Revert "chore: use improved levelization script with threading and argparse"

This reverts commit 5c1d7d9ae9.

.github/workflows/levelization.yml

@@ -10,7 +10,7 @@ jobs:
     steps:
     - uses: actions/checkout@v3
     - name: Check levelization
-      run: Builds/levelization/levelization.sh
+      run: python Builds/levelization/levelization.py
     - name: Check for differences
       id: assert
       run: |
@@ -40,7 +40,7 @@ jobs:
           To fix it, you can do one of two things:
           1. Download and apply the patch generated as an artifact of this
              job to your repo, commit, and push.
-          2. Run './Builds/levelization/levelization.sh' in your repo,
+          2. Run 'python Builds/levelization/levelization.py' in your repo,
              commit, and push.
 
           See Builds/levelization/README.md for more info.

.gitignore

@@ -53,6 +53,9 @@ Builds/levelization/results/paths.txt
 Builds/levelization/results/includes/
 Builds/levelization/results/includedby/
 
+# Python
+__pycache__
+
 # Ignore tmp directory.
 tmp
 

Builds/levelization/README.md

@@ -50,7 +50,7 @@ that `test` code should *never* be included in `ripple` code.)
 
 ## Validation
 
-The [levelization.sh](levelization.sh) script takes no parameters,
+The [levelization.py](levelization.py) script takes no parameters,
 reads no environment variables, and can be run from any directory,
 as long as it is in the expected location in the rippled repo.
 It can be run at any time from within a checked out repo, and will
@@ -84,7 +84,7 @@ It generates many files of [results](results):
   Github Actions workflow to test that levelization loops haven't
   changed.  Unfortunately, if changes are detected, it can't tell if
   they are improvements or not, so if you have resolved any issues or
-  done anything else to improve levelization, run `levelization.sh`,
+  done anything else to improve levelization, run `levelization.py`,
   and commit the updated results.
 
 The  `loops.txt` and `ordering.txt` files relate the modules
@@ -108,7 +108,7 @@ The committed files hide the detailed values intentionally, to
 prevent false alarms and merging issues, and because it's easy to
 get those details locally.
 
-1. Run `levelization.sh`
+1. Run `levelization.py`
 2. Grep the modules in `paths.txt`.
    * For example, if a cycle is found `A ~= B`, simply `grep -w
      A Builds/levelization/results/paths.txt | grep -w B`

Builds/levelization/levelization.py

@@ -0,0 +1,283 @@
+#!/usr/bin/env python3
+
+"""
+Usage: levelization.py
+This script takes no parameters, and can be called from any directory in the file system.
+"""
+
+import os
+import re
+import sys
+from collections import defaultdict
+from pathlib import Path
+
+# Compile regex patterns once at module level
+INCLUDE_PATTERN = re.compile(r"^\s*#include.*/.*\.h")
+INCLUDE_PATH_PATTERN = re.compile(r'[<"]([^>"]+)[>"]')
+
+
+def dictionary_sort_key(s):
+    """
+    Create a sort key that mimics 'sort -d' (dictionary order).
+    Dictionary order only considers blanks and alphanumeric characters.
+    """
+    return "".join(c for c in s if c.isalnum() or c.isspace())
+
+
+def get_level(file_path):
+    """
+    Extract the level from a file path (second and third directory components).
+    Equivalent to bash: cut -d/ -f 2,3
+
+    Examples:
+        src/ripple/app/main.cpp -> ripple.app
+        src/test/app/Import_test.cpp -> test.app
+    """
+    parts = file_path.split("/")
+
+    if len(parts) >= 3:
+        level = f"{parts[1]}/{parts[2]}"
+    elif len(parts) >= 2:
+        level = f"{parts[1]}/toplevel"
+    else:
+        level = file_path
+
+    # If the "level" indicates a file, cut off the filename
+    if "." in level.split("/")[-1]:
+        # Use the "toplevel" label as a workaround for `sort`
+        # inconsistencies between different utility versions
+        level = level.rsplit("/", 1)[0] + "/toplevel"
+
+    return level.replace("/", ".")
+
+
+def extract_include_level(include_line):
+    """
+    Extract the include path from an #include directive.
+    Gets the first two directory components from the include path.
+    Equivalent to bash: cut -d/ -f 1,2
+
+    Examples:
+        #include <ripple/basics/base_uint.h> -> ripple.basics
+        #include "ripple/app/main/Application.h" -> ripple.app
+    """
+    match = INCLUDE_PATH_PATTERN.search(include_line)
+    if not match:
+        return None
+
+    include_path = match.group(1)
+    parts = include_path.split("/")
+
+    if len(parts) >= 2:
+        include_level = f"{parts[0]}/{parts[1]}"
+    else:
+        include_level = include_path
+
+    # If the "includelevel" indicates a file, cut off the filename
+    if "." in include_level.split("/")[-1]:
+        include_level = include_level.rsplit("/", 1)[0] + "/toplevel"
+
+    return include_level.replace("/", ".")
+
+
+def find_repository_directories(start_path, depth_limit=10):
+    """
+    Find the repository root by looking for src or include folders.
+    Walks up the directory tree from the start path.
+    """
+    current = start_path.resolve()
+
+    for _ in range(depth_limit):
+        src_path = current / "src"
+        include_path = current / "include"
+        has_src = src_path.exists()
+        has_include = include_path.exists()
+
+        if has_src or has_include:
+            dirs = []
+            if has_src:
+                dirs.append(src_path)
+            if has_include:
+                dirs.append(include_path)
+            return current, dirs
+
+        parent = current.parent
+        if parent == current:
+            break
+        current = parent
+
+    raise RuntimeError(
+        "Could not find repository root. "
+        "Expected to find a directory containing 'src' and/or 'include' folders."
+    )
+
+
+def main():
+    script_dir = Path(__file__).parent.resolve()
+    os.chdir(script_dir)
+
+    # Clean up and create results directory.
+    results_dir = script_dir / "results"
+    if results_dir.exists():
+        import shutil
+
+        shutil.rmtree(results_dir)
+    results_dir.mkdir()
+
+    # Find the repository root.
+    try:
+        repo_root, scan_dirs = find_repository_directories(script_dir)
+        print(f"Found repository root: {repo_root}")
+        for scan_dir in scan_dirs:
+            print(f"  Scanning: {scan_dir.relative_to(repo_root)}")
+    except RuntimeError as e:
+        print(f"Error: {e}", file=sys.stderr)
+        sys.exit(1)
+
+    # Find all #include directives.
+    print("\nScanning for raw includes...")
+    raw_includes = []
+    rawincludes_file = results_dir / "rawincludes.txt"
+
+    with open(rawincludes_file, "w", buffering=8192) as raw_f:
+        for dir_path in scan_dirs:
+            for file_path in dir_path.rglob("*"):
+                if not file_path.is_file():
+                    continue
+                try:
+                    rel_path_str = str(file_path.relative_to(repo_root))
+                    with open(
+                        file_path, "r", encoding="utf-8", errors="ignore", buffering=8192
+                    ) as f:
+                        for line in f:
+                            if "#include" not in line or "boost" in line:
+                                continue
+                            if INCLUDE_PATTERN.match(line):
+                                line_stripped = line.strip()
+                                entry = f"{rel_path_str}:{line_stripped}\n"
+                                print(entry, end="")
+                                raw_f.write(entry)
+                                raw_includes.append((rel_path_str, line_stripped))
+                except Exception as e:
+                    print(f"Error reading {file_path}: {e}", file=sys.stderr)
+
+    # Build levelization paths and count directly.
+    print("Build levelization paths")
+    path_counts = defaultdict(int)
+
+    for file_path, include_line in raw_includes:
+        include_level = extract_include_level(include_line)
+        if not include_level:
+            continue
+        level = get_level(file_path)
+        if level != include_level:
+            path_counts[(level, include_level)] += 1
+
+    # Sort and deduplicate paths.
+    print("Sort and deduplicate paths")
+    sorted_items = sorted(
+        path_counts.items(),
+        key=lambda x: (dictionary_sort_key(x[0][0]), dictionary_sort_key(x[0][1])),
+    )
+
+    paths_file = results_dir / "paths.txt"
+    with open(paths_file, "w") as f:
+        for (level, include_level), count in sorted_items:
+            line = f"{count:7} {level} {include_level}\n"
+            print(line.rstrip())
+            f.write(line)
+
+    # Split into flat-file database.
+    print("Split into flat-file database")
+    includes_dir = results_dir / "includes"
+    includedby_dir = results_dir / "includedby"
+    includes_dir.mkdir()
+    includedby_dir.mkdir()
+
+    includes_data = defaultdict(list)
+    includedby_data = defaultdict(list)
+
+    for (level, include_level), count in sorted_items:
+        includes_data[level].append((include_level, count))
+        includedby_data[include_level].append((level, count))
+
+    for level in sorted(includes_data.keys(), key=dictionary_sort_key):
+        with open(includes_dir / level, "w") as f:
+            for include_level, count in includes_data[level]:
+                line = f"{include_level} {count}\n"
+                print(line.rstrip())
+                f.write(line)
+
+    for include_level in sorted(includedby_data.keys(), key=dictionary_sort_key):
+        with open(includedby_dir / include_level, "w") as f:
+            for level, count in includedby_data[include_level]:
+                line = f"{level} {count}\n"
+                print(line.rstrip())
+                f.write(line)
+
+    # Search for loops.
+    print("Search for loops")
+    loops_file = results_dir / "loops.txt"
+    ordering_file = results_dir / "ordering.txt"
+
+    # Pre-load all include files into memory for fast lookup.
+    includes_cache = {}
+    includes_lookup = {}
+
+    for include_file in sorted(includes_dir.iterdir(), key=lambda p: p.name):
+        if not include_file.is_file():
+            continue
+        includes_cache[include_file.name] = []
+        includes_lookup[include_file.name] = {}
+        with open(include_file, "r") as f:
+            for line in f:
+                parts = line.strip().split()
+                if len(parts) >= 2:
+                    name, count = parts[0], int(parts[1])
+                    includes_cache[include_file.name].append((name, count))
+                    includes_lookup[include_file.name][name] = count
+
+    loops_found = set()
+
+    with open(loops_file, "w", buffering=8192) as loops_f, open(
+        ordering_file, "w", buffering=8192
+    ) as ordering_f:
+        for source in sorted(includes_cache.keys()):
+            for include, include_freq in includes_cache[source]:
+                if include not in includes_lookup:
+                    continue
+
+                source_freq = includes_lookup[include].get(source)
+
+                if source_freq is not None:
+                    loop_key = tuple(sorted([source, include]))
+                    if loop_key in loops_found:
+                        continue
+                    loops_found.add(loop_key)
+
+                    loops_f.write(f"Loop: {source} {include}\n")
+
+                    diff = include_freq - source_freq
+                    if diff > 3:
+                        loops_f.write(f"  {source} > {include}\n\n")
+                    elif diff < -3:
+                        loops_f.write(f"  {include} > {source}\n\n")
+                    elif source_freq == include_freq:
+                        loops_f.write(f"  {include} == {source}\n\n")
+                    else:
+                        loops_f.write(f"  {include} ~= {source}\n\n")
+                else:
+                    ordering_f.write(f"{source} > {include}\n")
+
+    # Print results.
+    print("\nOrdering:")
+    with open(ordering_file, "r") as f:
+        print(f.read(), end="")
+
+    print("\nLoops:")
+    with open(loops_file, "r") as f:
+        print(f.read(), end="")
+
+
+if __name__ == "__main__":
+    main()

Builds/levelization/levelization.sh

@@ -1,130 +0,0 @@
-#!/bin/bash
-
-# Usage: levelization.sh
-# This script takes no parameters, reads no environment variables,
-# and can be run from any directory, as long as it is in the expected
-# location in the repo.
-
-pushd $( dirname $0 )
-
-if [ -v PS1 ]
-then
-  # if the shell is interactive, clean up any flotsam before analyzing
-  git clean -ix
-fi
-
-# Ensure all sorting is ASCII-order consistently across platforms.
-export LANG=C
-
-rm -rfv results
-mkdir results
-includes="$( pwd )/results/rawincludes.txt"
-pushd ../..
-echo Raw includes:
-grep -r '^[ ]*#include.*/.*\.h' include src | \
-    grep -v boost | tee ${includes}
-popd
-pushd results
-
-oldifs=${IFS}
-IFS=:
-mkdir includes
-mkdir includedby
-echo Build levelization paths
-exec 3< ${includes} # open rawincludes.txt for input
-while read -r -u 3 file include
-do
-    level=$( echo ${file} | cut -d/ -f 2,3 )
-    # If the "level" indicates a file, cut off the filename
-    if [[ "${level##*.}" != "${level}" ]]
-    then
-        # Use the "toplevel" label as a workaround for `sort`
-        # inconsistencies between different utility versions
-        level="$( dirname ${level} )/toplevel"
-    fi
-    level=$( echo ${level} | tr '/' '.' )
-
-    includelevel=$( echo ${include} | sed 's/.*["<]//; s/[">].*//' | \
-        cut -d/ -f 1,2 )
-    if [[ "${includelevel##*.}" != "${includelevel}" ]]
-    then
-        # Use the "toplevel" label as a workaround for `sort`
-        # inconsistencies between different utility versions
-        includelevel="$( dirname ${includelevel} )/toplevel"
-    fi
-    includelevel=$( echo ${includelevel} | tr '/' '.' )
-
-    if [[ "$level" != "$includelevel" ]]
-    then
-        echo $level $includelevel | tee -a paths.txt
-    fi
-done
-echo Sort and dedup paths
-sort -ds paths.txt | uniq -c | tee sortedpaths.txt
-mv sortedpaths.txt paths.txt
-exec 3>&- #close fd 3
-IFS=${oldifs}
-unset oldifs
-
-echo Split into flat-file database
-exec 4<paths.txt # open paths.txt for input
-while read -r -u 4 count level include
-do
-    echo ${include} ${count} | tee -a includes/${level}
-    echo ${level} ${count} | tee -a includedby/${include}
-done
-exec 4>&- #close fd 4
-
-loops="$( pwd )/loops.txt"
-ordering="$( pwd )/ordering.txt"
-pushd includes
-echo Search for loops
-# Redirect stdout to a file
-exec 4>&1
-exec 1>"${loops}"
-for source in *
-do
-  if [[ -f "$source" ]]
-  then
-    exec 5<"${source}" # open for input
-    while read -r -u 5 include includefreq
-    do
-      if [[ -f $include ]]
-      then
-        if grep -q -w $source $include
-        then
-          if grep -q -w "Loop: $include $source" "${loops}"
-          then
-            continue
-          fi
-          sourcefreq=$( grep -w $source $include | cut -d\  -f2 )
-          echo "Loop: $source $include"
-          # If the counts are close, indicate that the two modules are
-          # on the same level, though they shouldn't be
-          if [[ $(( $includefreq - $sourcefreq )) -gt 3 ]]
-          then
-              echo -e "  $source > $include\n"
-          elif [[ $(( $sourcefreq - $includefreq )) -gt 3 ]]
-          then
-              echo -e "  $include > $source\n"
-          elif [[ $sourcefreq -eq $includefreq ]]
-          then
-              echo -e "  $include == $source\n"
-          else
-              echo -e "  $include ~= $source\n"
-          fi
-        else
-          echo "$source > $include" >> "${ordering}"
-        fi
-      fi
-    done
-    exec 5>&- #close fd 5
-  fi
-done
-exec 1>&4 #close fd 1
-exec 4>&- #close fd 4
-cat "${ordering}"
-cat "${loops}"
-popd
-popd
-popd