mongoexport: การส่งออกคอลเลกชันเป็นไฟล์

mongoexport is a MongoDB Database Tools command-line utility that reads documents from a collection and writes them to a file. It outputs data in JSON (JSONL format, one document per line) or CSV format. Common uses include creating data backups for analysis, migrating data to another system, sharing snapshots for debugging, or seeding other environments from a production sample.

The simplest export command specifies the source database, collection, and output file. By default mongoexport writes JSONL—one JSON document per line—which is ideal for streaming large exports and re-importing with mongoimport. Use --out to name the output file; without it, documents print to stdout.

Use --type csv to export as comma-separated values. CSV exports require you to list the fields to include with --fields—mongoexport cannot infer all fields from a schema-flexible collection automatically. The first row of the CSV output contains the column headers matching the fields you specified.

Use the --query flag with a JSON filter document to export only the documents that match a condition, instead of the entire collection. This is useful for extracting a subset—such as all orders from a specific month, or all users with a certain role—without exporting data you don't need.

Use --fields (or -f) with a comma-separated list of field names to include in the output. Fields not listed are excluded. This is equivalent to an inclusion projection in a find query. Combined with --query, you can export precisely the data slice you need, keeping file sizes small and sharing only relevant fields.

Use --sort with a JSON sort document to export documents in a specific order. This is useful when you need the output file to be ordered for downstream processing—for example, exporting events in chronological order for replay, or exporting users alphabetically for a report.

Use --limit to export at most N documents and --skip to skip the first N documents. These flags let you export a sample—for example, the 1,000 most recent orders—or export in batches by paginating through the collection with repeated calls using skip offsets. Use with a --sort to make the limit meaningful.

mongoexport outputs Extended JSON by default, which uses special wrapper objects to represent BSON types not native to JSON—for example, {"$oid":"..."} for ObjectId and {"$date":"..."} for dates. Use --relaxed to output relaxed Extended JSON where dates become ISO strings and numbers are plain JSON numbers—more readable but loses some type fidelity for re-import.

Use the Atlas cluster's SRV connection string in the --uri flag to export from an Atlas cluster. Atlas requires TLS (automatically enabled with SRV URIs). For very large Atlas collections, prefer the Atlas UI's 'Export Collection' button, or use Atlas Data Federation for analytical exports to S3 or other data lakes.

Choose the right tool for your use case:

• mongoexport: outputs human-readable JSON or CSV; suitable for data sharing, analysis, or selective migration; can filter and project
• mongodump: outputs binary BSON; full collection or database backups; faster and preserves all BSON types losslessly; not human-readable

Without the --out flag, mongoexport prints to stdout, enabling Unix pipe compositions. You can pipe the output to wc -l to count documents, to grep to filter lines, to jq for JSON processing, or to another command entirely. This makes mongoexport a flexible component in data pipeline scripts.

Test your understanding of MongoDB & NoSQL Databases concepts from this lesson.

In this lesson you learned: mongoexport outputs JSONL or CSV from a collection with optional filter, field selection, and sort, --query lets you export a targeted subset of documents, and mongodump is preferred for full backups while mongoexport is better for selective, human-readable exports. Next up we explore mongodump and mongorestore for full binary database backups.