mapsd

Because sometimes sed needs an explicit key/value map.

Quickstart

1. Download latest release for your system

Or you can also build it from source using cargo (see Installation). Then move the binary somewhere like in /usr/local/bin (just make sure it is some folder already in your $PATH)

2. Create a `.csv` file with 2 columns:

old_string1,new_string1
old_string2,new_string2
old_string3,new_string3

WARNING: beware of the spaces and separator you put here !! Every character counts, even spaces

3. Replace all occurrences of the 1st column with the 2nd, over all `.html` files in the `./Documents` directory

mapsd "./Documents/**/*.html" -m ./my_map.csv

(No worries, by default it applies the replacement in a copy of the files. When you feel you are ready you can add the --DANGEROUSLY-REPLACE-INPLACE flag to actually replace things in place.)

Usage

mapsd 0.0.1
Find & Replace text in multiple files using an explicit CSV of Before/After pairs.

USAGE:
    mapsd [FLAGS] [OPTIONS] <FILES>

FLAGS:
        --has-headers                    CSV has headers
    -h, --help                           Prints help information
        --DANGEROUSLY-REPLACE-INPLACE    Replace files in-place (USE WITH CAUTION)
        --silent                         Suppress output
    -V, --version                        Prints version information

OPTIONS:
    -d, --delimiter <delimiter>    CSV delimiter [default: ,]
    -m, --map <map>                Path to the CSV file containing key/value pairs [default: map.csv]
    -p, --prefix <prefix>          Prefix to use for the resulting copied file [default: replaced.]

ARGS:
    <FILES>    Files to process (glob pattern)

The quickstart is quite explicit, but here are some tips & tricks.

Beware of spaces in the .csv file: old_string1,new_string1 is not the same thing as old_string1, new_string1 nor old_string1,new_string1 . All characters count and WILL be matched/replaced as-is.
Pick the right delimiter: the default delimiter is , but of course that means both your old key AND new value cannot contain the same character ,. In some cases this cannot work, so you will need to get creative and maybe invent some new delimiters to be more explicit and avoid conflicts. In such case, just provide it with the --delimiter or -d flag, for example this could your delimiter if you wanted to --delimiter "===", but if so, your .csv file should look like this:
```
old_string1===new_string1
old_string2=== new_string2
old_string3===new_string3
```
(NOTE: in the above example, all occurences of the string old_string2 will be replaced with the string new_string2 (with a space at the beginning)
How to delete? simply leaving the new value part empty, like this:
```
old_string1===new_string1
old_string2=== new_string2
old_string3===
```
in this example, all occurences of the string old_string3 will be removed from all matching files.
My csv has headers: add the --has-headers flag, and mapsd will skip the first line of your csv.
Change files directly without copy: add the --DANGEROUSLY-REPLACE-INPLACE flag which, as its name suggests... comes with great responsibilities ;)

Installation

git clone git@github.com:Jonarod/mapsd.git
cd mapsd
cargo build --release

# Move it somewhere, like this
sudo mv ./target/release/mapsd /usr/local/bin/

Check everything is fine:

mapsd --version

cargo install mapsd

Roadmap

add a --regex flag to interpret each key in the .csv as a regex instead of a litteral string

mapsd 0.0.1