1
|
|
|
<?php declare(strict_types = 1); |
2
|
|
|
/** |
3
|
|
|
* Finds intervals inside strings |
4
|
|
|
* |
5
|
|
|
* PHP version 7+ |
6
|
|
|
* |
7
|
|
|
* @category IntervalParser |
8
|
|
|
* @author Ekin H. Bayar <[email protected]> |
9
|
|
|
* @version 0.2.0 |
10
|
|
|
*/ |
11
|
|
|
namespace IntervalParser; |
12
|
|
|
|
13
|
|
|
class IntervalFinder |
14
|
|
|
{ |
15
|
|
|
# Leading separator |
|
|
|
|
16
|
|
|
const LEADING_SEPARATOR = "(?<leadingSeparator>\s?(?:in)\s?)"; |
|
|
|
|
17
|
|
|
|
18
|
|
|
# Regex to match a valid interval and any trailing string, holds the interval in $matches['interval'], the rest in $matches['trailing'] |
|
|
|
|
19
|
|
|
const INTERVAL_WITH_TRAILING_DATA = "^(?<interval>(?&timepart)++)(?<trailing>.+)*?$/uix"; |
|
|
|
|
20
|
|
|
|
21
|
|
|
# Regex to handle an input that may have multiple intervals along with leading and/or trailing data |
|
|
|
|
22
|
|
|
const MULTIPLE_INTERVALS = <<<'REGEX' |
23
|
|
|
^(?<leading>.*?)? |
24
|
|
|
(?<sep>(?&leadingSeparator))? |
25
|
|
|
(?<interval>(?&timepart)++) |
26
|
|
|
(?<trailing>.*) |
27
|
|
|
/uix |
28
|
|
|
REGEX; |
29
|
|
|
|
30
|
|
|
/** |
31
|
|
|
* @var ParserSettings |
32
|
|
|
*/ |
33
|
|
|
private $settings; |
34
|
|
|
|
35
|
|
|
/** |
36
|
|
|
* @var Normalizer |
37
|
|
|
*/ |
38
|
|
|
private $normalizer; |
39
|
|
|
|
40
|
|
|
/** |
41
|
|
|
* IntervalFinder constructor. |
42
|
|
|
* |
43
|
|
|
* Default settings are : |
44
|
|
|
* |
45
|
|
|
* string $symbolSeparator = ',', |
46
|
|
|
* string $wordSeparator = null |
47
|
|
|
* |
48
|
|
|
* @param \IntervalParser\ParserSettings $settings |
49
|
|
|
* @param \IntervalParser\Normalizer $normalizer |
50
|
|
|
*/ |
51
|
4 |
|
public function __construct(ParserSettings $settings, Normalizer $normalizer) |
52
|
|
|
{ |
53
|
4 |
|
$this->settings = $settings; |
54
|
4 |
|
$this->normalizer = $normalizer; |
55
|
|
|
} |
56
|
|
|
|
57
|
|
|
/** |
58
|
|
|
* Looks for a valid interval along with leading and/or trailing data IF the respective flags are set. |
59
|
|
|
* TimeInterval is essentially DateInterval with extra information such as interval offset & length, leading/trailing data. |
|
|
|
|
60
|
|
|
* |
61
|
|
|
* @param string $input |
62
|
|
|
* @param int $flags |
63
|
|
|
* @return TimeInterval|array |
64
|
|
|
* @throws InvalidFlagException |
65
|
|
|
* @throws FormatException |
66
|
|
|
*/ |
67
|
4 |
|
public function find(string $input, int $flags = IntervalFlags::INTERVAL_ONLY) |
68
|
|
|
{ |
69
|
|
|
if ($flags |
70
|
4 |
|
& ~IntervalFlags::INTERVAL_ONLY |
71
|
4 |
|
& ~IntervalFlags::REQUIRE_TRAILING |
72
|
4 |
|
& ~IntervalFlags::REQUIRE_LEADING |
73
|
4 |
|
& ~IntervalFlags::MULTIPLE_INTERVALS |
74
|
|
|
) { throw new InvalidFlagException("You have tried to use an invalid flag combination."); } |
|
|
|
|
75
|
|
|
|
76
|
4 |
|
if ($flags & IntervalFlags::INTERVAL_ONLY) { |
77
|
|
|
|
78
|
|
|
$input = $this->normalizer->normalize($input); |
79
|
|
|
|
80
|
|
|
$definition = Pattern::DEFINE . Pattern::INTEGER . Pattern::TIME_PART . ')'; |
81
|
|
|
$expression = $definition . Pattern::INTERVAL_ONLY; |
82
|
|
|
|
83
|
|
|
if (preg_match($expression, $input)) { |
84
|
|
|
$intervalOffset = 0; |
85
|
|
|
$intervalLength = strlen($input); |
86
|
|
|
|
87
|
|
|
# create and return the interval object |
|
|
|
|
88
|
|
|
$interval = \DateInterval::createFromDateString($input); |
89
|
|
|
return new TimeInterval($interval, $intervalOffset, $intervalLength); |
90
|
|
|
} |
91
|
|
|
|
92
|
|
|
throw new FormatException("Given input is not a valid interval."); |
|
|
|
|
93
|
|
|
} |
94
|
|
|
|
95
|
4 |
|
if ($flags == (IntervalFlags::REQUIRE_LEADING | IntervalFlags::REQUIRE_TRAILING)) { |
96
|
|
|
|
97
|
1 |
|
$expression = $this->settings->getLeadingSeparatorExpression(); |
98
|
|
|
|
99
|
1 |
|
$leadingSeparation = preg_match($expression, $input, $matches, PREG_OFFSET_CAPTURE); |
100
|
1 |
|
if (!$leadingSeparation) { |
101
|
|
|
throw new FormatException("Allowing leading data requires using a separator. Ie. foo in <interval>"); |
|
|
|
|
102
|
|
|
} |
103
|
|
|
|
104
|
1 |
|
$leadingData = $matches[1][0] ?? null; |
|
|
|
|
105
|
1 |
|
$intervalAndTrailingData = $matches[2][0] ?? null; |
106
|
|
|
|
107
|
|
|
# throw early for missing parts |
|
|
|
|
108
|
1 |
|
if (!$leadingData) { |
|
|
|
|
109
|
|
|
throw new FormatException("Given input does not contain a valid leading data."); |
|
|
|
|
110
|
|
|
} |
111
|
1 |
|
if (!$intervalAndTrailingData) { |
|
|
|
|
112
|
|
|
throw new FormatException("Given input does not contain a valid interval and/or trailing data."); |
|
|
|
|
113
|
|
|
} |
114
|
|
|
|
115
|
1 |
|
$intervalOffset = $matches[2][1] ?? null; |
116
|
|
|
|
117
|
|
|
# If interval contains non-strtotime-compatible abbreviations, replace 'em |
|
|
|
|
118
|
1 |
|
$intervalAndTrailingData = $this->normalizer->normalize($intervalAndTrailingData); |
119
|
|
|
|
120
|
1 |
|
$definition = Pattern::DEFINE . Pattern::INTEGER . Pattern::TIME_PART . ')'; |
121
|
1 |
|
$expression = $definition . self::INTERVAL_WITH_TRAILING_DATA; |
122
|
|
|
|
123
|
1 |
|
if (preg_match($expression, $intervalAndTrailingData, $parts)) { |
124
|
|
|
|
125
|
1 |
|
$interval = $parts['interval']; |
|
|
|
|
126
|
1 |
|
$trailingData = $parts['trailing']; |
127
|
1 |
|
$intervalLength = strlen($interval); |
128
|
|
|
|
129
|
|
|
# create and return the interval object |
|
|
|
|
130
|
1 |
|
$interval = \DateInterval::createFromDateString($interval); |
131
|
1 |
|
return new TimeInterval($interval, $intervalOffset, $intervalLength, $leadingData, $trailingData); |
132
|
|
|
} |
133
|
|
|
|
134
|
|
|
throw new FormatException("Given input does not contain a valid interval and/or trailing data."); |
|
|
|
|
135
|
|
|
} |
136
|
|
|
|
137
|
3 |
|
if ($flags & IntervalFlags::REQUIRE_LEADING) { |
138
|
|
|
|
139
|
1 |
|
$expression = $this->settings->getLeadingSeparatorExpression(); |
140
|
|
|
|
141
|
1 |
|
$leadingSeparation = preg_match($expression, $input, $matches, PREG_OFFSET_CAPTURE); |
142
|
1 |
|
if (!$leadingSeparation) { |
143
|
|
|
throw new FormatException("Allowing leading data requires using a separator. Ie. foo in <interval>"); |
|
|
|
|
144
|
|
|
} |
145
|
|
|
|
146
|
1 |
|
$leadingData = $matches[1][0] ?? null; |
|
|
|
|
147
|
1 |
|
$intervalAndPossibleTrailingData = $matches[2][0] ?? null; |
148
|
|
|
|
149
|
1 |
|
if (!$leadingData) { |
|
|
|
|
150
|
|
|
throw new FormatException("Could not find any valid leading data."); |
|
|
|
|
151
|
|
|
} |
152
|
|
|
|
153
|
1 |
|
if (!$intervalAndPossibleTrailingData) { |
|
|
|
|
154
|
|
|
throw new FormatException("Could not find any valid interval and/or leading data."); |
|
|
|
|
155
|
|
|
} |
156
|
|
|
|
157
|
1 |
|
$intervalOffset = $matches[2][1] ?? null; |
158
|
|
|
|
159
|
|
|
# If interval contains non-strtotime-compatible abbreviations, replace 'em |
|
|
|
|
160
|
1 |
|
$safeInterval = $this->normalizer->normalize($intervalAndPossibleTrailingData); |
161
|
|
|
|
162
|
|
|
# since above normalization is expected to not return any trailing data, only check for a valid interval |
|
|
|
|
163
|
1 |
|
$definition = Pattern::DEFINE . Pattern::INTEGER . Pattern::TIME_PART . ')'; |
164
|
1 |
|
$expression = $definition . Pattern::INTERVAL_ONLY; |
165
|
|
|
|
166
|
1 |
|
if (preg_match($expression, $safeInterval, $parts)) { |
167
|
1 |
|
$interval = $parts['interval']; |
|
|
|
|
168
|
1 |
|
$intervalLength = strlen($interval); |
169
|
|
|
|
170
|
|
|
# create the interval object |
|
|
|
|
171
|
1 |
|
$interval = \DateInterval::createFromDateString($interval); |
172
|
1 |
|
return new TimeInterval($interval, $intervalOffset, $intervalLength, $leadingData); |
173
|
|
|
} |
174
|
|
|
|
175
|
|
|
throw new FormatException("Given input does not contain a valid interval. Keep in mind trailing data is not allowed with current flag."); |
|
|
|
|
176
|
|
|
} |
177
|
|
|
|
178
|
2 |
|
if ($flags & IntervalFlags::REQUIRE_TRAILING) { |
179
|
|
|
|
180
|
2 |
|
$definition = Pattern::DEFINE . Pattern::INTEGER . Pattern::TIME_PART . ')'; |
181
|
2 |
|
$expression = $definition . self::INTERVAL_WITH_TRAILING_DATA; |
182
|
|
|
|
183
|
|
|
# If interval contains non-strtotime-compatible abbreviations, replace 'em |
|
|
|
|
184
|
2 |
|
$safeInterval = $this->normalizer->normalize($input); |
185
|
|
|
|
186
|
|
|
# Separate interval from trailing data |
|
|
|
|
187
|
2 |
|
if (preg_match($expression, $safeInterval, $parts)) { |
188
|
2 |
|
$trailingData = $parts['trailing'] ?? null; |
189
|
2 |
|
$interval = $parts['interval'] ?? null; |
|
|
|
|
190
|
|
|
|
191
|
2 |
|
if (!$interval) { |
|
|
|
|
192
|
|
|
throw new FormatException("Could not find any valid interval."); |
|
|
|
|
193
|
|
|
} |
194
|
|
|
|
195
|
2 |
|
if (!$trailingData) { |
|
|
|
|
196
|
|
|
throw new FormatException("Could not find any valid trailing data."); |
|
|
|
|
197
|
|
|
} |
198
|
|
|
|
199
|
2 |
|
$intervalLength = strlen($interval); |
200
|
2 |
|
$intervalOffset = 0; # since we don't allow leading data here |
|
|
|
|
201
|
|
|
|
202
|
|
|
# create the interval object |
|
|
|
|
203
|
2 |
|
$interval = \DateInterval::createFromDateString($interval); |
204
|
2 |
|
return new TimeInterval($interval, $intervalOffset, $intervalLength, null, $trailingData); |
205
|
|
|
} |
206
|
|
|
|
207
|
|
|
throw new FormatException("Given input does not contain a valid interval. Keep in mind leading data is not allowed with current flag."); |
|
|
|
|
208
|
|
|
} |
209
|
|
|
|
210
|
|
|
if ($flags & IntervalFlags::MULTIPLE_INTERVALS) { |
211
|
|
|
|
212
|
|
|
$payload = []; |
|
|
|
|
213
|
|
|
$separator = ($this->settings->getSeparationType() == 'symbol') |
214
|
|
|
? $this->settings->getSymbolSeparator() |
215
|
|
|
: $this->settings->getWordSeparator(); |
216
|
|
|
|
217
|
|
|
$expression = "/(?J)\b(?:(?<match>.*?)\s?{$separator})\s?|\b(?<match>.*)/ui"; |
|
|
|
|
218
|
|
|
|
219
|
|
|
if (preg_match_all($expression, $input, $intervals, PREG_SET_ORDER)) { |
220
|
|
|
|
221
|
|
|
$intervalSet = array_filter(array_map(function($set) { |
222
|
|
|
foreach ($iter = new IntervalIterator($set) as $key => $interval) { |
223
|
|
|
if ($iter->key() === 'match') { |
224
|
|
|
return $interval; |
225
|
|
|
} |
226
|
|
|
} |
227
|
|
|
}, $intervals)); |
228
|
|
|
|
229
|
|
|
foreach ($intervalSet as $key => $interval) { |
230
|
|
|
|
231
|
|
|
$definition = Pattern::DEFINE . self::LEADING_SEPARATOR . Pattern::INTEGER . Pattern::TIME_PART . ')'; |
|
|
|
|
232
|
|
|
$expression = $definition . self::MULTIPLE_INTERVALS; |
233
|
|
|
|
234
|
|
|
preg_match($expression, $interval, $matches); |
235
|
|
|
$matches = array_filter($matches); |
236
|
|
|
|
237
|
|
|
$leadingData = $matches['leading'] ?? null; |
238
|
|
|
$leadingSep = $matches['sep'] ?? null; |
239
|
|
|
$interval = $matches['interval'] ?? null; |
240
|
|
|
$trailing = $matches['trailing'] ?? null; |
241
|
|
|
|
242
|
|
|
if (!$leadingData) $leadingData = $leadingSep ?? ""; |
|
|
|
|
243
|
|
|
|
244
|
|
|
$intervalOffset = (!$leadingSep) ? 0 : strlen($leadingData) + strlen($leadingSep); |
245
|
|
|
|
246
|
|
|
# If interval contains non-strtotime-compatible abbreviations, replace them |
|
|
|
|
247
|
|
|
$safeInterval = $this->normalizer->normalize($interval . $trailing); |
248
|
|
|
|
249
|
|
|
# Separate intervals from trailing data |
|
|
|
|
250
|
|
|
if (preg_match($expression, $safeInterval, $parts)) { |
251
|
|
|
$trailingData = $parts['trailing'] ?? null; |
252
|
|
|
$interval = $parts['interval'] ?? null; |
|
|
|
|
253
|
|
|
if (!$interval) continue; |
|
|
|
|
254
|
|
|
|
255
|
|
|
$intervalLength = strlen($interval); |
256
|
|
|
# create the interval object |
|
|
|
|
257
|
|
|
$interval = \DateInterval::createFromDateString($interval); |
|
|
|
|
258
|
|
|
$payload[] = new TimeInterval($interval, $intervalOffset, $intervalLength, $leadingData, $trailingData); |
|
|
|
|
259
|
|
|
} |
260
|
|
|
} |
261
|
|
|
|
262
|
|
|
if ($payload) return $payload; |
|
|
|
|
263
|
|
|
} |
264
|
|
|
} |
265
|
|
|
} |
266
|
|
|
} |
267
|
|
|
|