Source: Thyagaraja Vaibhavam — Ramaa Ramana Rara, fetched 2026-09-05.
rama_ramana_rara_explicit_charanams.txt contains the six lyric blocks from
HtmlTextExtractor.extract → normalize_garbled_diacritics. Navigation,
pronunciation guides, and commentary are omitted. The English language header
is added for the language-block parser path. All original section markers remain.rama_ramana_rara_internal_echo.txt is the same text with only the four C5
prefixes (च5., చ5., ಚ5., ച5.) removed. This reproduces the merged C4+C5
shape documented at the end of TRACK-133; it is not an unmodified historical
source capture. No lyric text, closing refrain, or English/Tamil marker is changed.rama_ramana_rara_echo.expected.json records exact type, order, label, and text
from the pre-fix parser on the complete downloaded HTML. Every language has
P + 6C. The live source currently includes explicit C5 markers and already parsed
correctly before this fix; the missing-marker fixture exercises the recovery.tests/test_pallavi_echo_split.py uses these fixtures for both ordinary language
blocks and Indic blocks after a metadata boundary. Negative cases cover ordinary
refrains, inline parentheses, non-charanam blocks, and ambiguous deficits.
Downloaded HTML SHA-256:
4c6199883b2a9a2c94db7c45a5602e7511ca7c420da9be6bf840c3681e21f3be.