Add Logiqx XML DAT parsing (Redump) and clear stale file matches on re-ingest

app/dat_parser.py

@@ -1,6 +1,8 @@
-"""Parser for clrmamepro (.dat) files.
+"""Parser for clrmamepro / Logiqx XML (.dat) files.
 
-Handles the clrmamepro text DAT format used by Redump and No-Intro::
+Supports both DAT layouts the project uses:
+
+* clrmamepro text DAT (No-Intro): parenthesised blocks, e.g.::
 
       clrmamepro (
           name "Sony - PlayStation Portable"
@@ -13,11 +15,15 @@ Handles the clrmamepro text DAT format used by Redump and No-Intro::
           rom ( name "WipEout Pure (Europe).iso" size 123 crc ABCD1234 md5 ... sha1 ... )
       )
 
-``parse_dat_text`` returns ``(header, entries)``.  Each entry is a dict with
+* Logiqx XML DAT (Redump): ``<game>`` elements with ``<rom>`` children.
-scalar keys (``name``, ``description``, ``region``, ``serial``, ...) and
+
-repeated nested blocks exposed as lists under their block key (``rom``).
+Both ``parse_dat_text`` and ``parse_dat_xml`` return ``(header, entries)``.
+Each entry is a dict with scalar keys (``name``, ``description``, ``region``,
+``serial``, ...) and repeated nested blocks exposed as lists under their block
+key (``rom``).
 """
 import re
+import xml.etree.ElementTree as ET
 from pathlib import Path
 
 _TOKEN_RE = re.compile(
@@ -124,5 +130,55 @@ def parse_dat_text(text):
     return header, entries
 
 
+def _xml_text(element, key):
+    child = element.find(key)
+    if child is not None and child.text:
+        return child.text.strip()
+    return None
+
+
+def parse_dat_xml(text):
+    """Parse a Logiqx XML DAT (the format redump.org uses).
+
+    Returns ``(header, entries)`` in the same shape as :func:`parse_dat_text`.
+    """
+    root = ET.fromstring(text)
+
+    header = {}
+    header_el = root.find("header")
+    if header_el is not None:
+        for child in header_el:
+            if child.text:
+                header[child.tag] = child.text.strip()
+
+    entries = []
+    for game in root.iter():
+        if game.tag not in ("game", "machine"):
+            continue
+        entry = {"name": (game.get("name") or "").strip()}
+        for key in ("description", "serial", "region", "category"):
+            value = _xml_text(game, key)
+            if value:
+                entry[key] = value
+        entry["rom"] = [
+            {
+                "name": rom.get("name", ""),
+                "size": rom.get("size"),
+                "crc": rom.get("crc"),
+                "md5": rom.get("md5"),
+                "sha1": rom.get("sha1"),
+                "serial": rom.get("serial"),
+            }
+            for rom in game.findall("rom")
+        ]
+        entries.append(entry)
+
+    return header, entries
+
+
 def parse_dat_file(path):
-    return parse_dat_text(Path(path).read_text(encoding="utf-8-sig"))
+    text = Path(path).read_text(encoding="utf-8-sig")
+    stripped = text.lstrip()
+    if stripped.startswith("<?xml") or stripped.startswith("<"):
+        return parse_dat_xml(text)
+    return parse_dat_text(text)

app/ingest.py

@@ -37,6 +37,12 @@ def ingest_dat(conn, path):
     existing = conn.execute("SELECT id FROM dats WHERE name = ?", (dat_name,)).fetchone()
     if existing:
         dat_id = existing["id"]
+        # Files that resolved to a release in this DAT must be re-identified
+        # against the new catalog (the release ids are about to be recycled).
+        conn.execute(
+            "UPDATE files SET release_id = NULL WHERE release_id IN (SELECT id FROM releases WHERE dat_id = ?)",
+            (dat_id,),
+        )
         conn.execute("DELETE FROM roms WHERE release_id IN (SELECT id FROM releases WHERE dat_id = ?)", (dat_id,))
         conn.execute("DELETE FROM releases WHERE dat_id = ?", (dat_id,))
         conn.execute(