TestSandhiCOutv1

[funderburkjim]

This is testfiles/TestSandhiCOutv1.txt

Three lines were changed regarding compound sandhi with consonants.

These reflect a difference between the pythonv3 and earlier versions and the pythonv4 version.

The testsuite.sh script is changed in pythonv4 so that TestSandhiCOutv1 is taken as the standard for TestSandhiC.

TestSandhiCOut.txt remains the standard for the earlier versions.

The differences occur when one of the nasals N, R, or Y (guttural, cerebral, palatal) is joined by compound sandhi to one or another sibilant. The table of the following comment enumerates the differences, with reference to the Paninian sutra involved.

[funderburkjim]

Table of differences

Spellings in SLP1 transliteration of Sanskrit.

pre-sandhi	pythonv4 sandhi	pythonv3 sandhi	Sutra
N-S	NkS	NS	8.3.28 NRoH kuk wuk Sari
N-z	Nkz	Nz	same
N-s	Nks	Ns	same
R-S	RwS	RS	same
R-s	Rws	Rs	same
N-Sa	NkCa	NSa	8.3.28 NRoH kuk wuk Sari, etc.
n-Sa	YCa	YSa	...

[funderburkjim]

How to verify

To verify the pythonv4 sandhis, cd to pythonv4 and use ScharfSandhiArg.py program. e.g. python ScharfSandhiArg.py C N-S provides output:

sandhi: START: "N-S"
sandhimain: START: "N-S"
nnohkuktuksari: "Nk-S"
non_acsandhi: "Nk-S"
sandhimain: DONE: "NkS"
sandhi: DONE: "NkS"

Similarly, for the pythonv3 results, cd to python3; be sure 'python' is a Python2.xx version (see #5), and run the same command: python ScharfSandhiArg.py C N-S which provides output:

sandhi: START="N-S"
sandhi1: START="N-S"
sandhimain: START Linary=" N-S    "
sandhimain: Linary=" NS    "
sandhi1: Linary=" NS    "
sandhi: Linary=" NS    "
ScharfSandhiArg: ans="NS"

[funderburkjim]

various examples

Here is a list of various examples, using the current python4 code.

python ../ScharfSandhiArg.py C N-S
sandhi: START: "N-S"
sandhimain: START: "N-S"
nnohkuktuksari: "Nk-S"
non_acsandhi: "Nk-S"
sandhimain: DONE: "NkS"
sandhi: DONE: "NkS"
ScharfSandhiArg: ans="NkS"

python ../ScharfSandhiArg.py C N-z
sandhi: START: "N-z"
sandhimain: START: "N-z"
nnohkuktuksari: "Nk-z"
non_acsandhi: "Nk-z"
sandhimain: DONE: "Nkz"
sandhi: DONE: "Nkz"
ScharfSandhiArg: ans="Nkz"

python ../ScharfSandhiArg.py C N-s
sandhi: START: "N-s"
sandhimain: START: "N-s"
nnohkuktuksari: "Nk-s"
non_acsandhi: "Nk-s"
sandhimain: DONE: "Nks"
sandhi: DONE: "Nks"
ScharfSandhiArg: ans="Nks"

python ../ScharfSandhiArg.py C R-S
sandhi: START: "R-S"
sandhimain: START: "R-S"
nnohkuktuksari: "Rw-S"
non_acsandhi: "Rw-S"
sandhimain: DONE: "RwS"
sandhi: DONE: "RwS"
ScharfSandhiArg: ans="RwS"

python ../ScharfSandhiArg.py C R-z
sandhi: START: "R-z"
sandhimain: START: "R-z"
nnohkuktuksari: "Rw-z"
jharojharisavarne: "R-z"
sandhimain: DONE: "Rz"
sandhi: DONE: "Rz"
ScharfSandhiArg: ans="Rz"

python ../ScharfSandhiArg.py C R-s
sandhi: START: "R-s"
sandhimain: START: "R-s"
nnohkuktuksari: "Rw-s"
non_acsandhi: "Rw-s"
sandhimain: DONE: "Rws"
sandhi: DONE: "Rws"
ScharfSandhiArg: ans="Rws"

python ../ScharfSandhiArg.py C n-S
sandhi: START: "n-S"
sandhimain: START: "n-S"
situk: "nt-S"
stohscunascuh: "Yc-S"
jharojharisavarne: "Y-S"
non_acsandhi: "Y-S"
sandhimain: DONE: "YS"
sandhi: DONE: "YS"
ScharfSandhiArg: ans="YS"

python ../ScharfSandhiArg.py C n-z
sandhi: START: "n-z"
sandhimain: START: "n-z"
sandhimain: DONE: "nz"
sandhi: DONE: "nz"
ScharfSandhiArg: ans="nz"

python ../ScharfSandhiArg.py C n-s
sandhi: START: "n-s"
sandhimain: START: "n-s"
sandhimain: DONE: "ns"
sandhi: DONE: "ns"
ScharfSandhiArg: ans="ns"

python ../ScharfSandhiArg.py C N-Sa
sandhi: START: "N-Sa"
sandhimain: START: "N-Sa"
nnohkuktuksari: "Nk-Sa"
saschoti: "Nk-Ca"
non_acsandhi: "Nk-Ca"
sandhimain: DONE: "NkCa"
sandhi: DONE: "NkCa"
ScharfSandhiArg: ans="NkCa"

python ../ScharfSandhiArg.py C n-Sa
sandhi: START: "n-Sa"
sandhimain: START: "n-Sa"
situk: "nt-Sa"
stohscunascuh: "Yc-Sa"
saschoti: "Yc-Ca"
jharojharisavarne: "Y-Ca"
non_acsandhi: "Y-Ca"
sandhimain: DONE: "YCa"
sandhi: DONE: "YCa"
ScharfSandhiArg: ans="YCa"

[funderburkjim]

Rationale for changes in 'standard'

(comment due to @scharfp )

The change to the “standard” is the adoption of close sandhi between a final nasal and initial stop. A stop is first inserted after the nasal by the two rules (Note 1 below) I mention in my comments. That stop gets deleted by 8.4.65 JaraH Jari savarRe where it is of the same series as the following, but not if it is of a different series. So we get the k and w remaining in some cases. Although 8.3.28 is rarely conditioned in actual Sanskrit, the reason for adopting it is that it is parallel to 8.3.31 which is quite common. That sandhi is what we decided was more ‘standard’. Panini doesn’t provide a way to get rid of the ‘k’ and ‘w’ in the other cases of difference.

Note1: The two rules are

• 8.3.28 NRoH kuk wuk Sari (for N- and R-)
• 8.3.31 Si tuk (for n-)