---
title: "Querying"
output: rmarkdown::html_vignette
vignette: >
  %\VignetteIndexEntry{querying}
  %\VignetteEngine{knitr::rmarkdown}
  %\VignetteEncoding{UTF-8}
---

```{r, include = FALSE}
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)
```

```{r setup}
library(ODataQuery)
service <- ODataQuery$new("https://services.odata.org/V4/TripPinServiceRW")
people_entity <- service$path("People")
```

There are three ways to query data using this package.

### Directly write OData query

The most direct way is to directly write your query in OData.

- Fast
- Can express any query
- You need to sanitize your inputs
- Can be hard to read if mixing with user code

```{R}
people_entity$filter("Concurrency gt 500")
people_entity$filter("Friends/any(f: f/FirstName eq 'John')")
```


### Using `and_query`, `or_query` and `not_query`

- Automatically sanitized
- Easy to read
- Not every query can be written this way

```{R}
people_entity$filter(Concurrency.gt = 500)
people_entity$filter(or_query(Concurrency.lt = 500, Concurrency.gt = 1500))
```

### Using `to_odata`

- Most readable
- Can express any query
- Automatically sanitized, use !! to unquote
- Slower to process than the other options

```{R}
people_entity$filter(to_odata(Concurrency > 500))
name <- 'John'
people_entity$filter(to_odata(Friends$any(f ~ f$FirstName == !!name)))
```