---
title: "Introduction to rosmium: R bindings for Osmium Tool"
output: rmarkdown::html_vignette
vignette: >
  %\VignetteIndexEntry{Introduction to rosmium: R bindings for Osmium Tool} 
  %\VignetteEngine{knitr::rmarkdown} 
  %\VignetteEncoding{UTF-8}
---

```{r, include = FALSE}
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  eval = identical(tolower(Sys.getenv("NOT_CRAN")), "true")
)
```

**rosmium** is a package that allows one to use [Osmium
Tool](https://osmcode.org/osmium-tool/) from R. Osmium is a multipurpose command
line tool that enables one to manipulate and analyze OpenStreetMap (OSM) files
through several different commands. Currently, the package does not aim to
cover the entirety of Osmium's API, instead making available functions that wrap
only a very limited set of Osmium's features. This vignette briefly overviews
the package with a few reproducible examples.

# Installation

Before using **rosmium** please make sure that you have it installed in
your computer. You can download either the most stable version from CRAN...

```{r, eval = FALSE}
install.packages("rosmium")
```

...or the development version from GitHub.

```{r, eval = FALSE}
# install.packages("remotes")
remotes::install_github("ipeaGIT/rosmium")
```

Please note that **rosmium** requires Osmium to be installed and added to the
PATH environment variable of your local system. For instructions on how to
install Osmium, please check [its official
website](https://osmcode.org/osmium-tool/#download).

# Usage

The package currently includes only three entrypoints to Osmium's API. To
demonstrate them, we will use some sample data bundled with the package.

```{r, eval = requireNamespace("ggplot2", quietly = TRUE) && identical(tolower(Sys.getenv("NOT_CRAN")), "true"), fig.width = 6, fig.height = 5.5}
library(rosmium)
library(ggplot2)

cur_pbf <- system.file("extdata/cur.osm.pbf", package = "rosmium")

cur_pbf_lines <- sf::st_read(cur_pbf, layer = "lines", quiet = TRUE)

ggplot(cur_pbf_lines) + geom_sf()
```

### `extract()`

`extract()` creates geographical extracts from OSM files. In its most basic
form, the function takes the path to the OSM file whose geographical extent
should be extracted from, the extent, either as a bounding box or as a
(multi)polygon, and the path to the file where the output should be written to.
Additionally, the function can also take the strategy to be used when creating
the extract, which defaults to `"complete_ways"`. Please check the function
documentation for details on the available strategies and their behavior.

```{r, eval = requireNamespace("ggplot2", quietly = TRUE) && identical(tolower(Sys.getenv("NOT_CRAN")), "true"), fig.width = 6, fig.height = 4.5}
# buffering the pbf bounding box 4000 meters inward and using the result
# extent to extract the osm data inside it. transforming the crs because
# inward buffers only work with projected crs

bbox <- sf::st_bbox(cur_pbf_lines)
bbox_polygon <- sf::st_as_sf(sf::st_as_sfc(bbox))
smaller_bbox_poly <- sf::st_buffer(sf::st_transform(bbox_polygon, 5880), -4000)
smaller_bbox_poly <- sf::st_transform(smaller_bbox_poly, 4326)

output_path <- extract(
  cur_pbf,
  smaller_bbox_poly,
  tempfile(fileext = ".osm.pbf"),
  spinner = FALSE
)

extracted_pbf_lines <- sf::st_read(output_path, layer = "lines", quiet = TRUE)

ggplot() +
  geom_sf(data = extracted_pbf_lines) +
  geom_sf(data = smaller_bbox_poly, color = "red", fill = NA)
```

### `tags_filter()`

`tags_filter()` filters OSM files, keeping objects matching at least one of the
specified expressions from the input. In its most basic form, the function takes
the path to the OSM file to which the filters should be applied, the filter
expressions that should be applied and the path to the file where the output
should be written to. Please check the function documentation for a description
of the filter expression format.

By default, not only the objects matching the expressions will be kept in the
output, but also the objects referenced by them. This behavior can be changed
with the `omit_referenced` parameter. The function also includes the
`invert_match` parameter, that inverts the sense of matching (excluding objects
that match the filters), and the `remove_tags` parameter, used to remove tags
from objects that are referenced by objects matching the filters, but which do
not match the filter themselves. Both arguments default to `FALSE`.

```{r}
# get all amenity nodes
output <- tags_filter(cur_pbf, "n/amenity", tempfile(fileext = ".osm.pbf"))
nodes <- sf::st_read(output, layer = "points", quiet = TRUE)
head(nodes$other_tags)

# get all objects (nodes, ways or relations) with an addr:* tag
output <- tags_filter(
  cur_pbf,
  "addr:*",
  tempfile(fileext = ".osm.pbf"),
  omit_referenced = TRUE,
  spinner = FALSE
)
nodes <- sf::st_read(output, layer = "points", quiet = TRUE)
head(nodes$other_tags)
```

### `show_content()`

Finally, `show_content()` displays the content of an OSM file either in `.html`,
`.xml` or `.opl` format. The function takes as input the path to the OSM file
whose content should be shown, the output format in which the content should be
displayed (defaulting to `.html`, the most human readable format, although also
the slowest to open and inspect, depending on the size of the input file) and
the type of objects that should be included in the output (defaulting to all
existing objects in the input). The function returns the path to the temporary
file in which the OSM file content was saved and opens the content in the web
browser or the most appropriate application, depending on the output format.

```{r}
# displays the content of the previous tags_filter() output in html format
show_content(output, spinner = FALSE)
```

```{r, echo = FALSE}
knitr::include_graphics("../man/figures/filtered_file_content.png")
```

## Acknowledgement

**rosmium** is developed by a team at the Institute for Applied Economic
Research (Ipea), Brazil. We would like to thank the authors and contributors of
Osmium for the development of [Osmium
Tool](https://github.com/osmcode/osmium-tool).