# download_pg_pre1950.sh
# Dependencies: wget, bzip2, tar, python3, rsync
set -Eeuo pipefail

ROOT="$HOME/projects/pre-1950s-text/data"
META_DIR="$ROOT/meta"
LIST="$ROOT/pre1950_ids.txt"
TXT_DIR="$ROOT/text"
MIRROR="rsync://rsync.gutenberg.org/gutenberg"   # any listed mirror is acceptable

mkdir -p "$META_DIR" "$TXT_DIR"

# 1. obtain full metadata archive (≈ 400 MB compressed, ~2 GB uncompressed)
# wget -N -P "$META_DIR" http://www.gutenberg.org/cache/epub/feeds/rdf-files.tar.bz2        # :contentReference[oaicite:0]{index=0}
# bzip2 -dk "$META_DIR/rdf-files.tar.bz2"
# tar -xf "$META_DIR/rdf-files.tar" -C "$META_DIR"

# 2. build list of eBook IDs satisfying:
#    – dcterms:issued ≤ 1949  (best available proxy; original print date is absent) :contentReference[oaicite:1]{index=1}
python3 - <<'PY'
import pathlib, re, xml.etree.ElementTree as ET, sys, datetime, itertools
META_ROOT = pathlib.Path("/Users/brk/projects/pre-1950s-text/data")
YEAR_LIMIT = 1949
ids = []
for rdf in META_ROOT.rglob("*.rdf"):
    # try:
    tree = ET.parse(rdf)
    issued = tree.find(".//{http://purl.org/dc/terms/}issued")
    if issued is None:
        print("issued is none")
        continue
    year = int(issued.text[:4])
    print(f"year is {year}")
    if year <= YEAR_LIMIT:
        match = re.search(r'pg(\d+)\.rdf$', rdf.name)
        if match:
            ids.append(int(match.group(1)))
            print(f"Adding id {ids[-1]} to list")
    # except Exception:
    #     print('Failed wtih exception')
    #     continue
ids.sort()
with open("/Users/brk/projects/pre-1950s-text/data/pre1950_ids.txt", "w") as f:
    for i in ids:
        # construct rsync path fragment  e.g. id 1342 ➜ 1/3/4/1342/1342-0.txt
        path = "/".join(list(str(i)[:-1]))  # '', '1', '3', '4' for 1342
        if path:
            path += "/"
        path += f"{i}/{i}-*.txt"
        f.write(f"{path}\n")
print(f"IDs written: {len(ids)}", file=sys.stderr)
PY

# 3. download only the plain‑text files matching those IDs
rsync -av --delete --include="*/" --include-from="$LIST" --exclude="*" "$MIRROR/" "$TXT_DIR/"    # :contentReference[oaicite:2]{index=2}