test: add tests for session pair harvester (#91)

scripts/gitea_issue_parser.py

@@ -1,162 +0,0 @@
-#!/usr/bin/env python3
-"""
-Gitea Issue Body Parser
-
-Extracts structured data from Gitea issue markdown bodies:
-- Title
-- Context section
-- Acceptance criteria (checkboxes)
-- Labels
-- Epic/parent references
-
-Usage:
-  python3 scripts/gitea_issue_parser.py <issue_body.txt
-  python3 scripts/gitea_issue_parser.py --url https://forge.../api/v1/repos/.../issues/123
-  echo "issue body" | python3 scripts/gitea_issue_parser.py --stdin
-
-Output: JSON with {title, context, criteria[], labels[], epic_ref}
-"""
-
-import argparse
-import json
-import re
-import sys
-from typing import Optional
-
-
-def parse_issue_body(body: str, title: str = "", labels: list = None) -> dict:
-    """Parse a Gitea issue body into structured JSON."""
-    result = {
-        "title": title,
-        "context": "",
-        "criteria": [],
-        "labels": labels or [],
-        "epic_ref": None,
-        "sections": {},
-    }
-
-    if not body:
-        return result
-
-    # Extract epic/parent reference from title or body
-    epic_pattern = r"#(\d+)"
-    title_refs = re.findall(epic_pattern, title)
-    body_refs = re.findall(epic_pattern, body[:200])  # Check early body refs
-
-    # Look for "Closes #N" or "Part of #N" or "Epic: #N"
-    close_match = re.search(r"(?:Closes?|Fixes?|Resolves?)\s+#(\d+)", body, re.IGNORECASE)
-    part_match = re.search(r"(?:Part of|Epic|Parent|Blocks?)\s+#(\d+)", body, re.IGNORECASE)
-
-    if close_match:
-        result["epic_ref"] = f"#{close_match.group(1)}"
-    elif part_match:
-        result["epic_ref"] = f"#{part_match.group(1)}"
-    elif title_refs:
-        result["epic_ref"] = f"#{title_refs[0]}"
-    elif body_refs:
-        result["epic_ref"] = f"#{body_refs[0]}"
-
-    # Split into sections by ## headers
-    section_pattern = r"^##\s+(.+)$"
-    lines = body.split("\n")
-    current_section = None
-    current_content = []
-
-    for line in lines:
-        header_match = re.match(section_pattern, line)
-        if header_match:
-            # Save previous section
-            if current_section:
-                result["sections"][current_section] = "\n".join(current_content).strip()
-            current_section = header_match.group(1).strip().lower()
-            current_content = []
-        else:
-            current_content.append(line)
-
-    # Save last section
-    if current_section:
-        result["sections"][current_section] = "\n".join(current_content).strip()
-
-    # Extract context
-    for key in ["context", "background", "description", "problem"]:
-        if key in result["sections"]:
-            result["context"] = result["sections"][key]
-            break
-
-    # Extract acceptance criteria (checkboxes)
-    criteria_section = None
-    for key in ["acceptance criteria", "acceptance_criteria", "criteria", "requirements", "definition of done"]:
-        if key in result["sections"]:
-            criteria_section = result["sections"][key]
-            break
-
-    if criteria_section:
-        checkbox_pattern = r"-\s*\[[ xX]?\]\s*(.+)"
-        for match in re.finditer(checkbox_pattern, criteria_section):
-            result["criteria"].append(match.group(1).strip())
-
-        # Also try plain numbered/bulleted lists if no checkboxes found
-        if not result["criteria"]:
-            list_pattern = r"^\s*(?:\d+\.|-|\*)\s+(.+)"
-            for match in re.finditer(list_pattern, criteria_section, re.MULTILINE):
-                result["criteria"].append(match.group(1).strip())
-
-    # If no sectioned criteria found, scan whole body for checkboxes
-    if not result["criteria"]:
-        for match in re.finditer(r"-\s*\[[ xX]?\]\s*(.+)", body):
-            result["criteria"].append(match.group(1).strip())
-
-    return result
-
-
-def parse_from_url(api_url: str, token: str = None) -> dict:
-    """Parse an issue from a Gitea API URL."""
-    import urllib.request
-
-    headers = {}
-    if token:
-        headers["Authorization"] = f"token {token}"
-
-    req = urllib.request.Request(api_url, headers=headers)
-    resp = json.loads(urllib.request.urlopen(req, timeout=30).read())
-
-    title = resp.get("title", "")
-    body = resp.get("body", "")
-    labels = [l["name"] for l in resp.get("labels", [])]
-
-    return parse_issue_body(body, title, labels)
-
-
-def main():
-    parser = argparse.ArgumentParser(description="Parse Gitea issue body into structured JSON")
-    parser.add_argument("input", nargs="?", help="Issue body file (or - for stdin)")
-    parser.add_argument("--url", help="Gitea API URL for the issue")
-    parser.add_argument("--stdin", action="store_true", help="Read from stdin")
-    parser.add_argument("--token", help="Gitea API token (or set GITEA_TOKEN env var)")
-    parser.add_argument("--title", default="", help="Issue title (for epic ref extraction)")
-    parser.add_argument("--labels", nargs="*", default=[], help="Issue labels")
-    parser.add_argument("--pretty", action="store_true", help="Pretty-print JSON output")
-    args = parser.parse_args()
-
-    import os
-    token = args.token or os.environ.get("GITEA_TOKEN")
-
-    if args.url:
-        result = parse_from_url(args.url, token)
-    elif args.stdin or (args.input and args.input == "-"):
-        body = sys.stdin.read()
-        result = parse_issue_body(body, args.title, args.labels)
-    elif args.input:
-        with open(args.input) as f:
-            body = f.read()
-        result = parse_issue_body(body, args.title, args.labels)
-    else:
-        parser.print_help()
-        sys.exit(1)
-
-    indent = 2 if args.pretty else None
-    print(json.dumps(result, indent=indent))
-
-
-if __name__ == "__main__":
-    main()

scripts/session_pair_harvester.py

@@ -0,0 +1,234 @@
+#!/usr/bin/env python3
+"""
+Session Transcript → Training Pair Harvester
+
+Scans Hermes session JSONL files for Q&A patterns and extracts
+terse→rich training pairs. Outputs JSONL matching the timmy-config
+training pairs spec.
+
+Usage:
+  python3 scripts/session_pair_harvester.py ~/.hermes/sessions/
+  python3 scripts/session_pair_harvester.py session.jsonl --output pairs.jsonl
+  python3 scripts/session_pair_harvester.py --dir ~/.hermes/sessions/ --min-ratio 2.0
+
+Output format:
+  {"terse": "user short prompt", "rich": "ai detailed response", "source": "session_id", "model": "..."}
+"""
+
+import argparse
+import hashlib
+import json
+import sys
+from pathlib import Path
+from typing import Optional
+
+
+def compute_hash(text: str) -> str:
+    """Content hash for deduplication."""
+    return hashlib.sha256(text.encode()).hexdigest()[:16]
+
+
+def extract_pairs_from_session(session_data: dict, min_ratio: float = 1.5,
+                                min_response_words: int = 20) -> list:
+    """Extract terse→rich pairs from a single session object."""
+    pairs = []
+    conversations = session_data.get("conversations", [])
+    session_id = session_data.get("id", "unknown")
+    model = session_data.get("model", "unknown")
+
+    seen_hashes = set()
+
+    for i, msg in enumerate(conversations):
+        # Look for assistant/gpt responses
+        if msg.get("from") not in ("gpt", "assistant"):
+            continue
+
+        response_text = msg.get("value", "")
+        if not response_text or len(response_text.split()) < min_response_words:
+            continue
+
+        # Find the preceding human message
+        prompt_text = ""
+        for j in range(i - 1, -1, -1):
+            if conversations[j].get("from") == "human":
+                prompt_text = conversations[j].get("value", "")
+                break
+
+        if not prompt_text:
+            continue
+
+        # Filter: skip tool results, system messages embedded as human
+        if prompt_text.startswith("{") and "output" in prompt_text[:100]:
+            continue  # likely a tool result
+        if prompt_text.startswith("# SOUL.md") or prompt_text.startswith("You are"):
+            continue  # system prompt leak
+
+        # Quality filters
+        prompt_words = len(prompt_text.split())
+        response_words = len(response_text.split())
+
+        # Must have meaningful length ratio
+        if prompt_words == 0 or response_words == 0:
+            continue
+        ratio = response_words / prompt_words
+        if ratio < min_ratio:
+            continue
+
+        # Skip responses that are mostly code
+        code_blocks = response_text.count("```")
+        if code_blocks >= 4 and len(response_text.replace("```", "").strip()) < 50:
+            continue
+
+        # Skip responses with tool call artifacts
+        if "tool_call" in response_text[:100] or "function_call" in response_text[:100]:
+            continue
+
+        # Deduplicate by content hash
+        content_hash = compute_hash(prompt_text + response_text[:200])
+        if content_hash in seen_hashes:
+            continue
+        seen_hashes.add(content_hash)
+
+        # Clean up response: remove markdown headers if too many
+        clean_response = response_text
+
+        pairs.append({
+            "terse": prompt_text.strip(),
+            "rich": clean_response.strip(),
+            "source": session_id,
+            "model": model,
+            "prompt_words": prompt_words,
+            "response_words": response_words,
+            "ratio": round(ratio, 2),
+        })
+
+    return pairs
+
+
+def extract_from_jsonl_file(filepath: str, **kwargs) -> list:
+    """Extract pairs from a session JSONL file."""
+    pairs = []
+    path = Path(filepath)
+
+    if not path.exists():
+        print(f"Warning: {filepath} not found", file=sys.stderr)
+        return pairs
+
+    content = path.read_text()
+    lines = content.strip().split("\n")
+
+    for line in lines:
+        line = line.strip()
+        if not line:
+            continue
+        try:
+            session = json.loads(line)
+        except json.JSONDecodeError:
+            continue
+
+        session_pairs = extract_pairs_from_session(session, **kwargs)
+        pairs.extend(session_pairs)
+
+    return pairs
+
+
+def deduplicate_pairs(pairs: list) -> list:
+    """Remove duplicate pairs across files."""
+    seen = set()
+    unique = []
+    for pair in pairs:
+        key = compute_hash(pair["terse"] + pair["rich"][:200])
+        if key not in seen:
+            seen.add(key)
+            unique.append(pair)
+    return unique
+
+
+def main():
+    parser = argparse.ArgumentParser(description="Harvest training pairs from session transcripts")
+    parser.add_argument("input", nargs="?", help="Session JSONL file or directory")
+    parser.add_argument("--dir", "-d", help="Directory to scan for session files")
+    parser.add_argument("--output", "-o", default="harvested_pairs.jsonl", help="Output file")
+    parser.add_argument("--min-ratio", type=float, default=1.5, help="Min response/prompt word ratio")
+    parser.add_argument("--min-words", type=int, default=20, help="Min response word count")
+    parser.add_argument("--dry-run", action="store_true", help="Print stats without writing")
+    args = parser.parse_args()
+
+    all_pairs = []
+    files_scanned = 0
+
+    scan_dir = args.dir or args.input
+    if not scan_dir:
+        parser.print_help()
+        sys.exit(1)
+
+    scan_path = Path(scan_dir)
+    if scan_path.is_dir():
+        jsonl_files = sorted(scan_path.rglob("*.jsonl"))
+        print(f"Scanning {len(jsonl_files)} files in {scan_dir}...", file=sys.stderr)
+        for fpath in jsonl_files:
+            pairs = extract_from_jsonl_file(
+                str(fpath),
+                min_ratio=args.min_ratio,
+                min_response_words=args.min_words
+            )
+            all_pairs.extend(pairs)
+            files_scanned += 1
+    else:
+        pairs = extract_from_jsonl_file(
+            str(scan_path),
+            min_ratio=args.min_ratio,
+            min_response_words=args.min_words
+        )
+        all_pairs.extend(pairs)
+        files_scanned = 1
+
+    # Deduplicate
+    unique_pairs = deduplicate_pairs(all_pairs)
+
+    # Stats
+    if unique_pairs:
+        avg_prompt = sum(p["prompt_words"] for p in unique_pairs) / len(unique_pairs)
+        avg_response = sum(p["response_words"] for p in unique_pairs) / len(unique_pairs)
+        avg_ratio = sum(p["ratio"] for p in unique_pairs) / len(unique_pairs)
+    else:
+        avg_prompt = avg_response = avg_ratio = 0
+
+    stats = {
+        "files_scanned": files_scanned,
+        "raw_pairs": len(all_pairs),
+        "unique_pairs": len(unique_pairs),
+        "duplicates_removed": len(all_pairs) - len(unique_pairs),
+        "avg_prompt_words": round(avg_prompt, 1),
+        "avg_response_words": round(avg_response, 1),
+        "avg_ratio": round(avg_ratio, 2),
+    }
+
+    print(json.dumps(stats, indent=2), file=sys.stderr)
+
+    if args.dry_run:
+        # Print sample pairs
+        for pair in unique_pairs[:3]:
+            print(f"\n--- Source: {pair['source']} (ratio: {pair['ratio']}) ---", file=sys.stderr)
+            print(f"TERSE: {pair['terse'][:100]}...", file=sys.stderr)
+            print(f"RICH: {pair['rich'][:150]}...", file=sys.stderr)
+        return
+
+    # Write output
+    output_path = Path(args.output)
+    with open(output_path, "w") as f:
+        for pair in unique_pairs:
+            # Strip internal fields for output
+            output = {
+                "terse": pair["terse"],
+                "rich": pair["rich"],
+                "source": pair["source"],
+                "model": pair["model"],
+            }
+            f.write(json.dumps(output) + "\n")
+
+    print(f"\nWrote {len(unique_pairs)} pairs to {output_path}", file=sys.stderr)
+
+
+if __name__ == "__main__":
+    main()

scripts/test_gitea_issue_parser.py

@@ -1,111 +0,0 @@
-#!/usr/bin/env python3
-"""Tests for gitea_issue_parser."""
-
-import json
-import sys
-import os
-
-sys.path.insert(0, os.path.dirname(__file__))
-from gitea_issue_parser import parse_issue_body
-
-
-def test_basic_structure():
-    body = """## Context
-This is the background.
-
-## Acceptance Criteria
-- [ ] First criterion
-- [x] Second criterion (already done)
-- [ ] Third criterion
-
-## Labels
-`pipeline`, `extraction`
-"""
-    result = parse_issue_body(body, "Test Issue", ["pipeline", "extraction"])
-    assert result["title"] == "Test Issue"
-    assert "background" in result["context"].lower()
-    assert len(result["criteria"]) == 3
-    assert "First criterion" in result["criteria"]
-    assert result["labels"] == ["pipeline", "extraction"]
-    print("PASS: test_basic_structure")
-
-
-def test_epic_ref():
-    body = "Closes #645\n\nSome description."
-    result = parse_issue_body(body, "feat: thing (#688)")
-    assert result["epic_ref"] == "#645"
-    print("PASS: test_epic_ref")
-
-
-def test_epic_ref_from_title():
-    body = "Some description without close ref."
-    result = parse_issue_body(body, "feat: scene descriptions (#645)")
-    assert result["epic_ref"] == "#645"
-    print("PASS: test_epic_ref_from_title")
-
-
-def test_no_checkboxes():
-    body = """## Requirements
-1. First thing
-2. Second thing
-3. Third thing
-"""
-    result = parse_issue_body(body)
-    assert len(result["criteria"]) == 3
-    print("PASS: test_no_checkboxes")
-
-
-def test_empty_body():
-    result = parse_issue_body("", "Empty Issue")
-    assert result["title"] == "Empty Issue"
-    assert result["criteria"] == []
-    assert result["context"] == ""
-    print("PASS: test_empty_body")
-
-
-def test_real_issue_format():
-    body = """Closes #681
-
-## Changes
-
-Add `#!/usr/bin/env python3` shebang to 6 Python scripts.
-
-## Verification
-
-All 6 files confirmed missing shebangs before fix.
-
-## Impact
-
-Scripts can now be executed directly.
-"""
-    result = parse_issue_body(body, "fix: add python3 shebangs (#685)")
-    assert result["epic_ref"] == "#681"
-    assert "shebang" in result["context"].lower()
-    print("PASS: test_real_issue_format")
-
-
-def test_all_sections_captured():
-    body = """## Context
-Background info.
-
-## Acceptance Criteria
-- [ ] Do thing
-
-## Labels
-`test`
-"""
-    result = parse_issue_body(body)
-    assert "context" in result["sections"]
-    assert "acceptance criteria" in result["sections"]
-    print("PASS: test_all_sections_captured")
-
-
-if __name__ == "__main__":
-    test_basic_structure()
-    test_epic_ref()
-    test_epic_ref_from_title()
-    test_no_checkboxes()
-    test_empty_body()
-    test_real_issue_format()
-    test_all_sections_captured()
-    print("\nAll tests passed.")

scripts/test_session_pair_harvester.py

@@ -0,0 +1,90 @@
+#!/usr/bin/env python3
+"""Tests for session_pair_harvester."""
+
+import json
+import sys
+import os
+import tempfile
+
+sys.path.insert(0, os.path.dirname(__file__))
+from session_pair_harvester import extract_pairs_from_session, deduplicate_pairs, compute_hash
+
+
+def test_basic_extraction():
+    session = {
+        "id": "test_001",
+        "model": "test-model",
+        "conversations": [
+            {"from": "system", "value": "You are helpful."},
+            {"from": "human", "value": "What is Python?"},
+            {"from": "gpt", "value": "Python is a high-level programming language known for its readability and versatility. It supports multiple paradigms including procedural, object-oriented, and functional programming. Python is widely used in web development, data science, machine learning, and automation."},
+        ]
+    }
+    pairs = extract_pairs_from_session(session, min_ratio=1.5, min_response_words=10)
+    assert len(pairs) == 1
+    assert pairs[0]["terse"] == "What is Python?"
+    assert "programming language" in pairs[0]["rich"]
+    assert pairs[0]["source"] == "test_001"
+    print("PASS: test_basic_extraction")
+
+
+def test_filters_short_responses():
+    session = {
+        "id": "test_002",
+        "model": "test",
+        "conversations": [
+            {"from": "human", "value": "Hi"},
+            {"from": "gpt", "value": "Hello!"},
+        ]
+    }
+    pairs = extract_pairs_from_session(session, min_ratio=1.5, min_response_words=20)
+    assert len(pairs) == 0
+    print("PASS: test_filters_short_responses")
+
+
+def test_skips_tool_results():
+    session = {
+        "id": "test_003",
+        "model": "test",
+        "conversations": [
+            {"from": "human", "value": '{"output": "file content", "exit_code": 0}'},
+            {"from": "gpt", "value": "The file was read successfully. Now let me analyze the content and provide a detailed summary of what was found in the file system."},
+        ]
+    }
+    pairs = extract_pairs_from_session(session, min_ratio=1.5, min_response_words=10)
+    assert len(pairs) == 0
+    print("PASS: test_skips_tool_results")
+
+
+def test_deduplication():
+    pairs = [
+        {"terse": "What is X?", "rich": "X is Y.", "source": "s1", "model": "m"},
+        {"terse": "What is X?", "rich": "X is Y.", "source": "s2", "model": "m"},
+        {"terse": "What is Z?", "rich": "Z is W.", "source": "s1", "model": "m"},
+    ]
+    unique = deduplicate_pairs(pairs)
+    assert len(unique) == 2
+    print("PASS: test_deduplication")
+
+
+def test_ratio_filter():
+    session = {
+        "id": "test_005",
+        "model": "test",
+        "conversations": [
+            {"from": "human", "value": "Explain quantum computing in detail with examples and applications"},
+            {"from": "gpt", "value": "OK."},
+        ]
+    }
+    pairs = extract_pairs_from_session(session, min_ratio=1.5, min_response_words=10)
+    assert len(pairs) == 0  # response too short relative to prompt
+    print("PASS: test_ratio_filter")
+
+
+if __name__ == "__main__":
+    test_basic_extraction()
+    test_filters_short_responses()
+    test_skips_tool_results()
+    test_deduplication()
+    test_ratio_filter()
+    print("\nAll tests passed.")