BROWSER-BASED TABLE TOOL

Concatenate CSV Columns

Concatenate CSV columns in your chosen order with explicit separators and empty handling. Work locally in your browser and review results before export.

File contents are processed locally in your browser. Refreshing or closing clears the session. The website still requests static resources.

1Import data—2Configure—3Review—4Export

Drop your table here

Drag and drop a CSV, TSV, or delimited text file

UTF-8 · Up to 10 MB per file

Start with your data

File content is processed in your browser, without uploading it.
Kept in page memory · Continue across tools · Refreshing or closing clears the sessionCSV / TSV · UTF-8
Sample library · Normal / Boundary / Error

Clear this session?

Imported data, steps, and results will be released. Export any needed copies first. Original files are unaffected.

Combine field contents in a deliberate order

Concatenate CSV columns when several fields belong in one readable value. A city and street can become an address line; an item identifier and color can become a label; a title and reference can become an internal description. The operation reads two or more fields from the same record and adds one destination. It does not find a matching record in another file, choose a preferred nonempty source, or merge the contents of different records.

The default preserves every source column and keeps the record order unchanged. Text stays text: 00123 retains its zeros, a long order number is not rounded, and a date-like value is not reformatted. Spaces already present in a source value are retained. This matters when a description uses intentional spacing, but it also means that an existing trailing space can combine with a space separator. Review source text before assuming that the separator alone caused an unexpected gap.

Start with confirmed cells

Choose files accepts CSV, TSV, and delimited text in UTF-8. Paste a table works with copied spreadsheet cells or CSV text. The parsing dialog separates file delimiters from the separator used for concatenation. First select Parse and preview, check the columns and header names, then Confirm import. A quoted comma or newline inside a field remains part of that field. The operation joins parsed values, never fragments cut directly from the raw CSV file.

In Source fields and order, select at least two fields. The numbered list records the order in which their contents will appear. Use the up and down buttons to rearrange it, or drag an item when using a pointer. The buttons provide the same ordering action from the keyboard. Selection order is independent of the visual order of columns in the original file. An identifier can appear before a description even when it was originally the last column.

Choose a Delimiter preset, or type a literal custom separator. An empty separator joins text without adding anything between values. A line break creates a multiline cell, not multiple output records. Prefix and Suffix add fixed text around the joined content. Supply a distinct Destination name and choose where to insert the new column. Keep source columns starts enabled. If you disable it, only the result copy drops the selected sources; the imported table remains available in the workflow history.

Three different ways to handle empty fields

Keep slots preserves a separator position for each selected source. With A, an empty string, and B joined by a hyphen, the output is A--B. This can be useful when a downstream process interprets positions, such as a middle component that may be absent. Skip slots omits empty sources before inserting separators, giving A-B. It is often more readable for a name or address, but it deliberately removes information about which component was absent.

Keep slots and flag for review produces the positional form while putting records with an empty source into a review report. This allows a team to inspect incomplete address components before importing a combined address into another system. The main result still contains those records. A review flag does not prove that the data is wrong; an apartment number or secondary name can be legitimately missing. Decide what is required for your actual destination rather than treating every blank as an error.

Empty definition distinguishes a missing cell from an empty string. Both are selected initially. Whitespace-only text is not empty unless you enable that setting. You can also list exact additional missing markers, one per line. Neither zero nor false is empty by default, and the literal strings NULL and N/A keep their content unless explicitly listed. These choices affect emptiness decisions; they do not generally trim, lowercase, or replace the original nonempty text.

All-empty records and fixed text

When every selected source is empty, Leave result empty is the default. It prevents a blank address from turning into a line of punctuation and prevents Prefix or Suffix from creating a misleading label on an otherwise empty record. This all-empty rule is evaluated before ordinary joining. It applies even if you supplied visible affixes. The All sources empty count helps identify these cases separately from successful nonempty combinations and records flagged for review.

Emit affixes and configured separators is an explicit alternative. If you keep slots, empty positions may still contribute separators between the affixes. If you skip slots, only the affixes remain. For example, a prefix TEST[ and suffix ] can produce TEST[] when all sources are skipped. This can be appropriate for a required placeholder format. It is usually inappropriate when an empty value must remain truly empty for the receiving system's missing-value logic.

Preview a complete address example

Consider fields city, street, and number. A normal record contains Hangzhou, Test Road, and 12. Select them in that order, choose a space separator, and name the destination address. The generated text is Hangzhou Test Road 12. Retaining all three source columns lets you confirm that a record with a missing street has not silently shifted its information into another field. Prefix and Suffix can add a test label without modifying the three source values.

Now compare a record containing Hangzhou, an empty street, and 12. Keep slots inserts two spaces between the nonempty values. Skip slots inserts one. If the city itself ends in a space, that source space survives either policy, so the visible gap can be larger. Use the cleaning tool in a separate, explicit step if trimming is actually wanted. Do not rely on concatenation to infer whether a space belongs to the text or to its presentation.

A second example uses the fields id, color, and size with values 00123, Blue, and L. A hyphen separator yields 00123-Blue-L. Move size before color and the output becomes 00123-L-Blue. The destination should follow the configuration immediately after a new run. An unchanged result is also valid: if two chosen sources happen to have equal text, exchanging their positions may have no visible effect for those records.

Review counts, preserve a rule, continue processing

Preview sample and estimate shows a small demonstration using current records. Run and review calculates the full result. Inspect Successful records, All sources empty, and Needs review with the actual output table. These counters describe different properties and may overlap. The total output record count stays equal to the input count. View search only changes what the grid displays; it does not silently narrow the exported result. Choose a report explicitly in the export dialog when you need review records alone.

Save rules stores the selected field bindings, order, separator, affixes, and empty settings. Rule files may reveal field names and fixed business text, so inspect them before sharing. Loading a rule against another file requires the expected column structure; an incompatible structure is rejected. Within the current workflow, stable column identities protect references through renaming or rearrangement. Apply to workflow passes the reviewed table to column mapping, filtering, or another existing tool, with undo and redo available for applied steps.

Export and troubleshoot the destination

Export new copy offers CSV or TSV, headers, and UTF-8 with or without BOM. A comma or newline inside a newly combined value is escaped by the shared serializer. Verification parses the serialized result again and checks its shape and values. A multiline address remains a single field in that check. Formula-prefix protection is explicit and reports changes, including protection added to text beginning with a plus sign or minus sign. It is not a promise that every spreadsheet application interprets every value identically.

A destination name conflict blocks the run instead of replacing another column. A combined cell that exceeds the field-length limit also blocks processing rather than cutting off the end of the address. Reduce the selected text, choose a more suitable task, or process a smaller source if a broader limit is reached. Cancel task stops a long-running worker. Changed settings mark the old result stale, preserving it for inspection while disabling export. Refreshing or closing the page clears the in-memory session, so download required results and rules before leaving.

Frequently asked questions

How do I combine first and last name columns with a space?

Select the columns in the order you want, use a space as the separator, and choose how empty values should be handled. Inspect names with a missing part before exporting. This combines the fields you provide; it does not infer a person’s name order or reliably split an existing full name into its components.

How do I avoid extra separators around missing fields?

Choose Skip slots and verify the empty definition. Spaces already inside nonempty source text remain; trim them in a separate cleaning step only when that is appropriate.

Can I add a fixed label before the result?

Yes, use Prefix and Suffix. All-empty records still produce an empty result by default. Change the all-empty option explicitly if those records should contain the label.

Is concatenation the same as selecting the first nonempty column?

No. Concatenation combines the selected contents in order. Skipping empty slots removes blanks from that sequence but still includes every selected nonempty field.

Why is a missing address listed for review?

The review policy identifies records with at least one empty selected source. It does not decide whether that component is required by your business. Inspect the report before removing records.