; the text itself was not available), and the Sri Lanka Sinhala Encyclopedia article "අක්ෂරමාලාව (සිංහල)" at encyclopedia.gov.lk (the site was unavailable). Anything attributed below to "school grammar" is therefore corroborated only by secondary Sinhala sources. It is given medium confidence pending a check against the textbook.

2. Table 1: Unicode Sinhala block, complete (U+0D80–U+0DFF)

Columns: ID = letter ID used in this repository. Unicode name is the normative name (with its = alias). Sinhala name is the traditional name. For letters these follow the "-යන්න" pattern the Unicode names were built from (WP-EN, SLS Table 4). The Sinhala spelling is a rendering of the Unicode/SLS romanized name made for this study; where B-BN/B-AK give a different everyday form, that form is shown too. Cat.: SIGN = ayogavaha/various sign; IV = independent vowel; C = consonant; VS = dependent vowel sign; D = digit; P = punctuation. Set: Ś = in the Śuddha (Eḷu) set, M = Miśra-only, N = added in the modern/NIE era. Status: live / rare / obsolete. "Since" = first Unicode version (CP, U-SP).

Unassigned in the block: 0D80, 0D84, 0D97–0D99, 0DB2, 0DBC, 0DBE–0DBF, 0DC7–0DC9, 0DCB–0DCE, 0DD5, 0DD7, 0DE0–0DE5, 0DF0–0DF1, 0DF5–0DFF (U-CH). SLS says 0D97–0D99 and 0DC7–0DC9 were held back deliberately for future vowels and consonants.

2a. Various signs (ayogavaha)

CP	Glyph	ID	Unicode name (= alias)	Sinhala name	Cat.	Set	Status
0D81	ඁ	-	SINHALA SIGN CANDRABINDU	චන්ද්‍රබින්දු (candrabindu)	SIGN	-	obsolete; Unicode 13.0 (2020). Spec says it is for archaic Sanskrit texts only, not modern Sinhala (U-SP).
0D82	ං	-	SINHALA SIGN ANUSVARAYA (= anusvara)	අනුස්වාරය / බින්දුව	SIGN	Ś	live (very common: සිංහල, ලංකාව)
0D83	ඃ	-	SINHALA SIGN VISARGAYA (= visarga)	විසර්ගය	SIGN	M	rare (Sanskrit loans: දුඃඛ, පුනඃ)

2b. Independent vowels (ස්වර / ප්‍රාණාක්ෂර)

CP	Glyph	ID	Unicode name (= alias)	Sinhala name	Set	Status
0D85	අ	a	SINHALA LETTER AYANNA (= a)	අයන්න	Ś	live
0D86	ආ	aa	AAYANNA (= aa)	ආයන්න	Ś	live
0D87	ඇ	ae	AEYANNA (= ae)	ඇයන්න	Ś (post-Sidat-Sangarā)	live
0D88	ඈ	aee	AEEYANNA (= aae)	ඈයන්න	Ś (post-Sidat-Sangarā)	live
0D89	ඉ	i	IYANNA (= i)	ඉයන්න	Ś	live
0D8A	ඊ	ii	IIYANNA (= ii)	ඊයන්න	Ś	live
0D8B	උ	u	UYANNA (= u)	උයන්න	Ś	live
0D8C	ඌ	uu	UUYANNA (= uu)	ඌයන්න	Ś	live
0D8D	ඍ	ru	IRUYANNA (= vocalic r)	ඉරුයන්න	M	live-rare (ඍතුව, ඍෂි)
0D8E	ඎ	ruu	IRUUYANNA (= vocalic rr)	ඉරූයන්න	M	obsolete as a letter. Its sign ෲ survives (SLS note 3).
0D8F	ඏ	ilu	ILUYANNA (= vocalic l)	ඉලුයන්න	M	obsolete (SLS: "do not occur in present usage")
0D90	ඐ	iluu	ILUUYANNA (= vocalic ll)	ඉලූයන්න	M	obsolete
0D91	එ	e	EYANNA (= e)	එයන්න	Ś	live
0D92	ඒ	ee	EEYANNA (= ee)	ඒයන්න	Ś	live
0D93	ඓ	ai	AIYANNA (= ai)	ඓයන්න	M	live-rare (ඓතිහාසික)
0D94	ඔ	o	OYANNA (= o)	ඔයන්න	Ś	live
0D95	ඕ	oo	OOYANNA (= oo)	ඕයන්න	Ś	live
0D96	ඖ	au	AUYANNA (= au)	ඖයන්න	M	live-rare (ඖෂධ)

2c. Consonants (ව්‍යඤ්ජන / ගාත්‍රාක්ෂර)

CP	Glyph	ID	Unicode name (= alias)	Sinhala name	Set	Status
0D9A	ක	ka	ALPAPRAANA KAYANNA (= ka)	අල්පප්‍රාණ කයන්න	Ś	live
0D9B	ඛ	kha	MAHAAPRAANA KAYANNA (= kha)	මහාප්‍රාණ කයන්න (also "බයානු කයන්න", WP-EN)	M	live (loans)
0D9C	ග	ga	ALPAPRAANA GAYANNA (= ga)	අල්පප්‍රාණ ගයන්න	Ś	live
0D9D	ඝ	gha	MAHAAPRAANA GAYANNA (= gha)	මහාප්‍රාණ ගයන්න	M	live (loans: සංඝ, මේඝ)
0D9E	ඞ	nga	KANTAJA NAASIKYAYA (= nga)	කණ්ඨජ නාසිකය	M	rare. SLS: never takes a vowel, appears only as ඞ්.
0D9F	ඟ	nnga	SANYAKA GAYANNA (= nnga)	සඤ්ඤක ගයන්න	Ś (sanyaka)	live (අඟල, හඟින)
0DA0	ච	ca	ALPAPRAANA CAYANNA (= ca)	අල්පප්‍රාණ චයන්න	Ś (modern); absent from Sidat Sangarā list	live
0DA1	ඡ	cha	MAHAAPRAANA CAYANNA (= cha)	මහාප්‍රාණ චයන්න	M	live (loans: ඡායා)
0DA2	ජ	ja	ALPAPRAANA JAYANNA (= ja)	අල්පප්‍රාණ ජයන්න	Ś	live
0DA3	ඣ	jha	MAHAAPRAANA JAYANNA (= jha)	මහාප්‍රාණ ජයන්න	M	very rare (Pali ඣාන)
0DA4	ඤ	nya	TAALUJA NAASIKYAYA (= nya)	තාලුජ නාසිකය	M	live (ඤාණ, සඤ්ඤා)
0DA5	ඥ	jnya	TAALUJA SANYOOGA NAAKSIKYAYA (= jnya)	තාලුජ සංයෝග නාසිකය	M (not in NIE 1989 list; in SLS/Unicode)	live (ඥාති, විශේෂඥ, ප්‍රඥා)
0DA6	ඦ	nyja	SANYAKA JAYANNA (= nyja)	සඤ්ඤක ජයන්න	N (sanyaka)	obsolete/never used (SLS: not in contemporary writing)
0DA7	ට	tta	ALPAPRAANA TTAYANNA (= tta)	අල්පප්‍රාණ ටයන්න	Ś	live
0DA8	ඨ	ttha	MAHAAPRAANA TTAYANNA (= ttha)	මහාප්‍රාණ ටයන්න	M	rare (ශ්‍රේෂ්ඨ)
0DA9	ඩ	dda	ALPAPRAANA DDAYANNA (= dda)	අල්පප්‍රාණ ඩයන්න	Ś	live
0DAA	ඪ	ddha	MAHAAPRAANA DDAYANNA (= ddha)	මහාප්‍රාණ ඩයන්න	M	very rare
0DAB	ණ	nna	MUURDHAJA NAYANNA (= nna)	මූර්ධජ ණයන්න	Ś	live (spelling-only distinction from න)
0DAC	ඬ	nndda	SANYAKA DDAYANNA (= nndda)	සඤ්ඤක ඩයන්න	Ś (sanyaka)	live (කඬ, හඬ)
0DAD	ත	ta	ALPAPRAANA TAYANNA (= ta)	අල්පප්‍රාණ තයන්න	Ś	live
0DAE	ථ	tha	MAHAAPRAANA TAYANNA (= tha)	මහාප්‍රාණ තයන්න	M	live (loans: ස්ථාන, කථා)
0DAF	ද	da	ALPAPRAANA DAYANNA (= da)	අල්පප්‍රාණ දයන්න	Ś	live
0DB0	ධ	dha	MAHAAPRAANA DAYANNA (= dha)	මහාප්‍රාණ දයන්න	M	live (loans: ධර්ම, බුද්ධ)
0DB1	න	na	DANTAJA NAYANNA (= na)	දන්තජ නයන්න	Ś	live
0DB3	ඳ	nda	SANYAKA DAYANNA (= nda)	සඤ්ඤක දයන්න	Ś (sanyaka)	live (සඳ, ඳ is very frequent)
0DB4	ප	pa	ALPAPRAANA PAYANNA (= pa)	අල්පප්‍රාණ පයන්න	Ś	live
0DB5	ඵ	pha	MAHAAPRAANA PAYANNA (= pha)	මහාප්‍රාණ පයන්න	M	live-rare (ඵල)
0DB6	බ	ba	ALPAPRAANA BAYANNA (= ba)	අල්පප්‍රාණ බයන්න	Ś	live
0DB7	භ	bha	MAHAAPRAANA BAYANNA (= bha)	මහාප්‍රාණ බයන්න	M	live (භාෂාව)
0DB8	ම	ma	MAYANNA (= ma)	මයන්න	Ś	live
0DB9	ඹ	mba	AMBA BAYANNA (= mba)	අඹ බයන්න / සඤ්ඤක බයන්න	Ś (sanyaka)	live (අඹ, තඹ)
0DBA	ය	ya	YAYANNA (= ya)	යයන්න	Ś	live
0DBB	ර	ra	RAYANNA (= ra)	රයන්න	Ś	live
0DBD	ල	la	DANTAJA LAYANNA (= la; dental)	දන්තජ ලයන්න	Ś	live
0DC0	ව	va	VAYANNA (= va)	වයන්න	Ś	live
0DC1	ශ	sha	TAALUJA SAYANNA (= sha)	තාලුජ ශයන්න	M	live (ශ්‍රී, ශාලා)
0DC2	ෂ	ssa	MUURDHAJA SAYANNA (= ssa; retroflex)	මූර්ධජ ෂයන්න	M	live (භාෂා; English "sh": ෂෝ)
0DC3	ස	sa	DANTAJA SAYANNA (= sa; dental)	දන්තජ සයන්න	Ś	live
0DC4	හ	ha	HAYANNA (= ha)	හයන්න	Ś	live
0DC5	ළ	lla	MUURDHAJA LAYANNA (= lla; retroflex)	මූර්ධජ ළයන්න	Ś	live (spelling-only distinction from ල)
0DC6	ෆ	fa	FAYANNA (= fa)	ෆයන්න	N (1989)	live (English loans). Unicode 3.0 (1999).

2d. Al-lakuna and dependent vowel signs (පිලි / පිල්ලම්). See §6 for names.

CP	Glyph	ID	Unicode name (= alias)	Decomposition (U-CH)
0DCA	්	hal	SINHALA SIGN AL-LAKUNA (= virama)	-
0DCF	ා	aa	VOWEL SIGN AELA-PILLA (= aa)	-
0DD0	ැ	ae	VOWEL SIGN KETTI AEDA-PILLA (= ae)	-
0DD1	ෑ	aee	VOWEL SIGN DIGA AEDA-PILLA (= aae)	-
0DD2	ි	i	VOWEL SIGN KETTI IS-PILLA (= i)	-
0DD3	ී	ii	VOWEL SIGN DIGA IS-PILLA (= ii)	-
0DD4	ු	u	VOWEL SIGN KETTI PAA-PILLA (= u)	-
0DD6	ූ	uu	VOWEL SIGN DIGA PAA-PILLA (= uu)	-
0DD8	ෘ	ru	VOWEL SIGN GAETTA-PILLA (= vocalic r)	-
0DD9	ෙ	e	VOWEL SIGN KOMBUVA (= e)	- (pre-base glyph)
0DDA	ේ	ee	VOWEL SIGN DIGA KOMBUVA (= ee)	≡ 0DD9 0DCA
0DDB	ෛ	ai	VOWEL SIGN KOMBU DEKA (= ai)	none (not ≡ ෙ+ෙ)
0DDC	ො	o	VOWEL SIGN KOMBUVA HAA AELA-PILLA (= o)	≡ 0DD9 0DCF
0DDD	ෝ	oo	VOWEL SIGN KOMBUVA HAA DIGA AELA-PILLA (= oo)	≡ 0DDC 0DCA
0DDE	ෞ	au	VOWEL SIGN KOMBUVA HAA GAYANUKITTA (= au)	≡ 0DD9 0DDF
0DDF	ෟ	ilu	VOWEL SIGN GAYANUKITTA (= vocalic l)	-
0DF2	ෲ	ruu	VOWEL SIGN DIGA GAETTA-PILLA (= vocalic rr)	-
0DF3	ෳ	iluu	VOWEL SIGN DIGA GAYANUKITTA (= vocalic ll)	-

2e. Digits and punctuation

CP	Glyph	Unicode name	Notes
0DE6–0DEF	෦ ෧ ෨ ෩ ෪ ෫ ෬ ෭ ෮ ෯	SINHALA LITH DIGIT ZERO … NINE	Astrological ("Lith Illakkam"), positional, has a zero. Unicode 7.0 (2014).
0DF4	෴	SINHALA PUNCTUATION KUNDDALIYA	කුණ්ඩලිය. Cross-ref U+11FFF TAMIL PUNCTUATION END OF TEXT.

2f. Characters outside the block that Sinhala needs

CP	Name	Sinhala role (U-SP, U-NS, SLS §4.3/§5)
200D	ZERO WIDTH JOINER	Required to form any conjunct (bandi akuru), yansaya, rakaransaya, repaya, or touching letter. Without ZWJ, al-lakuna is always visible.
200C	ZERO WIDTH NON-JOINER	SLS only: ZWNJ + vowel sign shows a sign on its own (`‌ා`). ZWNJ ් ZWJ ය gives a stand-alone yansaya. ර ් ZWJ ZWNJ gives a stand-alone repaya. The Unicode Sinhala section does not mention ZWNJ (U-SP).
00A0	NO-BREAK SPACE	SLS lists it. The Unicode convention is NBSP or dotted circle as the base for showing a lone combining mark.
0964/0965	DEVANAGARI DANDA / DOUBLE DANDA	Not Sinhala-specific. Sometimes seen in Pali texts. Use in Sinhala text is not confirmed by a published source (open), and it is not encoded in the Sinhala block.

3. Table 2: Sinhala Archaic Numbers (U+111E0–U+111FF), "Sinhala Illakkam"

Non-positional, no zero (U-AN). 111E0 and 111F5–111FF are unassigned. Unicode 7.0 (2014) per WP-EN (CP not checked for this block).

CP	Name	CP	Name
111E1	SINHALA ARCHAIC DIGIT ONE	111EA	SINHALA ARCHAIC NUMBER TEN
111E2	… DIGIT TWO	111EB	… NUMBER TWENTY
111E3	… DIGIT THREE	111EC	… NUMBER THIRTY
111E4	… DIGIT FOUR	111ED	… NUMBER FORTY
111E5	… DIGIT FIVE	111EE	… NUMBER FIFTY
111E6	… DIGIT SIX	111EF	… NUMBER SIXTY
111E7	… DIGIT SEVEN	111F0	… NUMBER SEVENTY
111E8	… DIGIT EIGHT	111F1	… NUMBER EIGHTY
111E9	… DIGIT NINE	111F2	… NUMBER NINETY
		111F3	… NUMBER ONE HUNDRED
		111F4	… NUMBER ONE THOUSAND

4. Rules and facts: inventory and encoding (INV-001 … INV-014)

INV-001: Unicode letter inventory

Statement: The Sinhala block encodes 18 independent vowels (0D85–0D96) and 41 consonants (0D9A–0DC6, with gaps). Every consonant is one code point, including the 5 sanyaka letters, ඥ, and ෆ.
Examples: ඟ = U+0D9F (not ඞ්+ග); ඥ = U+0DA5 (not ඤ්‍ජ).
Exceptions: None.
Confidence: high. Sources: U-CH, SLS §5.2.

INV-002: Independent vowels are atomic

Statement: An independent vowel must be stored as its own code point, never as අ + vowel sign.
Examples: ආ is U+0D86, not 0D85 0DCF. ඒ is U+0D92, not එ + ්.
Exceptions: SLS describes ආ as අ + ා and ඒ as එ + ් in terms of how the letters are written. This is writing order only; the stored form must be the atomic code point.
Confidence: high. Sources: SLS §5.1 note, §6.2.

INV-003: Composite vowel signs and normalization

Statement: ේ ො ෝ ෞ have canonical decompositions (table 2d), so NFC and NFD differ for them. Store the single precomposed sign.
Examples: කෝ = 0D9A 0DDD. The sequence 0D9A 0DD9 0DCF 0DCA is permitted by Unicode but SLS discourages it.
Exceptions: ෛ (0DDB) has no decomposition. Two kombuvas (ෙෙ) are not canonically equal to ෛ, so any conversion or normalization step must map "kombuva twice" to 0DDB itself.
Confidence: high. Sources: U-CH, SLS §5.4 note 2.

INV-004: Al-lakuna and conjunct formation

Statement: Al-lakuna (්, 0DCA) is always visible and does not by itself form a conjunct. A consonant cluster becomes a ligature or reduced form only with C + ් + ZWJ + C.
Examples: ක්‍ෂ = ක ් ZWJ ෂ (kssa). න්‍ද = න ් ZWJ ද. ක්‍ව = ක ් ZWJ ව.
Exceptions: Touching letters use a different order (INV-005).
Confidence: high. Sources: U-SP §13.2, SLS §5.8.

INV-005: Touching letters (bandi akuru, Pali style)

Statement: Unicode encodes a touching conjunct as C + ZWJ + ් + C, with ZWJ before al-lakuna.
Examples: Pali text written in Sinhala script, e.g. a ka touching a ka in ධම්මචක්ක (glyph only).
Exceptions: Conflict: the 2004 SLS draft says touching letters use 0DCA 200D, the same order as conjuncts.
Confidence: medium (Unicode high; SLS final text not seen). Sources: U-SP, SLS §5.8 note.

INV-006: Named sequences for yansaya, rakaransaya, repaya

Statement: Three reduced forms have fixed sequences:
- yansaya (ය after a consonant) = 0DCA 200D 0DBA
- rakaransaya (ර after a consonant) = 0DCA 200D 0DBB
- repaya (ර before a consonant) = 0DBB 0DCA 200D, placed before the following consonant
Examples: ක්‍ය kya. ක්‍ර kra. ක්‍රෙ kre. කර්‍ම karma (repaya form).
Exceptions:
- Repaya is optional: කර්ම and කර්‍ම are both valid (SLS §3.5).
- SLS says yansaya is not written after ර; it marks its own example spelling with ර + yansaya incorrect.
Confidence: high. Sources: U-NS, SLS §5.6–5.7.

INV-007: ZWJ omission is meaningful

Statement: Leaving out the ZWJ gives an explicit-hal spelling. This is a legitimate spelling choice, not an error.
Examples: ක්ය (no ZWJ) vs ක්‍ය.
Exceptions: None.
Confidence: high. Sources: SLS §5.6 note 2.

INV-008: Ayogavaha are trailing combining marks

Statement: ං and ඃ are combining marks. They follow a vowel, a consonant with its inherent a, or a vowel sign, and they are always the last character of the cluster.
Examples:
- අං = 0D85 0D82
- කං = 0D9A 0D82
- කොං = 0D9A 0DDD 0D82 (SLS's own example)
- කුඃ = 0D9A 0DD4 0D83
Exceptions: They never follow a pure (hal) consonant.
Confidence: high. Sources: SLS §3.3, §5.5.

INV-009: Contextual vowel-sign shapes share one code

Statement: Shape variants of al-lakuna and of the u/uu signs are all encoded with the same code point.
Examples:
- The hal on ක් looks different from the hal on ට, but both are 0DCA.
- කු (0D9A 0DD4) has a different u-shape from නු (0DB1 0DD4).
- u/uu take alternative forms after ක ග ඟ ත භ ශ, and again after a rakaransaya.
Exceptions: None.
Confidence: high. Sources: SLS §5.4, U-SP.

INV-010: Irregular ligatures are ordinary sequences

Statement: ර and ළ have irregular ligatures with the u and ae signs, but they are still stored as plain consonant + sign.
Examples: රු = 0DBB 0DD4. රූ = 0DBB 0DD6. රැ = 0DBB 0DD0. ළු = 0DC5 0DD4.
Exceptions: SLS lists ළු ("muurdhaja lu") as a distinct form for convenience, but there is no separate code for it.
Confidence: high. Sources: SLS §5.4, §6.1.

INV-011: Candrabindu is not modern Sinhala

Statement: U+0D81 is used only for archaic Sanskrit texts. Transliteration of ordinary modern text should not produce it.
Examples: none
Exceptions: None.
Confidence: high. Sources: U-SP, CP.

INV-012: Two numeral systems; modern text uses European digits

Statement:
- Sinhala Illakkam (111E1–111F4): archaic, no zero, separate signs for 10–90, 100 and 1000. Used before 1815; the Kandyan Convention clause numbers are an example.
- Sinhala Lith Illakkam (0DE6–0DEF): astrological, positional, has a zero. Used for horoscopes into the 20th century.
- Everyday Sinhala uses 0–9.
Examples: none
Exceptions: Disanayaka (for ICTA/SLSI, 2008) argued that the "archaic" numerals also had a zero and a place-holder concept, and opposed encoding the 11 compound numerals (10…1000). Unicode encoded them anyway.
Confidence: high (encoding); medium (historical claims). Sources: U-CH, U-AN, U-L2, WP-NUM.

INV-013: Kunddaliya

Statement: ෴ (කුණ්ඩලිය, 0DF4) is a historical full stop or ornament. Today it appears mainly to close a paragraph or as decoration, including on social media. Modern punctuation is Western (. , ? !).
Examples: none
Exceptions: None.
Confidence: high. Sources: U-SP, SLS §4.2, WP-EN.

INV-014: Unicode history

Statement:
- Sinhala block (incl. ෆ and ෴): Unicode 3.0 (1999)
- Lith digits 0DE6–0DEF and the Archaic Numbers block: Unicode 7.0 (2014)
- Candrabindu 0D81: Unicode 13.0 (2020)
- SLS 1134 and its 2001 revision were the basis for ISO/IEC 10646 Sinhala.
Examples: none
Exceptions: None.
Confidence: high (CP for 0D81/0DE6/0DC6); medium (archaic block). Sources: CP, SLS Foreword.

5. Table 3: Hodiya: alphabets through history

Stage	Name	Vowels	Consonants	Total	What changed	Sources	Conf.
13th c. (Dambadeniya)	සිදත් සඟරා හෝඩිය	10: අ ආ ඉ ඊ උ ඌ එ ඒ ඔ ඕ	20: ක ග ජ ට ඩ ණ ත ද න ප බ ම ය ර ල ව ස හ ළ අං	30	Classical Eḷu. No ඇ ඈ, no ච, ං counted as a consonant.	KLN, WP-SI, B-IH, B-S4	high
(later)	එළු හෝඩිය / ශුද්ධ සිංහල (අමිශ්‍ර) හෝඩිය	12 (adds ඇ ඈ)	20	32	Adds ඇ ඈ. KLN says ඇ ඈ were added but still gives 30 (internal inconsistency).	B-IH, B-S4, WP-SI; KLN conflicting	medium
Kandyan (Mahanuwara) period	වදන් කවි (වඩන කවි) හෝඩිය	16	34	50	Adds Pali/Sanskrit letters. Used in temple teaching ("pansal hodiya") into the 20th c.	KLN, B-IH	medium
1891 (A. M. Gunasekara, Comprehensive Grammar)	මිශ්‍ර සිංහල හෝඩිය	18	36	54	Adds 4 to Vadan-kavi (ඍ ඎ ඏ ඐ ඓ ඖ ශ ෂ … per KLN). Gunasekara also proposed ෆ.	KLN, LD, WP-SI	medium
1989 (NIE, Maharagama; Sinhala Lekhana Rītiya committee)	නූතන / සම්මත සිංහල හෝඩිය	18	42	60	Adds ෆ and five sanyaka (ඟ ඦ ඬ ඳ ඹ). ං ඃ written අං අඃ. ඥ not included.	KLN, WP-SI, B-S4, LD, YM, ST	medium-high
1990 (J. B. Disanayaka)	සමකාලීන සිංහල හෝඩිය	-	-	(≈60)	Drops ඏ ඐ and ඦ; adds ඥ; KLN also mentions new "closed" vowel notations. Details unclear.	KLN, WP-SI	low
2004 (SLS 1134 rev. 2) / Unicode	encoding standard	18	41 + 2 semi-consonants	61	All 41 Unicode consonants, incl. both ඥ and ඦ.	SLS §3	high

INV-015: Śuddha vs Miśra (modern definition)

Statement: "Śuddha" (Eḷu) letters are enough for native Sinhala words. "Miśra" adds letters needed only for tatsama (Sanskrit/Pali) or English loans: the aspirates, extra sibilants, ඍ-series and ඏ-series, ඓ ඖ, ඃ, ඞ ඤ ඥ, and ෆ.
- Śuddha (modern sense) = අ ආ ඇ ඈ ඉ ඊ උ ඌ එ ඒ ඔ ඕ + ක ග ච ජ ට ඩ ණ ත ද න ප බ ම ය ර ල ව ස හ ළ + ඟ ඬ ඳ ඹ + ං
- Miśra = the 60/61-letter set
Examples: Native ගඟ (ganga, "river") uses only Śuddha letters. Tatsama ගංගා (gangaa) uses ං. Loan ධර්මය (dharmaya) needs Miśra ධ.
Exceptions:
- The classical Sidat Sangarā list has no ච and no sanyaka letters. WP-EN's "Śuddha" includes ච and the 4 sanyaka, because it describes modern phonemes.
- ණ and ළ are Śuddha even though they are no longer distinct phonemes.
- Using Miśra letters is partly a matter of prestige or etymology (WP-EN).
Confidence: medium. Sources: WP-EN, WP-SI, KLN.

INV-016: Why "60 vs 61"

Statement: The modern count is reconciled as follows:
- NIE 60 = 18 vowels + 42 consonants, where the 42 are 40 consonant letters (25 varga + 5 sanyaka + ය ර ල ව + ශ ෂ ස හ ළ ෆ) plus ං and ඃ.
- SLS 61 = 18 vowels + 41 consonants (the same 40 + ඥ) + 2 semi-consonants.
- So SLS 61 = NIE 60 + ඥ.
Examples: none
Exceptions:
- WP-SI's 42-list, as consulted in summary form, is one letter short. ඦ is presumed to be the missing letter because KLN says 1989 added ඦ (see Q2).
- Some sources say "18 + 42" while also listing ඥ.
Confidence: medium. Sources: SLS §3, KLN, WP-SI, YM.

INV-017: Alphabet order

Statement: Traditional order (followed by Unicode/SLS):
1. Vowels: අ ආ ඇ ඈ ඉ ඊ උ ඌ ඍ ඎ ඏ ඐ එ ඒ ඓ ඔ ඕ ඖ
2. Ayogavaha: ං ඃ (NIE places them after the vowels as අං අඃ)
3. Varga rows, each followed by its sanyaka: ක ඛ ග ඝ ඞ ඟ / ච ඡ ජ ඣ ඤ ඥ ඦ / ට ඨ ඩ ඪ ණ ඬ / ත ථ ද ධ න ඳ / ප ඵ බ භ ම ඹ
4. ය ර ල ව ශ ෂ ස හ ළ ෆ
Examples: SLS put ං ඃ at the start of the code page (0D82–0D83) to help collation.
Exceptions: SLS warns that sorting still needs a dedicated collation algorithm.
Confidence: high. Sources: U-CH, SLS §3.3 note 2, §4.

6. Table 4: Pili (vowel signs and other strokes): names

Every vowel sign is written after the consonant in memory, even when it is drawn before it (ෙ). The Sinhala names come from SLS Table 1 and §6.1 (romanized), Unicode names, and B-BN. Alternate everyday names are separated by "/".

CP	Sign	Vowel ID	Sinhala name(s)	Romanized (SLS/Unicode)	Position	Notes
0DCA	්	hal	හල් ලකුණ / හල් කිරීම / ඇල (al)	al-lakuna (virama)	above / attached	"Hal kirīma" names the act of removing the vowel; the sign is the al/hal lakuna. There are two glyph forms (on ක vs on ට), but one code.
0DCF	ා	aa	ඇලපිල්ල	aela-pilla	right
0DD0	ැ	ae	කෙටි ඇදය / කෙටි ඇදපිල්ල	ketti aeda-pilla	right-lower
0DD1	ෑ	aee	දිග ඇදය / දිග (දික්) ඇදපිල්ල	diga aeda-pilla	right-lower
0DD2	ි	i	කෙටි ඉස්පිල්ල	ketti is-pilla	above
0DD3	ී	ii	දිග (දික්) ඉස්පිල්ල	diga is-pilla	above
0DD4	ු	u	කෙටි පාපිල්ල (the hooked form: කෙටි වක් පාපිල්ල)	ketti paa-pilla 1 / 2	below	SLS lists two shapes (7, 7a). B-BN uses "වක් පාපිල්ල" for the hooked form after ක ග ත …
0DD6	ූ	uu	දිග පාපිල්ල (hooked: දිග වක් පාපිල්ල)	diga paa-pilla 1 / 2	below
0DD8	ෘ	ru	ගැටපිල්ල / කෙටි ගැටපිල්ල	gaetta-pilla	right
0DF2	ෲ	ruu	දිග ගැටපිල්ල / ගැටපිලි දෙක	diga gaetta-pilla	right	Rare (පිතෲ).
0DD9	ෙ	e	කොම්බුව	kombuva	left (pre-base)	Written before the consonant but stored after it.
0DDA	ේ	ee	කොම්බුව හා හල් ලකුණ / දිග කොම්බුව (B-BN: කොම්බුව හා උස්පිල්ල)	diga kombuva	left + above	≡ ෙ + ්
0DDB	ෛ	ai	කොම්බු දෙක (ද්විත්ව කොම්බුව)	kombu deka	left
0DDC	ො	o	කොම්බුව හා ඇලපිල්ල	kombuva haa aela-pilla	two-part	≡ ෙ + ා
0DDD	ෝ	oo	කොම්බුව හා දිග ඇලපිල්ල	kombuva haa diga aela-pilla	two-part	≡ ො + ්
0DDE	ෞ	au	කොම්බුව හා ගයනුකිත්ත	kombuva haa gayanukitta	two-part	≡ ෙ + ෟ
0DDF	ෟ	ilu	ගයනුකිත්ත	gayanukitta	right	Alone it means vocalic l (obsolete). It is used as part of ෞ and ඖ.
0DF3	ෳ	iluu	දිග ගයනුකිත්ත / ගයනුකිති දෙක	diga gayanukitta	right	Unused (SLS note 4).
0D82	ං	-	බින්දුව / අනුස්වාරය	anusvaraya	right	ayogavaha
0D83	ඃ	-	විසර්ගය	visargaya	right	ayogavaha
(seq.)	්‍ය	-	යංශය	yansaya	right	"non-vocalic stroke" (SLS)
(seq.)	්‍ර	-	රකාරාංශය	rakaransaya	below
(seq.)	ර්‍	-	රේඵය	repaya	above next consonant

INV-018: Pili in SLS table 2

Statement: Each consonant can combine with 17 vocalic forms: hal, inherent a, and 15 vowel signs. ෟ and ෳ are excluded as obsolete.
- With yansaya: 8 valid vowel combinations. With rakaransaya: 12.
- Adding ං or ඃ gives SLS's figure of "109 possible letters" per consonant.
Examples: ක් ක කා කැ කෑ කි කී කු කූ කෘ කෲ කෙ කේ කෛ කො කෝ කෞ. Yansaya set: ක්‍ය ක්‍යා ක්‍යු ක්‍යූ ක්‍යෙ ක්‍යේ ක්‍යො ක්‍යෝ.
Exceptions:
- SLS: not every combination is valid for every consonant. ඞ appears only as ඞ්.
- SLS's own yansaya table lists kyu/kyuu, which are rare in practice.
Confidence: high (as SLS defines it). Sources: SLS §3.4–3.5, Tables 1–3.

7. Table 5: Classification of letters

7a. Vowels (ස්වර)

Class	Sinhala term	Members (IDs)	Conf.
Short	හ්‍රස්ව	අ ඇ ඉ උ ඍ ඏ එ ඔ (a ae i u ru ilu e o)	high
Long	දීර්ඝ	ආ ඈ ඊ ඌ ඎ ඐ ඒ ඕ (aa aee ii uu ruu iluu ee oo)	high
Diphthong	සන්ධ්‍යක්ෂර / සංයුක්ත ස්වර	ඓ ඖ (ai au). Usually counted with the long vowels.	medium
Unique to Sinhala among Indic scripts	-	ඇ ඈ (ae aee). SLS: "unique to the Sinhala language… since the 7th century".	high (SLS)
Sounds vs letters	-	SLS: the 61 symbols represent 40 sounds (14 vowel + 26 consonant).	high (as claim)

7b. Consonants: varga grid (place × manner)

Varga (Sinhala)	Place (Sinhala)	aghosha alpaprana	aghosha mahaprana	ghosha alpaprana	ghosha mahaprana	nasal (වර්ගාන්ත / අනුනාසික)	sanyaka (prenasalised)
ක වර්ගය	කණ්ඨජ (velar)	ක ka	ඛ kha	ග ga	ඝ gha	ඞ nga	ඟ nnga
ච වර්ගය	තාලුජ (palatal)	ච ca	ඡ cha	ජ ja	ඣ jha	ඤ nya	ඦ nyja
ට වර්ගය	මූර්ධජ (retroflex)	ට tta	ඨ ttha	ඩ dda	ඪ ddha	ණ nna	ඬ nndda
ත වර්ගය	දන්තජ (dental)	ත ta	ථ tha	ද da	ධ dha	න na	ඳ nda
ප වර්ගය	ඕෂ්ඨජ (labial)	ප pa	ඵ pha	බ ba	භ bha	ම ma	ඹ mba

Non-varga consonants:

Class	Sinhala term	Members	Place	Conf.
Semivowels / liquids	අන්තස්ථ	ය ya (palatal), ර ra (retroflex/alveolar), ල la (dental), ව va (dento-labial)	-	high
Sibilants / fricatives	ඌෂ්ම	ශ sha (palatal), ෂ ssa (retroflex), ස sa (dental), හ ha (velar/glottal). B-AK also lists ඃ and ෆ (6 in all).	-	medium (4-member list high; 6-member list from one blog)
Retroflex lateral	-	ළ lla (මූර්ධජ). Traditionally grouped apart from the varga letters.	මූර්ධජ	high
Dento-labial	දන්තෝෂ්ඨජ	ව va, ෆ fa	-	medium
Velar-palatal / velar-labial vowels	කණ්ඨතාලුජ / කණ්ඨෝෂ්ඨජ	එ ඒ ඓ / ඔ ඕ ඖ (Sanskrit tradition)	-	medium
ඥ jnya	තාලුජ සංයෝග නාසිකය	Historically the cluster ජ්+ඤ. Pronounced [gn]/[gɲ] in modern speech.	palatal	medium

INV-019: Alpaprana / mahaprana

Statement: Each varga has two unaspirated (අල්පප්‍රාණ) and two aspirated (මහාප්‍රාණ) stops.
- Mahaprana = ඛ ඝ ඡ ඣ ඨ ඪ ථ ධ ඵ භ (10)
- Alpaprana = ක ග ච ජ ට ඩ ත ද ප බ (10)
- Modern Sinhala does not pronounce aspiration. The choice of letter is etymological (spelling only).
Examples: ධර්මය is pronounced like darmaya. කථාව and කතාව are both seen; the spelling is debated.
Exceptions: Nasals, semivowels, sibilants and ha are classified as neither (or as alpaprana in some grammars).
Confidence: high (classes); high (no phonemic aspiration, WP-EN). Sources: B-AK, U-CH names, WP-EN.

INV-020: Ghosha / aghosha (voicing)

Statement:
- Aghosha (voiceless) = the first two letters of each varga (ක ඛ ච ඡ ට ඨ ත ථ ප ඵ) + ශ ෂ ස (+ ෆ, ඃ by extension)
- Ghosha (voiced) = the remaining varga letters, the nasals, the sanyaka letters, ය ර ල ව, හ, ළ, and all vowels
Examples: none
Exceptions: හ is traditionally ghosha in Sanskrit phonetics; Sinhala sources follow that. Not confirmed in a primary Sinhala textbook (open; see Q10).
Confidence: medium. Sources: B-AK (corroborative only); Sanskrit tradition.

INV-021: Sanyaka (සඤ්ඤක / අර්ධ නාසික) letters

Statement: There are 5 prenasalised stops: ඟ ඦ ඬ ඳ ඹ ("half-nasal", romanized by SLS as nng, ndj, nnd, nd, mb). Each is a single letter and a single code point. Sinhala treats them as one segment, distinct from a full nasal plus a stop.
Examples:
- අඟල (anngala) vs අංග (anga)
- සඳ (sanda, "moon") vs සන්ද (sanda, a name)
- කඬ / හඬ (hanndda); අඹ (amba)
Exceptions: ඦ is never used. Native spellings use ඳ etc.; tatsama words use න්ද etc.
Confidence: high. Sources: SLS §3, U-CH, WP-EN.

8. Table 6: Rare, obsolete, Pali/Sanskrit-only, and foreign-sound letters

Letter	ID	Status	Typical domain	Example	Conf.	Source
ඏ ඐ	ilu iluu	obsolete; SLS keeps them only "for completeness"; Disanayaka 1990 dropped them	Sanskrit grammar	(ඏකාරය as a citation only)	high	SLS, KLN, YM
ෟ (alone), ෳ	ilu iluu signs	obsolete (ෟ survives inside ෞ/ඖ)	-	-	high	SLS note 4
ඎ	ruu	letter obsolete; sign ෲ used in a handful of Sanskrit words	Sanskrit	මාතෲ (NLPC 65; ෘ in මාතෘ is commoner, NLPC 2,103), පිතෲ (rare: NLPC 0 exact, 4 with endings)	medium	SLS note 3; NLPC
ඍ / ෘ	ru	live in tatsama words	Sanskrit	ඍතුව, කෘෂිකර්ම, වෘක්ෂ, ගෘහ	high	SLS, WP-EN
ඓ ඖ / ෛ ෞ	ai au	live, low frequency	Sanskrit	ඓතිහාසික, වෛද්‍ය, ඖෂධ, පෞද්ගලික	high	-
ඃ	-	rare	Sanskrit	දුඃඛ, අතඃපුර	medium	-
ඁ	-	not for modern Sinhala	archaic Sanskrit	-	high	U-SP
ඞ	nga	rare; only as ඞ් before a velar	Pali/Sanskrit	වාඞ්මය, සඞ්ඝ (Pali)	medium	SLS
ඤ	nya	live	Pali/Sanskrit + native	ඤාණ, සඤ්ඤා	high	SLS
ඥ	jnya	live (high frequency in -ඥ "expert" words)	Sanskrit	ඥාති, විශේෂඥ, නීතිඥ, ප්‍රඥා	high	SLS, YM
ඦ	nyja	never used	-	-	high	SLS, WP-EN
ඣ ඪ	jha ddha	very rare	Pali/Sanskrit	ඣාන (jhāna)	medium	-
ඨ ඵ	ttha pha	rare	Sanskrit	ශ්‍රේෂ්ඨ, ඵල	high	-
ෆ	fa	live; English/foreign /f/	English	ෆැෂන්, ෆ්‍රාන්සය	high	LD, WP-EN
ඇ ඈ / ැ ෑ	ae aee	native vowels, also the default for English /æ/	English loans	බැංකුව (bank), කැෆේ	high	SLS §3
ෂ	ssa	in loans; also the usual letter for English /ʃ/	English	ෂෝ (show, NLPC 2,006), ෂර්ට් (shirt, NLPC 2,195)	medium	NLPC (usage); not codified in a published standard
(none)	z	no letter for /z/. Usually written with ස (sometimes ශ/ජ).	English	සූ (zoo), සීරෝ (zero, NLPC 104)	medium	YM; NLPC (usage)

INV-022: History of ෆ

Statement: A. M. Gunasekara first proposed ෆ in his 1891 grammar and used it in print in 1897. The NIE writing-rules committee formally adopted it in 1989. Before that, writers used ප.
Examples: ප්‍රැන්සිස් → ෆ්‍රැන්සිස් (Francis).
Exceptions: One Lankadeepa commenter claims government offices still do not recognize ෆ in official documents. Not confirmed by an official source (open).
Confidence: medium-high. Sources: LD, KLN, WP-SI.

INV-023: Pronunciation mergers among letters

Statement: Several letter pairs sound the same in modern Sinhala; only the spelling differs:
- ණ/න and ළ/ල are pronounced alike
- ශ/ෂ are both pronounced [ʃ]
- each mahaprana letter is pronounced like its alpaprana partner
- ඤ/ඥ sound the same only word-initially. Elsewhere ඥ behaves as two consonant sounds (SLS).
Examples: ඥාන ~ ඤාණ (word-initial). ප්‍රඥා (non-initial).
Exceptions: None.
Confidence: high (ණ/ළ/aspirates, WP-EN, YM, SLS). Sources: SLS §3.2 note 2, WP-EN, YM.

9. Numerals and punctuation (summary)

System	Range	Zero	Positional	Period of use	Status
Sinhala Illakkam (archaic)	111E1–111F4	no (Unicode). Disanayaka 2008 disputed this.	no	to 1815 and some palm-leaf use	historical
Sinhala Lith Illakkam	0DE6–0DEF	yes (the glyph resembles hal lakuna)	yes	horoscopes into the 20th c.	specialist
European digits	0030–0039	yes	yes	modern	default
Kunddaliya ෴	0DF4	-	-	historical full stop	decorative

10. Implications for romanization and transliteration

These points apply to anyone converting between a Latin-script romanization and Sinhala script, or validating stored Sinhala text.

Use atomic code points.
- Independent vowels, the composite signs ේ ො ෝ ෞ ෛ, and sanyaka consonants should each be produced and stored as their single code point (INV-002, INV-003).
- අ+ා, ෙ+ා, ෙ+ෙ and ඞ්+ග are not correct stored forms.
- Normalize to NFC; this composes ෙ+ා into ො, etc. It does not fix ෙ+ෙ.
ZWJ is part of spelling.
- A romanization needs explicit, predictable correspondences for:
  - conjuncts (C ් ZWJ C)
  - yansaya (C ් ZWJ ය)
  - rakaransaya (C ් ZWJ ර)
  - repaya (ර ් ZWJ C)
  - optionally touching letters (C ZWJ ් C)
- Common modern practice: use ZWJ for C+ya/C+ra and for ක්‍ෂ, but not for arbitrary clusters.
- An explicit-hal spelling (no ZWJ) must remain representable (INV-007).
Sanyaka vs nasal+stop is a real contrast.
- nnga, nndda, nda, mba must be kept distinct from n+g, n+d, m+b (INV-021).
- For example, a romanization must distinguish අඟල from අංගය, and සඳ from සන්ද.
Ayogavaha ං comes after the vowel. aṁ corresponds to ං after any vowel sign. ං and ඃ never follow a hal consonant (INV-008). ං is very frequent; ඃ is rare.
Many letters are spelling-only distinctions (INV-019, INV-023): aspirates, ණ/න, ළ/ල, ශ/ෂ.
- A romanization that is to be reversible needs distinct symbols for them (e.g. kh, N, L, sh/Sh); otherwise converting to Sinhala requires a dictionary.
- Pure phonetics cannot recover them.
Tier the inventory. The live set covers ordinary modern text.
- Low frequency: ඍ/ෘ, ඓ ඖ, ඃ, ඞ, ඣ ඪ, ෲ.
- ඏ ඐ ෟ(alone) ෳ ඁ ඦ should not appear in transliterations of ordinary modern text (INV-011, Table 6); they belong to archaic or Sanskrit material.
Foreign sounds.
- /f/ → ෆ
- English /æ/ → ඇ ඈ / ැ ෑ
- sh → ෂ or ශ, decided per word
- /z/ has no target letter. A romanization must state its policy (e.g. map to ස). Note that some existing romanization schemes use z for ඇ.
Irregular glyphs are ordinary sequences. රු රූ රැ ළු need no special codes; the font handles them (INV-010).
Kombuva is written first but stored after. A romanization is naturally in logical order (ke = ක + ෙ), which matches storage order.
Digits. Modern text uses European digits. Lith digits and ෴ are specialist or decorative characters.

11. Open questions / conflicting sources

#	Issue	Sources in conflict	Status
Q1	Is the modern alphabet 60 or 61 letters, and is ඥ part of it?	NIE 1989 = 60 (no ඥ, per KLN/WP-SI/YM) vs SLS 1134 = 61 (with ඥ)	Reconciled hypothesis INV-016. Verify against the NIE Grade 6 textbook.
Q2	Is ඦ in the 60?	KLN: added by NIE 1989 as one of five sanyaka; WP-SI 42-list (consulted in summary form) omits it; Disanayaka 1990 reportedly removed it	Open
Q3	Śuddha / Eḷu count: 30 or 32? Does it include ච and the sanyaka?	Sidat Sangarā list has no ච/sanyaka (WP-SI, KLN). WP-EN's "śuddha" includes ච and ඟ ඬ ඳ ඹ. KLN says Eḷu = 30 even after adding ඇ ඈ; B-IH/B-S4 say 32.	Medium. Different definitions: classical vs phonemic.
Q4	Mishra 54: did it already contain the sanyaka letters?	B-IH: yes (ඟ ඦ ඬ ඳ ඹ). KLN: sanyaka added only in 1989.	Open
Q5	Vadan-kavi hodiya letter list (16 + 34)	Sources give only counts, no list	Open
Q6	Touching-letter encoding: ZWJ before or after al-lakuna	Unicode: `C ZWJ ් C`. SLS 2004 draft: `C ් ZWJ C`	Unicode governs. Final SLS text unseen.
Q7	Disanayaka 1990 "samakaleena" alphabet: exact contents and the "closed vowel" additions	KLN only, and the available text is unclear on this point	Low confidence
Q8	Ushma set: 4 (ශ ෂ ස හ) or 6 (+ ඃ ෆ)?	Sanskrit tradition: 4. B-AK blog: 6	Open (not confirmed in a textbook)
Q9	Place of articulation for ඇ ඈ, and whether ය ර ල ව have assigned places in school grammar	No source located	Open
Q10	ghosha/aghosha membership of හ, sanyaka letters, and nasals in Sinhala school grammar	Inferred from the Sanskrit system	Open
Q11	Sinhala-script traditional names (e.g. "මහාප්‍රාණ කයන්න" vs "බයානු කයන්න"; "තාලුජ නාසිකය" vs "තාලුජ නාසික්‍යය")	Unicode/SLS give only romanized forms; the Sinhala spellings in Table 1 are renderings made for this study	Medium
Q12	Pili alternate names: ඇදය vs ඇදපිල්ල; දිග vs දික්; whether ේ is "කොම්බුව හා හල් ලකුණ" or "දිග කොම්බුව" in textbooks	SLS/Unicode vs B-BN	Medium
Q13	Did archaic Illakkam have a zero?	Unicode chart: no. Disanayaka/ICTA 2008: yes (palm-leaf evidence)	Historical dispute; encoding is settled
Q14	Unicode version for the Archaic Numbers block (7.0?)	WP-EN only; not checked against DerivedAge	Likely 7.0
Q15	Common orthography for English /z/ and /ʃ/	Only popular sources; no standard	Open. Needs corpus check.
Q16	Whether `ර ් ZWJ ය` renders as repaya + ya or as ra + yansaya, and which is correct for words like කාර්‍ය	SLS: yansaya not used after ර, and gives a special ZWNJ sequence for repaya + yansaya. Font behaviour varies.	Needs font testing (Noto Sinhala, Iskoola Pota)

01: Inventory of the Sinhala Writing System ​

0. Summary ​

1. Sources (cited by key in the tables) ​

2. Table 1: Unicode Sinhala block, complete (U+0D80–U+0DFF) ​

2a. Various signs (ayogavaha) ​

2b. Independent vowels (ස්වර / ප්‍රාණාක්ෂර) ​

2c. Consonants (ව්‍යඤ්ජන / ගාත්‍රාක්ෂර) ​

2d. Al-lakuna and dependent vowel signs (පිලි / පිල්ලම්). See §6 for names. ​

2e. Digits and punctuation ​

2f. Characters outside the block that Sinhala needs ​

3. Table 2: Sinhala Archaic Numbers (U+111E0–U+111FF), "Sinhala Illakkam" ​

4. Rules and facts: inventory and encoding (INV-001 … INV-014) ​

5. Table 3: Hodiya: alphabets through history ​

6. Table 4: Pili (vowel signs and other strokes): names ​

7. Table 5: Classification of letters ​

7a. Vowels (ස්වර) ​

7b. Consonants: varga grid (place × manner) ​

8. Table 6: Rare, obsolete, Pali/Sanskrit-only, and foreign-sound letters ​

9. Numerals and punctuation (summary) ​

10. Implications for romanization and transliteration ​

11. Open questions / conflicting sources ​