feat: add SNFI volume prediction module with compressed RDA coefficients and automatic fallback

[aitorvv]

Description

This PR implements the SNFI tree volume prediction module for the silviculture package, satisfying the requirements of PR 5.

Key additions:

• Consolidated approximately 90 raw CSV files of SNFI3 and SNFI4 coefficients into package-level .rda data files, compressed using .xz to optimize disk footprint (~380 KB).
• Added scripts/regenerate_volume_models.R to provide a reproducible compilation pipeline from source data.
• Implemented silv_predict_snfi_volume(), which accepts vectors of species, provinces, DBH, height, and qualities and returns a consistent dataframe containing the four volume components (vcc, vsc, iavc, vle). It automatically falls back to SNFI3 parameters for provinces that do not have SNFI4 equations.
• Implemented silv_snfi_provinces() and silv_snfi_species() to allow users to inspect available parameters directly in memory.
• Created tests/testthat/test-predict-snfi-volume.R executing comprehensive combination checks (covering species, quality levels, and province lookup pathways).
• Cleaned all non-ASCII character warnings in R/predict-snfi-volume.R and updated dataset documentation in R/data.R to comply with CRAN checks.

Verification

# Inspect new/modified datasets and check documentation definitions
devtools::load_all()
str(snfi3_volume_coefficients)   # 37730 rows x 16 cols
str(snfi4_volume_coefficients)   # 20925 rows x 15 cols
# Check documentation descriptions for F.c., Parametro, and D.n.m.
?snfi3_volume_coefficients
?snfi4_volume_coefficients

# Test public listing functions
head(silv_snfi_provinces())      # Shows province_id, province_name, snfi_version
head(silv_snfi_species("SNFI4")) # Shows species codes and names

# Verify prediction function with single tree (Pinus radiata in Álava)
res_single <- silv_predict_snfi_volume(
  province = 1,
  species  = 28,
  dbh      = 25,
  h        = 18
)
print(res_single)
# Expected structure: 1 row, columns: vcc, vsc, iavc, vle, snfi_version (SNFI4)

# Verify automatic fallback to SNFI3 for Cantabria (province 39)
res_fallback <- silv_predict_snfi_volume(
  province = c(1, 3),
  species  = c(28, 28),
  dbh      = c(25, 30),
  h        = c(18, 20)
)
print(res_fallback)
# Expected: row 1 uses SNFI4, row 2 falls back to SNFI3