CSV to Fixed Width Converter

What CSV to fixed-width conversion actually does

CSV stores fields as delimited cells. Fixed-width stores fields as character ranges. This conversion maps each cell into a reserved width, pads or truncates to fit, and writes lines with no commas between fields — only positions.

You keep the logical columns from the spreadsheet, expressed the way a mainframe print file or portal upload expects them. The tool does not invent a COBOL copybook, packed-decimal encoding, or EBCDIC binary layout; it produces readable positional text from already-decoded CSV.

Success means every data line has fields starting at the same indices, empty fields still occupy their full width, and the total record shape matches what the importer’s parser was written for.

When to convert CSV to fixed-width

Convert when a legacy portal, bank, payroll vendor, or government system only accepts positional uploads; when a batch job’s JCL or control card describes fixed record layouts; or when you edited data as CSV and must return a column-aligned file.

Skip conversion if the consumer already accepts CSV, JSON, or an API — prefer the modern path. Skip if you still lack authoritative widths; guessing from one row risks silent misalignment across an entire batch.

A practical pattern: Fixed Width to CSV for exploration, Online CSV Editor for cleanup, then CSV to Fixed Width with documented lengths for the final drop.

• Good fit: mainframe uploads, COBOL batch inputs, space-padded gov filings, vendor FTP drops.
• Poor fit: systems that already take CSV/TSV, binary VSAM, or EBCDIC packed records without a text stage.
• Better as a script: nightly production loads with a published copybook and CI checks on record length.

Padding, truncation, and field alignment

Padding fills unused positions so the next field always starts at the same character index. Right-padding with spaces is the common default for alphanumeric fields in print-style layouts. Some numeric conventions left-pad with spaces or zeros; follow the spec when it differs.

Truncation enforces the width when a value is too long. Without truncation, a long city name would push AGE or ZIP into the wrong columns for that row only — catastrophic for positional parsers. Prefer fixing source data or using the layout’s overflow fields over silently losing characters when compliance depends on full values.

Internal spaces (New York, compound surnames) stay inside one field; they are not delimiters. Only the assigned width decides where that field ends.

Matching widths to copybooks and data dictionaries

COBOL copybooks and flat-file dictionaries list picture clauses or start/length pairs. Translate those lengths into the width list you apply here. Include filler fields even if your CSV omits them — you may need empty padded columns to keep later fields on the correct offsets.

Government layouts often publish PDF or spreadsheet dictionaries with position and length. Build your CSV column order to match that sequence before converting; reordering columns after padding is harder than sorting the sheet first.

When the dictionary and a vendor sample disagree, trust the dictionary for production and ask the vendor about the sample. Do not split the difference on widths.

Headers: emit, omit, or label-only

Some human-oriented fixed reports include a header line of labels padded like data. Many machine importers reject any non-data row. Read the upload guide before including headers in the output.

If you emit headers, pad labels to the same widths as data columns so the header itself is positional. A comma-separated header line above fixed-width data is usually wrong for legacy jobs.

For debugging, keep a headed version for humans and a data-only version for upload. Do not mix them in the file you submit.

Excel, Sheets, and preparing the CSV

Spreadsheet apps mutate data on open: leading zeros vanish, long IDs become scientific notation, and dates reformat. Import CSV with columns forced to Text when those values matter, or maintain the sheet with apostrophe/text formatting before export.

Remove blank trailing columns, consistent quote issues, and merged cells before export. Fixed-width conversion assumes a rectangular CSV; ghost columns become empty padded fields and shift everything if widths are assigned by position.

Save as UTF-8 CSV when the target allows Unicode. If the mainframe expects a legacy code page, transcode after generating positional text with a dedicated encoding step — do not assume UTF-8 bytes will be accepted.

Empty fields, fillers, and numeric signs

Empty CSV cells should become a run of spaces equal to the field width. That preserves alignment for following columns. Deleting those spaces by hand breaks the contract.

Layouts often include FILLER or unused reserved zones. Represent them as empty CSV columns with the correct widths, or concatenate constant spaces in a preprocessing step if your sheet should not show them to editors.

Signed numerics, implied decimal places, and overpunch conventions are outside simple space-padding. If the importer expects those encodings, apply them in a specialized step; this converter focuses on character-oriented positional text.

Line endings, record length, and file shape

Many legacy systems expect a fixed record length including or excluding the newline. Confirm whether CRLF (Windows) or LF is required. Wrong line endings can make an otherwise correct payload fail validation.

After conversion, check that every data line has the same character count (before newline). Variable lengths usually mean inconsistent widths, missing fillers, or unexpected multiline CSV fields that were not flattened.

Multiline quoted CSV cells do not map cleanly to single-line positional records. Flatten addresses and comments into one line per field before converting.

Round-trip with Fixed Width to CSV

Fixed Width to CSV helps you explore a sample dump and clean data in a spreadsheet. CSV to Fixed Width rebuilds positional text for upload. Together they support edit-in-modern-tools, submit-to-legacy workflows.

Round-trips are not guaranteed bit-identical. Detection approximates boundaries; padding style, truncation, and header policy may differ from the original producer. For compliance-sensitive feeds, compare your output to a golden vendor sample in a monospace diff.

When detection inferred widths that differ from the official dictionary, always prefer the dictionary on the way out.

Mainframe, COBOL, and portal quirks

Mainframe transfers may wrap files with RDW, block headers, or require a specific RECFM. Your desktop fixed-width text is often an intermediate form that a transfer tool or middleware re-wraps — confirm the handoff format with ops.

Carriage-control columns and printer spacing appear in some reports but not in upload layouts. Do not copy report chrome into an inbound positional file unless the spec says so.

Portal validators may check checksum rows, trailer counts, or file naming patterns in addition to field positions. Passing visual alignment is necessary but not always sufficient.

Typical workflows

Analyst edit: receive a fixed dump, convert to CSV, fix rows in Sheets, convert back with documented widths, upload.

Vendor onboarding: build a CSV template matching the dictionary column order, fill data, generate fixed-width for the first certification file.

Test fixtures: maintain small CSV golden sets in git and generate fixed-width artifacts in CI for importer tests.

Incident recovery: regenerate yesterday’s positional file from the archived CSV when a corrupted upload must be resent.

Browser limits, privacy, and sensitive uploads

Parsing and padding run locally. CSV content is not sent to Convert CSV servers for the transform. Site analytics and advertising follow the privacy policy separately.

Very large sheets can slow the tab. Convert a pilot batch first; use offline scripts for millions of rows once widths are proven.

Payroll, tax, health, and case files often contain PII. Control who receives both the CSV and the fixed-width output, and clear downloads on shared machines after the portal accepts the file.

Quality checklist before you upload

Confirm widths match the current dictionary version. Confirm header policy. Spot-check longest values and empty fields in monospace.

Verify record length, line endings, and encoding. Compare against a known-good sample when available. Note the source CSV and generation time for support tickets.

• Widths from the official spec, not a guess.
• Monospace preview shows stable columns.
• Empty fields still occupy full width.
• No unintended truncation on critical fields.
• Encoding and line endings match the portal.

Related tools and next steps

Need to spreadsheet a positional dump first? Use Fixed Width to CSV. Cleaning headers, blanks, and column order is easiest in the Online CSV Editor before you pad for upload.

If the destination can accept delimited files, prefer CSV or the site’s Excel converters and skip positional formatting. Use this guide when position — not commas — is the contract the importer was written for.