Merge pull request #75 from seladb/do-not-skip-headers-in-preprocessor

Add option to retain #include directives in preprocessor

cxxheaderparser/preprocessor.py

@@ -16,10 +16,15 @@ class PreprocessorError(Exception):
 
 
 class _CustomPreprocessor(Preprocessor):
-    def __init__(self, encoding: typing.Optional[str]):
+    def __init__(
+        self,
+        encoding: typing.Optional[str],
+        passthru_includes: typing.Optional["re.Pattern"],
+    ):
         Preprocessor.__init__(self)
         self.errors: typing.List[str] = []
         self.assume_encoding = encoding
+        self.passthru_includes = passthru_includes
 
     def on_error(self, file, line, msg):
         self.errors.append(f"{file}:{line} error: {msg}")
@@ -58,12 +63,15 @@ def make_pcpp_preprocessor(
     include_paths: typing.List[str] = [],
     retain_all_content: bool = False,
     encoding: typing.Optional[str] = None,
+    passthru_includes: typing.Optional["re.Pattern"] = None,
 ) -> PreprocessorFunction:
     """
     Creates a preprocessor function that uses pcpp (which must be installed
     separately) to preprocess the input text.
 
     :param encoding: If specified any include files are opened with this encoding
+    :param passthru_includes: If specified any #include directives that match the
+                              compiled regex pattern will be part of the output.
 
     .. code-block:: python
 
@@ -75,7 +83,7 @@ def make_pcpp_preprocessor(
     """
 
     def _preprocess_file(filename: str, content: str) -> str:
-        pp = _CustomPreprocessor(encoding)
+        pp = _CustomPreprocessor(encoding, passthru_includes)
         if include_paths:
             for p in include_paths:
                 pp.add_path(p)

tests/test_preprocessor.py

@@ -1,9 +1,16 @@
 import os
 import pathlib
+import re
 
 from cxxheaderparser.options import ParserOptions
 from cxxheaderparser.preprocessor import make_pcpp_preprocessor
-from cxxheaderparser.simple import NamespaceScope, ParsedData, parse_file, parse_string
+from cxxheaderparser.simple import (
+    NamespaceScope,
+    ParsedData,
+    parse_file,
+    parse_string,
+    Include,
+)
 from cxxheaderparser.types import (
     FundamentalSpecifier,
     NameSpecifier,
@@ -135,3 +142,23 @@ def test_preprocessor_encoding(tmp_path: pathlib.Path) -> None:
             ]
         )
     )
+
+
+def test_preprocessor_passthru_includes(tmp_path: pathlib.Path) -> None:
+    """Ensure that all #include pass through"""
+    h_content = '#include "t2.h"\n'
+
+    with open(tmp_path / "t1.h", "w") as fp:
+        fp.write(h_content)
+
+    with open(tmp_path / "t2.h", "w") as fp:
+        fp.write("")
+
+    options = ParserOptions(
+        preprocessor=make_pcpp_preprocessor(passthru_includes=re.compile(".+"))
+    )
+    data = parse_file(tmp_path / "t1.h", options=options)
+
+    assert data == ParsedData(
+        namespace=NamespaceScope(), includes=[Include(filename='"t2.h"')]
+    )