# rename-ext

> changes file-extension based on mime-type

### also

- removes `*?|^:"<>` characters from the file-name
- reducing multiple spaces, dots, and underscores
- adds rename-date suffix to ensure uniquness of the filename
- strips exec-flag from non-executive types (usual leftover from FAT/NTFS)
- optimizes images while at it (shows reduction size in `kb` and `%`)


## install

1. look at [install](./install.sh) file for instructions
2. adjust `PREFIX` at wish
3. make sure the final directory is on `$PATH`


## use

`rename-ext /Data/Pictures/_unsorted/**/*`

to only change extension and nothing else:

`rename-ext --only-ext /Data/Photos/2020/*`

to only print a report:

`rename-ext --test-run /Data/Docs/_dump/*`


## requires

- [`file`](https://darwinsys.com/file/) - to determine file type
- [`sd`](https://github.com/chmln/sd) – better sed
- [`jpegoptim`](https://github.com/tjko/jpegoptim) – to compress JPEGs
- [`imagemagick`](https://imagemagick.org/)
  or [`graphicsmagic`](http://www.graphicsmagick.org/) – to mogrify PNGs
- [`pngcrush`](https://github.com/Kjuly/pngcrush) – to compress PNGs
<!-- - [`pngquant`](https://github.com/kornelski/pngquant) – to compress PNGs -->


## todo
- [ ] find better determination tool for `application/octet-stream`
- [ ] leverage `fd` to do recursive paralel execution if some of parameters are folders


## q&a

**Q**: why not just rename files to hashes of their contents, and store original filenames in some database?

**A**: good point. that database is in my case a filesystem itself.