dataflow/api/routes/mappings.js
Paul Trowbridge 3be5ccc435 Add TSV export/import backend and update unmapped sample column
- Restore export.tsv and import-csv endpoints to mappings routes
- sample column is always last in export and discarded on import
- get_unmapped_values now returns distinct source field values as sample instead of full raw records

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-04-04 20:19:51 -04:00

315 lines
11 KiB
JavaScript

/**
* Mappings Routes
* Manage value mappings
*/
const express = require('express');
const multer = require('multer');
const { parse } = require('csv-parse/sync');
const upload = multer({ storage: multer.memoryStorage() });
const SYSTEM_COLS = new Set(['source_name', 'rule_name', 'input_value', 'record_count', 'sample']);
module.exports = (pool) => {
const router = express.Router();
// List all mappings for a source
router.get('/source/:source_name', async (req, res, next) => {
try {
const { rule_name } = req.query;
let query = 'SELECT * FROM mappings WHERE source_name = $1';
const params = [req.params.source_name];
if (rule_name) {
query += ' AND rule_name = $2';
params.push(rule_name);
}
query += ' ORDER BY rule_name, input_value';
const result = await pool.query(query, params);
res.json(result.rows);
} catch (err) {
next(err);
}
});
// Get unmapped values
router.get('/source/:source_name/unmapped', async (req, res, next) => {
try {
const { rule_name } = req.query;
const result = await pool.query(
'SELECT * FROM get_unmapped_values($1, $2)',
[req.params.source_name, rule_name || null]
);
res.json(result.rows);
} catch (err) {
next(err);
}
});
// Export unmapped values + existing mappings as TSV
// Columns: source_name, rule_name, input_value, record_count, <output keys...>, sample
// sample is always last and is discarded on import
router.get('/source/:source_name/export.tsv', async (req, res, next) => {
try {
const { rule_name } = req.query;
const source_name = req.params.source_name;
const [unmappedResult, mappedResult] = await Promise.all([
pool.query('SELECT * FROM get_unmapped_values($1, $2)', [source_name, rule_name || null]),
pool.query(
'SELECT * FROM mappings WHERE source_name = $1' + (rule_name ? ' AND rule_name = $2' : '') + ' ORDER BY rule_name, input_value',
rule_name ? [source_name, rule_name] : [source_name]
)
]);
// Collect output keys from existing mappings
const outputKeys = [];
for (const row of mappedResult.rows) {
for (const key of Object.keys(row.output || {})) {
if (!outputKeys.includes(key)) outputKeys.push(key);
}
}
const escape = (val) => String(val ?? '').replace(/\t/g, ' ');
// sample is always last
const allCols = ['source_name', 'rule_name', 'input_value', 'record_count', ...outputKeys, 'sample'];
const dataRows = [];
for (const row of unmappedResult.rows) {
const r = {
source_name,
rule_name: row.rule_name,
input_value: Array.isArray(row.extracted_value) ? JSON.stringify(row.extracted_value) : String(row.extracted_value ?? ''),
record_count: row.record_count,
sample: Array.isArray(row.sample) ? row.sample.join(' | ') : String(row.sample ?? '')
};
for (const key of outputKeys) r[key] = '';
dataRows.push(r);
}
for (const row of mappedResult.rows) {
const r = {
source_name: row.source_name,
rule_name: row.rule_name,
input_value: Array.isArray(row.input_value) ? JSON.stringify(row.input_value) : String(row.input_value ?? ''),
record_count: '',
sample: ''
};
for (const key of outputKeys) r[key] = row.output?.[key] ?? '';
dataRows.push(r);
}
const tsv = [
allCols.map(escape).join('\t'),
...dataRows.map(r => allCols.map(c => escape(r[c])).join('\t'))
].join('\n');
res.setHeader('Content-Type', 'text/tab-separated-values');
res.setHeader('Content-Disposition', `attachment; filename="mappings_${source_name}.tsv"`);
res.send(tsv);
} catch (err) {
next(err);
}
});
// Import mappings from uploaded TSV
// Any column that isn't a system field (source_name, rule_name, input_value, record_count, sample)
// is treated as an output key. sample is discarded wherever it appears.
router.post('/source/:source_name/import-csv', upload.single('file'), async (req, res, next) => {
const client = await pool.connect();
try {
if (!req.file) {
return res.status(400).json({ error: 'No file uploaded. Send TSV as multipart field named "file".' });
}
const records = parse(req.file.buffer, { columns: true, skip_empty_lines: true, trim: true, delimiter: '\t' });
if (records.length === 0) {
return res.status(400).json({ error: 'File is empty.' });
}
const outputKeys = Object.keys(records[0]).filter(k => !SYSTEM_COLS.has(k));
const mappings = [];
for (const row of records) {
const { source_name, rule_name, input_value } = row;
const output = {};
for (const key of outputKeys) {
if (row[key] && row[key].trim() !== '') output[key] = row[key].trim();
}
if (Object.keys(output).length === 0) continue;
let parsedInput;
try { parsedInput = JSON.parse(input_value); } catch { parsedInput = input_value; }
mappings.push({ source_name, rule_name, input_value: parsedInput, output });
}
if (mappings.length === 0) {
return res.status(400).json({ error: 'No rows with output values filled in.' });
}
await client.query('BEGIN');
const results = [];
for (const { source_name, rule_name, input_value, output } of mappings) {
const result = await client.query(
`INSERT INTO mappings (source_name, rule_name, input_value, output)
VALUES ($1, $2, $3, $4)
ON CONFLICT (source_name, rule_name, input_value)
DO UPDATE SET output = EXCLUDED.output
RETURNING *`,
[source_name, rule_name, JSON.stringify(input_value), JSON.stringify(output)]
);
results.push(result.rows[0]);
}
await client.query('COMMIT');
res.status(201).json({ count: results.length, mappings: results });
} catch (err) {
await client.query('ROLLBACK');
next(err);
} finally {
client.release();
}
});
// Get single mapping
router.get('/:id', async (req, res, next) => {
try {
const result = await pool.query(
'SELECT * FROM mappings WHERE id = $1',
[req.params.id]
);
if (result.rows.length === 0) {
return res.status(404).json({ error: 'Mapping not found' });
}
res.json(result.rows[0]);
} catch (err) {
next(err);
}
});
// Create mapping
router.post('/', async (req, res, next) => {
try {
const { source_name, rule_name, input_value, output } = req.body;
if (!source_name || !rule_name || !input_value || !output) {
return res.status(400).json({
error: 'Missing required fields: source_name, rule_name, input_value, output'
});
}
const result = await pool.query(
`INSERT INTO mappings (source_name, rule_name, input_value, output)
VALUES ($1, $2, $3, $4)
RETURNING *`,
[source_name, rule_name, JSON.stringify(input_value), JSON.stringify(output)]
);
res.status(201).json(result.rows[0]);
} catch (err) {
if (err.code === '23505') { // Unique violation
return res.status(409).json({ error: 'Mapping already exists' });
}
if (err.code === '23503') { // Foreign key violation
return res.status(404).json({ error: 'Source or rule not found' });
}
next(err);
}
});
// Bulk create mappings
router.post('/bulk', async (req, res, next) => {
const client = await pool.connect();
try {
const { mappings } = req.body;
if (!Array.isArray(mappings)) {
return res.status(400).json({ error: 'Expected array of mappings' });
}
await client.query('BEGIN');
const results = [];
for (const mapping of mappings) {
const { source_name, rule_name, input_value, output } = mapping;
const result = await client.query(
`INSERT INTO mappings (source_name, rule_name, input_value, output)
VALUES ($1, $2, $3, $4)
ON CONFLICT (source_name, rule_name, input_value)
DO UPDATE SET output = EXCLUDED.output
RETURNING *`,
[source_name, rule_name, JSON.stringify(input_value), JSON.stringify(output)]
);
results.push(result.rows[0]);
}
await client.query('COMMIT');
res.status(201).json({ count: results.length, mappings: results });
} catch (err) {
await client.query('ROLLBACK');
next(err);
} finally {
client.release();
}
});
// Update mapping
router.put('/:id', async (req, res, next) => {
try {
const { input_value, output } = req.body;
const result = await pool.query(
`UPDATE mappings
SET input_value = COALESCE($2, input_value),
output = COALESCE($3, output)
WHERE id = $1
RETURNING *`,
[req.params.id, input_value, output ? JSON.stringify(output) : null]
);
if (result.rows.length === 0) {
return res.status(404).json({ error: 'Mapping not found' });
}
res.json(result.rows[0]);
} catch (err) {
next(err);
}
});
// Delete mapping
router.delete('/:id', async (req, res, next) => {
try {
const result = await pool.query(
'DELETE FROM mappings WHERE id = $1 RETURNING id',
[req.params.id]
);
if (result.rows.length === 0) {
return res.status(404).json({ error: 'Mapping not found' });
}
res.json({ success: true, deleted: result.rows[0].id });
} catch (err) {
next(err);
}
});
return router;
};