archives: fixed bugs with serving archives from non-ascii repos, and also deliver archives at much bigger reading blocks for faster downloads

rhodecode/apps/repository/views/repo_files.py

@@ -38,7 +38,7 b' from rhodecode.lib import diffs, helpers'
 from rhodecode.lib import audit_logger
 from rhodecode.lib.hash_utils import sha1_safe
 from rhodecode.lib.rc_cache.archive_cache import get_archival_cache_store, get_archival_config, ReentrantLock
-from rhodecode.lib.str_utils import safe_bytes
+from rhodecode.lib.str_utils import safe_bytes, convert_special_chars
 from rhodecode.lib.view_utils import parse_path_ref
 from rhodecode.lib.exceptions import NonRelativePathError
 from rhodecode.lib.codeblocks import (
@@ -63,17 +63,17 b' from rhodecode.model.db import Repositor'
 log = logging.getLogger(__name__)
 
 
-def get_archive_name(db_repo_name, commit_sha, ext, subrepos=False, path_sha='', with_hash=True):
+def get_archive_name(db_repo_id, db_repo_name, commit_sha, ext, subrepos=False, path_sha='', with_hash=True):
     # original backward compat name of archive
-    clean_name = safe_str(db_repo_name.replace('/', '_'))
+    clean_name = safe_str(convert_special_chars(db_repo_name).replace('/', '_'))
 
-    # e.g vcsserver-sub-1-abcfdef-archive-all.zip
-    # vcsserver-sub-0-abcfdef-COMMIT_SHA-PATH_SHA.zip
-
+    # e.g vcsserver-id-abcd-sub-1-abcfdef-archive-all.zip
+    # vcsserver-id-abcd-sub-0-abcfdef-COMMIT_SHA-PATH_SHA.zip
+    id_sha = sha1_safe(str(db_repo_id))[:4]
     sub_repo = 'sub-1' if subrepos else 'sub-0'
     commit = commit_sha if with_hash else 'archive'
     path_marker = (path_sha if with_hash else '') or 'all'
-    archive_name = f'{clean_name}-{sub_repo}-{commit}-{path_marker}{ext}'
+    archive_name = f'{clean_name}-id-{id_sha}-{sub_repo}-{commit}-{path_marker}{ext}'
 
     return archive_name
 
@@ -396,7 +396,7 b' class RepoFilesView(RepoAppView):'
 
         # used for cache etc, consistent unique archive name
         archive_name_key = get_archive_name(
-            self.db_repo_name, commit_sha=commit.short_id, ext=ext, subrepos=subrepos,
+            self.db_repo.repo_id, self.db_repo_name, commit_sha=commit.short_id, ext=ext, subrepos=subrepos,
             path_sha=path_sha, with_hash=True)
 
         if not with_hash:
@@ -404,7 +404,7 b' class RepoFilesView(RepoAppView):'
 
         # what end client gets served
         response_archive_name = get_archive_name(
-            self.db_repo_name, commit_sha=commit.short_id, ext=ext, subrepos=subrepos,
+            self.db_repo.repo_id, self.db_repo_name, commit_sha=commit.short_id, ext=ext, subrepos=subrepos,
             path_sha=path_sha, with_hash=with_hash)
 
         # remove extension from our archive directory name
@@ -444,9 +444,10 b' class RepoFilesView(RepoAppView):'
         if not reader:
             raise ValueError('archive cache reader is empty, failed to fetch file from distributed archive cache')
 
-        def archive_iterator(_reader):
+        def archive_iterator(_reader, block_size: int = 4096*512):
+            # 4096 * 64 = 64KB
             while 1:
-                data = _reader.read(1024)
+                data = _reader.read(block_size)
                 if not data:
                     break
                 yield data

rhodecode/lib/vcs/backends/base.py

@@ -1222,7 +1222,7 b' class BaseCommit(object):'
         Creates an archive containing the contents of the repository.
 
         :param archive_name_key: unique key under this archive should be generated
-        :param kind: one of following: ``"tbz2"``, ``"tgz"``, ``"zip"``.
+        :param kind: one of the following: ``"tbz2"``, ``"tgz"``, ``"zip"``.
         :param archive_dir_name: name of root directory in archive.
             Default is repository name and commit's short_id joined with dash:
             ``"{repo_name}-{short_id}"``.
@@ -1238,8 +1238,7 b' class BaseCommit(object):'
         allowed_kinds = [x[0] for x in settings.ARCHIVE_SPECS]
         if kind not in allowed_kinds:
             raise ImproperArchiveTypeError(
-                'Archive kind (%s) not supported use one of %s' %
-                (kind, allowed_kinds))
+                f'Archive kind ({kind}) not supported use one of {allowed_kinds}')
 
         archive_dir_name = self._validate_archive_prefix(archive_dir_name)
         mtime = mtime is not None or time.mktime(self.date.timetuple())