Reccmp comparison engine refactor (#405)

* Reccmp comparison engine refactor

* Remove redundant references to 'entry' symbol

tools/isledecomp/isledecomp/parser/__init__.py

@@ -1,2 +1,3 @@
+from .codebase import DecompCodebase
 from .parser import DecompParser
 from .linter import DecompLinter

tools/isledecomp/isledecomp/parser/codebase.py

@@ -0,0 +1,44 @@
+"""For aggregating decomp markers read from an entire directory and for a single module."""
+from typing import Iterable, Iterator, List
+from .parser import DecompParser
+from .node import (
+    ParserSymbol,
+    ParserFunction,
+    ParserVtable,
+    ParserVariable,
+)
+
+
+class DecompCodebase:
+    def __init__(self, filenames: Iterable[str], module: str) -> None:
+        self._symbols: List[ParserSymbol] = []
+
+        parser = DecompParser()
+        for filename in filenames:
+            parser.reset()
+            with open(filename, "r", encoding="utf-8") as f:
+                parser.read_lines(f)
+
+            for sym in parser.iter_symbols(module):
+                sym.filename = filename
+                self._symbols.append(sym)
+
+    def iter_line_functions(self) -> Iterator[ParserFunction]:
+        """Return lineref functions separately from nameref. Assuming the PDB matches
+        the state of the source code, a line reference is a guaranteed match, even if
+        multiple functions share the same name. (i.e. polymorphism)"""
+        return filter(
+            lambda s: isinstance(s, ParserFunction) and not s.is_nameref(),
+            self._symbols,
+        )
+
+    def iter_name_functions(self) -> Iterator[ParserFunction]:
+        return filter(
+            lambda s: isinstance(s, ParserFunction) and s.is_nameref(), self._symbols
+        )
+
+    def iter_vtables(self) -> Iterator[ParserVtable]:
+        return filter(lambda s: isinstance(s, ParserVtable), self._symbols)
+
+    def iter_variables(self) -> Iterator[ParserVariable]:
+        return filter(lambda s: isinstance(s, ParserVariable), self._symbols)

tools/isledecomp/isledecomp/parser/parser.py

@@ -1,6 +1,6 @@
 # C++ file parser
 
-from typing import List, Iterable, Iterator
+from typing import List, Iterable, Iterator, Optional
 from enum import Enum
 from .util import (
     is_blank_or_comment,
@@ -122,6 +122,11 @@ class DecompParser:
     def variables(self) -> List[ParserSymbol]:
         return [s for s in self._symbols if isinstance(s, ParserVariable)]
 
+    def iter_symbols(self, module: Optional[str] = None) -> Iterator[ParserSymbol]:
+        for s in self._symbols:
+            if module is None or s.module == module:
+                yield s
+
     def _recover(self):
         """We hit a syntax error and need to reset temp structures"""
         self.state = ReaderState.SEARCH