Lint checker for 1.x/2.x terms (#933)

Signed-off-by: Michael Oviedo <[email protected]>

Author: OVI3D0

.ci/scripts/check_deprecated_terms.py

@@ -1,75 +1,113 @@
 #!/usr/bin/env python3
-
-import os
-import re
-import sys
-
-DEPRECATED_TERMS = [
+# .ci/scripts/check_deprecated_terms.py
+ 
+import os, re, sys, argparse
+ 
+# ---- Term sets ----
+# Block these when you're on the *2.x* branch (i.e., forbid legacy 1.x names):
+TERMS_1X = [
     "provision-configs",
+    "provision-config-instances",
+    "results-publishing",
     "results-publisher",
     "load-worker-coordinator-hosts",
     "execute-test",
 ]
-
-SKIP_DIRS = [".git", "venv", "__pycache__", ".pytest_cache"]
+ 
+# Block these when you're on the *1.x* branch (i.e., forbid 2.x names):
+TERMS_2X = [
+    "cluster-configs",
+    "reporting",
+    "worker-hosts",
+    "run-test",
+    "test-run",
+]
+ 
+SKIP_DIRS = {".git", "venv", "__pycache__", ".pytest_cache", ".ci", "tests"}
 VALID_EXTENSIONS = (".py", ".yml", ".yaml", ".md", ".sh", ".json", ".txt")
-
-VARIANT_PATTERNS = []
-
-def generate_variants(term):
+ 
+SUPRESS_MARKERS = {
+    "block-1x": "check-deprecated-terms-disable-1x",
+    "block-2x": "check-deprecated-terms-disable-2x"
+}
+ 
+def generate_variants(term: str) -> set[str]:
     base = term.replace("-", " ").replace("_", " ")
     words = base.split()
     variants = set()
-
-    # kebab-case, snake_case, PascalCase, camelCase
+    # kebab, snake, Pascal, camel
     variants.add("-".join(words))
     variants.add("_".join(words))
-    variants.add("".join([w.capitalize() for w in words]))        # PascalCase
+    variants.add("".join([w.capitalize() for w in words]))  # PascalCase
     variants.add(words[0] + "".join([w.capitalize() for w in words[1:]]))  # camelCase
-
-    # Word order flip for 2-word terms
-    if len(words) == 2:
+ 
+    # Optional: flip order for 2-word terms, but avoid silly "-ip" flips creating noise
+    if len(words) == 2 and not words[1].lower() == "ip":
         variants.add("-".join(words[::-1]))
         variants.add("_".join(words[::-1]))
         variants.add(words[1] + words[0].capitalize())  # camelCase reverse
-
     return variants
-
-for term in DEPRECATED_TERMS:
-    for variant in generate_variants(term):
-        VARIANT_PATTERNS.append(re.compile(re.escape(variant), re.IGNORECASE))
-
-def should_check_file(filename):
-    return filename.endswith(VALID_EXTENSIONS)
-
-def main():
-    error_found = False
-
+ 
+def build_patterns(terms: list[str]) -> list[re.Pattern]:
+    pats = []
+    for t in terms:
+        for v in generate_variants(t):
+            pats.append(re.compile(re.escape(v), re.IGNORECASE))
+    return pats
+ 
+def should_check_file(path: str) -> bool:
+    return path.endswith(VALID_EXTENSIONS)
+ 
+def walk_and_check(patterns: list[re.Pattern], mode: str) -> int:
+    error_found = 0
+    suppress_marker = SUPRESS_MARKERS.get(mode)
     for root, _, files in os.walk("."):
-        if any(skip in root for skip in SKIP_DIRS):
+        if any(skip in root.split(os.sep) for skip in SKIP_DIRS):
             continue
-
         for f in files:
             full_path = os.path.join(root, f)
             if not should_check_file(full_path):
                 continue
-
             try:
-                with open(full_path, "r", encoding="utf-8") as file:
-                    for i, line in enumerate(file, 1):
-                        for pattern in VARIANT_PATTERNS:
-                            if pattern.search(line):
-                                print(f"[Deprecated Term] {full_path}:{i}: {line.strip()}")
-                                error_found = True
+                with open(full_path, "r", encoding="utf-8") as fh:
+                    previous_line = ""
+                    for i, line in enumerate(fh, 1):
+                        if suppress_marker in previous_line or suppress_marker in line:
+                            previous_line = line
+                            continue
+                        for patt in patterns:
+                            if patt.search(line):
+                                print(f"[Forbidden Term] {full_path}:{i}: {line.strip()}")
+                                error_found = 1
+                                break
             except Exception as e:
                 print(f"[Warning] Skipped file {full_path}: {e}")
-
-    if error_found:
-        print("\n❌ Deprecated terms found. Please remove or rename them.")
+    return error_found
+ 
+def main():
+    p = argparse.ArgumentParser(description="Check forbidden term set by mode or env.")
+    p.add_argument("--mode", choices=["block-1x", "block-2x"], default=os.getenv("OSB_TERM_MODE"))
+    args = p.parse_args()
+ 
+    mode = args.mode
+    if not mode:
+        print("No mode provided (use --mode block-1x | block-2x or set OSB_TERM_MODE). Exiting 0.")
+        sys.exit(0)
+ 
+    if mode == "block-1x":
+        terms = TERMS_1X
+        banner = "❌ 1.x terms found in 2.x branch. Replace with 2.x names."
+    else:
+        terms = TERMS_2X
+        banner = "❌ 2.x terms found in 1.x branch. Replace with 1.x names."
+ 
+    patterns = build_patterns(terms)
+    failed = walk_and_check(patterns, mode)
+    if failed:
+        print("\n" + banner)
         sys.exit(1)
-
-    print("✅ No deprecated terms found.")
+    print("✅ No forbidden terms found for", mode)
     sys.exit(0)
-
+ 
 if __name__ == "__main__":
     main()

.github/workflows/unit-test.yml

@@ -25,6 +25,26 @@ jobs:
       - name: Run the CI build script
         run: bash .ci/build.sh build_and_unit_test
 
-      - name: Run deprecated term lint checker
-        run: .ci/scripts/check_deprecated_terms.py
-        if: contains(github.event.pull_request.labels.*.name, '2.x-terms-only')
+      - name: Detect term-check mode from labels
+        id: detect_mode
+        shell: bash
+        run: |
+          labels='${{ toJson(github.event.pull_request.labels.*.name) }}'
+          echo "PR labels: $labels"
+          mode=""
+
+          if echo "$labels" | grep -qi '"check-1.x-terms"'; then
+            mode="block-1x"
+          elif echo "$labels" | grep -qi '"check-2.x-terms"'; then
+            mode="block-2x"
+          fi
+
+          echo "mode=$mode" >> "$GITHUB_OUTPUT"
+
+      # 2) Run the checker only if a mode was selected
+      - name: Run term lint checker
+        if: steps.detect_mode.outputs.mode != ''
+        env:
+          OSB_TERM_MODE: ${{ steps.detect_mode.outputs.mode }}
+        run: |
+          python3 .ci/scripts/check_deprecated_terms.py --mode "$OSB_TERM_MODE"

RELEASE_GUIDE.md

@@ -73,7 +73,7 @@ Add backport labels to PRs and commits so that changes can be added to `main` br
     3. Run `opensearch-benchmark --version` to ensure that it is the correct version
     4. Run `opensearch-benchmark --help`
     5. Run `opensearch-benchmark list workloads`
-    6. Run a basic workload on Linux and MacOS:  `opensearch-benchmark execute-test --workload pmc --test-mode`
+    6. Run a basic workload on Linux and MacOS:  `opensearch-benchmark run --workload pmc --test-mode`
     7. If you are fastidious, you can check the installed source files at `` `python3 -m site --user-site`/osbenchmark `` to verify that a recent change is indeed present.
 
 8. Verify Docker Hub Staging OSB Image Works:

osbenchmark/benchmark.py

@@ -1292,6 +1292,7 @@ def handle_command_suggestions():
     Check for common command mistakes and provide helpful suggestions
     Returns True if suggestion was provided, False otherwise
     """
+    # check-deprecated-terms-disable-1x
     DEPRECATED_SUBCOMMANDS = ["execute-test", "execute"]
     if len(sys.argv) > 1 and sys.argv[1] in DEPRECATED_SUBCOMMANDS:
         console.info("Did you mean 'run'?")

osbenchmark/publisher.py

@@ -52,7 +52,7 @@ class Throughput(Enum):
     MEDIAN = "median"
 
 def summarize(results, cfg):
-    SummaryResultsPublisher(results, cfg).publish()
+    SummaryResultsPublisher(results, cfg).publish() # check-deprecated-terms-disable-1x
 
 
 def compare(cfg, baseline_id, contender_id):
@@ -115,7 +115,7 @@ def comma_separated_string_to_number_list(string_list):
     return results
 
 
-
+# check-deprecated-terms-disable-1x
 class SummaryResultsPublisher:
     def __init__(self, results, config):
         self.results = results

osbenchmark/utils/console.py

@@ -168,12 +168,12 @@ def println(msg, console_prefix=None, end="\n", flush=False, force=False, logger
 
 
 def progress(width=90):
-    return CmdLineProgressResultsPublisher(width, plain_output=PLAIN)
-
+    return CmdLineProgressResultsPublisher(width, plain_output=PLAIN) # check-deprecated-terms-disable-1x
 
+# check-deprecated-terms-disable-1x
 class CmdLineProgressResultsPublisher:
     """
-    CmdLineProgressResultsPublisher supports displaying an updating progress indication together with an information message.
+    This class supports displaying an updating progress indication together with an information message.
 
     :param printer: allow use of a different print method to assist with patching in unittests
     """