forked from overtrue/pinyin
-
Notifications
You must be signed in to change notification settings - Fork 0
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
- Loading branch information
Showing
24 changed files
with
712,935 additions
and
326 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,100 @@ | ||
| #!/usr/bin/env php | ||
| <?php | ||
|
|
||
| $polyphones = explode(',', file_get_contents(__DIR__ . '/../sources/polyphones.txt')); | ||
| $charsSouce = __DIR__ . '/../sources/chars.txt'; | ||
| $wordsSouce = __DIR__ . '/../sources/words.txt'; | ||
| $surnamesSource = file(__DIR__.'/../sources/surnames.txt'); | ||
|
|
||
|
|
||
| if (!file_exists($charsSouce)) { | ||
| file_put_contents($charsSouce, file_get_contents('https://raw.githubusercontent.com/mozillazg/pinyin-data/master/pinyin.txt')); | ||
| } | ||
|
|
||
| if (!file_exists($wordsSouce)) { | ||
| file_put_contents($wordsSouce, file_get_contents('https://raw.githubusercontent.com/mozillazg/phrase-pinyin-data/master/large_pinyin.txt')); | ||
| } | ||
|
|
||
|
|
||
| // ------------------------------------------------ | ||
| $surnames = []; | ||
| foreach ($surnamesSource as $line) { | ||
| [$surname, $pinyin] = explode(',', trim($line)); | ||
|
|
||
| $surnames[trim($surname)] = preg_split('/\s+/', trim($pinyin)); | ||
| } | ||
|
|
||
|
|
||
| // ------------------------------------------------ | ||
|
|
||
| $chars = []; | ||
| foreach (file($charsSouce) as $line) { | ||
| // U+4E2D: zhōng,zhòng # 中 | ||
| preg_match('/^U\+(?<code>[0-9A-Z]+):\s+(?<pinyin>\S+)\s+#\s*(?<char>\S+)/', $line, $matched); | ||
|
|
||
| if ($matched && !empty($matched['pinyin'])) { | ||
| $pinyin = explode(',', $matched['pinyin']); | ||
| $chars[$matched['char']] = $pinyin; | ||
| } elseif (!str_starts_with($line, '#')) { | ||
| throw new Exception("行解析错误：$line"); | ||
| } | ||
| } | ||
|
|
||
| // ------------------------------------------------ | ||
| $words = []; | ||
| foreach (file($wordsSouce) as $line) { | ||
| // 㞎㞎: bǎ ba # 注释 | ||
| preg_match('/^(?<word>[^#\s]+):\s+(?<pinyin>[\p{L} ]+)#?/u', $line, $matched); | ||
|
|
||
| if ($matched && !empty($matched['pinyin'])) { | ||
| $pinyin = explode(' ', trim($matched['pinyin'])); | ||
|
|
||
| $wordChars = preg_split('//u', $matched['word'], -1, PREG_SPLIT_NO_EMPTY); | ||
|
|
||
| try { | ||
| $pinyinSegments = array_combine($wordChars, $pinyin); | ||
| } catch (Throwable $e) { | ||
| throw new Exception("行解析错误：$line"); | ||
| } | ||
|
|
||
| // 多音字处理 | ||
| $polyphoneChars = array_intersect_key($wordChars, $polyphones); | ||
|
|
||
| foreach ($polyphoneChars as $char) { | ||
| // 如果词里的任何一个多音字在词里的读音和常用读音不一致，则需要加入词典，否则抛弃该词 | ||
| if (isset($chars[$char]) && $pinyinSegments[$char] != $chars[$char][0]) { | ||
| $words[$matched['word']] = join("\t", ["", ...$pinyin, ""]); | ||
| break; | ||
| } | ||
| } | ||
| } | ||
| } | ||
|
|
||
| // 清理 | ||
| exec('rm -rf ' . __DIR__ . '/../data/*'); | ||
|
|
||
| // 姓氏 | ||
| file_put_contents(__DIR__ . '/../data/surnames.php', "<?php\nreturn ".var_export($surnames, true).";\n"); | ||
| echo count($surnames)." surnames saved.\n"; | ||
|
|
||
| // 单字：带多音 | ||
| file_put_contents(__DIR__ . '/../data/char-with-polyphones.php', "<?php\nreturn ".var_export($chars, true).";\n"); | ||
| echo count($chars)." chars with polyphones saved.\n"; | ||
|
|
||
| // 单字：不带多音 | ||
| $charsNoPolyphones = []; | ||
| foreach ($chars as $char => $pinyin) { | ||
| $charsNoPolyphones[$char] = "\t{$pinyin[0]}\t"; | ||
| } | ||
|
|
||
| file_put_contents(__DIR__ . '/../data/chars.php', "<?php\nreturn ".var_export($charsNoPolyphones, true).";\n"); | ||
| echo count($charsNoPolyphones)." chars saved.\n"; | ||
|
|
||
| // 词：从长到短 + 单字 | ||
| $words = array_merge($words, $charsNoPolyphones); | ||
| uksort($words, fn ($a, $b) => strlen($b) <=> strlen($a)); | ||
|
|
||
| foreach (array_chunk($words, 8000, true) as $index => $group) { | ||
| file_put_contents(__DIR__ . "/../data/words-{$index}.php", "<?php\nreturn ".var_export($group, true).";\n"); | ||
| echo count($group)." words saved in ".__DIR__ . "/../data/words-{$index}.php \n"; | ||
| } |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -1,68 +1,68 @@ | ||
| { | ||
| "name": "overtrue/pinyin", | ||
| "description": "Chinese to pinyin translator.", | ||
| "keywords": [ | ||
| "chinese", | ||
| "pinyin", | ||
| "cn2pinyin" | ||
| "name": "overtrue/pinyin", | ||
| "description": "Chinese to pinyin translator.", | ||
| "keywords": [ | ||
| "chinese", | ||
| "pinyin", | ||
| "cn2pinyin" | ||
| ], | ||
| "homepage": "https://github.com/overtrue/pinyin", | ||
| "license": "MIT", | ||
| "authors": [ | ||
| { | ||
| "name": "overtrue", | ||
| "homepage": "http://github.com/overtrue", | ||
| "email": "[email protected]" | ||
| } | ||
| ], | ||
| "autoload": { | ||
| "psr-4": { | ||
| "Overtrue\\Pinyin\\": "src/" | ||
| } | ||
| }, | ||
| "autoload-dev": { | ||
| "psr-4": { | ||
| "Overtrue\\Pinyin\\Tests\\": "tests/" | ||
| } | ||
| }, | ||
| "require": { | ||
| "php": ">=7.1" | ||
| }, | ||
| "require-dev": { | ||
| "phpunit/phpunit": "~9.5", | ||
| "brainmaestro/composer-git-hooks": "^2.7", | ||
| "friendsofphp/php-cs-fixer": "^3.2" | ||
| }, | ||
| "extra": { | ||
| "hooks": { | ||
| "pre-commit": [ | ||
| "composer test", | ||
| "composer fix-style" | ||
| ], | ||
| "pre-push": [ | ||
| "composer test", | ||
| "composer check-style" | ||
| ] | ||
| } | ||
| }, | ||
| "scripts": { | ||
| "post-update-cmd": [ | ||
| "cghooks update" | ||
| ], | ||
| "homepage": "https://github.com/overtrue/pinyin", | ||
| "license": "MIT", | ||
| "authors": [ | ||
| { | ||
| "name": "overtrue", | ||
| "homepage": "http://github.com/overtrue", | ||
| "email": "[email protected]" | ||
| } | ||
| "post-merge": "composer install", | ||
| "post-install-cmd": [ | ||
| "cghooks add --ignore-lock", | ||
| "cghooks update" | ||
| ], | ||
| "autoload": { | ||
| "psr-4": { | ||
| "Overtrue\\Pinyin\\": "src/" | ||
| }, | ||
| "files": ["src/const.php"] | ||
| }, | ||
| "autoload-dev": { | ||
| "psr-4": { | ||
| "Overtrue\\Pinyin\\Test\\": "tests/" | ||
| } | ||
| }, | ||
| "require": { | ||
| "php":">=7.1" | ||
| }, | ||
| "require-dev": { | ||
| "phpunit/phpunit": "~9.5", | ||
| "brainmaestro/composer-git-hooks": "^2.7", | ||
| "friendsofphp/php-cs-fixer": "^3.2" | ||
| }, | ||
| "extra": { | ||
| "hooks": { | ||
| "pre-commit": [ | ||
| "composer test", | ||
| "composer fix-style" | ||
| ], | ||
| "pre-push": [ | ||
| "composer test", | ||
| "composer check-style" | ||
| ] | ||
| } | ||
| }, | ||
| "scripts": { | ||
| "post-update-cmd": [ | ||
| "cghooks update" | ||
| ], | ||
| "post-merge": "composer install", | ||
| "post-install-cmd": [ | ||
| "cghooks add --ignore-lock", | ||
| "cghooks update" | ||
| ], | ||
| "cghooks": "vendor/bin/cghooks", | ||
| "check-style": "php-cs-fixer fix --using-cache=no --diff --config=.php_cs --dry-run --ansi", | ||
| "fix-style": "php-cs-fixer fix --using-cache=no --config=.php_cs --ansi", | ||
| "test": "vendor/bin/phpunit --colors=always" | ||
| }, | ||
| "scripts-descriptions": { | ||
| "test": "Run all tests.", | ||
| "check-style": "Run style checks (only dry run - no fixing!).", | ||
| "fix-style": "Run style checks and fix violations." | ||
| } | ||
| "cghooks": "vendor/bin/cghooks", | ||
| "check-style": "php-cs-fixer fix --using-cache=no --diff --dry-run --ansi", | ||
| "fix-style": "php-cs-fixer fix --using-cache=no --ansi", | ||
| "test": "vendor/bin/phpunit --colors=always", | ||
| "build": "php ./bin/build" | ||
| }, | ||
| "scripts-descriptions": { | ||
| "test": "Run all tests.", | ||
| "check-style": "Run style checks (only dry run - no fixing!).", | ||
| "fix-style": "Run style checks and fix violations." | ||
| } | ||
| } |
Oops, something went wrong.