fix: less brittle meta parsing

dzgui/api/mods.py

@@ -1,6 +1,6 @@
 import hashlib
 import logging
-import shlex
+import re
 
 from concurrent.futures import wait
 from concurrent.futures import ThreadPoolExecutor
@@ -53,39 +53,25 @@ def is_mission(path: Path) -> bool:
     file = path / "init.c"
     return file.exists()
 
-def parse_meta(file: Path) -> ModMeta | None:
+def tokenize(file: str) -> dict[Any] | None:
-    mod = file / "meta.cpp"
+    delimiter=r"\s*=\s*"
-    if mod.exists() is False:
+    modmeta = {}
-        return None
     try:
-        with open(file / "meta.cpp", "r") as f:
+        with open(file, "r", encoding="utf-8") as file:
-            st = f.read()
+            for line in file:
-            lex = shlex.shlex(st)
+                line = line.strip().rstrip(";")
-            lex.whitespace += "=;"
+                if not line:
-            v = []
+                    continue
-            while True:
+                els = re.split(delimiter, line, maxsplit=1)
-                tok = lex.get_token()
+                if len(els) == 2:
-                if not tok:
+                    key, value = els
-                    break
+                    print(value)
-                if tok == "protocol" or tok == "publishedid":
+                    modmeta[key.strip()] = value.strip('"')
-                    ntok = lex.get_token()
+        return modmeta
-                elif tok == "timestamp":
-                    # NOTE: some malformed .NET tick conversions result in numbers < 0
-                    ntok = lex.get_token()
-                    if ntok == "-":
-                        ntok += str(lex.get_token())
-                elif tok == "name":
-                    ntok = lex.get_token()
-                    if ntok is not None:
-                        ntok = ntok.split('"')[1]
-                if ntok is not None:
-                    v.append(ntok)
-            return ModMeta(*v)
     except Exception as e:
         logger.critical(e)
         return None
 
-
 def get_mod_size(path: Path) -> float:
     s = 0
     for f in path.rglob("*"):
@@ -105,12 +91,12 @@ def parse_mods(mods: list[Path]) -> list[Any]:
     for mod in mods:
         mod_dir = mod.name
         symlink = _hash(mod_dir)
-        meta = parse_meta(mod)
+        meta = tokenize(mod)
         if meta is None:
             continue
         size = get_mod_size(mod)
         # NOTE: final col is cell renderer highlight toggle
-        clean.append([meta.name, symlink, mod_dir, size, False])
+        clean.append([meta["name"], symlink, mod_dir, size, False])
     clean.sort(key=lambda row: str(row[0]).casefold())
     return clean
 

tests/fixtures/cpp/meta1.cpp

@@ -0,0 +1,5 @@
+protocol = 1;
+publishedid = 1234567890;
+name = "ModName";
+timestamp = 638278452000000000;
+hash = "a1b2c3d4e5f6g7h8i9j0";

tests/fixtures/cpp/meta2.cpp

@@ -0,0 +1,5 @@
+protocol = 1
+publishedid = 1234567890
+name = "ModName"
+timestamp = 638278452000000000
+hash = "a1b2c3d4e5f6g7h8i9j0"

tests/fixtures/cpp/meta3.cpp

@@ -0,0 +1,5 @@
+protocol=1
+publishedid=1234567890
+name="ModName"
+timestamp=638278452000000000
+hash="a1b2c3d4e5f6g7h8i9j0"

tests/fixtures/cpp/meta4.cpp

@@ -0,0 +1,5 @@
+protocol=1
+publishedid=1234567890
+name=ModName
+timestamp=638278452000000000
+hash=a1b2c3d4e5f6g7h8i9j0

tests/fixtures/cpp/meta5.cpp

@@ -0,0 +1,5 @@
+protocol=1;
+publishedid=1234567890;
+name=ModName;
+timestamp=638278452000000000;
+hash=a1b2c3d4e5f6g7h8i9j0;

tests/fixtures/cpp/meta6.cpp

@@ -0,0 +1,5 @@
+protocol=1
+publishedid=1234567890;
+name="ModName";
+timestamp=638278452000000000;
+hash= a1b2c3d4e5f6g7h8i9j0;

tests/fixtures/cpp/test.py

@@ -0,0 +1,3 @@
+from dzgui.api.mods import tokenize
+
+print(tokenize("meta4.cpp"))

tests/test_modmeta.py

@@ -0,0 +1,17 @@
+import pytest
+from dzgui.api.mods import tokenize
+from tests.fixtures import fixture_path
+
+@pytest.mark.mods
+@pytest.mark.parametrize("fixture", [
+    fixture_path("cpp/meta1.cpp"),
+    fixture_path("cpp/meta2.cpp"),
+    fixture_path("cpp/meta3.cpp"),
+    fixture_path("cpp/meta4.cpp"),
+    fixture_path("cpp/meta5.cpp"),
+    fixture_path("cpp/meta6.cpp"),
+    ]
+)
+def test_modmeta(fixture: str) -> None:
+    meta = tokenize(fixture)
+    assert meta["name"] == "ModName"