filtering custom extractor that uses the js extractor for svelte files (

#5302)

* use custom extractor for svelte files

* use a different, and perhaps more-solid approach

babeljs.cfg

@@ -1,8 +1,9 @@
 [extractors]
-jinja2_custom = kitsune.lib.babel:extract_jinja 
+jinja2_custom = kitsune.lib.babel:extract_jinja
+svelte_custom = kitsune.lib.babel:extract_svelte
 
 [ignore: kitsune/**/static/**/js/*-all.js]
 [ignore: kitsune/**/static/**/js/*-min.js]
 [javascript: kitsune/**/static/**/js/*.js]
-[javascript: svelte/**/*.svelte]
+[svelte_custom: svelte/**/*.svelte]
 [jinja2_custom: kitsune/**/static/**/tpl/**.njk]

kitsune/lib/babel.py

@@ -1,8 +1,14 @@
+from io import BytesIO
+
+from babel.messages.extract import extract_javascript
 import django
 from django.conf import settings
 from jinja2.ext import babel_extract
 
 
+SVELTE_SPECIAL_TAG_PREFIXES = set([b"{#", b"{:", b"{/", b"{@"])
+
+
 def generate_option(value):
  """
  Generate option to pass to babel_extract() from a TEMPLATES['OPTION'] value
@@ -18,6 +24,36 @@ def generate_option(value):
  return value
 
 
+def extract_svelte(fileobj, keywords, comment_tags, options):
+ """
+ Custom extractor for Svelte files.
+
+ Since Babel's built-in JavaScript extractor is designed to handle JSX
+ files as well plain JavaScript files, it can also handle nearly all of
+ the syntax within a Svelte file. What it doesn't always handle well are
+ the statements that open, continue, and close Svelte logic blocks. They
+ sometimes confuse it such that it fails to recognize some localizable
+ strings. This custom extractor essentially hides those statements from
+ the underlying Babel JavaScript extractor, while maintaining the same
+ line numbers as the original Svelte file.
+
+ NOTE: This approach assumes that any special Svelte tags always exist
+ on their own line without anything else on that line. Of course,
+ this approach is not as elegant as using a full-blown parser for
+ Svelte files, but it's much easier and will work for our current
+ practical purposes.
+ """
+ filtered = BytesIO()
+ for line in fileobj:
+ if line.lstrip()[:2] in SVELTE_SPECIAL_TAG_PREFIXES:
+ # Write a new line to maintain equivalent line numbers.
+ filtered.write(b"\n")
+ else:
+ filtered.write(line)
+ filtered.seek(0)
+ return extract_javascript(filtered, keywords, comment_tags, options)
+
+
 def extract_jinja(fileobj, keywords, comment_tags, options):
  """
  Wrapper around jinja2's babel_extract() that sets the relevant options by