1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226
//! Introduction to `odbc-api` (documentation only) //! //! # About ODBC //! //! ODBC is an open standard which allows you to connect to various data sources. Mostly these data //! sources are databases, but ODBC drivers are also available for various file types like Excel or //! CSV. //! //! Your application does not does not link against a driver, but will link against an ODBC driver //! manager which must be installed on the system you intend to run the application. On modern //! Windows Platforms ODBC is always installed, on OS-X or Linux distributions a driver manager like //! [unixODBC](http://www.unixodbc.org/) must be installed by whomever manages the system. //! //! To connect to a data source a driver for the specific data source in question must be installed. //! On windows you can type 'ODBC Data Sources' into the search box to start a little GUI which //! shows you the various drivers and preconfigured data sources on your system. //! //! This however is not a guide on how to configure and setup ODBC. This is a guide on how to use //! the Rust bindings for applications which want to utilize ODBC data sources. //! //! # Quickstart //! //! ```no_run //! //! A program executing a query and printing the result as csv to standard out. Requires //! //! `anyhow` and `csv` crate. //! //! use anyhow::Error; //! use odbc_api::{buffers::TextRowSet, Cursor, Environment}; //! use std::{ //! io::{stdout, Write}, //! path::PathBuf, //! }; //! //! /// Maximum number of rows fetched with one row set. Fetching batches of rows is usually much //! /// faster than fetching individual rows. //! const BATCH_SIZE: u32 = 100000; //! //! fn main() -> Result<(), Error> { //! // Write csv to standard out //! let out = stdout(); //! let mut writer = csv::Writer::from_writer(out); //! //! // We know this is going to be the only ODBC environment in the entire process, so this is //! // safe. //! let environment = unsafe { Environment::new() }?; //! //! // Connect using a DSN. Alternatively we could have used a connection string //! let mut connection = environment.connect( //! "DataSourceName", //! "Username", //! "Password", //! )?; //! //! // Execute a one of query without any parameters. //! match connection.execute("SELECT * FROM TableName", ())? { //! Some(cursor) => { //! // Write the column names to stdout //! let mut headline : Vec<String> = cursor.column_names()?.collect::<Result<_,_>>()?; //! writer.write_record(headline)?; //! //! // Use schema in cursor to initialize a text buffer large enough to hold the largest //! // possible strings for each column. //! let mut buffers = TextRowSet::for_cursor(BATCH_SIZE, &cursor)?; //! // Bind the buffer to the cursor. It is now being filled with every call to fetch. //! let mut row_set_cursor = cursor.bind_buffer(&mut buffers)?; //! //! // Iterate over batches //! while let Some(batch) = row_set_cursor.fetch()? { //! // Within a batch, iterate over every row //! for row_index in 0..batch.num_rows() { //! // Within a row iterate over every column //! let record = (0..batch.num_cols()) //! .map(|col_index| batch.at(col_index, row_index).unwrap_or(&[])); //! // Writes row as csv //! writer.write_record(record)?; //! } //! } //! } //! None => { //! eprintln!( //! "Query came back empty. No output has been created." //! ); //! } //! } //! //! Ok(()) //! } //! ``` //! //! # 32 Bit and 64 Bit considerations. //! //! To consider wether you want to work with 32 Bit or 64 Bit data sources is especially important //! for windows users, as driver managers (and possibly drivers) may both exist at the same time //! in the same system. //! //! In any case, depending on the platform part of your target tripple either 32 Bit or 64 Bit //! drivers are going to work, but not both. On a private windows machine (even on a modern 64 Bit //! Windows) it is not unusual to find lots of 32 Bit drivers installed on the system, but none for //! 64 Bits. So for windows users it is worth thinking about not using the default toolchain which //! is likely 64 Bits and to switch to a 32 Bit one. On other platforms you are usually fine //! sticking with 64 Bits, as there are not usually any drivers preinstalled anyway, 64 Bit or //! otherwise. //! //! No code changes are required, so you can also just build both if you want to. //! //! # Connect to a data source //! //! ## Setting up the ODBC Environment //! //! To connect with a data source we need a connection. To create a connection we need an ODBC //! environment. //! //! ```no_run //! use odbc_api::Environment; //! //! // I herby solemnly swear that this is the only ODBC environment in the entire process, thus //! // making this call safe. //! unsafe { //! let env = Environment::new()?; //! } //! # Ok::<(), odbc_api::Error>(()) //! ``` //! //! Oh dear! Aren't we of to a bad start. First step in using this API and already a piece of unsafe //! code. Well we are talking with a C API those contract explicitly demands that there MUST be at //! most one ODBC Environment in the entire process. This requirement can only be verified in //! application code. If you write a library you MUST NOT wrap the creation of an ODBC environment //! in a safe function call. If another libray would do the same and an application were to use //! both of these, it might create two environments in safe code and thus causing undefined //! behaviour, which is clearly a violation of Rusts safety guarantees. On the other hand in //! application code it is pretty easy to get right. You call it, and you call it only once. //! //! Apart from that. This is it. Our ODBC Environment is ready for action. //! //! These bindings currently support two ways of creating a connections: //! //! ## Connect using a connection string //! //! Connection strings do not require that the data source is preconfigured by the driver manager //! this makes them very flexible. //! //! ```no_run //! use odbc_api::Environment; //! //! // I herby solemnly swear that this is the only ODBC environment in the entire process, thus //! // making this call safe. //! let env = unsafe { //! Environment::new()? //! }; //! //! let connection_string = " //! Driver={ODBC Driver 17 for SQL Server};\ //! Server=localhost;\ //! UID=SA;\ //! PWD=<YourStrong@Passw0rd>;\ //! "; //! //! let mut conn = env.connect_with_connection_string(connection_string)?; //! # Ok::<(), odbc_api::Error>(()) //! ``` //! //! There is a syntax to these connection strings, but few people go through the trouble to learn //! it. Most common strategy is to google one that works for with your data source. The connection //! borrows the environment, so you will get a compiler error, if your environment goes out of scope //! before the connection does. //! //! ## Connect using a Data Source Name (DSN) //! //! Should a data source be known by the driver manager we can access it using its name and //! credentials. This is more convinient for the user or application developer, but requires a //! configuration of the ODBC driver manager. Think of it as shifting work from users to //! administrators. //! //! ```no_run //! use odbc_api::Environment; //! //! // I herby solemnly swear that this is the only ODBC environment in the entire process, thus //! // making this call safe. //! let env = unsafe { //! Environment::new()? //! }; //! //! let mut conn = env.connect("YourDatabase", "SA", "<YourStrong@Passw0rd>")?; //! # Ok::<(), odbc_api::Error>(()) //! ``` //! //! How to configure such data sources is not the scope of this guide, and depends on the driver //! manager in question. //! //! ## Lifetime considerations for Connections //! //! An ODBC connection MUST NOT outlive the ODBC environment. This is modeled as the connection //! borrowing the environment. It is a shared borrow, to allow for more than one connection per //! environment. This way the compiler will catch programming errors early. The most popular among //! them seems to be returning a `Connection` from a function which also creates the environment. //! //! # Executing a statement //! //! With our ODBC connection all set up and ready to go, we can execute an SQL query: //! //! ```no_run //! use odbc_api::Environment; //! //! let env = unsafe { //! Environment::new()? //! }; //! //! let mut conn = env.connect("YourDatabase", "SA", "<YourStrong@Passw0rd>")?; //! if let Some(cursor) = conn.execute("SELECT year, name FROM Birthdays;", ())? { //! // Use cursor to process query results. //! } //! # Ok::<(), odbc_api::Error>(()) //! ``` //! //! The first parameter of `execute` is the SQL statement text. The second parameter is used to pass //! arguments of the SQL Statements itself (more on that later). Ours has none, so we use `()` to //! not bind any arguments to the statement. You can learn all about passing parameters from the //! [`parameter module level documentation`](`crate::parameter`). It may feature an example for //! your usecase. //! //! Note that the result of the operation is an `Option`. This reflects that not every statement //! returns a [`Cursor`](crate::Cursor). `INSERT` statements usually do not, but even `SELECT` //! queries which would return zero rows can depending on the driver return either an empty cursor //! or no cursor at all. Should a cursor exists, it must be consumed or closed. The `drop` handler //! of Cursor will close it for us. If the `Option` is `None` there is nothing to close, so is all //! taken care of, nice.