Import and Export Data

Bulk import and export in Data tab → Database — Excel, CSV, multimedia files, and account creation.

Supports importing up to approximately 10 million rows, at up to 100,000 rows per minute.

Import Data

Prepare Data Files

Table file requirements:

UTF-8 encoded Excel files (.xlsx only) or CSV files.
First row must be headers.
Cells intended to be blank must be completely empty (no hidden spaces or line breaks).
Must contain at least one data record.
No formula calculations are supported.

You can download template tables from the bottom left of the upload interface.

excel_sample

For importing images, videos, and files, prepare an additional ZIP package:

Place media files anywhere in the ZIP — at the root or inside any sub-folder. Dedicated image, video, and file folders are no longer required.
Each file is imported as the type of the column you map it to: map it to an image column and it imports as an image, map it to a file column and it imports as a file. The same image can therefore be imported as both an image and a file (attachment), and one file referenced by several columns of different types is imported into each.
In the CSV or Excel file, each multimedia column value must be the file’s path exactly as it appears inside the ZIP, e.g. cover0.png, products/cover0.png, api.pdf.
Compress the files (or folders) directly into a ZIP. Do not wrap everything in an extra top-level folder before compressing.

database

Upload Files

Click Import in the bottom-left corner of Database. Upload prepared tables and optional multimedia ZIP. Multiple tables per run (mixed CSV and Excel); one multimedia ZIP only.

entry

upload_file

Select Import Type

Choose Insert or Update. Insert means directly inserting data into the database, while Update updates existing records based on the primary key (id).

Configure Mapping

Mapping consists of:

Table mapping: Maps uploaded tables to database tables. System tables (account, payment, etc.) are not supported in this flow — use Import accounts below for bulk account creation. Each Excel sheet and each CSV file maps to one table.
Field Mapping: Maps columns from uploaded tables to database table fields. Columns that don’t need to be imported can be ignored using the Ignore option on the right side of each row.

mapping

Related Field Mapping

If other tables are related to the current table, related table mapping needs to be configured.

How it works: after a relation is established, the current table stores the related record ID. During import, match a column in your file to a field on the related table (such as username or email) to resolve the corresponding unique ID.

For example, if account and post have a one-to-many relation, the author column in post stores the account id. During import, match an author name in your file to username on the account table to resolve the ID.

To prevent ambiguous ID resolution when related lookup values are duplicated (for example, multiple Account rows share the same display name), configure a unique constraint so the field you match on (such as username) is unique in the related table.

unique_constraint

After configuring the unique constraint, when mapping related tables, map the author column from the Excel table to the username field in the Account table. If the constraint includes multiple fields, configure matching for multiple fields.

relation_mapping

Multimedia Field Mapping

To import multimedia files (images, videos, and files), each multimedia column value should be the file’s path inside the uploaded ZIP, such as cover0.png, media/intro.mp4, or api.pdf. Each file is imported as the type of the column it is mapped to.

During import, the system will automatically extract these files from the uploaded ZIP package and store them in the database.

Geographic Coordinate Field Mapping

To import coordinates, first select the coordinate type (WGS84, GCJ-02, or BD-09), then configure latitude and longitude mapping separately. Therefore, the import table needs two columns to store latitude and longitude.

geo_mapping

Data Conflict Handling

When the import mode is Insert and fields have unique constraints, a conflict occurs if duplicate values appear in the imported data (or between imported data and existing rows, depending on your configuration). Conflict handling applies only in these unique-constraint conflict cases.

First, select the constraint:

Any Unique Constraint: Triggers conflict handling if any field with a unique constraint conflicts.
Specific Unique Constraint: Select a specific unique constraint as the basis for conflict judgment.

The system provides three conflict handling methods:

Do nothing: Ignore rows that conflict with the selected unique constraint and continue importing non-conflicting rows. Use case: You can tolerate skipping conflicting rows and prioritize importing the rest first.
Update: Use the newly imported data to overwrite existing rows that conflict under the selected unique constraint. Use case: Imported data is the latest source of truth and should overwrite existing values.
Terminate: Immediately stop the entire import process and record detailed error information. This option is only available when Any is selected. Use case: You require strict consistency and cannot accept any unresolved unique-constraint conflict.

constraint

Trigger Configuration

If the data table has configured triggers (such as INSERT, UPDATE, or DELETE), you can choose whether to trigger them during import. Import triggers use a whitelist mechanism. During import:

Newly added triggers are not triggered unless they are added to the whitelist.
Modified triggers are triggered with their updated logic if they remain in the whitelist.
Deleted triggers are removed from the whitelist and will no longer be triggered.

Preview Data

After configuration, you can preview the import data. The system displays the first 10 rows of the table for users to confirm data format and content. Multimedia file previews only show filenames, not specific content.

preview

Execute Import and View Results

After confirmation, click the Import button to start the import process. The import process uses temporary storage, which is released automatically when the import completes.

After import, the system will indicate success or failure. Detailed import results, including successful entries, can be viewed in History. If records were skipped (due to conflict handling) or failed (due to data or system errors), the system provides a downloadable log file (retained for 7 days) with detailed records and error reasons. If import fails, first check failure details in History, then use Import Error Handling to quickly locate the issue.

To ensure data integrity, if the import encounters a system or data error (for example, a type mismatch), the operation is atomic: none of the data from this batch will be written to the database. Records intentionally skipped via conflict handling do not trigger this rollback.

import_result

log

Import accounts

Bulk account creation — insert only; does not update existing accounts by id.

Bulk account creation is essentially a bulk call to the account registration API, so it directly depends on the sign-in methods enabled in your project.

In Settings tab → Sign in, enable username, email, and/or phone sign-in.

When multiple sign in methods are enabled, imported data can include any one sign in method, or multiple sign in methods at the same time.

Prepare Account File

Prepare a table file containing fields related to the sign in methods. File format requirements are the same as Prepare Data Files:

UTF-8 encoded Excel files (.xlsx only) or CSV files.
First row must be headers.
Cells intended to be blank must be completely empty (no hidden spaces or line breaks).
Must contain at least one data record.
No formula calculations are supported.

You can also fill in a template file downloaded from the lower-left corner of the Import accounts upload interface.

Only four fields are supported for import: username, email, phone_number, and password. Other fields (such as profile_image) are not supported.