将文本解析为音素-Javascript
我有一个问题，在过去几天里我一直在广泛地研究这个问题，正如你可以从标题中看出的，这个问题与将字符串解析为音素的文本到语音类型算法有关 我不打算将其用于文本到语音的转换，主要是为我即将进行的项目进行文本分析。我在这里找到了一些有用的信息： 利用这些信息，我希望创建一个类似的工具。以下是上述软件的工作方式： 本质上，从字符串的第一个字符开始，我们检查字符串左右两侧的字符 （对于单词开头的字符，左边的字符将是空白） 我们根据44个英语音素和每个音素下的字母设置规则 这里是我获得数据来制定这些规则的地方： -我相信这与之前文章中的一些规则是不同的 以下是我迄今为止创建的规则对象的一个示例：将文本解析为音素-Javascript,javascript,string,algorithm,sorting,parsing,Javascript,String,Algorithm,Sorting,Parsing,我有一个问题，在过去几天里我一直在广泛地研究这个问题，正如你可以从标题中看出的，这个问题与将字符串解析为音素的文本到语音类型算法有关 我不打算将其用于文本到语音的转换，主要是为我即将进行的项目进行文本分析。我在这里找到了一些有用的信息： 利用这些信息，我希望创建一个类似的工具。以下是上述软件的工作方式： 本质上，从字符串的第一个字符开始，我们检查字符串左右两侧的字符 （对于单词开头的字符，左边的字符将是空白） 我们根据44个英语音素和每个音素下的字母设置规则 这里是我获得数据来制定这些规则的地
var rules_syntax = {
cons_b: {case:["b", "bb"], pro: "b"},
cons_d: {case:["d", "dd", "ed"], pro: "d"},
cons_f: {case:["f", "ph"], pro: "f"},
cons_g: {case:["g", "gg"], pro: "g"},
cons_h: {case:["h"], pro: "h"},
cons_j: {case:["j", "g", "ge", "dge"], pro: "j"},
cons_k: {case:["c", "k", "ck", "ch", "cc", "que"], pro: "k"},
cons_l: {case:["l", "ll"], pro: "l"},
cons_m: {case:["m", "mm", "mb"], pro: "m"},
cons_n: {case:["n", "nn", "kn", "gn"], pro: "n"},
cons_p: {case:["p", "pp"], pro: "p"},
cons_r: {case:["r", "rr", "wr"], pro: "r"},
cons_s: {case:["s", "se", "ss", "ci", "ce", "sc"], pro: "s"},
cons_t: {case:["t", "tt", "ed"], pro: "t"},
cons_v: {case:["v", "ve"], pro: "v"},
cons_w: {case:["w"], pro: "w"},
cons_y: {case:["y", "oi", "io"], pro: "y"},
cons_z: {case:["z", "zz", "ze", "se", "x"], pro: "z"},
cons_diag_th: {case:["th"], pro: "th"},
cons_diag_th_voice: {case:["th"], pro: "thh"},
cons_diag_ng: {case:["ng", "n"], pro: "n"},
cons_diag_sh: {case:["sh", "ss", "ch", "ti", "ci"], pro: "sh"},
cons_diag_ch: {case:["ch", "tch"], pro: "ch"},
cons_diag_zh: {case:["ge", "s"], pro: "zh"},
cons_diag_wh: {case:["wh"], pro: "w"},
shor_vowe_a: {case: ["a", "au"], pro: "a"},
shor_vowe_e: {case: ["e", "ea"], pro: "e"},
shor_vowe_i: {case: ["i"], pro: "i"},
shor_vowe_o: {case: ["o", "a", "au", "aw", "ough"], pro: "o"},
shor_vowe_u: {case: ["u", "o"], pro: "u"},
}
var test01 = function(word)
{
var arr = [];
var arr2 = [];
//for every letter in the word
for(var x = 0; x < word.length; x ++)
{
//grab the letter
var index = word[x];
var pert = [];
//push the letter to an array
arr.push(index);
//create an object which holds the neighboring letters
var pro = new Object();
pro.left = word[x - 1];
pro.org = index;
pro.right = word[x + 1];
//store this object in an array
arr2.push(pro);
//check to see if any spaces exist (undefined) and replace them with a ""
for(b in pro)
{
if(!pro[b])
{
pro[b] = "";
}
}
//loop through each letter object
for(var y = 0; y < arr2.length; y ++)
{
var letter = arr2[y];
letter.rules = [];
//loop through each rule that exists
for (z in rules_syntax)
{
var rule = rules_syntax[z];
//for each rule, loop through its graphemes
for(var a = 0; a < rule.case.length; a ++)
{
var grapheme = rule.case[a];
//combine the original letter, and the letter to the left
var letterCheck = letter.left + letter.org;
//check if the above combo (left + original) exists in the list of graphemes
if(letterCheck === grapheme)
{
//here is where I would put any other conditions to check for any other rules.
pert.push(rule.pro);
letter.rules.push(rule.pro);
}
if(letter.org === grapheme)
{
//here is where I would put any other conditions to check for any other rules.
pert.push(rule.pro);
letter.rules.push(rule.pro);
}
}
}
}
console.log(pert);
}
}
test01("cautious");
谢谢。我认为您误读了文章中的表2： “#”一个或多个元音 “*”（星）一个或多个辅音 “*”（圈）B、D、V、G、J、L、M、N、R、W和Z中的一个：a浊音 辅音 “$”一个辅音后跟一个E或I （ER，E，ES，ED，ING，ELY）中的“%”一个后缀 （S，C，G，Z，X，J，CH，SH）中的一个：sibilant （T，S，R，D，L，Z，N，J，TH，CH，SH）之一：辅音 影响后面长u的声音（参见rule和mule） “^”一个辅音 （E，I，Y）中的“+”一个：前元音 “：”零个或多个辅音 该表的标题是： 英文到IPA翻译规则中出现的特殊符号 换句话说，这些不是规则，而是对符号的描述 真正的规则从文章的第51页开始。根据我对算法的理解，它将从解析“t”开始，然后转到TRULE.ENG部分（第58页）。这里唯一适用的规则是
[T]=/T/[AU]=/AO///T/至于“笑”，根据规则，我会说“au”的发音和“teached”一样。因为规则是
#[GH]=/(16) ["k", "k", "a", "o", "a", "o", "u", "t", "sh", "i", "y", "o", "u", "u", "s", "zh"]