Switch to dash encoding for table/database/row-pk in paths

* Dash encoding functions, tests and docs, refs #1439
* dash encoding is now like percent encoding but with dashes
* Use dash-encoding for row PKs and ?_next=, refs #1439
* Use dash encoding for table names, refs #1439
* Use dash encoding for database names, too, refs #1439

See also https://simonwillison.net/2022/Mar/5/dash-encoding/
This commit is contained in:
Simon Willison 2022-03-07 07:38:29 -08:00 committed by GitHub
commit 1baa030eca
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
13 changed files with 173 additions and 53 deletions

View file

@ -17,6 +17,8 @@ from datasette.utils import (
InvalidSql,
LimitedWriter,
call_with_supported_arguments,
dash_decode,
dash_encode,
path_from_row_pks,
path_with_added_args,
path_with_removed_args,
@ -203,17 +205,17 @@ class DataView(BaseView):
async def resolve_db_name(self, request, db_name, **kwargs):
hash = None
name = None
db_name = urllib.parse.unquote_plus(db_name)
if db_name not in self.ds.databases and "-" in db_name:
decoded_name = dash_decode(db_name)
if decoded_name not in self.ds.databases and "-" in db_name:
# No matching DB found, maybe it's a name-hash?
name_bit, hash_bit = db_name.rsplit("-", 1)
if name_bit not in self.ds.databases:
if dash_decode(name_bit) not in self.ds.databases:
raise NotFound(f"Database not found: {name}")
else:
name = name_bit
name = dash_decode(name_bit)
hash = hash_bit
else:
name = db_name
name = decoded_name
try:
db = self.ds.databases[name]
@ -233,9 +235,7 @@ class DataView(BaseView):
return await db.table_exists(t)
table, _format = await resolve_table_and_format(
table_and_format=urllib.parse.unquote_plus(
kwargs["table_and_format"]
),
table_and_format=dash_decode(kwargs["table_and_format"]),
table_exists=async_table_exists,
allowed_formats=self.ds.renderers.keys(),
)
@ -243,11 +243,11 @@ class DataView(BaseView):
if _format:
kwargs["as_format"] = f".{_format}"
elif kwargs.get("table"):
kwargs["table"] = urllib.parse.unquote_plus(kwargs["table"])
kwargs["table"] = dash_decode(kwargs["table"])
should_redirect = self.ds.urls.path(f"{name}-{expected}")
if kwargs.get("table"):
should_redirect += "/" + urllib.parse.quote_plus(kwargs["table"])
should_redirect += "/" + dash_encode(kwargs["table"])
if kwargs.get("pk_path"):
should_redirect += "/" + kwargs["pk_path"]
if kwargs.get("as_format"):
@ -467,7 +467,7 @@ class DataView(BaseView):
return await db.table_exists(t)
table, _ext_format = await resolve_table_and_format(
table_and_format=urllib.parse.unquote_plus(args["table_and_format"]),
table_and_format=dash_decode(args["table_and_format"]),
table_exists=async_table_exists,
allowed_formats=self.ds.renderers.keys(),
)
@ -475,7 +475,7 @@ class DataView(BaseView):
args["table"] = table
del args["table_and_format"]
elif "table" in args:
args["table"] = urllib.parse.unquote_plus(args["table"])
args["table"] = dash_decode(args["table"])
return _format, args
async def view_get(self, request, database, hash, correct_hash_provided, **kwargs):

View file

@ -12,6 +12,7 @@ from datasette.utils import (
MultiParams,
append_querystring,
compound_keys_after_sql,
dash_encode,
escape_sqlite,
filters_should_redirect,
is_url,
@ -142,7 +143,7 @@ class RowTableShared(DataView):
'<a href="{base_url}{database}/{table}/{flat_pks_quoted}">{flat_pks}</a>'.format(
base_url=base_url,
database=database,
table=urllib.parse.quote_plus(table),
table=dash_encode(table),
flat_pks=str(markupsafe.escape(pk_path)),
flat_pks_quoted=path_from_row_pks(row, pks, not pks),
)
@ -199,8 +200,8 @@ class RowTableShared(DataView):
link_template.format(
database=database,
base_url=base_url,
table=urllib.parse.quote_plus(other_table),
link_id=urllib.parse.quote_plus(str(value)),
table=dash_encode(other_table),
link_id=dash_encode(str(value)),
id=str(markupsafe.escape(value)),
label=str(markupsafe.escape(label)) or "-",
)
@ -765,7 +766,7 @@ class TableView(RowTableShared):
if prefix is None:
prefix = "$null"
else:
prefix = urllib.parse.quote_plus(str(prefix))
prefix = dash_encode(str(prefix))
next_value = f"{prefix},{next_value}"
added_args = {"_next": next_value}
if sort: