Knowledge-clustering

- Implemented 28: The `anchor` command was improved: now the introduced knowledge is displayed.
- Added a ``--version`` command.

**Bug corrections**:
- Fixed 25 (`clustering`): accents are now taken into in the distance computation (bug reported by Antonio Casares)
- Fixed 29 (`clustering`): plurals/suffixes are taken into account even for scope inference
- Fixed 30 (`addquotes`): scope should be handled correctly

**Misc**:
- The code is now *typed*!
- Introduced tests (run `make test` in the development environment).

Fixes 27.

Fixing 23.

Fixing bugs 17, 19, 20 and 21 concerning the `addquotes` command.

- The **addquote** feature was improved:
- Instead of just printing the line number when it finds a match, it prints the content of the line and emphasizes the substring corresponding to the knowledge. (Suggested by Aliaume Lopez)
- Suggests defining synonyms when finding a string in the TeX document that looks like a defined knowledge. (10, suggested by [LeonPyramid](https://github.com/LeonPyramid))

- New feature: **anchor** that finds missing anchor points. (16, suggested by [Nicolas Waldburger](https://github.com/nicolas-waldburger))