csvmerge: an interactive merge tool for .csv files

This is an interactive merge tool specifically designed for .csv files and written in Python. It performs a line-by-line comparison and when it detects a conflict it prompts the user to select the source file from which to pick the preferred value.

Basic usage

Run the script as:

python3 csvmerge.py -i1 <input-file-1> -i2 <input-file-2> -o <output-file>

In case of conflict the user is prompted to select the source file: they need to press 1 for file n. 1 or 2 for file n. 2, and confirm the choice with <Enter>.

If the user realizes that a specific conflict is going to recur consistently, they can automate their decision by entering 1+ or 2+. This adds the choice to a list of automatic responses, preventing future prompts for that conflict. A series of asterisks on the screen indicates the total number of automatic choices that have been saved.

Iris data set from different sources¹

Other options

• Use --sort to sort alphabetically the columns of both input files. This option is useful when the two files share some, but not all, columns.
• Use --skip1 <column-name> and --skip2 <column-name> to remove the specified columns from file n. 1 and file n. 2 before the comparison begins. You can specify multiple columns for both files. Removed columns are added back to the output file, on the right, file n. 1 before file n. 2. This option, too, can be useful when the two files share only some columns.
• Use --drop1 <column-name> and --drop2 <column-name> to remove the specified columns from file n. 1 and file n. 2 before the comparison begins. You can specify multiple columns for both files. Removed columns are not added back to the output file.
• Use --case-insensitive to perform case-insensitive comparisons. If two values differ only by case, the value from file n. 1 will be selected.
• Use --skip-empty to ignore empty values. If only one of the conflicting cells is empty, the value of the other cell will be automatically selected.
• Use --always <id> if you already know what to do in case of conflict and you don't need to interact with the tool:
  • --always 1 to always pick the value from file n. 1.
  • --always 2 to always pick the value from file n. 2.
  • --always l to always pick the longest value. If the two values have the same length, the value from file n. 1 will be selected.
  • --always s to always pick the shortest value. If the two values have the same length, the value from file n. 1 will be selected.
• Use -q to activate quiet mode.
• Use --no-color to highlight the differences using asterisks rather than red ANSI formatting.
• Use --info to display the information about the input files and exit.
• Use --delimiter <char> to specify the character used to separate the values in the .csv files. It defaults to ,.

All these options can be displayed by running the helper with -h.

--sort, --skip1 and --skip2 work only if the input files begin with a header row.

Contributing

Contributions are most welcome by forking the repository and sending a pull request. Errors and new features proposals can be reported opening an issue as well.

References

1. J. C. Bezdek, J. M. Keller, R. Krishnapuram, L. I. Kuncheva and N. R. Pal, "Will the real iris data please stand up?", in IEEE Transactions on Fuzzy Systems, vol. 7, no. 3, pp. 368-369, June 1999, doi: 10.1109/91.771092.