[EN Currency] ISO 4217 codes not recognised as currency prefixes; crore/lakh multipliers missing from currency context

[nikitabuxy]

Is your feature request related to a problem? Please describe.

Two gaps in English currency recognition that affect financial document
processing (SEC filings, annual reports, earnings releases):

1. ISO 4217 codes are recognised as currency suffixes (27 million GBP
   works) but not as currency prefixes (GBP 27 million returns no
   result). Leading ISO codes are the dominant notation in financial
   documents from the UK, EU, Scandinavia, Southeast Asia, and most
   non-US markets.

2. crore and lakh are already correctly defined with numeric values
   in English-Numbers.yaml (crore: 10000000, lakh: 100000) but
   are absent from MultiplierRegex in English-NumbersWithUnit.yaml,
   so they are never applied when parsing currency amounts. crore is
   the standard unit in Indian financial reporting and appears in SEC
   filings from Indian subsidiaries of multinational companies.

To Reproduce

from recognizers_text import Culture
from recognizers_number_with_unit import NumberWithUnitRecognizer

model = NumberWithUnitRecognizer(Culture.English).get_currency_model()

# Gap 1 — ISO prefix not recognised
model.parse("GBP 27 million")    # []  — no result
model.parse("USD 20 million")    # []  — no result
model.parse("SEK 60,500,000")    # []  — no result
model.parse("JPY 50 billion")    # []  — no result
model.parse("CAD$1,700,000")     # []  — no result

# Gap 2 — crore/lakh multiplier not applied
model.parse("Rs 660 crore")      # [{'value': '660', 'unit': 'Rupee'}]
# Expected: {'value': '6600000000', 'unit': 'Rupee'}

Describe the solution you'd like

Gap 1 — ISO codes as currency prefixes

Add ISO 4217 codes to CurrencyPrefixList in
Patterns/English/English-NumbersWithUnit.yaml. For currencies that
already have a prefix entry (symbol-based), merge the ISO code alongside
the existing patterns. For currencies with no prefix entry, add a new
entry.

Expected behaviour after fix:

model.parse("GBP 27 million")
# [{'value': '27000000', 'unit': 'British pound', 'isoCurrency': 'GBP'}]

model.parse("SEK 60,500,000")
# [{'value': '60500000', 'unit': 'Swedish krona'}]

model.parse("CAD$1,700,000")
# [{'value': '1700000', 'unit': 'Canadian dollar', 'isoCurrency': 'CAD'}]

ISO codes to cover (commonly seen as prefixes in financial documents):
GBP, USD, EUR, JPY, CAD, AUD, CHF, HKD, SGD, KRW, INR, MXN, BRL, ZAR,
NOK, SEK, DKK, VND, CNY, RMB, and short-form variants A$ (AUD) and SG$.

Gap 2 — crore/lakh in currency context

Add lakh and crore to MultiplierRegex in
Patterns/English/English-NumbersWithUnit.yaml:

# before
MultiplierRegex: !simpleRegex
  def: \s*\b(thousand|million|billion|trillion)s?\b

# after
MultiplierRegex: !simpleRegex
  def: \s*\b(thousand|million|billion|trillion|lakh|crore)s?\b

This is a one-line change — the numeric values are already defined in
English-Numbers.yaml and only need to be included in the unit-context
multiplier pattern.

Expected behaviour after fix:

model.parse("Rs 660 crore")
# [{'value': '6600000000', 'unit': 'Rupee'}]

model.parse("Rs 5 lakh")
# [{'value': '500000', 'unit': 'Rupee'}]

Describe alternatives you've considered

The normalisation for ISO prefixes could alternatively be applied in
QueryProcessor.preprocess() (shared across all recogniser types), but
that risks unintended side-effects on date, dimension, and other models.
Scoping it to CurrencyModel.parse() is safer.

Additional context

- The related bug (wrong values when ISO code is concatenated to a
digit, e.g. USD34.6 million → 6000000) is filed as issue #3211 
Both issues share the same root area but have different fixes — the
bug requires a normalisation step in CurrencyModel.parse(); this
enhancement requires pattern additions to the YAML resource.