mac-cain13 / R.swift

Strong typed, autocompleted resources like images, fonts and segues in Swift projects
MIT License
9.49k stars 763 forks source link

Add xcstrings support #864

Closed fthdgn closed 10 months ago

fthdgn commented 10 months ago

I did a simple development for xcstrings support.

It is backward compatible for for .xcstrings files converted from .strings files. It means the same R.string codes will be generated.

However, it is not backward compatible for .xcstrings files converted from .stringsdict files, because of named arguments.

About named arguments

My implementation strips names of substitutions from generated arguments. I found using them problematic on some cases.

In my opinion, named arguments does not worth the implementation. Xcode generates substituons only if there are more than one pluralable parameter on new xcstrings files. The most of the string values will not be have any information to generate named arguments.

Problematic case 1

Some key information is lost on the xcstring convertion of stringsdict files. Original stringsdict content.

    <key>x_users</key>
    <dict>
        <key>NSStringLocalizedFormatKey</key>
        <string>%#@users@</string>
        <key>users</key>
        <dict>
            <key>NSStringFormatSpecTypeKey</key>
            <string>NSStringPluralRuleType</string>
            <key>NSStringFormatValueTypeKey</key>
            <string>d</string>
            <key>one</key>
            <string>%d user</string>
            <key>other</key>
            <string>%d users</string>
        </dict>
    </dict>

Generated xcstrings content

{
  "x_users" : {
    "localizations" : {
      "en" : {
        "variations" : {
          "plural" : {
            "one" : {"stringUnit" : {"value" : "%d user"}},
            "other" : {"stringUnit" : {"value" : "%d users"}
}}}}}}}

R.swift creates .x_users(users: Int) for stringsdict file. However, xcstrings file does not have the argument name information any more.

If string value of the original content is <string>Add %#@users@</string>, generated xcstrings would keep the substitution information.

Problematic case 2

{
  "example": {
    "localizations": {
      "en": {
        "substitutions": {
          "device_iphone": {
            "argNum": 1,
            "formatSpecifier": "lld",
            "variations": {
              "plural": {
                "one": {"stringUnit": {"value": "%arg iPhone"}},
                "other": {"stringUnit": {"value": "%arg iPhones"}}
              }}},
          "device_mac": {
            "argNum": 1,
            "formatSpecifier": "lld",
            "variations": {
              "plural": {
                "one": {"stringUnit": {"value": "%arg Mac"}},
                "other": {"stringUnit": {"value": "%arg Macs"}}
              }}},
          "input_iphone": {
            "argNum": 2,
            "formatSpecifier": "lld",
            "variations": {
              "plural": {
                "one": {"stringUnit": {"value": "%arg touch"}},
                "other": {"stringUnit": {"value": "%arg touches"}}
              }}},
          "input_mac": {
            "argNum": 2,
            "formatSpecifier": "lld",
            "variations": {
              "plural": {
                "one": {"stringUnit": {"value": "%arg key"}},
                "other": {"stringUnit": {"value": "%arg keys"}}
              }}},
        "variations": {
          "device": {
            "iphone": {
              "stringUnit": {
                "value": "%#@device_iphone@ and %#@input_iphone@"
              }},
            "mac": {
              "stringUnit": {
                "value": "%#@device_mac@ and %#@input_mac@"
              }}}}}}}}

This strings value is "1 iPhone and 10 touches" on iPhones and "1 Mac and 10 keys" on Mac. There are 2 substitutions for each device variation. Which signature should we use? .example(device_iphone: Int, input_iphone: Int) .example(device_mac: Int, inputmac: Int) .example( arg1: Int, _ arg2: Int) .example(device Int, input: Int) //With some extra coding to detect shared parts of the names.

About algorithm

The alghorithm tries to convert localization of source language to single string with basic format parameters, then it uses FormatPart.formatParts(formatString:)` on this string to extract parameters.

This convertion works like that:

Substitutions replacement works like that:

Example 1

  "account": {
    "localizations": {
      "en": {
        "stringUnit": {"value": "Account"}
      }}}

Example 2

  "x_sets": {
    "localizations": {
      "en": {
        "variations": {
          "plural": {
            "one": {"stringUnit": {"value": "a set"}},
            "other": {"stringUnit": {"value": "%d sets"}}
          }}}}}

Example 3

{
  "example": {
    "localizations": {
      "en": {
        "stringUnit": {
          "value": "%#@books@ and %#@pens@"
        },
        "substitutions": {
          "books": {
            "argNum": 1,
            "formatSpecifier": "lld",
            "variations": {
              "plural": {
                "one": {"stringUnit": {"value": "%arg book"}},
                "other": {"stringUnit": {"value": "%arg books"}}
              }}},
          "pens": {
            "argNum": 2,
            "formatSpecifier": "lld",
            "variations": {
              "plural": {
                "one": {"stringUnit": {"value": "%arg pen"}},
                "other": {"stringUnit": {"value": "%arg pens"}}
              }}}}}}}}