Support CLCACHE_BASEDIR in nodirect mode

Do a simple case-insensitive find-and-replace to transform absolute
paths into relative paths within the preprocessor output that is used
to compute the hash. This makes CLCACHE_NODIRECT mode usable in the
presence of the __FILE__ macro.

Fixes a bug in _normalizedCommandLine that caused the source filename
to be included in the hash computation.

Author: oktal3700

README.asciidoc

@@ -86,10 +86,11 @@ CLCACHE_NODIRECT::
     key. Use this if you experience problems with direct mode or if you need
     built-in macroses like \__TIME__ to work correctly.
 CLCACHE_BASEDIR::
-    Has effect only when direct mode is on. Set this to path to root directory
-    of your project. This allows clcache to cache relative paths, so if you
-    move your project to different directory, clcache will produce cache hits as
-    before.
+    Set this to path to root directory of your project. In direct mode, this allows
+    clcache to cache relative paths, so if you move your project to different directory,
+    clcache will produce cache hits as before. When direct mode is disabled, clcache will
+    translate any absolute paths in the preprocessor output into relative paths before
+    computing the hash.
 CLCACHE_OBJECT_CACHE_TIMEOUT_MS::
     Overrides the default ObjectCacheLock timeout (Default is 10 * 1000 ms).
     The ObjectCacheLock is used to give exclusive access to the cache, which is

clcache/__main__.py

@@ -118,6 +118,19 @@ def normalizeBaseDir(baseDir):
         return None
 
 
+def caseInsensitiveReplace(data, old, new):
+    dataLower = data.lower()
+    oldLower = old.lower()
+    output = b''
+    index = 0
+    while True:
+        nextIndex = dataLower.find(oldLower, index)
+        if nextIndex < 0:
+            return output + data[index:]
+        output += data[index:nextIndex] + new
+        index = nextIndex + len(old)
+
+
 def getCachedCompilerConsoleOutput(path):
     try:
         with open(path, 'rb') as f:
@@ -469,6 +482,11 @@ def computeKeyNodirect(compilerBinary, commandLine, environment):
         compilerHash = getCompilerHash(compilerBinary)
         normalizedCmdLine = CompilerArtifactsRepository._normalizedCommandLine(commandLine)
 
+        if "CLCACHE_BASEDIR" in os.environ:
+            baseDir = normalizeBaseDir(os.environ["CLCACHE_BASEDIR"]).replace("\\", "\\\\").encode("UTF-8")
+            baseDirReplacement = BASEDIR_REPLACEMENT.encode("UTF-8")
+            preprocessedSourceCode = caseInsensitiveReplace(preprocessedSourceCode, baseDir, baseDirReplacement)
+
         h = HashAlgorithm()
         h.update(compilerHash.encode("UTF-8"))
         h.update(' '.join(normalizedCmdLine).encode("UTF-8"))
@@ -495,7 +513,7 @@ def _normalizedCommandLine(cmdline):
         argsToStrip += ("MP",)
 
         return [arg for arg in cmdline
-                if not (arg[0] in "/-" and arg[1:].startswith(argsToStrip))]
+                if arg[0] in "/-" and not arg[1:].startswith(argsToStrip)]
 
 class CacheFileStrategy:
     def __init__(self, cacheDirectory=None):

tests/integrationtests/basedir/nodirect.cpp

@@ -0,0 +1,8 @@
+#include <iostream>
+
+int main()
+{
+    std::cout << __FILE__ << '\n';
+}
+
+

tests/test_integration.py

@@ -1069,12 +1069,14 @@ def tearDown(self):
         os.chdir(self.savedCwd)
         self.tempDir.cleanup()
 
-    def _runCompiler(self, cppFile, extraArgs=None):
+    def _runCompiler(self, cppFile, extraArgs=None, direct=True):
         cmd = CLCACHE_CMD + ["/nologo", "/EHsc", "/c"]
         if extraArgs:
             cmd.extend(extraArgs)
         cmd.append(cppFile)
         env = dict(os.environ, CLCACHE_DIR=self.clcacheDir, CLCACHE_BASEDIR=os.getcwd())
+        if not direct:
+            env["CLCACHE_NODIRECT"] = "1"
         self.assertEqual(subprocess.call(cmd, env=env), 0)
 
     def expectHit(self, runCompiler):
@@ -1140,6 +1142,11 @@ def runCompiler():
             self._runCompiler("main.cpp", ["/DRESOURCES_DIR={}".format(os.getcwd())])
         self.expectMiss([runCompiler, runCompiler])
 
+    def testBasedirNoDirect(self):
+        def runCompiler():
+            self._runCompiler(os.path.join(os.getcwd(), "nodirect.cpp"), direct=False)
+        self.expectHit([runCompiler, runCompiler])
+
     def testBasedirRelativeIncludeArg(self):
         basedir = os.getcwd()
 

tests/test_unit.py

@@ -109,6 +109,13 @@ def testFilesBeneathRecursive(self):
             self.assertIn(r".\d\4.txt", files)
             self.assertIn(r".\d\e\5.txt", files)
 
+    def testCaseInsensitiveReplace(self):
+        replace = lambda s: clcache.caseInsensitiveReplace(s, b"PaTtErN", b"replacement")
+        self.assertEqual(replace(b"pattern may appear"), b"replacement may appear")
+        self.assertEqual(replace(b"infix paTTErn embed"), b"infix replacement embed")
+        self.assertEqual(replace(b"Pattern and PATTERN and pattern"), b"replacement and replacement and replacement")
+        self.assertEqual(replace(b"and finally PATTERN"), b"and finally replacement")
+
 
 class TestExtendCommandLineFromEnvironment(unittest.TestCase):
     def testEmpty(self):