odbc_api/parameter/
blob.rs

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
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
use odbc_sys::{len_data_at_exec, CDataType, DATA_AT_EXEC};

use crate::{
    handles::{DelayedInput, HasDataType, Statement},
    DataType, Error, ParameterCollection, ParameterTupleElement,
};
use std::{
    ffi::c_void,
    fs::File,
    io::{self, BufRead, BufReader},
    num::NonZeroUsize,
    path::Path,
};

/// A `Blob` can stream its contents to the database batch by batch and may therefore be used to
/// transfer large amounts of data, exceeding the drivers capabilities for normal input parameters.
///
/// # Safety
///
/// If a hint is implemented for `blob_size` it must be accurate before the first call to
/// `next_batch`.
pub unsafe trait Blob: HasDataType {
    /// CData type of the binary data returned in the batches. Likely to be either
    /// [`crate::sys::CDataType::Binary`], [`crate::sys::CDataType::Char`] or
    /// [`crate::sys::CDataType::WChar`].
    fn c_data_type(&self) -> CDataType;

    /// Hint passed on to the driver regarding the combined size of all the batches. This hint is
    /// passed then the parameter is bound to the statement, so its meaning is only defined before
    /// the first call to `next_batch`. If `None` no hint about the total length of the batches is
    /// passed to the driver and the indicator will be set to [`crate::sys::DATA_AT_EXEC`].
    fn size_hint(&self) -> Option<usize>;

    /// Retrieve the next batch of data from the source. Batches may not be empty. `None` indicates
    /// the last batch has been reached.
    fn next_batch(&mut self) -> io::Result<Option<&[u8]>>;

    /// Convinience function. Same as calling [`self::BlobParam::new`].
    fn as_blob_param(&mut self) -> BlobParam
    where
        Self: Sized,
    {
        BlobParam::new(self)
    }
}

/// Parameter type which can be used to bind a [`self::Blob`] as parameter to a statement in order
/// for its contents to be streamed to the database at statement execution time.
pub struct BlobParam<'a> {
    /// Should be [`crate::sys::DATA_AT_EXEC`] if no size hint is given, or the result of
    /// [`crate::sys::len_data_at_exec`].
    indicator: isize,
    /// Trait object to be bound as a delayed parameter.
    blob: &'a mut dyn Blob,
}

impl<'a> BlobParam<'a> {
    pub fn new(blob: &'a mut impl Blob) -> Self {
        let indicator = if let Some(size) = blob.size_hint() {
            len_data_at_exec(size.try_into().unwrap())
        } else {
            DATA_AT_EXEC
        };
        Self { indicator, blob }
    }
}

unsafe impl DelayedInput for BlobParam<'_> {
    fn cdata_type(&self) -> CDataType {
        self.blob.c_data_type()
    }

    fn indicator_ptr(&self) -> *const isize {
        &self.indicator as *const isize
    }

    fn stream_ptr(&mut self) -> *mut c_void {
        // Types must have the same size for the transmute to work in the reverse cast.
        debug_assert_eq!(
            std::mem::size_of::<*mut &mut dyn Blob>(),
            std::mem::size_of::<*mut c_void>()
        );
        &mut self.blob as *mut &mut dyn Blob as *mut c_void
    }
}

impl HasDataType for BlobParam<'_> {
    fn data_type(&self) -> DataType {
        self.blob.data_type()
    }
}

unsafe impl ParameterCollection for BlobParam<'_> {
    fn parameter_set_size(&self) -> usize {
        1
    }

    unsafe fn bind_parameters_to(&mut self, stmt: &mut impl Statement) -> Result<(), Error> {
        stmt.bind_delayed_input_parameter(1, self).into_result(stmt)
    }
}

unsafe impl ParameterTupleElement for &mut BlobParam<'_> {
    unsafe fn bind_to(
        &mut self,
        parameter_number: u16,
        stmt: &mut impl Statement,
    ) -> Result<(), Error> {
        stmt.bind_delayed_input_parameter(parameter_number, *self)
            .into_result(stmt)
    }
}

/// Wraps borrowed bytes with a batch_size and implements [`self::Blob`]. Use this type to send long
/// array of bytes to the database.
pub struct BlobSlice<'a> {
    /// If `true` the blob is going to be bound as [`DataType::LongVarbinary`] and the bytes are
    /// interpreted as [`CDataType::Binary`]. If false the blob is going to be bound as
    /// [`DataType::LongVarchar`] and the bytes are interpreted as [`CDataType::Char`].
    pub is_binary: bool,
    /// Maximum number of bytes transferred to the database in one go. May be largere than the
    /// remaining blob size.
    pub batch_size: usize,
    /// Remaining bytes to transfer to the database.
    pub blob: &'a [u8],
}

impl<'a> BlobSlice<'a> {
    /// Construct a Blob from a byte slice. The blob is going to be bound as a `LongVarbinary` and
    /// will be transmitted in one batch.
    ///
    /// # Example
    ///
    /// ```
    /// use odbc_api::{Connection, parameter::{Blob, BlobSlice}, IntoParameter, Error};
    ///
    /// fn insert_image(
    ///     conn: &Connection<'_>,
    ///     id: &str,
    ///     image_data: &[u8]
    /// ) -> Result<(), Error>
    /// {
    ///     let mut blob = BlobSlice::from_byte_slice(image_data);
    ///
    ///     let insert = "INSERT INTO Images (id, image_data) VALUES (?,?)";
    ///     let parameters = (&id.into_parameter(), &mut blob.as_blob_param());
    ///     conn.execute(&insert, parameters)?;
    ///     Ok(())
    /// }
    /// ```
    pub fn from_byte_slice(blob: &'a [u8]) -> Self {
        Self {
            is_binary: true,
            batch_size: blob.len(),
            blob,
        }
    }

    /// Construct a Blob from a text slice. The blob is going to be bound as a `LongVarchar` and
    /// will be transmitted in one batch.
    ///
    /// # Example
    ///
    /// This example insert `title` as a normal input parameter but streams the potentially much
    /// longer `String` in `text` to the database as a large text blob. This allows to circumvent
    /// the size restrictions for `String` arguments of many drivers (usually around 4 or 8 KiB).
    ///
    /// ```
    /// use odbc_api::{Connection, parameter::{Blob, BlobSlice}, IntoParameter, Error};
    ///
    /// fn insert_book(
    ///     conn: &Connection<'_>,
    ///     title: &str,
    ///     text: &str
    /// ) -> Result<(), Error>
    /// {
    ///     let mut blob = BlobSlice::from_text(&text);
    ///
    ///     let insert = "INSERT INTO Books (title, text) VALUES (?,?)";
    ///     let parameters = (&title.into_parameter(), &mut blob.as_blob_param());
    ///     conn.execute(&insert, parameters)?;
    ///     Ok(())
    /// }
    /// ```
    pub fn from_text(text: &'a str) -> Self {
        Self {
            is_binary: false,
            batch_size: text.len(),
            blob: text.as_bytes(),
        }
    }
}

impl HasDataType for BlobSlice<'_> {
    fn data_type(&self) -> DataType {
        if self.is_binary {
            DataType::LongVarbinary {
                length: NonZeroUsize::new(self.blob.len()),
            }
        } else {
            DataType::LongVarchar {
                length: NonZeroUsize::new(self.blob.len()),
            }
        }
    }
}

unsafe impl Blob for BlobSlice<'_> {
    fn c_data_type(&self) -> CDataType {
        if self.is_binary {
            CDataType::Binary
        } else {
            CDataType::Char
        }
    }

    fn size_hint(&self) -> Option<usize> {
        Some(self.blob.len())
    }

    fn next_batch(&mut self) -> io::Result<Option<&[u8]>> {
        if self.blob.is_empty() {
            return Ok(None);
        }

        if self.blob.len() >= self.batch_size {
            let (head, tail) = self.blob.split_at(self.batch_size);
            self.blob = tail;
            Ok(Some(head))
        } else {
            let last_batch = self.blob;
            self.blob = &[];
            Ok(Some(last_batch))
        }
    }
}

/// Wraps an [`std::io::BufRead`] and implements [`self::Blob`]. Use this to stream contents from an
/// [`std::io::BufRead`] to the database. The blob implementation is going to directly utilize the
/// Buffer of the [`std::io::BufRead`] implementation, so the batch size is likely equal to that
/// capacity.
pub struct BlobRead<R> {
    /// `true` if `size` is to interpreted as the exact ammount of bytes contained in the reader, at
    /// the time of binding it as a parameter. `false` if `size` is to be interpreted as an upper
    /// bound.
    exact: bool,
    size: usize,
    consume: usize,
    buf_read: R,
}

impl<R> BlobRead<R> {
    /// Construct a blob read from any [`std::io::BufRead`]. The `upper bound` is used in the type
    /// description then binding the blob as a parameter.
    ///
    /// # Examples
    ///
    /// This is more flexible than [`Self::from_path`]. Note however that files provide metadata
    /// about the length of the data, which `io::BufRead` does not. This is not an issue for most
    /// drivers, but some can perform optimization if they know the size in advance. In the tests
    /// SQLite has shown a bug to only insert empty data if no size hint has been provided.
    ///
    /// ```
    /// use std::io::BufRead;
    /// use odbc_api::{Connection, parameter::{Blob, BlobRead}, IntoParameter, Error};
    ///
    /// fn insert_image_to_db(
    ///     conn: &Connection<'_>,
    ///     id: &str,
    ///     image_data: impl BufRead) -> Result<(), Error>
    /// {
    ///     const MAX_IMAGE_SIZE: usize = 4 * 1024 * 1024;
    ///     let mut blob = BlobRead::with_upper_bound(image_data, MAX_IMAGE_SIZE);
    ///
    ///     let sql = "INSERT INTO Images (id, image_data) VALUES (?, ?)";
    ///     let parameters = (&id.into_parameter(), &mut blob.as_blob_param());
    ///     conn.execute(sql, parameters)?;
    ///     Ok(())
    /// }
    /// ```
    pub fn with_upper_bound(buf_read: R, upper_bound: usize) -> Self {
        Self {
            exact: false,
            consume: 0,
            size: upper_bound,
            buf_read,
        }
    }

    /// Construct a blob read from any [`std::io::BufRead`]. The `upper bound` is used in the type
    /// description then binding the blob as a parameter and is also passed to indicate the size
    /// of the actual value to the ODBC driver.
    ///
    /// # Safety
    ///
    /// The ODBC driver may use the exact size hint to allocate buffers internally. Too short may
    /// lead to invalid writes and too long may lead to invalid reads, so to be save the hint must
    /// be exact.
    pub unsafe fn with_exact_size(buf_read: R, exact_size: usize) -> Self {
        Self {
            exact: true,
            consume: 0,
            size: exact_size,
            buf_read,
        }
    }
}

impl BlobRead<BufReader<File>> {
    /// Construct a blob from a Path. The metadata of the file is used to give the ODBC driver a
    /// size hint.
    ///
    /// # Example
    ///
    /// [`BlobRead::from_path`] is the most convenient way to turn a file path into a [`Blob`]
    /// parameter. The following example also demonstrates that the streamed blob parameter can be
    /// combined with reqular input parmeters like `id`.
    ///
    /// ```
    /// use std::{error::Error, path::Path};
    /// use odbc_api::{Connection, parameter::{Blob, BlobRead}, IntoParameter};
    ///
    /// fn insert_image_to_db(
    ///     conn: &Connection<'_>,
    ///     id: &str,
    ///     image_path: &Path) -> Result<(), Box<dyn Error>>
    /// {
    ///     let mut blob = BlobRead::from_path(&image_path)?;
    ///
    ///     let sql = "INSERT INTO Images (id, image_data) VALUES (?, ?)";
    ///     let parameters = (&id.into_parameter(), &mut blob.as_blob_param());
    ///     conn.execute(sql, parameters)?;
    ///     Ok(())
    /// }
    /// ```
    pub fn from_path(path: &Path) -> io::Result<Self> {
        let file = File::open(path)?;
        let size = file.metadata()?.len().try_into().unwrap();
        let buf_read = BufReader::new(file);
        Ok(Self {
            consume: 0,
            exact: true,
            size,
            buf_read,
        })
    }
}

impl<R> HasDataType for BlobRead<R>
where
    R: BufRead,
{
    fn data_type(&self) -> DataType {
        DataType::LongVarbinary {
            length: NonZeroUsize::new(self.size),
        }
    }
}

unsafe impl<R> Blob for BlobRead<R>
where
    R: BufRead,
{
    fn c_data_type(&self) -> CDataType {
        CDataType::Binary
    }

    fn size_hint(&self) -> Option<usize> {
        if self.exact {
            Some(self.size)
        } else {
            None
        }
    }

    fn next_batch(&mut self) -> io::Result<Option<&[u8]>> {
        if self.consume != 0 {
            self.buf_read.consume(self.consume);
        }
        let batch = self.buf_read.fill_buf()?;
        self.consume = batch.len();
        if batch.is_empty() {
            Ok(None)
        } else {
            Ok(Some(batch))
        }
    }
}