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
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
// Copyright 2015-2016 Brian Smith.
//
// Permission to use, copy, modify, and/or distribute this software for any
// purpose with or without fee is hereby granted, provided that the above
// copyright notice and this permission notice appear in all copies.
//
// THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHORS DISCLAIM ALL WARRANTIES
// WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
// MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHORS BE LIABLE FOR ANY
// SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
// WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN ACTION
// OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF OR IN
// CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.

//! Testing framework.
//!
//! Unlike the rest of *ring*, this testing framework uses panics pretty
//! liberally. It was originally designed for internal use--it drives most of
//! *ring*'s internal tests, and so it is optimized for getting *ring*'s tests
//! written quickly at the expense of some usability. The documentation is
//! lacking. The best way to learn it is to look at some examples. The digest
//! tests are the most complicated because they use named sections. Other tests
//! avoid named sections and so are easier to understand.
//!
//! # Examples
//!
//! ## Writing Tests
//!
//! Input files look like this:
//!
//! ```text
//! # This is a comment.
//!
//! HMAC = SHA1
//! Input = "My test data"
//! Key = ""
//! Output = 61afdecb95429ef494d61fdee15990cabf0826fc
//!
//! HMAC = SHA256
//! Input = "Sample message for keylen<blocklen"
//! Key = 000102030405060708090A0B0C0D0E0F101112131415161718191A1B1C1D1E1F
//! Output = A28CF43130EE696A98F14A37678B56BCFCBDD9E5CF69717FECF5480F0EBDF790
//! ```
//!
//! Test cases are separated with blank lines. Note how the bytes of the `Key`
//! attribute are specified as a quoted string in the first test case and as
//! hex in the second test case; you can use whichever form is more convenient
//! and you can mix and match within the same file. The empty sequence of bytes
//! can only be represented with the quoted string form (`""`).
//!
//! Here's how you would consume the test data:
//!
//! ```ignore
//! use ring::test;
//!
//! test::from_file("src/hmac_tests.txt", |section, test_case| {
//!     assert_eq!(section, ""); // This test doesn't use named sections.
//!
//!     let digest_alg = test_case.consume_digest_alg("HMAC");
//!     let input = test_case.consume_bytes("Input");
//!     let key = test_case.consume_bytes("Key");
//!     let output = test_case.consume_bytes("Output");
//!
//!     // Do the actual testing here
//! });
//! ```
//!
//! Note that `consume_digest_alg` automatically maps the string "SHA1" to a
//! reference to `digest::SHA1`, "SHA256" to `digest::SHA256`, etc.
//!
//! ## Output When a Test Fails
//!
//! When a test case fails, the framework automatically prints out the test
//! case. If the test case failed with a panic, then the backtrace of the panic
//! will be printed too. For example, let's say the failing test case looks
//! like this:
//!
//! ```text
//! Curve = P-256
//! a = 2b11cb945c8cf152ffa4c9c2b1c965b019b35d0b7626919ef0ae6cb9d232f8af
//! b = 18905f76a53755c679fb732b7762251075ba95fc5fedb60179e730d418a9143c
//! r = 18905f76a53755c679fb732b7762251075ba95fc5fedb60179e730d418a9143c
//! ```
//! If the test fails, this will be printed (if `$RUST_BACKTRACE` is `1`):
//!
//! ```text
//! src/example_tests.txt: Test panicked.
//! Curve = P-256
//! a = 2b11cb945c8cf152ffa4c9c2b1c965b019b35d0b7626919ef0ae6cb9d232f8af
//! b = 18905f76a53755c679fb732b7762251075ba95fc5fedb60179e730d418a9143c
//! r = 18905f76a53755c679fb732b7762251075ba95fc5fedb60179e730d418a9143c
//! thread 'example_test' panicked at 'Test failed.', src\test.rs:206
//! stack backtrace:
//!    0:     0x7ff654a05c7c - std::rt::lang_start::h61f4934e780b4dfc
//!    1:     0x7ff654a04f32 - std::rt::lang_start::h61f4934e780b4dfc
//!    2:     0x7ff6549f505d - std::panicking::rust_panic_with_hook::hfe203e3083c2b544
//!    3:     0x7ff654a0825b - rust_begin_unwind
//!    4:     0x7ff6549f63af - std::panicking::begin_panic_fmt::h484cd47786497f03
//!    5:     0x7ff654a07e9b - rust_begin_unwind
//!    6:     0x7ff654a0ae95 - core::panicking::panic_fmt::h257ceb0aa351d801
//!    7:     0x7ff654a0b190 - core::panicking::panic::h4bb1497076d04ab9
//!    8:     0x7ff65496dc41 - from_file<closure>
//!                         at C:\Users\Example\example\<core macros>:4
//!    9:     0x7ff65496d49c - example_test
//!                         at C:\Users\Example\example\src\example.rs:652
//!   10:     0x7ff6549d192a - test::stats::Summary::new::ha139494ed2e4e01f
//!   11:     0x7ff6549d51a2 - test::stats::Summary::new::ha139494ed2e4e01f
//!   12:     0x7ff654a0a911 - _rust_maybe_catch_panic
//!   13:     0x7ff6549d56dd - test::stats::Summary::new::ha139494ed2e4e01f
//!   14:     0x7ff654a03783 - std::sys::thread::Thread::new::h2b08da6cd2517f79
//!   15:     0x7ff968518101 - BaseThreadInitThunk
//! ```
//!
//! Notice that the output shows the name of the data file
//! (`src/example_tests.txt`), the test inputs that led to the failure, and the
//! stack trace to the line in the test code that panicked: entry 9 in the
//! stack trace pointing to line 652 of the file `example.rs`.

use {digest, error};
use std;
use std::string::String;
use std::vec::Vec;
use std::io::BufRead;

/// A test case. A test case consists of a set of named attributes. Every
/// attribute in the test case must be consumed exactly once; this helps catch
/// typos and omissions.
pub struct TestCase {
    attributes: Vec<(String, String, bool)>,
}

impl TestCase {
    /// Maps the strings "SHA1", "SHA256", "SHA384", and "SHA512" to digest
    /// algorithms, maps "SHA224" to `None`, and panics on other (erroneous)
    /// inputs. "SHA224" is mapped to None because *ring* intentionally does
    /// not support SHA224, but we need to consume test vectors from NIST that
    /// have SHA224 vectors in them.
    pub fn consume_digest_alg(&mut self, key: &str)
                              -> Option<&'static digest::Algorithm> {
        let name = self.consume_string(key);
        match name.as_ref() {
            "SHA1" => Some(&digest::SHA1),
            "SHA224" => None, // We actively skip SHA-224 support.
            "SHA256" => Some(&digest::SHA256),
            "SHA384" => Some(&digest::SHA384),
            "SHA512" => Some(&digest::SHA512),
            _ => panic!("Unsupported digest algorithm: {}", name)
        }
    }

    /// Returns the value of an attribute that is encoded as a sequence of an
    /// even number of hex digits, or as a double-quoted UTF-8 string. The
    /// empty (zero-length) value is represented as "".
    pub fn consume_bytes(&mut self, key: &str) -> Vec<u8> {
        let mut s = self.consume_string(key);
        if s.starts_with("\"") {
            // The value is a quoted strong.
            // XXX: We don't deal with any inner quotes.
            if !s.ends_with("\"") {
                panic!("expected quoted string, found {}", s);
            }
            let _ = s.pop();
            let _ = s.remove(0);
            Vec::from(s.as_bytes())
        } else {
            // The value is hex encoded.
            match from_hex(&s) {
                Ok(s) => s,
                Err(ref err_str) => {
                    panic!("{} in {}", err_str, s);
                }
            }
        }
    }

    /// Returns the value of an attribute that is an integer, in decimal
    /// notation.
    pub fn consume_usize(&mut self, key: &str) -> usize {
        let s = self.consume_string(key);
        s.parse::<usize>().unwrap()
    }

    /// Returns the raw value of an attribute, without any unquoting or
    /// other interpretation.
    pub fn consume_string(&mut self, key: &str) -> String {
        self.consume_optional_string(key)
            .unwrap_or_else(|| panic!("No attribute named \"{}\"", key))
    }

    /// Like `consume_string()` except it returns `None` if the test case
    /// doesn't have the attribute.
    pub fn consume_optional_string(&mut self, key: &str) -> Option<String> {
        for &mut (ref name, ref value, ref mut consumed) in &mut self.attributes {
            if key == name {
                if *consumed {
                    panic!("Attribute {} was already consumed", key);
                }
                *consumed = true;
                return Some(value.clone());
            }
        }
        None
    }
}


/// Reads test cases out of the file with the path given by
/// `test_data_relative_file_path`, calling `f` on each vector until `f` fails
/// or until all the test vectors have been read. `f` can indicate failure
/// either by returning `Err()` or by panicking.
pub fn from_file<F>(test_data_relative_file_path: &str, mut f: F)
                    where F: FnMut(&str, &mut TestCase)
                    -> Result<(), error::Unspecified> {
    let path = std::path::PathBuf::from(test_data_relative_file_path);
    let file = std::fs::File::open(path).unwrap();
    let mut lines = std::io::BufReader::new(&file).lines();

    let mut current_section = String::from("");
    let mut failed = false;

    loop {
        let mut test_case =
                match parse_test_case(&mut current_section, &mut lines) {
            Some(test_case) => test_case,
            None => { break; },
        };

        let result = std::panic::catch_unwind(std::panic::AssertUnwindSafe(|| {
            f(&current_section, &mut test_case)
        }));
        let result = match result {
            Ok(Ok(())) => {
                if !test_case.attributes.iter().any(
                        |&(_, _, ref consumed)| !consumed) {
                    Ok(())
                } else {
                    failed = true;
                    Err("Test didn't consume all attributes.")
                }
            },
            Ok(Err(_)) => Err("Test returned Err(error::Unspecified)."),
            Err(_) => Err("Test panicked."),
        };

        if let Err(msg) = result {
            failed = true;

            println!("{}: {}", test_data_relative_file_path, msg);
            for (ref name, ref value, ref consumed) in test_case.attributes {
                let consumed_str = if *consumed { "" } else { " (unconsumed)" };
                println!("{}{} = {}", name, consumed_str, value);
            }
        };
    }

    if failed {
        panic!("Test failed.")
    }
}

/// Decode an string of hex digits into a sequence of bytes. The input must
/// have an even number of digits.
pub fn from_hex(hex_str: &str) -> Result<Vec<u8>, String> {
    if hex_str.len() % 2 != 0 {
        return Err(
            String::from("Hex string does not have an even number of digits"));
    }

    fn from_hex_digit(d: u8) -> Result<u8, String> {
        if d >= b'0' && d <= b'9' {
            Ok(d - b'0')
        } else if d >= b'a' && d <= b'f' {
            Ok(d - b'a' + 10u8)
        } else if d >= b'A' && d <= b'F' {
            Ok(d - b'A' + 10u8)
        } else {
            Err(format!("Invalid hex digit '{}'", d as char))
        }
    }

    let mut result = Vec::with_capacity(hex_str.len() / 2);
    for digits in hex_str.as_bytes().chunks(2) {
        let hi = try!(from_hex_digit(digits[0]));
        let lo = try!(from_hex_digit(digits[1]));
        result.push((hi * 0x10) | lo);
    }
    Ok(result)
}

type FileLines<'a> = std::io::Lines<std::io::BufReader<&'a std::fs::File>>;

fn parse_test_case(current_section: &mut String,
                   lines: &mut FileLines) -> Option<TestCase> {
    let mut attributes = Vec::new();

    let mut is_first_line = true;
    loop {
        let line = match lines.next() {
            None => None,
            Some(result) => Some(result.unwrap()),
        };

        if cfg!(feature = "test_logging") {
            if let Some(ref text) = line {
                println!("Line: {}", text);
            }
        }

        match line {
            // If we get to EOF when we're not in the middle of a test case,
            // then we're done.
            None if is_first_line => {
                return None;
            },

            // End of the file on a non-empty test cases ends the test case.
            None => {
                return Some(TestCase {
                    attributes: attributes,
                });
            },

            // A blank line ends a test case if the test case isn't empty.
            Some(ref line) if line.len() == 0 => {
                if !is_first_line {
                    return Some(TestCase {
                        attributes: attributes,
                    });
                }
                // Ignore leading blank lines.
            },

            // Comments start with '#'; ignore them.
            Some(ref line) if line.starts_with("#") => { },

            Some(ref line) if line.starts_with("[") => {
                assert!(is_first_line);
                assert!(line.ends_with("]"));
                current_section.truncate(0);
                current_section.push_str(line);
                let _ = current_section.pop();
                let _ = current_section.remove(0);
            },

            Some(ref line) => {
                is_first_line = false;

                let parts: Vec<&str> = line.splitn(2, " = ").collect();
                if parts.len() != 2 {
                    panic!("Syntax error: Expected Key = Value.");
                };

                let key = parts[0].trim();
                let value = parts[1].trim();

                // Don't allow the value to be ommitted. An empty value can be
                // represented as an empty quoted string.
                assert!(value.len() != 0);

                // Checking is_none() ensures we don't accept duplicate keys.
                attributes.push((String::from(key), String::from(value), false));
            }
        }
    }
}


#[cfg(test)]
mod tests {
    use {error, test};

    #[test]
    fn one_ok() {
        test::from_file("src/test_1_tests.txt", |_, test_case| {
            let _ = test_case.consume_string("Key");
            Ok(())
        });
    }

    #[test]
    #[should_panic(expected = "Test failed.")]
    fn one_err() {
        test::from_file("src/test_1_tests.txt", |_, test_case| {
            let _ = test_case.consume_string("Key");
            Err(error::Unspecified)
        });
    }

    #[test]
    #[should_panic(expected = "Test failed.")]
    fn one_panics() {
        test::from_file("src/test_1_tests.txt", |_, test_case| {
            let _ = test_case.consume_string("Key");
            panic!("");
        });
    }

    #[test] #[should_panic(expected = "Test failed.")] fn first_err()  { err_one(0) }
    #[test] #[should_panic(expected = "Test failed.")] fn middle_err() { err_one(1) }
    #[test] #[should_panic(expected = "Test failed.")] fn last_err()   { err_one(2) }

    fn err_one(test_to_fail: usize) {
        let mut n = 0;
        test::from_file("src/test_3_tests.txt", |_, test_case| {
            let _ = test_case.consume_string("Key");
            let result = if n != test_to_fail {
                Ok(())
            } else {
                Err(error::Unspecified)
            };
            n += 1;
            result
        });
    }

    #[test] #[should_panic(expected = "Test failed.")] fn first_panic()  { panic_one(0) }
    #[test] #[should_panic(expected = "Test failed.")] fn middle_panic() { panic_one(1) }
    #[test] #[should_panic(expected = "Test failed.")] fn last_panic()   { panic_one(2) }

    fn panic_one(test_to_fail: usize) {
        let mut n = 0;
        test::from_file("src/test_3_tests.txt", |_, test_case| {
            let _ = test_case.consume_string("Key");
            if n == test_to_fail {
                panic!("Oh Noes!");
            };
            n += 1;
            Ok(())
        });
    }

    #[test]
    #[should_panic(expected = "Syntax error: Expected Key = Value.")]
    fn syntax_error() {
        test::from_file("src/test_1_syntax_error_tests.txt", |_, _| {
            Ok(())
        });
    }

    #[test]
    #[should_panic]
    fn file_not_found() {
        test::from_file("src/test_file_not_found_tests.txt", |_, _| {
            Ok(())
        });
    }
}