README.Rmd

---
output: github_document
---

<!-- README.md is generated from README.Rmd. Please edit that file -->

```{r, include = FALSE}
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.path = "man/figures/README-",
  out.width = "100%"
)
```

# rmorphodita

<!-- badges: start -->
[![R-CMD-check](https://github.com/skvrnami/rmorphodita/workflows/R-CMD-check/badge.svg)](https://github.com/skvrnami/rmorphodita/actions)
[![Codecov test coverage](https://codecov.io/gh/skvrnami/rmorphodita/branch/main/graph/badge.svg)](https://codecov.io/gh/skvrnami/rmorphodita?branch=main)
<!-- badges: end -->

The goal of rmorphodita is to enable morphological analysis, tagging and generation using [MorphoDiTa's](https://github.com/ufal/morphodita) Python bindings (contained in 
the [`ufal.morphodita` Python package](https://pypi.org/project/ufal.morphodita/)).

## Installation

And the development version from [GitHub](https://github.com/) with:

``` r
# install.packages("devtools")
devtools::install_github("skvrnami/rmorphodita")
```
## Example

First you need to install morphodita by running `install_morphodita()`.
```{r example}
library(rmorphodita)
```

```{r, eval=FALSE}
install_morphodita()
```

Then you need to download a language model to use for tagging etc.
There are three languages available: Czech (`CZ`), Slovak (`SK`), and English (`EN`).
The `download_models` function downloads a .zip file with models from [LINDAT/CLARIAH-CZ repository](https://lindat.mff.cuni.cz/) to a specified directory, unzips them and returns
list of files with morphological taggers and dictionaries.

```{r download, message=FALSE}
cz_models <- download_models(lang = "CZ", dest_folder = "tmp")
cz_models
```

Then it is necessary to load tagger:
```{r}
cz_tagger <- load_tagger(cz_models[8])
```

```{r}
tagged_text <- morpho_tag(cz_tagger, "Já bych všechny ty počítače zakázala.", NULL)
tagged_text
```

Function `morpho_analyze` returns all possible forms of a word.
```{r}
morpho_analyze(cz_tagger, "kout")
```

And function `morpho_generate` returns all possible forms of a given lemma that
complies with the specified wildcard. In the case below, it returns all nouns in second case.
```{r}
morpho_generate(cz_tagger, "kout", tag_wildcard = "N???2?")
```

As the tags are quite unintelligible, it is possible to extract and recode them like this.
The `extract_hm_tags` function splits the tag into columns indicating particular grammatical categories such as part of speech (`pos`), gender, number, case etc.
The `recode_tags` function then recode the tag marks into factor with a full description of the tag meaning (using the `TAGS` list which stores the meaning of the tag values).

```{r}
tagged_text %>%
    extract_hm_tags() %>%
    recode_tags(., tags_df = TAGS)
```

```{r, echo=FALSE}
unlink("tmp", recursive = TRUE)
```