Spell out the operation of the quantified comparison bare names

[estebanzimanyi]

MobilityDB declares the functions behind the ever, always and temporal
comparison operators with their operation spelled out: eEqual, eNotEqual,
eLessThan, eLessEqual, eGreaterThan and eGreaterEqual, and the same behind a
for always and t for the temporal comparison (MobilityDB#2710), since eLt is
the built-in elt of Apache Spark and Apache Flink, which read a function name
whatever its case.

The three comparison families of meta/portable-aliases.json carry these 18
names for the same 18 operators, so the catalog's byOperator and byBareName,
and every binding generated from them, name the functions as MobilityDB
declares them. The tests read the new names, and the catalog example of
docs/portable-aliases.md reads the values the catalog holds: tEqual for #=,
the sorted bare names from aEqual to tNotEqual, and 41 aliases.

Measured over MobilityDB master 129433dbfd with the catalog derived from the
repository root: 41 portable aliases, with ?= mapping to eEqual, %>= to
aGreaterEqual and #<> to tNotEqual; test_portable.py and
test_portable_parity.py read 16 of 16, the live parity gate included. With
the same catalog the full suite reads 6 failed, 321 passed and 24 skipped on
this branch and on master alike: the six, the coverage gate, the family
classification and four struct layouts, read a built libmeos, which the tests
workflow builds before it derives the catalog.