Compare two lists

Decide what you want the comparison to answer

List comparison tells you which entries are shared and which are missing from either side. It does not require the lists to be sorted. If List A contains the same three product codes as List B in a different order, their membership is identical. A document diff would show moved lines in that situation; this tool instead asks whether each entire value occurs on both sides.

Put a known baseline in List A and the version you want to check in List B. For example, put yesterday's assignment list in A and today's export in B. Only in A then shows assignments that disappeared from the export; Only in B shows assignments that were added. If you put the versions in the opposite panels, the answers reverse. Label your source files before copying so that the meaning stays clear.

For a report, describe the question before exporting a result: “Which IDs exist in the current inventory but were not present last week?” That is more useful than an unlabeled file of IDs. It also helps a reviewer know whether ordering, repeated appearances or values in other columns need a separate check.

Paste one exact item per line in each panel

Try this with three product codes in each list. Paste the first block into List A and the second block into List B. There are no comma separators or labels in the input; the line breaks tell the editor where each item ends.

List A

SKU-101
    SKU-205
    SKU-340

List B

SKU-340
    SKU-420
    SKU-101

Only in A returns SKU-205. Only in B returns SKU-420. In both lists returns SKU-101 then SKU-340, following their order in A. Seeing those two in a different order in B is not treated as a change. Try switching views without editing the input to see how the same pair of lists answers different questions.

If you have a comma-delimited row such as Paris,"New York, NY",Berlin, use the list formatter to split it before comparing. Pasting it as one line here makes it one item. If a spreadsheet export contains several columns, select the matching identifier column instead of comparing the full record.

Go back to the two list panels

Choose the result view that matches the task

Use Only in A to find a baseline value missing from the second list. Use Only in B to find a value present only in the second list. Use In both lists to check overlap. For example, when comparing enrolled participants with an attendance export, Only in A can identify names or IDs still missing from the attendance list, while Only in B can flag unexpected entries for review.

All unique items produces one list of everything seen on either side. It keeps the order of first appearances in A and appends items exclusive to B. Items in just one list combines the two differences, with the A-only group first. If direction matters, export the groups separately and label them. A combined output does not attach an A or B marker to each line.

Before treating any missing item as an error, make sure the two exports represent the same scope. A list for a different date, project, filter or permission level can be valid and still look incomplete beside the baseline. The tool cannot infer which records ought to exist.

Choose whether case and outer spaces matter

Default matching uses each complete line as written. ABC-09 and abc-09 are different items unless Ignore case is on. That option can be useful for a list of display names, but risky when case is significant in filenames or system identifiers. Start with exact matching and relax it only when your matching rule is deliberate.

Ignore outer spaces treats a value surrounded by whitespace as equivalent to the same value without those outer characters. It does not collapse whitespace inside an item. A name with one internal space and the same name with two can still differ. The output preserves the first original spelling and spacing from the relevant list. If you need a cleaned result as well as a comparison, prepare a separate normalized copy rather than assuming comparison flags rewrite the source.

If values look identical but remain different, inspect an affected pair in the invisible character inspector. Ordinary spaces and nonbreaking spaces can look similar. An accented character may also have more than one underlying representation. Review the effect of Unicode normalization before using it on keys that may need exact equality.

Compare a stable key instead of a whole record

A product label can change without changing the product. Comparing full records such as SKU-205,Small blue notebook and SKU-205,Blue pocket notebook treats them as different lines. If you want to know whether a product still exists, compare SKU-205 alone. Then inspect the description change separately for records that occur in both exports.

Use the CSV and TSV converter to extract the same column from each export. Verify that you picked the right position in each file: column 2 in one table may not have the same meaning as column 2 in the other. Skip headers in both inputs if they are labels, not records. Preserve leading zeros and review rows with missing IDs. An empty identifier cannot safely stand in for a unique record.

Do not use people's names as unique identifiers when you have a more stable key. Two different people can share a name, and the same name can appear in multiple forms. Use only the minimum necessary fields when investigating a discrepancy, especially if you will share screenshots or results outside the original team.

Do not use membership to verify quantities

A list of repeated values answers a different question from a set of unique values. If A lists part-17 twice and B lists it once, the item is in both lists. It is not a difference in membership, although the number of occurrences changed. The tool deliberately counts each matching key once and ignores blank lines. It is not suitable as a reconciliation of inventory counts or transaction quantities by itself.

When repeated appearances are meaningful, keep the original input and use an appropriate method to count entries by key. Compare those counts as separate data. If you only want one copy of each line in a single list, use Remove duplicate lines instead. You do not need a second panel for that job.

If you need to see reordering, insertions or edits in a document, use the text diff checker. It can show changes in line order and highlight edits inside words. Use the two tools in sequence if the job requires it: first check which IDs appeared or disappeared, then review actual edits to the records that stayed.

Review a result before acting on it

Start with a small known example with one item unique to each side and one shared item. This checks that you understand which side you are viewing. Then inspect an actual output value against both original files. A surprising absence may be a case difference, a hidden character, a differently formatted ID or a row that was filtered out before export. Changing matching options should be a deliberate decision, not a way to make a discrepancy disappear.

Record any matching rules you enabled when you share a result. For example: “Only in B, with outer spaces ignored and case preserved.” Note the time or version of each source. If the result will drive a bulk update, ask a reviewer to confirm the expected scope and test a few entries first. The preview reports matches, not permission to remove or change records.

When a list was assembled from several sources, use the lists and tables guide to check the separators and the record boundaries before repeating the comparison. A carefully prepared key list makes discrepancies easier to explain.

Run a final checklist on the full lists

Verify that each panel contains one meaningful value per line and that the two lists describe the same unit. Confirm that both use the same identifier convention, including punctuation and leading zeros. Check the first and last few lines of both inputs so that missing copied rows are not mistaken for real differences.

Look at shared and unique items together before choosing what to export. If nearly every item is unique, inspect a pair that you expected to match: the mismatch may come from entire rows being pasted instead of their keys. If every item matches even though you expected a difference, make sure you did not paste the same version twice or enable a case rule that erases a distinction you care about.

Finally, save the output using a filename that includes the selected view, such as inventory-only-in-current.txt. Leave the originals unchanged. This makes it possible to repeat the audit or explain an unexpected result later without reconstructing the input from a single list of differences.