Skip to main content
The upstream OpenFlights database uses different naming conventions and occasionally contains data quality issues. The normalize.js script applies corrections automatically to ensure consistent, accurate airline data.

​
Why Corrections Are Needed

The OpenFlights project is community-maintained and follows its own conventions. This package applies corrections for:
  • Country name standardization - Converting formal names to commonly-used names
  • Data corruption fixes - Correcting fields where data leaked between columns
  • Duplicate resolution - Marking defunct airlines that share codes with active ones
  • Code accuracy - Fixing incorrect IATA/ICAO codes in upstream data
All corrections are applied transparently via normalize.js after fetching upstream data. The original upstream data is never modified directly.

​
Country Name Corrections

The COUNTRY_CORRECTIONS map standardizes country names from formal/official names to commonly-recognized names:
normalize.js:14-28
var COUNTRY_CORRECTIONS = {
  'Macao':                                   'Macau',
  'Republic of Korea':                       'South Korea',
  "Democratic People's Republic of Korea":   'North Korea',
  'Russian Federation':                      'Russia',
  'Syrian Arab Republic':                    'Syria',
  'Hong Kong SAR of China':                  'Hong Kong',
  'Canadian Territories':                    'Canada',
  'Macedonia':                               'North Macedonia',
  'Myanmar':                                 'Myanmar (Burma)',
  'Burma':                                   'Myanmar (Burma)',
  'Congo (Kinshasa)':                        'Democratic Republic of the Congo',
  'Congo (Brazzaville)':                     'Republic of the Congo',
};

​
How It Works

During normalization, each airline’s country field is checked against the correction map:
normalize.js:102-106
// Apply country name corrections
var countryIdx = 6;
if (COUNTRY_CORRECTIONS[row[countryIdx]]) {
  row[countryIdx] = COUNTRY_CORRECTIONS[row[countryIdx]];
}

​
Adding Country Corrections

To add a new country name correction:
1

Edit normalize.js

Open normalize.js and locate the COUNTRY_CORRECTIONS object around line 15
2

Add Entry

Add your correction to the map:
var COUNTRY_CORRECTIONS = {
  'Macao': 'Macau',
  'Republic of Korea': 'South Korea',
  // Add your new correction here
  'Your Upstream Name': 'Your Preferred Name',
};
3

Re-run Pipeline

Apply the correction to your data:
node normalize.js
node convert.js
Always use the exact upstream country name as the key (left side) and your preferred name as the value (right side).

​
Per-Airline ID Corrections

The ID_CORRECTIONS map applies field-level corrections to specific airlines by their OpenFlights ID. This fixes:
  • Field corruption (data leaked between columns)
  • Incorrect active status
  • Wrong IATA/ICAO codes
  • Duplicate entries that cause code collisions

​
Structure

var ID_CORRECTIONS = {
  // Key: airline ID (as string)
  // Value: object with field overrides
  '439': { country: 'United States' },
  '2056': { country: 'Hong Kong' },
  '4374': { iata: 'GQ' },
};
Available fields for correction: name, alias, iata, icao, callsign, country, active

​
Example Corrections

Some airlines have callsigns that leaked into the country field:
normalize.js:34-37
var ID_CORRECTIONS = {
  // Alaska Airlines: upstream has country="ALASKA" (callsign leaked into country field)
  '439':   { country: 'United States' },
  // Dragonair: upstream has country="DRAGON" (callsign leaked into country field)
  '2056':  { country: 'Hong Kong' },
Airlines incorrectly marked active that share codes with current airlines:
normalize.js:38-44
  // Defunct/duplicate airlines incorrectly marked active upstream:
  '2988':  { active: 'N' },  // Japan Airlines Domestic (dup of JAL, same ICAO JAL)
  '3384':  { active: 'N' },  // Malmo Aviation (defunct)
  '3386':  { active: 'N' },  // Malmö Aviation (dup of above, same ICAO SCW)
  '4264':  { active: 'N' },  // Royal Nepal Airlines (defunct, replaced by Nepal Airlines)
  '4560':  { active: 'N' },  // Swissair (defunct, dup of Swiss Intl, same ICAO SWR)
Defunct airlines sharing codes with active ones (prevents lookup conflicts):
normalize.js:45-51
  // IATA code collision fixes - defunct airlines sharing code with active ones:
  '595':   { active: 'N' },  // Atlant-Soyuz Airlines (3G) - ceased 2011
  '19862': { active: 'N' },  // AsiaCargo Express (3G) - rebranded/ceased 2020
  '2051':  { active: 'N' },  // DonbassAero (5D) - ceased 2013
  '3432':  { active: 'N' },  // Maxair (8M) - ceased 2005
  '3570':  { active: 'N' },  // Myanmar Airways International (8M) - duplicate entry
Wrong codes in upstream data:
normalize.js:79-82
  // Wrong IATA codes in upstream data:
  '4374':  { iata: 'GQ' },  // Sky Express Greece - correct code is GQ not G3
  '19849': { iata: 'GX' },  // Guangxi Beibu Gulf Airlines - correct code is GX not UQ
};

​
How It Works

ID corrections are applied after country corrections:
normalize.js:108-116
// Apply per-ID corrections
if (ID_CORRECTIONS[id]) {
  var fixes = ID_CORRECTIONS[id];
  Object.keys(fixes).forEach(function(field) {
    var idx = columns.indexOf(field);
    if (idx !== -1) row[idx] = fixes[field];
  });
}
Each correction object can override any field for that specific airline ID.

​
Adding ID Corrections

1

Identify the Issue

Find the airline ID that needs correction. You can search airlines.json or airlines.dat by name or code.
2

Edit normalize.js

Open normalize.js and locate the ID_CORRECTIONS object around line 33. Add your correction:
var ID_CORRECTIONS = {
  // ... existing corrections ...
  
  // Your new correction
  '12345': { 
    active: 'N',           // Mark as inactive
    country: 'New Country'  // Fix country name
  },
};
3

Add a Comment

Document why the correction is needed:
// AirlineX (AB) - merged with AirlineY 2024
'12345': { active: 'N' },
4

Test & Apply

Re-run the normalization pipeline:
node normalize.js
node convert.js
Always include a comment explaining the correction. This helps future maintainers understand why the fix was applied.

​
Other Data Quality Fixes

Beyond the correction maps, normalize.js also applies general data cleaning:

​
Callsign Whitespace Trimming

Some upstream records have leading spaces in the callsign field:
normalize.js:97-100
// Strip leading spaces from callsign (upstream data corruption)
if (row[5] && row[5][0] === ' ') {
  row[5] = row[5].trim();
}

​
Tracking Corrections

The normalize script reports how many rows were modified: 
$ node normalize.js
Normalized 127 rows in airlines.dat
This count includes all corrections applied: country name changes, ID-specific fixes, and whitespace trimming.

​
Best Practices

Document Everything

Always add comments explaining why a correction is needed. Include dates and reasons.

Test First

Before committing corrections, test them locally to ensure they work as expected.

Upstream First

Consider submitting fixes to OpenFlights first. Use local corrections only when upstream changes aren’t feasible.

Version Control

Commit normalize.js changes separately from data updates for clarity in git history.

​
Correction Priority

Corrections are applied in this order:
  1. Callsign whitespace trimming
  2. Country name corrections (via COUNTRY_CORRECTIONS)
  3. Per-airline ID corrections (via ID_CORRECTIONS)
If you specify a country correction both in COUNTRY_CORRECTIONS and in an ID_CORRECTIONS entry, the ID_CORRECTIONS value will take precedence for that specific airline.

Build docs developers (and LLMs) love

Get started for freeTalk to us