# sis4

## Setup
### Prerequisites

* GCC   >= 7.1
* Boost >= 1.67
* CMake >= 3.6

### Build Process

tl;dr (with ninja):
```
git submodule update --init --recursive
make develop
```

In detail:

First pull the submodules:
```
git submodule update --init --recursive
```
Then choose your code generator. The fastest way is to build with ninja:
```
make develop
```
For compatibility reasons, there is also the option to \
build with the default (make) code generator:
```
make check
```
Note that both rules are executing quick tests after the building process.

Lastly, you can switch to your preferred compilers by putting them in the environment like:
```
make develop CXX=g++-8 CC=gcc-8
```

### Generate Documentation

```
pip3 install -r docs/sphinx/requirements.txt
make docs
```

## Usage

sis4 provides both a CLI and a Server Feature.

### Command Line Interface

#### Full-Text Search

Output index as file and query the index (suited for large indices):
```
./build/bin/sis-index --output=<index_filename> [doc1 doc2 ...]
./build/bin/sis-query --index=<index_filename> --width=<width> <query>
```
Example:
```
./build/bin/sis-index -o test_index.bin ./etc/Ts-213_OA_NORM_TEXT.txt
./build/bin/sis-query -i test_index.bin -w 5 mensch
```
Or do both in one step:
```
./build/bin/sis-query -q mensch -w 5 ./etc/Ts-213_OA_NORM_TEXT.txt
```

#### Lemma Search

Output lemma-index as file and query the index (suited for large indices):
```
./build/bin/sis-lemmaindex --output=<index_filename> [dict1 dict2 ...]
./build/bin/sis-lemmaindex --index=<index_filename> <query>
```
Example:
```
./build/bin/sis-lemmaindex -o test_lemma_index.bin ./etc/frequencies.lemma
./build/bin/sis-lemmaindex -i test_lemma_index.bin mensch
```
Or do both in one step:
```
./build/bin/sis-lemmaquery -q mensch ./etc/frequencies.lemma
```

### Server Feature

Generate lemma-indices specified by config and start server (default port = 8080)
```
./build/bin/sis-lemmaindex -i ./etc/test_config.json -f ./etc
./build/bin/sis-server -i ./etc/test_config.json
```

Query *without* dictionary specification (default dict = default)
```
http://localhost:<port>/lemmasearch?query=<query>
```
Query *with* dictionary specification
```
http://localhost:<port>/lemmasearch?query=<query>&dict=<dict>
```
\
\
For further usage, make sure to consult
```
./build/bin/sis-<extension> -h
```

## TODO

- [X] Case-insensitive lemma search
- [ ] Case-insensitive full text search
- [ ] Fuzzy search


[Background](https://arxiv.org/pdf/1301.0722.pdf)