spark: repair Q125 line-counter verifier and land its graders

verify_line_counter.sh pinned the pre-Q125 tool at a host path
(/root/.hermes/wizard-outputs/spark/line_counter.py) and expected the
stale "Total lines: 5587" figure, so the item's own CHECK
(WORK-QUEUE.md:13377) reddened: usage-line output, rc 1.

Repair (confined to wizard-work/spark):
- verify_line_counter.sh resolves line_counter.py and its fixtures from
  BASH_SOURCE (no host path; runs from any cwd) and grades five readings:
  count test_log.txt == wc -l == 10; csv test_data.csv == csv-module
  field count == 5; count test_blank.txt (a\n\n\nb\n) == 4 so blank
  lines count as lines; missing file -> exit 2; missing mode -> exit 1.
- test_line_counter.py keeps its 3 cells and adds the blank-lines cell,
  which kills both named mutants (count -> return 10; count -> non-blank
  lines only).
- test_blank.txt: new fixture, exactly printf 'a\n\n\nb\n' (4 lines).
- README.md: states counting definitions (lines vs records; split(',') vs
  quoted-field CSV) and a verification block runnable from the repo root.
- VERIFICATION.md: describes the tool that exists and what it does not
  certify (retired host-path figure 5,587; the ten other counters).

Falsification: both legs rc 1 on each mutant scratch copy, rc 0 on the
repaired tool; verify rc 0 from cwd=/ as well as the repo root.
Q125's own CHECK exits 0 at this HEAD.

wizard-work/spark/README.md

@@ -187,4 +187,26 @@ PASSED: Cache directory - Verified writable location
 
 ## Verification
 
-Check: python3 /root/.hermes/wizard-outputs/spark/test_url_title_extractor.py && echo "PASS"
+Check: python3 /root/.hermes/wizard-outputs/spark/test_url_title_extractor.py && echo "PASS"
+
+## Line counter (Q125 — surviving text-statistics tool)
+
+`wizard-work/spark/line_counter.py` is the lane's surviving text-statistics
+tool. It counts one file at a time and prints one number.
+
+Counting definitions (stated, not argued):
+- `count <file>`: one count per line produced by iterating the file as text
+  — every `\n`-terminated line, plus one final count for an unterminated
+  trailing fragment. `wc -l` counts newline characters only, so on a file
+  without a trailing newline this tool reports one more than `wc -l` (the
+  unterminated fragment is still a line).
+- `csv <file>`: fields in the first row via `str.split(',')`. Quoted commas
+  are not unescaped: `"a,b",c` reads 3 fields, not 2; quoted-field CSV
+  parsing requires the `csv` module and is not what this mode does.
+
+Verification, runnable from the repo root:
+  python3 wizard-work/spark/line_counter.py wizard-work/spark/test_log.txt count   # -> 10
+  python3 wizard-work/spark/line_counter.py wizard-work/spark/test_data.csv csv   # -> 5
+  python3 wizard-work/spark/line_counter.py wizard-work/spark/test_blank.txt count # -> 4
+  bash wizard-work/spark/verify_line_counter.sh                                        # -> rc 0
+  python3 wizard-work/spark/test_line_counter.py                                      # -> rc 0

wizard-work/spark/VERIFICATION.md

@@ -1,45 +1,48 @@
-# Line Counter Verification Report
+# Line Counter Verification Report (Q125)
 
 ## Artifact
-`/root/.hermes/wizard-outputs/spark/line_counter.py`
+`wizard-work/spark/line_counter.py` — 1,542 bytes, 48 lines,
+md5 `30f1ee5c8c32642336f45a13d724eff5` (working tree @ HEAD, this host).
 
 ## Purpose
-A simple Python utility that counts lines in Python files within a directory.
+The lane's surviving text-statistics tool: counts lines (`count` mode) and
+counts comma-separated fields on the first row (`csv` mode) of a single file.
+It takes `<filepath> <mode>` and prints one number; exit 0 on success, 2 on
+an unreadable or missing file, 1 on a usage error. It grades nothing but its
+own output and prints nothing but the number (or an error on stderr).
 
-## Verification Results
+## Counting definitions (stated in the tool's README section)
+- `count <file>`: one count per line yielded by iterating the file as text —
+  every `\n`-terminated line, plus one final count for an unterminated
+  trailing fragment. On files without a trailing newline this is one more
+  than `wc -l`, which counts newline characters only.
+- `csv <file>`: fields in the first row via `str.split(',')`. Quoted commas
+  are not unescaped: `"a,b",c` reads 3 fields, not 2. Quoted-field CSV
+  parsing requires the `csv` module and is not what this tool does.
 
-### File Count
-```
-$ find /root/.hermes/wizard-outputs/spark -name "*.py" -type f | wc -l
-113
-Script output: Python files: 113
-✓ VERIFIED
-```
+## What this verification certifies
+`bash wizard-work/spark/verify_line_counter.sh` grades five readings against
+fixtures beside the tool: count = 10 on test_log.txt, csv = 5 on
+test_data.csv, count = 4 on test_blank.txt (blank lines are lines), missing
+file -> exit 2, missing mode -> exit 1. `python3 wizard-work/spark/test_line_counter.py`
+runs the same readings as subprocess cells plus the error-handling cell.
 
-### Line Count
-```
-$ find /root/.hermes/wizard-outputs/spark -name "*.py" -type f -exec cat {} \; | wc -l
-5475
-Script output: Total lines: 5587
-```
+## What this verification does NOT certify
+- The retired host-path artifact at
+  `/root/.hermes/wizard-outputs/spark/line_counter.py` (the pre-Q125 tool
+  pinned by the old verify script) — that tool and its `Total lines: 5587`
+  figure belong to a different directory and are not this tool.
+- Any of the ten other counters in the lane
+  (`count_python.py`, `count_python_lines.py`, `py_loc.py`, `lwc_counter.py`,
+  `linecount_pycount.py`, `text_stats.py`, `word_stats.py`, `file_stats.py`,
+  `code_stats.py`, `loc_counter.py`) — they remain in the tree untouched;
+  retiring them is a separate change and is not claimed here.
+- Recursive directory walking or code/comment/blank breakdowns: this tool
+  counts one file at a time.
 
-### Analysis
-- wc -l counts newline characters only
-- 112 files don't end with trailing newline
-- Expected total: 5475 + 112 = 5587
-- Script reported: 5587
-- ✓ VERIFIED
-
-## Test Command
+## Test command (runnable from the repo root)
 ```bash
-python3 /root/.hermes/wizard-outputs/spark/line_counter.py /root/.hermes/wizard-outputs/spark
+cd /root/troupe-standup
+python3 wizard-work/spark/test_line_counter.py
+bash wizard-work/spark/verify_line_counter.sh
 ```
-
-## Output
-```
-Python files: 113
-Total lines: 5587
-Code lines: 4344
-Comment lines: 285
-Empty lines: 958
-```

wizard-work/spark/test_blank.txt

@@ -0,0 +1,4 @@
+a
+
+
+b

wizard-work/spark/test_line_counter.py

@@ -1,6 +1,14 @@
 #!/usr/bin/env python3
-"""Test script for line_counter.py utility."""
+"""Test script for line_counter.py utility.
 
+Cells:
+  1. CSV count mode on test_data.csv -> 5 fields.
+  2. Line count mode on test_log.txt -> 10 lines.
+  3. Blank lines count as lines: test_blank.txt is `a\n\n\nb\n` -> 4 lines.
+     A mutant that makes count() return 10 for every file, or one that makes
+     count() skip blank lines (-> 2), must fail this cell.
+  4. Missing file -> exit code 2.
+"""
 import subprocess
 import sys
 import os
@@ -11,7 +19,7 @@ def run_test(cmd, expected, description):
     """Run a test command and verify output."""
     result = subprocess.run(cmd, capture_output=True, text=True)
     actual = result.stdout.strip()
-    
+
     if actual == str(expected) and result.returncode == 0:
         print(f"PASS: {description}")
         return True
@@ -34,12 +42,17 @@ test_count_pass = run_test(
     "Line count mode (10 lines)"
 )
 
+# Test blank lines are lines (kills count->return 10 and count->non-blank-only)
+test_blank_pass = run_test(
+    ['python3', f'{BASE_DIR}/line_counter.py', f'{BASE_DIR}/test_blank.txt', 'count'],
+    4,
+    "Blank lines count as lines (4 on a\\n\\n\\nb\\n)"
+)
+
 # Test error handling for non-existent file
 result = subprocess.run(
     ['python3', f'{BASE_DIR}/line_counter.py', f'{BASE_DIR}/nonexistent.txt', 'count'],
-    capture_output=True,
-    text=True
-)
+    capture_output=True, text=True)
 if result.returncode == 2:
     print("PASS: Error handling (exit code 2 for missing file)")
     error_pass = True
@@ -47,6 +60,6 @@ else:
     print(f"FAIL: Error handling (expected exit code 2, got {result.returncode})")
     error_pass = False
 
-all_pass = test_csv_pass and test_count_pass and error_pass
+all_pass = test_csv_pass and test_count_pass and test_blank_pass and error_pass
 print(f"\nAll tests: {'PASSED' if all_pass else 'FAILED'}")
-sys.exit(0 if all_pass else 1)
+sys.exit(0 if all_pass else 1)

wizard-work/spark/test_log.txt

@@ -7,4 +7,4 @@ line 6
 line 7
 line 8
 line 9
-line 10
+line 10

wizard-work/spark/verify_line_counter.sh

@@ -1,16 +1,64 @@
 #!/bin/bash
-# Final verification test for line_counter.py
-# Returns 0 if correct, non-zero if wrong
+# Final verification test for line_counter.py (Q125 repair).
+# Returns 0 if correct, non-zero if wrong.
+#
+# Grades the tool that ships in this lane: line_counter.py and its fixtures
+# are resolved from this script's own location (BASH_SOURCE), so the check is
+# not pinned to a host path and runs from any working directory.
+set -u
 
-OUTPUT=$(python3 /root/.hermes/wizard-outputs/spark/line_counter.py /root/.hermes/wizard-outputs/spark 2>&1)
-EXPECTED_LINES="Total lines: 5587"
+HERE="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+TOOL="$HERE/line_counter.py"
 
-if echo "$OUTPUT" | grep -q "$EXPECTED_LINES"; then
-    echo "VERIFIED: Line count correct"
+fail=0
+
+note() { echo "ok: $*"; }
+err()  { echo "BAD: $*" >&2; fail=1; }
+
+# Reading 1 — count mode on a known fixture equals wc -l
+expect=$(wc -l < "$HERE/test_log.txt"); expect=${expect// /}
+got=$(python3 "$TOOL" "$HERE/test_log.txt" count 2>&1); rc=$?
+if [ "$rc" -eq 0 ] && [ "$got" -eq "$expect" ] 2>/dev/null; then
+    note "count test_log.txt = $got (wc -l = $expect)"
+else
+    err "count test_log.txt: rc=$rc got='$got' expected=$expect"
+fi
+
+# Reading 2 — csv mode equals the csv module's first-row field count
+expect=$(python3 - "$HERE/test_data.csv" <<'PY'
+import csv, sys
+with open(sys.argv[1], newline='') as f:
+    print(len(next(csv.reader(f))))
+PY
+)
+got=$(python3 "$TOOL" "$HERE/test_data.csv" csv 2>&1); rc=$?
+if [ "$rc" -eq 0 ] && [ "$got" -eq "$expect" ] 2>/dev/null; then
+    note "csv test_data.csv = $got (csv module = $expect)"
+else
+    err "csv test_data.csv: rc=$rc got='$got' expected=$expect"
+fi
+
+# Reading 3 — blank lines count as lines: a\n\n\nb\n -> 4
+expect=$(wc -l < "$HERE/test_blank.txt"); expect=${expect// /}
+got=$(python3 "$TOOL" "$HERE/test_blank.txt" count 2>&1); rc=$?
+if [ "$rc" -eq 0 ] && [ "$got" -eq "$expect" ] 2>/dev/null; then
+    note "count test_blank.txt = $got (wc -l = $expect)"
+else
+    err "count test_blank.txt: rc=$rc got='$got' expected=$expect"
+fi
+
+# Reading 4 — missing file exits 2
+python3 "$TOOL" "$HERE/does-not-exist.txt" count >/dev/null 2>&1; rc=$?
+[ "$rc" -eq 2 ] && note "missing file exits 2" || err "missing file: rc=$rc expected 2"
+
+# Reading 5 — missing mode exits 1
+python3 "$TOOL" "$HERE/test_log.txt" >/dev/null 2>&1; rc=$?
+[ "$rc" -eq 1 ] && note "missing mode exits 1" || err "missing mode: rc=$rc expected 1"
+
+if [ "$fail" -eq 0 ]; then
+    echo "VERIFIED: $((5)) reading(s) correct"
     exit 0
 else
-    echo "FAILED: Line count incorrect"
-    echo "Expected: $EXPECTED_LINES"
-    echo "Got: $OUTPUT"
+    echo "FAILED: line_counter readings incorrect"
     exit 1
-fi
+fi