Use custom data.

ICU is used to:
- List locales (`locales_tree`)
- Get information about those locales in different languages (`lang_tree`)
- Remove accents (`translit`)

Tracing ICU without custom data, we also load `likelySubtags` and
`metadata`. Not sure why and what happen when missing but let's be
conservatives and include them.

kiwixbuild/dependencies/icu4c.py

@@ -4,16 +4,41 @@
     MakeBuilder
 )
 
-from kiwixbuild.utils import SkipCommand, Remotefile
-from kiwixbuild._global import get_target_step
+from kiwixbuild.utils import pj, SkipCommand, Remotefile, extract_archive
+from kiwixbuild._global import get_target_step, neutralEnv
+import os, shutil
 
 class Icu(Dependency):
     name = "icu4c"
 
     class Source(ReleaseDownload):
-        archive = Remotefile('icu4c-73_3-src.tgz',
-                             '818a80712ed3caacd9b652305e01afc7fa167e6f2e94996da44b90c2ab604ce1',
-                             'https://github.com/unicode-org/icu/releases/download/release-73-2/icu4c-73_2-src.tgz')
+        archive_src = Remotefile('icu4c-73_3-src.tgz',
+                                 '818a80712ed3caacd9b652305e01afc7fa167e6f2e94996da44b90c2ab604ce1',
+                                 'https://github.com/unicode-org/icu/releases/download/release-73-2/icu4c-73_2-src.tgz')
+        archive_data = Remotefile('icu4c-73_2-data.zip',
+                                  'ca1ee076163b438461e484421a7679fc33a64cd0a54f9d4b401893fa1eb42701',
+                                  'https://github.com/unicode-org/icu/releases/download/release-73-2/icu4c-73_2-data.zip')
+
+
+        archives = [archive_src, archive_data]
+
+        def _extract(self, context):
+            context.try_skip(self.extract_path)
+            if os.path.exists(self.extract_path):
+                shutil.rmtree(self.extract_path)
+            extract_archive(
+                pj(neutralEnv('archive_dir'), self.archive_src.name),
+                neutralEnv('source_dir'),
+                topdir=None,
+                name=self.source_dir)
+            shutil.rmtree(pj(neutralEnv('source_dir'), self.source_dir, 'source', 'data'))
+            extract_archive(
+                pj(neutralEnv('archive_dir'), self.archive_data.name),
+                pj(neutralEnv('source_dir'), self.source_dir, 'source'),
+                topdir='data',
+                name='data'
+            )
+
         patches = [
         #           "icu4c_fix_static_lib_name_mingw.patch",
         #           "icu4c_android_elf64_st_info.patch",
@@ -49,3 +74,6 @@ def configure_option(self):
             if platformInfo.build in ('android', 'wasm'):
                 options += " --with-data-packaging=archive"
             return options
+
+        def set_env(self, env):
+            env['ICU_DATA_FILTER_FILE'] = pj(os.path.dirname(os.path.realpath(__file__)), "icu4c_data_filter.json")

kiwixbuild/dependencies/icu4c_data_filter.json

@@ -0,0 +1,14 @@
+{
+    "strategy": "additive",
+    "featureFilters": {
+        "lang_tree": "include",
+        "locales_tree": "include",
+        "translit": "include",
+        "misc": {
+            "includelist": [
+               "likelySubtags",
+               "metadata"
+            ]
+        }
+    }
+}