Files
addr2line
ahash
aho_corasick
arrayref
arrayvec
artemis_asset
artemis_core
artemis_erc20_app
artemis_eth_app
artemis_ethereum
backtrace
base58
bip39
bitmask
bitvec
blake2_rfc
block_buffer
block_padding
byte_slice_cast
byte_tools
byteorder
cfg_if
clear_on_drop
const_random
const_random_macro
constant_time_eq
crunchy
crypto_mac
curve25519_dalek
derive_more
digest
ed25519_dalek
either
environmental
ethabi_decode
ethbloom
ethereum_types
failure
failure_derive
fake_simd
fixed_hash
frame_metadata
frame_support
frame_support_procedural
frame_support_procedural_tools
frame_support_procedural_tools_derive
frame_system
futures
futures_channel
futures_core
futures_executor
futures_io
futures_macro
futures_sink
futures_task
futures_util
async_await
future
io
lock
sink
stream
task
generic_array
getrandom
gimli
hash256_std_hasher
hash_db
hashbrown
hex
hex_literal
hmac
hmac_drbg
impl_codec
impl_rlp
impl_serde
impl_trait_for_tuples
inflector
cases
camelcase
case
classcase
kebabcase
pascalcase
screamingsnakecase
sentencecase
snakecase
tablecase
titlecase
traincase
numbers
deordinalize
ordinalize
string
constants
deconstantize
demodulize
pluralize
singularize
suffix
foreignkey
integer_sqrt
itertools
keccak
lazy_static
libc
lock_api
log
memchr
memory_db
memory_units
merlin
nodrop
num_bigint
num_cpus
num_integer
num_rational
num_traits
object
once_cell
opaque_debug
pallet_bridge
pallet_verifier
parity_scale_codec
parity_scale_codec_derive
parity_util_mem
parity_util_mem_derive
parity_wasm
parking_lot
parking_lot_core
paste
paste_impl
pbkdf2
pin_project
pin_project_internal
pin_utils
ppv_lite86
primitive_types
proc_macro2
proc_macro_crate
proc_macro_hack
proc_macro_nested
quote
radium
rand
rand_chacha
rand_core
rand_pcg
ref_cast
ref_cast_impl
regex
regex_syntax
rental
rental_impl
rlp
rustc_demangle
rustc_hash
rustc_hex
schnorrkel
scopeguard
secp256k1
serde
serde_derive
sha2
slab
smallvec
sp_application_crypto
sp_arithmetic
sp_core
sp_debug_derive
sp_externalities
sp_inherents
sp_io
sp_panic_handler
sp_runtime
sp_runtime_interface
sp_runtime_interface_proc_macro
sp_state_machine
sp_std
sp_storage
sp_tracing
sp_trie
sp_version
sp_wasm_interface
stable_deref_trait
static_assertions
substrate_bip39
subtle
syn
synstructure
thread_local
tiny_keccak
toml
tracing
tracing_attributes
tracing_core
trie_db
trie_root
twox_hash
typenum
uint
unicode_normalization
unicode_xid
wasmi
wasmi_validation
zeroize
zeroize_derive
  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
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
use alloc::vec::Vec;
use crate::io;

use super::{Deserialize, Error, Serialize, VarUint32};

use alloc::vec;
use core::{
	cmp::min,
	iter::{FromIterator, IntoIterator},
	mem, slice
};

/// A map from non-contiguous `u32` keys to values of type `T`, which is
/// serialized and deserialized ascending order of the keys. Normally used for
/// relative dense maps with occasional "holes", and stored as an array.
///
/// **SECURITY WARNING:** This code is currently subject to a denial of service
/// attack if you create a map containing the key `u32::MAX`, which should never
/// happen in normal data. It would be pretty easy to provide a safe
/// deserializing mechanism which addressed this problem.
#[derive(Debug, Default)]
pub struct IndexMap<T> {
	/// The number of non-`None` entries in this map.
	len: usize,

	/// A vector of entries. Missing entries are represented as `None`.
	entries: Vec<Option<T>>,
}

impl<T> IndexMap<T> {
	/// Create an empty `IndexMap`, preallocating enough space to store
	/// `capacity` entries without needing to reallocate the underlying memory.
	pub fn with_capacity(capacity: usize) -> IndexMap<T> {
		IndexMap {
			len: 0,
			entries: Vec::with_capacity(capacity),
		}
	}

	/// Clear the map.
	pub fn clear(&mut self) {
		self.entries.clear();
		self.len = 0;
	}

	/// Return the name for the specified index, if it exists.
	pub fn get(&self, idx: u32) -> Option<&T> {
		match self.entries.get(idx as usize) {
			Some(&Some(ref value)) => Some(value),
			Some(&None) | None => None,
		}
	}

	/// Does the map contain an entry for the specified index?
	pub fn contains_key(&self, idx: u32) -> bool {
		match self.entries.get(idx as usize) {
			Some(&Some(_)) => true,
			Some(&None) | None => false,
		}
	}

	/// Insert a name into our map, returning the existing value if present.
	///
	/// Note: This API is designed for reasonably dense indices based on valid
	/// data. Inserting a huge `idx` will use up a lot of RAM, and this function
	/// will not try to protect you against that.
	pub fn insert(&mut self, idx: u32, value: T) -> Option<T> {
		let idx = idx as usize;
		let result = if idx >= self.entries.len() {
			// We need to grow the array, and add the new element at the end.
			for _ in 0..(idx - self.entries.len()) {
				// We can't use `extend(repeat(None)).take(n)`, because that
				// would require `T` to implement `Clone`.
				self.entries.push(None);
			}
			self.entries.push(Some(value));
			debug_assert_eq!(idx + 1, self.entries.len());
			self.len += 1;
			None
		} else {
			// We're either replacing an existing element, or filling in a
			// missing one.
			let existing = self.entries[idx].take();
			if existing.is_none() {
				self.len += 1;
			}
			self.entries[idx] = Some(value);
			existing
		};
		if mem::size_of::<usize>() > 4 {
			debug_assert!(self.entries.len() <= (u32::max_value() as usize) + 1);
		}
		#[cfg(slow_assertions)]
		debug_assert_eq!(self.len, self.slow_len());
		result
	}

	/// Remove an item if present and return it.
	pub fn remove(&mut self, idx: u32) -> Option<T> {
		let result = match self.entries.get_mut(idx as usize) {
			Some(value @ &mut Some(_)) => {
				self.len -= 1;
				value.take()
			}
			Some(&mut None) | None => None,
		};
		#[cfg(slow_assertions)]
		debug_assert_eq!(self.len, self.slow_len());
		result
	}

	/// The number of items in this map.
	pub fn len(&self) -> usize {
		#[cfg(slow_assertions)]
		debug_assert_eq!(self.len, self.slow_len());
		self.len
	}

	/// Is this map empty?
	pub fn is_empty(&self) -> bool {
		self.len == 0
	}

	/// This function is only compiled when `--cfg slow_assertions` is enabled.
	/// It computes the `len` value using a slow algorithm.
	///
	/// WARNING: This turns a bunch of O(n) operations into O(n^2) operations.
	/// We may want to remove it once the code is tested, or to put it behind
	/// a feature flag named `slow_debug_checks`, or something like that.
	#[cfg(slow_assertions)]
	fn slow_len(&self) -> usize {
		self.entries.iter().filter(|entry| entry.is_some()).count()
	}

	/// Create a non-consuming iterator over this `IndexMap`'s keys and values.
	pub fn iter(&self) -> Iter<T> {
		// Note that this does the right thing because we use `&self`.
		self.into_iter()
	}

	/// Custom deserialization routine.
	///
	/// We will allocate an underlying array no larger than `max_entry_space` to
	/// hold the data, so the maximum index must be less than `max_entry_space`.
	/// This prevents mallicious *.wasm files from having a single entry with
	/// the index `u32::MAX`, which would consume far too much memory.
	///
	/// The `deserialize_value` function will be passed the index of the value
	/// being deserialized, and must deserialize the value.
	pub fn deserialize_with<R, F>(
		max_entry_space: usize,
		deserialize_value: &F,
		rdr: &mut R,
	) -> Result<IndexMap<T>, Error>
	where
		R: io::Read,
		F: Fn(u32, &mut R) -> Result<T, Error>,
	{
		let len: u32 = VarUint32::deserialize(rdr)?.into();
		let mut map = IndexMap::with_capacity(len as usize);
		let mut prev_idx = None;
		for _ in 0..len {
			let idx: u32 = VarUint32::deserialize(rdr)?.into();
			if idx as usize >= max_entry_space {
				return Err(Error::Other("index is larger than expected"));
			}
			match prev_idx {
				Some(prev) if prev >= idx => {
					// Supposedly these names must be "sorted by index", so
					// let's try enforcing that and seeing what happens.
					return Err(Error::Other("indices are out of order"));
				}
				_ => {
					prev_idx = Some(idx);
				}
			}
			let val = deserialize_value(idx, rdr)?;
			map.insert(idx, val);
		}
		Ok(map)
	}

}

impl<T: Clone> Clone for IndexMap<T> {
	fn clone(&self) -> IndexMap<T> {
		IndexMap {
			len: self.len,
			entries: self.entries.clone(),
		}
	}
}

impl<T: PartialEq> PartialEq<IndexMap<T>> for IndexMap<T> {
	fn eq(&self, other: &IndexMap<T>) -> bool {
		if self.len() != other.len() {
			// If the number of non-`None` entries is different, we can't match.
			false
		} else {
			// This is tricky, because one `Vec` might have a bunch of empty
			// entries at the end which we want to ignore.
			let smallest_len = min(self.entries.len(), other.entries.len());
			self.entries[0..smallest_len].eq(&other.entries[0..smallest_len])
		}
	}
}

impl<T: Eq> Eq for IndexMap<T> {}

impl<T> FromIterator<(u32, T)> for IndexMap<T> {
	/// Create an `IndexMap` from an iterator.
	///
	/// Note: This API is designed for reasonably dense indices based on valid
	/// data. Inserting a huge `idx` will use up a lot of RAM, and this function
	/// will not try to protect you against that.
	fn from_iter<I>(iter: I) -> Self
	where
		I: IntoIterator<Item = (u32, T)>,
	{
		let iter = iter.into_iter();
		let (lower, upper_opt) = iter.size_hint();
		let mut map = IndexMap::with_capacity(upper_opt.unwrap_or(lower));
		for (idx, value) in iter {
			map.insert(idx, value);
		}
		map
	}
}

/// An iterator over an `IndexMap` which takes ownership of it.
pub struct IntoIter<T> {
	next_idx: u32,
	remaining_len: usize,
	iter: vec::IntoIter<Option<T>>,
}

impl<T> Iterator for IntoIter<T> {
	type Item = (u32, T);

	fn size_hint(&self) -> (usize, Option<usize>) {
		(self.remaining_len, Some(self.remaining_len))
	}

	fn next(&mut self) -> Option<Self::Item> {
		// Bail early if we know there are no more items. This also keeps us
		// from repeatedly calling `self.iter.next()` once it has been
		// exhausted, which is not guaranteed to keep returning `None`.
		if self.remaining_len == 0 {
			return None;
		}
		while let Some(value_opt) = self.iter.next() {
			let idx = self.next_idx;
			self.next_idx += 1;
			if let Some(value) = value_opt {
				self.remaining_len -= 1;
				return Some((idx, value));
			}
		}
		debug_assert_eq!(self.remaining_len, 0);
		None
	}
}

impl<T> IntoIterator for IndexMap<T> {
	type Item = (u32, T);
	type IntoIter = IntoIter<T>;

	fn into_iter(self) -> IntoIter<T> {
		IntoIter {
			next_idx: 0,
			remaining_len: self.len,
			iter: self.entries.into_iter(),
		}
	}
}

/// An iterator over a borrowed `IndexMap`.
pub struct Iter<'a, T: 'static> {
	next_idx: u32,
	remaining_len: usize,
	iter: slice::Iter<'a, Option<T>>,
}

impl<'a, T: 'static> Iterator for Iter<'a, T> {
	type Item = (u32, &'a T);

	fn size_hint(&self) -> (usize, Option<usize>) {
		(self.remaining_len, Some(self.remaining_len))
	}

	fn next(&mut self) -> Option<Self::Item> {
		// Bail early if we know there are no more items. This also keeps us
		// from repeatedly calling `self.iter.next()` once it has been
		// exhausted, which is not guaranteed to keep returning `None`.
		if self.remaining_len == 0 {
			return None;
		}
		while let Some(value_opt) = self.iter.next() {
			let idx = self.next_idx;
			self.next_idx += 1;
			if let &Some(ref value) = value_opt {
				self.remaining_len -= 1;
				return Some((idx, value));
			}
		}
		debug_assert_eq!(self.remaining_len, 0);
		None
	}
}

impl<'a, T: 'static> IntoIterator for &'a IndexMap<T> {
	type Item = (u32, &'a T);
	type IntoIter = Iter<'a, T>;

	fn into_iter(self) -> Iter<'a, T> {
		Iter {
			next_idx: 0,
			remaining_len: self.len,
			iter: self.entries.iter(),
		}
	}
}

impl<T> Serialize for IndexMap<T>
where
	T: Serialize,
	Error: From<<T as Serialize>::Error>,
{
	type Error = Error;

	fn serialize<W: io::Write>(self, wtr: &mut W) -> Result<(), Self::Error> {
		VarUint32::from(self.len()).serialize(wtr)?;
		for (idx, value) in self {
			VarUint32::from(idx).serialize(wtr)?;
			value.serialize(wtr)?;
		}
		Ok(())
	}
}

impl<T: Deserialize> IndexMap<T>
where
	T: Deserialize,
	Error: From<<T as Deserialize>::Error>,
{
	/// Deserialize a map containing simple values that support `Deserialize`.
	/// We will allocate an underlying array no larger than `max_entry_space` to
	/// hold the data, so the maximum index must be less than `max_entry_space`.
	pub fn deserialize<R: io::Read>(
		max_entry_space: usize,
		rdr: &mut R,
	) -> Result<Self, Error> {
		let deserialize_value: fn(u32, &mut R) -> Result<T, Error> = |_idx, rdr| {
			T::deserialize(rdr).map_err(Error::from)
		};
		Self::deserialize_with(max_entry_space, &deserialize_value, rdr)
	}
}

#[cfg(test)]
mod tests {
	use crate::io;
	use super::*;

	#[test]
	fn default_is_empty_no_matter_how_we_look_at_it() {
		let map = IndexMap::<String>::default();
		assert_eq!(map.len(), 0);
		assert!(map.is_empty());
		assert_eq!(map.iter().collect::<Vec<_>>().len(), 0);
		assert_eq!(map.into_iter().collect::<Vec<_>>().len(), 0);
	}

	#[test]
	fn with_capacity_creates_empty_map() {
		let map = IndexMap::<String>::with_capacity(10);
		assert!(map.is_empty());
	}

	#[test]
	fn clear_removes_all_values() {
		let mut map = IndexMap::<String>::default();
		map.insert(0, "sample value".to_string());
		assert_eq!(map.len(), 1);
		map.clear();
		assert_eq!(map.len(), 0);
	}

	#[test]
	fn get_returns_elements_that_are_there_but_nothing_else() {
		let mut map = IndexMap::<String>::default();
		map.insert(1, "sample value".to_string());
		assert_eq!(map.len(), 1);
		assert_eq!(map.get(0), None);
		assert_eq!(map.get(1), Some(&"sample value".to_string()));
		assert_eq!(map.get(2), None);
	}

	#[test]
	fn contains_key_returns_true_when_a_key_is_present() {
		let mut map = IndexMap::<String>::default();
		map.insert(1, "sample value".to_string());
		assert!(!map.contains_key(0));
		assert!(map.contains_key(1));
		assert!(!map.contains_key(2));
	}

	#[test]
	fn insert_behaves_like_other_maps() {
		let mut map = IndexMap::<String>::default();

		// Insert a key which requires extending our storage.
		assert_eq!(map.insert(1, "val 1".to_string()), None);
		assert_eq!(map.len(), 1);
		assert!(map.contains_key(1));

		// Insert a key which requires filling in a hole.
		assert_eq!(map.insert(0, "val 0".to_string()), None);
		assert_eq!(map.len(), 2);
		assert!(map.contains_key(0));

		// Insert a key which replaces an existing key.
		assert_eq!(
			map.insert(1, "val 1.1".to_string()),
			Some("val 1".to_string())
		);
		assert_eq!(map.len(), 2);
		assert!(map.contains_key(1));
		assert_eq!(map.get(1), Some(&"val 1.1".to_string()));
	}

	#[test]
	fn remove_behaves_like_other_maps() {
		let mut map = IndexMap::<String>::default();
		assert_eq!(map.insert(1, "val 1".to_string()), None);

		// Remove an out-of-bounds element.
		assert_eq!(map.remove(2), None);
		assert_eq!(map.len(), 1);

		// Remove an in-bounds but missing element.
		assert_eq!(map.remove(0), None);
		assert_eq!(map.len(), 1);

		// Remove an existing element.
		assert_eq!(map.remove(1), Some("val 1".to_string()));
		assert_eq!(map.len(), 0);
	}

	#[test]
	fn partial_eq_works_as_expected_in_simple_cases() {
		let mut map1 = IndexMap::<String>::default();
		let mut map2 = IndexMap::<String>::default();
		assert_eq!(map1, map2);

		map1.insert(1, "a".to_string());
		map2.insert(1, "a".to_string());
		assert_eq!(map1, map2);

		map1.insert(0, "b".to_string());
		assert_ne!(map1, map2);
		map1.remove(0);
		assert_eq!(map1, map2);

		map1.insert(1, "not a".to_string());
		assert_ne!(map1, map2);
	}

	#[test]
	fn partial_eq_is_smart_about_none_values_at_the_end() {
		let mut map1 = IndexMap::<String>::default();
		let mut map2 = IndexMap::<String>::default();

		map1.insert(1, "a".to_string());
		map2.insert(1, "a".to_string());

		// Both maps have the same (idx, value) pairs, but map2 has extra space.
		map2.insert(10, "b".to_string());
		map2.remove(10);
		assert_eq!(map1, map2);

		// Both maps have the same (idx, value) pairs, but map1 has extra space.
		map1.insert(100, "b".to_string());
		map1.remove(100);
		assert_eq!(map1, map2);

		// Let's be paranoid.
		map2.insert(1, "b".to_string());
		assert_ne!(map1, map2);
	}

	#[test]
	fn from_iterator_builds_a_map() {
		let data = &[
			// We support out-of-order values here!
			(3, "val 3"),
			(2, "val 2"),
			(5, "val 5"),
		];
		let iter = data.iter().map(|&(idx, val)| (idx, val.to_string()));
		let map = IndexMap::from_iter(iter);
		assert_eq!(map.len(), 3);
		assert_eq!(map.get(2), Some(&"val 2".to_string()));
		assert_eq!(map.get(3), Some(&"val 3".to_string()));
		assert_eq!(map.get(5), Some(&"val 5".to_string()));
	}

	#[test]
	fn iterators_are_well_behaved() {
		// Create a map with reasonably complex internal structure, making
		// sure that we have both internal missing elements, and a bunch of
		// missing elements at the end.
		let data = &[(3, "val 3"), (2, "val 2"), (5, "val 5")];
		let src_iter = data.iter().map(|&(idx, val)| (idx, val.to_string()));
		let mut map = IndexMap::from_iter(src_iter);
		map.remove(5);

		// Make sure `size_hint` and `next` behave as we expect at each step.
		{
			let mut iter1 = map.iter();
			assert_eq!(iter1.size_hint(), (2, Some(2)));
			assert_eq!(iter1.next(), Some((2, &"val 2".to_string())));
			assert_eq!(iter1.size_hint(), (1, Some(1)));
			assert_eq!(iter1.next(), Some((3, &"val 3".to_string())));
			assert_eq!(iter1.size_hint(), (0, Some(0)));
			assert_eq!(iter1.next(), None);
			assert_eq!(iter1.size_hint(), (0, Some(0)));
			assert_eq!(iter1.next(), None);
			assert_eq!(iter1.size_hint(), (0, Some(0)));
		}

		// Now do the same for a consuming iterator.
		let mut iter2 = map.into_iter();
		assert_eq!(iter2.size_hint(), (2, Some(2)));
		assert_eq!(iter2.next(), Some((2, "val 2".to_string())));
		assert_eq!(iter2.size_hint(), (1, Some(1)));
		assert_eq!(iter2.next(), Some((3, "val 3".to_string())));
		assert_eq!(iter2.size_hint(), (0, Some(0)));
		assert_eq!(iter2.next(), None);
		assert_eq!(iter2.size_hint(), (0, Some(0)));
		assert_eq!(iter2.next(), None);
		assert_eq!(iter2.size_hint(), (0, Some(0)));
	}

	#[test]
	fn serialize_and_deserialize() {
		let mut map = IndexMap::<String>::default();
		map.insert(1, "val 1".to_string());

		let mut output = vec![];
		map.clone()
			.serialize(&mut output)
			.expect("serialize failed");

		let mut input = io::Cursor::new(&output);
		let deserialized = IndexMap::deserialize(2, &mut input).expect("deserialize failed");

		assert_eq!(deserialized, map);
	}

	#[test]
	fn deserialize_requires_elements_to_be_in_order() {
		// Build a in-order example by hand.
		let mut valid = vec![];
		VarUint32::from(2u32).serialize(&mut valid).unwrap();
		VarUint32::from(0u32).serialize(&mut valid).unwrap();
		"val 0".to_string().serialize(&mut valid).unwrap();
		VarUint32::from(1u32).serialize(&mut valid).unwrap();
		"val 1".to_string().serialize(&mut valid).unwrap();
		let map = IndexMap::<String>::deserialize(2, &mut io::Cursor::new(valid))
			.expect("unexpected error deserializing");
		assert_eq!(map.len(), 2);

		// Build an out-of-order example by hand.
		let mut invalid = vec![];
		VarUint32::from(2u32).serialize(&mut invalid).unwrap();
		VarUint32::from(1u32).serialize(&mut invalid).unwrap();
		"val 1".to_string().serialize(&mut invalid).unwrap();
		VarUint32::from(0u32).serialize(&mut invalid).unwrap();
		"val 0".to_string().serialize(&mut invalid).unwrap();
		let res = IndexMap::<String>::deserialize(2, &mut io::Cursor::new(invalid));
		assert!(res.is_err());
	}

	#[test]
	fn deserialize_enforces_max_idx() {
		// Build an example with an out-of-bounds index by hand.
		let mut invalid = vec![];
		VarUint32::from(1u32).serialize(&mut invalid).unwrap();
		VarUint32::from(5u32).serialize(&mut invalid).unwrap();
		"val 5".to_string().serialize(&mut invalid).unwrap();
		let res = IndexMap::<String>::deserialize(1, &mut io::Cursor::new(invalid));
		assert!(res.is_err());
	}
}