South Indian matching

/v1/match/dashakoot and /v1/match/papasamyam follow common practice as published sources state it. We have not read a classical text for either, and regional almanacs differ, so every answer carries basis: convention. In both, person_a is the groom and person_b the bride: the rules are not the same in each direction.

The ten poruthams

Stars and signs are the Moon's. Counts run from the bride's to the groom's, with hers counted as 1. Without a birth time, the Moon is taken at 12:00 and a warning says so.

Porutham	Agrees	Moderate
Dina	The count divided by 9 leaves 2, 4, 6, 8 or 0
Gana	Same gana	Deva with Manushya
Mahendra	Count of 4, 7, 10, 13, 16, 19, 22 or 25
Stree Deergha	Count above 13	Count above 9
Yoni	Any pair not listed as hostile
Rasi	Count of 1, 7, 9, 10 or 11	Count of 12
Rasi Adhipati	Same lord, or friends both ways	Anything else, unless enemies both ways
Vashya	Same class	Neighbouring class
Rajju	Different rajju
Vedha	Not a vedha pair

Every other case does not agree. The rules for Dina, Mahendra, Stree Deergha, Rasi, Rajju and Vedha follow HinduTone's guide. The score of 1, 0.5 or 0 follows Tamilcube's calculator.

These choices are ours:

Gana, Rasi Adhipati: the guide only urges caution on Rakshasa pairs, and does not cover a friend paired with an enemy. We count Rakshasa with another gana as not agreeing, and the mixed pairs as moderate.
Rasi: count 12 is moderate because the guide says many astrologers are cautious about 2/12. Other sources accept 8; we follow the guide, which avoids 6/8.
Vashya: published tables of vashya signs conflict, and we could not confirm one. We use the same five classes as our Ashtakoot, so this differs from a Tamil panchangam's sign table.
Yoni, gana, lords: these use the tables our Ashtakoot already uses.
Counting: the guide says the bride's star counts as 1 only for Dina. We use the same count for the others.

meets_minimum needs Rajju to agree and at least minimum_agreeing poruthams to agree. Sources name 5, 6 or 7, and one names no minimum at all, so the default of 5 is a setting and not a ruling.

Papasamyam

Sources agree on the method. For each chart, take Sun, Mars, Saturn and Rahu, and see which fall in houses 1, 2, 4, 7, 8 or 12 from the lagna, the Moon and Venus, by whole signs. Compare the totals: the result is satisfactory when the groom's total is equal or higher. Both birth times are needed.

They disagree on the points, so you choose with method:

`method`	Points	From
`count`	1 for each placement	Astrologymag's worked table, which our test reproduces
`weighted`	House units (3, 3, 3, 5, 6, 1) times planet units (Rahu 1, Sun 2, Saturn 3, Mars 4), half from the Moon, a quarter from Venus, then scaled by dignity	A 2014 article by an astrologer

The weighted dignity shares are exalted 0, moolatrikona 1/4, own sign 1/2, friend's sign 3/4 and debilitated 5/4. The article gives no share for other signs, and no dignity for Rahu, so we use full points for both. The dignity tables themselves are the standard ones.

What we leave out, because the sources conflict:

Ketu: some sources count it. include_ketu adds it with count; the weighted article gives it no dosha.
Margin: one source allows a groom's total of up to twice the bride's, and another 1.25 times. We report ratio_groom_to_bride and judge only which total is higher.
Averaging, Jupiter's aspect, star exemptions: another source averages the three references and lets Jupiter's aspect cancel points. These are not applied.
Comparing methods: totals from count and weighted are on different scales.

Hindi covers the labels and sentences. Star, sign and planet names in the rows stay in English.