Transliteration practice (quiz)
Knowing which scheme you are typing in is the single most common fix for "why didn't my word match." This is a cross-dictionary skill — it applies to every dictionary on the site, not one dictionary's structure. The 28 questions below check the material on the Encoding & Transliteration page: the schemes and where they diverge, why CDSL stores everything in SLP1, Vedic-accent marking, and the transcoder.
Each answer is hidden until you reveal it. To practice for real, set an input scheme in any display and try the conversions there.
Schemes — and where they diverge
rāma looks the same (rAma) in SLP1, Harvard-Kyoto, and ITRANS; śiva is where they split
(Siva / ziva / shiva). The sounds that actually diverge are the sibilants, vocalic
ṛ/ḷ, and the retroflexes.
Quiz · schemes
7 questions
Which transliteration scheme does CDSL use internally to store all of its Sanskrit text?
Show answer
SLP1 (Sanskrit Library Phonetic Basic) — one ASCII character per phoneme.The word rāma is written identically in which three ASCII schemes (and what is that spelling)?
Show answer
SLP1, Harvard-Kyoto, and ITRANS — all write it rAma. None of rāma's sounds diverge across those ASCII schemes, so for this word the input scheme barely matters.Which classes of sound actually diverge across the ASCII schemes (so that choosing the right input scheme matters)?
Show answer
The sibilants (ś, ṣ), the vocalic ṛ / ḷ, and the retroflex consonants.In Harvard-Kyoto, the sound ś (as in śiva) is written:
Show answer
z. SLP1 writes it S, ITRANS writes it sh, Harvard-Kyoto writes it z — three ASCII spellings of one letter.In ITRANS, the sound ś (as in śiva) is written:
Show answer
shWhich scheme is described as “Roman with diacritics (IAST-style)”?
Show answer
romanTwo transcoders exist but are NOT offered in the display input/output menus. Name them and what each is.
Show answer
wx (used at Hyderabad University) and as (Anglicized Sanskrit, Thomas Malten's letter-number scheme).
Why SLP1 internally
Quiz · why SLP1
2 questions
Why does CDSL store everything in SLP1 internally?
Show answer
SLP1 is lossless and unambiguous — every Sanskrit phoneme maps to exactly one ASCII character, so sorting, searching, and round-tripping are deterministic.In what encoding is the downloadable XML for each dictionary?
Show answer
SLP1.
Convert between schemes
Give the exact characters you would type or see. Mind the case — in SLP1 capital S is ś
but lowercase s is the dental s.
Quiz · conversions
8 questions
Write śiva (IAST) in SLP1.
Show answer
SivaWrite śiva (IAST) in Harvard-Kyoto.
Show answer
zivaWrite śiva (IAST) in ITRANS.
Show answer
shivaWrite śiva in Devanāgarī.
Show answer
शिवWrite rāma (IAST) in SLP1.
Show answer
rAmaWrite rāma in Devanāgarī.
Show answer
रामThe input box is set to slp1. You want to look up śiva — what exactly do you type?
Show answer
Siva— the capital S is SLP1's ś; lowercase s would be the dental s.The input box is set to hk (Harvard-Kyoto). You want śiva — what do you type?
Show answer
ziva
Vedic accents
Several dictionaries mark Vedic pitch accents — but you never have to type one, because the search key is stored unaccented.
Quiz · accents
5 questions
How is the udātta (high pitch) accent written in the SLP1 source?
Show answer
As / attached to its vowel — e.g. Monier-Williams keys agni as agni/ (accent on the final -i, agní).Do you have to type the Vedic accent to make a word match?
Show answer
No. Lookup uses the plain search key <k1> (e.g. agni); the accent lives on a separate display key <k2> (agni/). The accent is for display, not matching — typing agni finds the word either way.Which of these dictionaries marks NO Vedic accents?
Show answer
Apte (AP90). MW, PWG/PW, and Grassmann all mark accents; Grassmann centrally, as a Ṛg-Veda dictionary. Apte does not.What does the accent=yes|no toggle (and API parameter) govern?
Show answer
Whether accents appear in the Devanāgarī output.How do anudātta and svarita appear in the entry body of the German and Grassmann sources, and how does an accent render in Roman?
Show answer
anudātta and svarita are written \ and ^ in the source body; in Roman an accent renders as a combining acute (á, ā́).
Typing & the transcoder
Quiz · typing & transcoder
6 questions
You have a diacritic Roman word (rāma). Which input mode should you select?
Show answer
roman.You only have an ASCII keyboard (no diacritics). Which input schemes let you avoid diacritics entirely?
Show answer
Harvard-Kyoto (hk), ITRANS (itrans), or SLP1 (slp1).Your lookup failed. What is the single most likely cause?
Show answer
An input-scheme mismatch (e.g. typing roman while the box expects hk). Re-check the selected input scheme first.Who originally developed the project's transcoder, and in what language (with later ports)?
Show answer
Ralph Bunker, in Java — later ported to PHP (transcoder.php) and Python (funderburkjim/sanskrit-transcoding).How does the transcoder define each scheme-to-scheme conversion?
Show answer
From per-pair XML tables named {X}_{Y}.xml (e.g. slp1_deva.xml), compiled into a finite-state machine at runtime.What does transcoder_processString('rAma', 'slp1', 'deva') return?
Show answer
राम
See also
- Encoding & Transliteration — the reference these questions test
- Reading Monier-Williams (with quizzes) — the MW-specific quiz set
- Search & Display — choosing input/output schemes in the four display modes
- API — the
input/output/accentparameters