Common Issues

How should I choose translation providers?

For most teams outside Mainland China, start with:

• DeepL
• OpenAI / ChatGPT
• Google

Then add regional providers only if needed for coverage, pricing, or network availability.

Recommended order is controlled by i18n-mage.translationServices.translateApiPriority.

Which i18n key patterns are supported?

The extension primarily detects keys from i18n function calls such as t(...).

Supported patterns include:

• direct keys, for example t('common.ok')
• dynamic composition, for example t("prefix." + key + ".suffix")

For dynamic patterns, usage analysis may match multiple keys. Inline hints show the first matched result.

Why were some undefined keys not fixed?

Check the following:

1. i18n-mage.translationServices.matchExistingKey and i18n-mage.translationServices.autoTranslateMissingKey are enabled as needed.
2. i18n-mage.translationServices.referenceLanguage is set correctly.
3. If language validation is enabled, source text matches expected language characteristics.
4. Your translation provider API key and quota are valid.

Why are translation calls failing intermittently?

Common causes are:

• API rate limits or quota exhaustion
• temporary provider-side errors
• network or proxy routing issues

You can reorder translateApiPriority, retry, or switch provider.

Why is performance slow in large repositories?

Tune analysis settings:

• configure i18n-mage.workspace.ignoredDirectories
• reduce scanned file types via i18n-mage.analysis.fileExtensions
• keep i18n-mage.analysis.fileSizeSkipThresholdKB at a reasonable value
• use decorationScope = visible for translation hints