Precompute upper bound for number of files to scan

Instead of calling the CodeBase iterator directly, finder now splits the scan
into two steps:

1. Build a list of all the files that might be in the code base.
2. Determine which files are in the code base.

The first step is a lot quicker than the second one, and tqdm can use the
length of the list to create a progress bar.

Signed-off-by: John Pennycook <john.pennycook@intel.com>

Author: Pennycook

codebasin/finder.py

@@ -163,16 +163,32 @@ def find(
     if isinstance(rootdir, Path):
         rootdir = str(rootdir)
 
-    # Identify all of the files in the codebase.
+    # Build up a list of potential source files.
+    def _potential_file_generator(codebase):
+        for directory in codebase._directories:
+            yield from Path(directory).rglob("*")
+
+    potential_files = []
+    for f in tqdm(
+        _potential_file_generator(codebase),
+        desc="Scanning current directory",
+        unit=" files",
+        leave=False,
+        disable=not show_progress,
+    ):
+        potential_files.append(f)
+
+    # Identify which files are in the code base.
     filenames = set()
     for f in tqdm(
-        codebase,
+        potential_files,
         desc="Identifying source files",
         unit=" files",
         leave=False,
         disable=not show_progress,
     ):
-        filenames.add(f)
+        if f in codebase:
+            filenames.add(f)
     for p in configuration:
         for e in configuration[p]:
             filenames.add(e["file"])