User guide

Timeline and recipes

Every change is a step you can undo, rewind to, or save as a recipe and replay on another file.

What the Timeline is

Every change you make is recorded as a step in plain words. This includes filtering, sorting, cleaning, masking, editing a cell, adding a column, and deleting rows. Steps appear in the Timeline panel on the right side of the screen. Toggle the panel with Cmd+B (Ctrl+B on Windows and Linux).

Non-destructive time travel

Click any step to jump back to that point. The data looks exactly as it did at that moment. Nothing is destroyed. Steps you leave behind are still there if you jump forward again.

If you go back to an earlier step and make a new change, BigXL creates a branch in the timeline rather than erasing your future work. You can jump between branches at any time.

Visual structure

The Timeline shows a tree graph with branches. Each branch represents a different path of changes. Each step has an icon showing what kind of action it was. Action types include filter, sort, edit, mask, delete, add column, fill, and colour.

Consecutive similar actions are grouped together automatically so the timeline does not get cluttered. For example, masking several columns in a row creates a single group. Click the group to expand it.

Renaming steps (milestones)

Double-click any step to give it a custom name. You can also right-click and choose Rename. Named steps appear as starred milestones. This is useful for marking important points like ‘before deduplication’ or ‘final clean version’.

Context menu on steps

Right-click any step to see options.

• Jump to this step goes back to this point in time.
• Rename step lets you add a custom label.
• Save recipe up to this step turns the history into a reusable script.
• Delete this step removes the action from your history.

Saving a recipe

Click Save in the Timeline panel, or right-click a step and choose Save recipe up to this step. This records the steps as a JSON file. A recipe captures what you did, not what the data looked like. This means it can run on a different file with the same column structure.

When saving, you can select which steps to include and which to skip.

Applying a recipe

• Apply to other tabs replays the current tab’s steps on other files you have open. Pick which tabs to apply to from a checklist. This is useful when you have a folder of files that all need the same cleanup.
• Import loads a saved recipe JSON file and replays it on the sheet you are currently looking at.

You can also pull steps from another open tab into the current one.

Batch processing a folder

From the File menu, select Apply these steps to a folder to apply a saved recipe to every file in a folder at once. This requires the Enterprise tier. BigXL processes the files outside the app and reports anything that failed a validation check instead of silently saving it.

Validation checks can include rules like ‘this column must be a number’ or ‘no duplicates in this column’.

When a file names its columns differently

Recurring exports rarely arrive with identical headers. A file that said name last month may say Full Name this month.

BigXL reads every file in the folder before it runs anything and tells you how many match the recipe, for example ‘2 of 3 files match this cleanup’. Any file that does not match is listed with the field it is missing and a dropdown of that file’s own columns.

• A column whose name normalises to the same letters, such as Full Name for name, is marked (looks right) as a suggestion. You still confirm it. BigXL never decides on your behalf that two differently named columns are the same thing.
• A file you have mapped stays in the list with a green marker and a note such as ‘Full Name used as name’, so you can see and change what you chose.
• The mapping is saved inside the recipe. Next month the answer is already there.

Files that are held back

A file BigXL cannot map is held back rather than partly processed. It appears in the results with the reason, and no output file is written for it. The summary reads, for example, ‘2 of 3 finished cleanly, 1 was not saved’.

This is deliberate. A run that applies the steps it understands and skips the rest produces an output you have no reason to distrust and cannot rely on.

The rule is strictest for privacy. If a recipe step masks a column and that column is not present in a file, the file fails. A masking step is never skipped because the column moved.

Combining a folder into one file

The same dialog offers two outcomes.

• Clean each file and save it on its own writes one output per input file into a folder you choose.
• Put every row into one combined file stacks every row into a single file instead.

Combining matches columns by name, not by position, so files whose columns are in a different order still line up. A column that exists in only some files is kept, and the rows that never had it are left blank. Tick Add a column saying which file each row came from to get a source_file column holding each row’s file name.

Combining does not apply your cleanup steps. The rows are stacked as they are, on the assumption that you would rather combine first and clean the combined file once.

Running the same folder twice

Results are replaced, not duplicated. Running a folder again overwrites the previous outputs rather than leaving report (2).csv beside report.csv. When you choose a results folder that already holds files the run would replace, BigXL says so before you start.