Fix migrate_to_sql.py (syntax error + bad import), add missing schema.sql, .gitignore, requirements.txt

[Copilot]

migrate_to_sql.py was unparseable and could never execute. db/schema.sql didn't exist despite being opened at startup. networkx was undeclared and __pycache__/.db files were tracked by git.

Fixes in migrate_to_sql.py

• Syntax error (line 61): stray nhm token inside subscript — imp["line" nhm] → imp["line"]
• Bad import (line 5): decompose_source doesn't exist in ast_decomposer.py — replaced with ASTDecomposer class and updated call site to ASTDecomposer(source).get_full_decomposition()
• Wrong dict key (lines 71, 76): _decompose_function returns self_explanatory_name, not name — would have raised KeyError at runtime after the above fixes landed

# Before
from src.ast_decomposer import decompose_source
...
logic = decompose_source(source)
...
(module_id, imp["module"], json.dumps(imp["names"]), imp["line" nhm]))
...
method["name"], ...
func["name"], ...

# After
from src.ast_decomposer import ASTDecomposer
...
decomposer = ASTDecomposer(source)
logic = decomposer.get_full_decomposition()
...
(module_id, imp["module"], json.dumps(imp["names"]), imp["line"]))
...
method["self_explanatory_name"], ...
func["self_explanatory_name"], ...

New files

• db/schema.sql — normalized schema (modules, imports, classes, functions, constants, dependency_map) that migrate_to_sql.py and dependency_resolver.py both query against; was referenced on line 18 but never existed
• requirements.txt — declares networkx>=3.0
• .gitignore — excludes __pycache__/, *.db, .nuanced/, standard Python build artifacts

Original prompt

Context

This project was working and then broke during recent changes. The owner needs the codebase back to a fully functional state. DO NOT delete any files. DO NOT redesign the architecture. Fix only what is broken. Match the owner's existing code style — human-readable, component-based, composed.

What Is Broken (fix ONLY these)

1. migrate_to_sql.py — Two bugs preventing execution

Bug A — Line 61: Syntax error

# BROKEN (line 61):
(module_id, imp["module"], json.dumps(imp["names"]), imp["line" nhm]))
# FIX:
(module_id, imp["module"], json.dumps(imp["names"]), imp["line"]))

nhm is a stray token that makes the file unparseable. Just remove it.

Bug B — Line 5: Import references a function that doesn't exist

# BROKEN:
from src.ast_decomposer import decompose_source
# FIX — use the ASTDecomposer class that actually exists in src/ast_decomposer.py:
from src.ast_decomposer import ASTDecomposer

Then update migrate() around line 43 to use the class:

# BROKEN:
logic = decompose_source(source)
# FIX:
decomposer = ASTDecomposer(source)
logic = decomposer.get_full_decomposition()

2. db/schema.sql — Missing file that migrate_to_sql.py tries to read

migrate_to_sql.py line 18 does with open(SCHEMA_PATH, "r") as f: conn.executescript(f.read()) where SCHEMA_PATH = "db/schema.sql". That file doesn't exist. Create it with the normalized schema that migrate_to_sql.py and src/dependency_resolver.py both expect. Looking at what those files query, the schema must be:

-- db/schema.sql
-- Normalized schema for the LogicAIn Codebase-as-Data registry.
-- Used by migrate_to_sql.py for the advanced cross-module pipeline.

CREATE TABLE IF NOT EXISTS modules (
    id INTEGER PRIMARY KEY AUTOINCREMENT,
    path TEXT UNIQUE NOT NULL,
    sha256 TEXT,
    last_synced TIMESTAMP DEFAULT CURRENT_TIMESTAMP
);

CREATE TABLE IF NOT EXISTS imports (
    id INTEGER PRIMARY KEY AUTOINCREMENT,
    module_id INTEGER NOT NULL,
    imported_module TEXT,
    imported_names TEXT,
    line_number INTEGER,
    FOREIGN KEY(module_id) REFERENCES modules(id)
);

CREATE TABLE IF NOT EXISTS classes (
    id INTEGER PRIMARY KEY AUTOINCREMENT,
    module_id INTEGER NOT NULL,
    name TEXT NOT NULL,
    bases TEXT,
    docstring TEXT,
    line_number INTEGER,
    FOREIGN KEY(module_id) REFERENCES modules(id)
);

CREATE TABLE IF NOT EXISTS functions (
    id INTEGER PRIMARY KEY AUTOINCREMENT,
    module_id INTEGER NOT NULL,
    class_id INTEGER,
    name TEXT NOT NULL,
    arguments TEXT,
    return_type TEXT,
    docstring TEXT,
    logic_calls TEXT,
    line_start INTEGER,
    line_end INTEGER,
    resolved_logic_id INTEGER,
    FOREIGN KEY(module_id) REFERENCES modules(id),
    FOREIGN KEY(class_id) REFERENCES classes(id),
    FOREIGN KEY(resolved_logic_id) REFERENCES functions(id)
);

CREATE TABLE IF NOT EXISTS constants (
    id INTEGER PRIMARY KEY AUTOINCREMENT,
    module_id INTEGER NOT NULL,
    name TEXT NOT NULL,
    value TEXT,
    line_number INTEGER,
    FOREIGN KEY(module_id) REFERENCES modules(id)
);

CREATE TABLE IF NOT EXISTS dependency_map (
    id INTEGER PRIMARY KEY AUTOINCREMENT,
    source_function_id INTEGER NOT NULL,
    target_function_id INTEGER NOT NULL,
    call_type TEXT DEFAULT 'direct',
    FOREIGN KEY(source_function_id) REFERENCES functions(id),
    FOREIGN KEY(target_function_id) REFERENCES functions(id)
);

3. Add .gitignore (missing — __pycache__/ dirs are committed)

Create .gitignore with standard Python entries:

__pycache__/
*.pyc
*.pyo
*.db
.env
.venv/
*.egg-info/
dist/
build/
.nuanced/

4. Add requirements.txt (missing — networkx dependency undeclared)

Create requirements.txt:

networkx>=3.0

CRITICAL RULES

• DO NOT delete any existing files — not codegen.py, not dependency_resolver.py, not migrate_to_sql.py, not anything
• DO NOT modify src/ast_decomposer.py, src/graph_engine.py, sync_codebase.py, logic.py, src/codegen.py, src/dependency_resolver.py, db/pdx_exploration.sql
• DO NOT modify any test files
• Only touch: migrate_to_sql.py (fix the two bugs), create db/schema.sql, create .gitignore, create requirements.txt
• Match the existing code style — clean, readable Python with comments

The following is the prior conversation context from the user's chat exploration (may be truncated):

Assistant: Now let me read the core files to fully understand the project:

Assistant: Now let me look at the remaining source files to get the full picture:

Assistant: Now let me check the remaining key files:

Assistant: OK — I've now read every file in the repo. Here's my honest, thorough assessment.

---

Understanding LogicAIn — The Full Picture

LogicAIn is a Codebase-as-Data toolkit: it uses Python's ast module to decompose source files into structured relational records (functions, classes, imports, constants), stores them in SQL...

This pull request was created from Copilot chat.

---

💬 We'd love your input! Share your thoughts on Copilot coding agent in our 2 minute survey.

[JoeyBe1]

This sees stalled @copilot

[Copilot]

This sees stalled @copilot

Not stalled — all fixes are complete and pushed in commit 609b973. Here's what was done:

• migrate_to_sql.py: syntax error (imp["line" nhm]) fixed, bad import (decompose_source → ASTDecomposer) fixed, wrong dict key (name → self_explanatory_name) fixed — the file now parses and runs cleanly
• db/schema.sql: created with the full normalized schema (modules, imports, classes, functions, constants, dependency_map)
• .gitignore and requirements.txt: created

All 9 existing tests pass. Ready to merge.